Antenatal screening for haemoglobinopathies in primary care: a cohort study and cluster randomised trial to inform a simulation model. The Screening for Haemoglobinopathies in First Trimester (SHIFT) trial

Article history

The research reported in this issue of the journal was commissioned by the HTA programme as project number 03/02/03. The contractual start date was in October 2004. The draft report began editorial review in August 2008 and was accepted for publication in July 2009. As the funder, by devising a commissioning brief, the HTA programme specified the research question and study design. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HTA editors and publisher have tried to ensure the accuracy of the authors’ report and would like to thank the referees for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

None

Copyright statement

© 2010 Queen’s Printer and Controller of HMSO. This journal is a member of and subscribes to the principles of the Committee on Publication Ethics (COPE) (http://www.publicationethics.org/). This journal may be freely reproduced for the purposes of private research and study and may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NETSCC, Health Technology Assessment, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

2010 Queen’s Printer and Controller of HMSO

Haemoglobinopathies

Haemoglobinopathies, including sickle cell disorders and thalassaemia, are inherited disorders of haemoglobin. It is estimated that 7% of the world’s population are carriers of a potentially pathological haemoglobin gene. 1 Haemoglobinopathies are among the commonest recessively inherited disorders in north-west Europe. There is a significant morbidity and mortality associated with haemoglobinopathies.

Global context

Each year up to half a million births worldwide are affected by a clinically significant form of haemoglobinopathy. 1 Haemoglobinopathies most commonly occur in populations originating from Africa, Asia and the Mediterranean. Due to patterns of migration, they are also seen in parts of the world such as the Caribbean, UK, Northern Europe and North America. 2

In 2006, the Secretariat of the World Health Organization (WHO) acknowledged the global burden of haemoglobinopathies, expressed concern that they were ‘not officially recognised as priorities in public health’, and ‘deplored the inequality of access to safe and appropriate genetic services’. 3 In response to this report, the 59th World Health Assembly urged member states to implement comprehensive programmes for the prevention and management of sickle cell anaemia. 4

Across the world, different screening strategies have been implemented. In Iran, for example, a preconceptual screening programme was introduced in 1996, followed by an antenatal programme in 2001;5 in Belgium and France, neonatal screening for sickle cell disease is offered,6 and in the Netherlands a pilot study of antenatal sickle cell and thalassaemia (SCT) screening has recently been completed. 7

Sickle cell and thalassaemia in the UK

Sickle cell disorders are estimated to affect more than 12,500 people in the UK, with approximately 240,000 healthy carriers. 8 The highest prevalence is among people of Caribbean and African ancestry. Beta thalassaemia is thought to affect more than 700 people, with approximately 214,000 healthy carriers. 8 The highest prevalence is among people of Eastern Mediterranean and Asian origin.

Screening policy in the UK

The NHS Sickle Cell and Thalassaemia Screening Programme (NHS SC&T Screening Programme) was set up in 2001, providing the world’s first linked antenatal and neonatal screening service. The overall aim of the programme is ‘to offer timely antenatal sickle cell and thalassaemia screening to all women (and couples) to facilitate informed decision making’. 8 The programme provides universal antenatal screening in high-prevalence areas (prevalence greater than 1.5/10,000 births) and selective screening in low-prevalence areas. 8–10 The programme aims to offer screening by 10 weeks’ gestation, to enable screening, prenatal diagnosis (PND) and any subsequent action to be completed by 12 weeks’ gestation. This is in line with current National Institute for Health and Clinical Excellence (NICE)11 and National Service Framework (NSF) guidelines. 12 NICE guidance explicitly states there should be two appointments, early in pregnancy, to ensure that women have the opportunity to make informed choices about pregnancy.

Currently, screening is offered to pregnant women in a variety of settings. The location of such appointments varies across the country: they may take place in hospital, in the woman’s home or in a community-based clinic. Use of a community-based clinic (which we have designated ‘community-based secondary care’) appears to form the most common pattern of screening in the UK, especially in Greater London, where about half of the UK minority ethnic population live. 13 We have therefore used this pattern of screening as the comparison against which screening in general practice will be judged.

The problem

Many at-risk pregnant women, and the fathers of their baby, are offered SCT carrier tests too late or the process of screening and diagnostic testing is taking too long to allow couples the full range of reproductive options. Reproductive options, such as prenatal diagnostic testing for carrier couples, and, when an affected pregnancy is identified, the offer of a termination, may be curtailed by the passage of time. Given the recent statutory requirements to reduce disparities between ethnic and social groups in the provision of health care,14 it is germane to consider how antenatal screening for SCT can be organised to minimise the failure to offer screening early in pregnancy, the basic premise upon which the policy of universal screening is based. 9

The variation in uptake of prenatal diagnostic tests is associated with the gestational age in pregnancy that the screening test is offered. For example, 73% of British Pakistani women accepted PND for thalassaemia when it was offered in the first trimester compared with 39% when offered in the second trimester. 15 This may be a particular problem for women whose pregnancies are at risk for sickle cell disease, with one retrospective study reporting that the mean gestation at which women who were carriers for sickle cell were first seen for counselling was 15 weeks, compared to 12.3 weeks for women who were carriers for thalassaemia. 16

There are several reasons why delays occur in couples learning about their carrier status. Some laboratories produce screening test results within 48 hours,17 whereas others take 5 days or more. 13 In addition, once results are known, it can take anything from 6 to 34 days for women to receive an appointment for counselling about diagnostic testing options. 18 Thus the speed at which laboratories process samples and report screening results can limit couples’ opportunities to make informed choices.

The NHS SC&T Screening Programme has set standards for the time taken for laboratories to report a result (3 working days) and for the time interval between informing women of their carrier status and offering an appointment for a consultation to discuss diagnostic testing options (5 working days).

Reducing delays in the screening process: the proposed trial

We propose to assess two ways of delivering antenatal screening for sickle cell and thalassaemia that have the potential to reduce delays in the screening process: offering screening at the time when women first report their pregnancies in primary care, and offering testing to the baby’s father at the same time as pregnant women are offering screening.

Cost-effectiveness

If the use of early testing in primary care is shown to be associated with improvements in the screening process (e.g. increases in the proportion of pregnant women undergoing screening by 10 weeks’ gestation) compared with those offered screening in community-based secondary care, then it is likely that there will be cost implications for both the health-care sector and for the screened women. For example, given that one in nine pregnancies spontaneously miscarry in the first trimester,20 some women who will undergo early testing in the primary care arm of the trial would not have been screened if the testing were delayed until the booking. In addition, parallel testing of fathers in primary care compared with the sequential testing of fathers, if accepted, will have cost implications. Again this is because of the increased risk of miscarriage in the first trimester and because the father is being tested before the results of the woman’s test is known, and it is likely that the woman may not be a carrier. Therefore, the economic evaluation will take a broad perspective and consider costs falling both on the NHS and on patients.

Future patterns of care

It is likely that as the NHS SC&T Screening Programme becomes established and the community becomes familiar with it, so the ways in which screening will be conducted will change. For example, carrier testing may take place before pregnancy, possibly linked with contraceptive advice. There will, however, always be a need to run several models of care, of which the early offer of screening in pregnancy will be an important one. 30 It is unclear whether this is best conducted by a GP, a practice nurse or a midwife. Discussions with members of primary care teams suggest that, given the way care is currently organised, the pattern we propose is the most feasible way of delivering an early offer of screening.

Aims and objectives

The trial aims to determine the viability of offering screening in primary care when women first report their pregnancy to their GP.

Objectives of SHIFT trial

• To assess and compare the effectiveness, acceptability and feasibility of offering antenatal SCT screening:

  1. – In primary care with parallel father testing, i.e. test offered to mothers and fathers at the same time.

  2. – In primary care with sequential father testing, i.e. test offered to mothers. The test is offered to father only if mother identified as carrier.

  3. – In secondary care with sequential father testing, i.e. test offered to mothers. The test is offered to father only if mother identified as carrier.

• To model the cost-effectiveness of the three patterns of care.

Outline of the monograph

Chapter 2 presents the population-based data that were used to estimate a trial baseline regarding the delay between pregnancy confirmation in primary care and time of antenatal SCT screening. Chapter 3 describes the training programme developed to enable HCPs to integrate the offer of SCT screening into pregnancy confirmation visits in primary care as required by the interventions that formed part of the SHIFT Trial.

The trial design and results are presented in Chapter 4 (clinical effectiveness), Chapter 5 (cost-effectiveness and modelling), Chapter 6 (informed choice), Chapter 7 (acceptability to women) and Chapter 8 (feasibility in primary care). Chapter 9 summarises the main findings of the trial and makes recommendations for future policy, practice and research.

Introduction

Using population-based data obtained prior to the intervention phase of the trial, we report the delay between pregnancy confirmation in primary care and time of antenatal SCT screening, and examine the extent to which patient and practice characteristics are associated with the delays. The data reported in this chapter are published. 31

Method

A cohort study was conducted of all pregnancies reported in 25 general practices from two primary care trusts (PCTs).

Procedure

Anonymised data were collected for the first day of the last menstrual period (LMP), the date of the pregnancy confirmation visit in primary care and the date of the test, as defined by the date of venesection. These were used to calculate (1) gestational age at time of pregnancy confirmation in primary care and (2) gestational age at time of SCT screening. Screening was ascertained up to 26 weeks’ gestation. Women who were not tested by 26 weeks’ gestation were classified as not tested for the purpose of prenatal screening of the fetus.

Routinely collected data for date of birth, parity, ethnicity, previously determined SCT carrier status, date of antenatal SCT screening, and termination of pregnancy or miscarriage were also extracted from primary care computer systems. The number of pregnancies reported at each practice was checked by comparing with the number of maternity referrals made by each practice. Date of antenatal SCT screening was obtained from maternity units and laboratories at hospitals if not available through primary care.

Data for times were skewed and are reported as medians [interquartile ranges (IQRs)]. The distribution for the median delay for each participating general practice was evaluated. A multiple linear regression model was implemented, with delay between pregnancy confirmation visit and screening for each patient as the dependent variable. Explanatory variables were maternal age (continuous), parity (primiparous, multiparous and ‘not known’), and ethnicity (North European, South European, African and African Caribbean, South Asian, other and mixed and not known). Our approach was informed broadly by census categories to enable comparison with previous studies. For analysis, women classified as: South European, African and African Caribbean, South Asian, other or mixed ethnicity were grouped as ‘higher risk’. A random effects model, with maximum likelihood estimation, fitted using the ‘xtreg’ command in stata version 9, was used to allow for clustering by practice. 33

Results

There were 2062 pregnancies from 25 practices that completed the run-in data collection phase.

Cases excluded from analysis:

• date of LMP unknown (n = 62)

• SCT carrier status already known, i.e. recorded in practice records prior to current pregnancy (n = 299)

• termination of pregnancy for reasons other than fetal abnormality (n = 157)

• miscarriage (n = 117)

• pregnancy confirmed after 26 weeks (n = 126).

There were 1441 eligible women who intended to proceed with their pregnancies. The women’s median (IQR) age at LMP was 28.5 (24.0 to 33.5), and 767 (53%) were primiparous. Routinely recorded data for ethnicity were available for 837 (58%) women, with 645 (45%) from ‘higher-risk’ groups, including ‘African Caribbean and Black African’ (228), ‘South European and Other European’ (99), ‘South and South East Asian’ (260), ‘Other ethnicity’ (47) and ‘Mixed ethnicity’ (11).

From the 1441 pregnancies eligible for analysis, there were 965 (67%) women who had SCT screening tests performed before 26 weeks’ gestation. The proportion screened did not vary by ethnicity. The median (IQR) gestational age at the pregnancy confirmation visit in primary care for all women was 7.6 weeks (6.0 to 10.1 weeks). The gestational age at the pregnancy confirmation visit for women who were screened by 26 weeks was 7.4 (5.9 to 9.9) weeks.

Among women with reported ethnicity available, there was a weak association between ethnicity and gestational age at booking after adjusting for maternal age, parity and clustering (p = 0.023). The median gestational age at booking ranged from 7.1 weeks in ‘Northern Europeans’ to 8.4 weeks in ‘African Caribbean and Black African’ groups.

Of the 965 women who underwent SCT screening, the gestational age at screening was 15.3 weeks (IQR 12.6 to 18.0 weeks), with 4.4% being screened by 10 weeks (Figure 1). The median delay between pregnancy confirmation in primary care and antenatal SCT screening in these women was 6.9 weeks (4.7 to 9.3 weeks).

FIGURE 1.

Distributions for gestational age at first pregnancy confirmation visit and sickle cell and thalassaemia screening for women whose pregnancies were confirmed and screened before 26 weeks’ gestation.

Variation between practices was greater than expected by chance (p < 0.001). At the practice with the shortest delays, the median time from pregnancy confirmation in primary care to screening was 3.7 weeks; at the practice with the highest delay, the median time from pregnancy confirmation in primary care to screening was 10.0 weeks.

After allowing for practice level variation, there was no association between delay times and the measured individual patient characteristics, including maternal age, parity and ethnic group. Specifically, ethnicity was not associated with delay times, and variation between practices was not accounted for by adjusting for ethnicity. In addition, there was no association between ethnicity and gestational age at screening (p = 0.351) or with the proportion screened by 26 weeks (p = 0.060).

Discussion

These population-based data show a 7-week delay between pregnancy confirmation in primary care and antenatal SCT screening. This delay undermines the policy and practice of facilitating informed reproductive choices early in pregnancy. These data are supported by a small retrospective study of women at risk of a sickle cell disorder that showed that some women did not have screening until 15 weeks’ gestation, many weeks after reporting their pregnancies in primary care. 19 Furthermore, one-third of women across all ethnic groups reported their pregnancies to their GPs early in pregnancy but did not undergo SCT screening before 26 weeks’ gestation. The data collected do not indicate whether this represents a service failure or individual choice not to undergo screening.

Even in this deprived inner city setting, women generally attended their general practices early in pregnancy for confirmation. The delay before screening uptake was not associated with individual women’s age, parity or ethnicity. There was, however, considerable variation in the extent of delays between general practices, with the best performing practices having a median delay of just over 3 weeks from pregnancy confirmation in primary care to screening. Practice level variation may result from delays in primary care referring women for antenatal care or delays in secondary care acting on these referrals. Individual patient factors did not appear to be associated with delays in screening uptake. The results suggest there is considerable scope for facilitating timely screening through improved organisation and delivery of antenatal care.

Comparison with existing literature

It has been suggested that the delay between pregnancy confirmation in primary care and the first midwifery appointment is greater for women from minority ethnic groups than for other women. 21,22,34 While there was some evidence that women from some minority ethnic groups attend primary care to confirm their pregnancies about 1 week later than other women, there was no evidence that the 7-week delay between pregnancy confirmation and testing was associated with ethnicity, indicating that delays within the health service, rather than women reporting late to confirm their pregnancies, are responsible for the large observed delay in testing.

Strengths and limitations

This is the first study to examine the delay between pregnancy confirmation in primary care and antenatal SCT screening using population-based data. Limitations of the data include:

• General practices observed were not randomly selected from the 123 general practices in the two PCTs studied and therefore may not be representative of the practices in these PCTs. The variables on which we have data show participating practices were similar in practice size, deprivation and ethnicity to other practices in the study PCTs.

• Study PCTs were chosen to represent geographical areas with a high prevalence of SCT disorders and may not be representative of other areas. They represent deprived inner city areas where general practice lists are longer but primary care services are generally less well organised than those in suburban or rural areas. 35

• Ethnicity data were not collected systematically across the practices and were available for about half of the women. This reflects a shortfall in the collecting of basic demographic data in primary care.

• Accuracy of the collected data was dependent on the reliability and validity of the data held in GP record systems. Following fidelity checks on the number of pregnancies, we are confident that the data presented here are an accurate representation of the number of pregnancies. The reliability of data collected about these pregnancies is not known.

Conclusion

The considerable delay between pregnancy confirmation in primary care and antenatal SCT screening may deprive couples of opportunities to opt for PND early in pregnancy. There is a need to evaluate models of care to reduce delays. The following chapter describes the development and evaluation of brief training for HCPs in primary care, in preparation for the trial aimed at evaluating primary-care-based models to reduce the delay observed in these population cohort data.

Introduction

Communication skills training is needed to help primary care HCPs integrate the offer of antenatal SCT screening into their pregnancy-confirmation consultations. Effective training requires HCP attendance, and acquiring the necessary knowledge, skills and attitudes. Offering training on-site increases the convenience of training for participants. Providing payments to practices to reimburse them for locum cover costs incurred should also maximise attendance. Additionally, putting measures in place for training to be disseminated to non-attenders could increase the reach of the training.

There is evidence that active trainee involvement and the use of simulated patients are effective methods for developing clinical communication skills. Furthermore, using simulated patients that behave in a repeated and standardised way (standardised patients) is an effective method of training.

The assessed outcomes are based on Cervano’s evaluation framework for continuing professional education. The framework includes the following dimensions:

• programme design and implementation

• learner participation (i.e. attendance)

• learner satisfaction (i.e. perceived usefulness of training)

• learner knowledge, skills and attitudes (i.e. self reported comfort and confidence with offer of screening)

• application of learning after the programme (i.e. offer of screening)

• impact of application of learning (i.e. test uptake and gestational age at test uptake).

Methods

Results

Learner participation

In total, 126 HCPs (87 GPs, 39 practice nurses) from the 17 practices were invited to attend a training session and 62% (78/126: 53 GPs, 25 practice nurses) attended. Those who joined the practices after the start of the trial (n = 49; 44 GPs, five nurses) were offered a training session but this did not involve simulated patients (Figure 2).

FIGURE 2.

Flow of health-care professionals through the training process.

Discussion

Despite efforts to maximise attendance (including provision of on-site training, the timing of training sessions being negotiated with practice staff, payments for locum cover, experts in the field leading the sessions), only 62% (78/126) of invited HCPs attended. Self-reported reasons for non-attendance included holiday, sickness, and joining the practice after the start of the trial. However, for some lack of attendance could reflect lack of motivation, low prioritisation or negative attitudes towards screening. The observed rate of attendance suggests the need for the continuous offer and availability of training to provide opportunities for non-attenders who are motivated but were unable to attend initial training. Ascertaining reasons for non-attendance more systematically would allow training to be tailored to overcome these barriers. In addition, there were 49 HCPs, not in post at the start of the trial, who were involved in pregnancy-confirmation consultations. While updates and reminders were sent to practices for new joiners and locums, the majority of this group did not take part in a training session with a simulated patient. Thought needs to be given to training of a mobile group.

Health-care professionals’ self-assessment of the usefulness of training, their confidence, knowledge and skill all increased with training. Given the study design, caution is needed in attributing these changes to the training per se, and in assuming that self-reported skills and knowledge strongly predict actual skills and knowledge. Screening was, however, offered more frequently by trained HCPs, suggesting that training had been effective in initiating the offer of screening. In addition, women who consulted a trained HCP were offered screened earlier than women consulting HCPs who were untrained and either in post at the time of training or who joined the practice subsequently.

Those attending for training may have differed from those who were invited but did not attend, in being more motivated to offer screening and this, not training, may have accounted for their greater likelihood of offering screening and the earlier gestational age at which women were screened. Evidence to suggest that training was at least in part responsible for this outcome comes from the observation that those who did not attend because they were not in post at the time of training behaved similarly to those who did not attend training, despite being in post at the time training was offered.

Training was identified as a major influence on the consultation and the opportunity to practice skills was found to be valuable. Perhaps not surprisingly, the time required for training was identified as a problem.

Strengths and limitations

The strength of this study was the use of mixed methods to evaluate training, in particular the objective measure of the clinical impact of training, i.e. the offer of screening and the gestational age of women at screening. The main limitation was the use of an observational before and after design, limiting our ability to attribute changes to the training.

Implications for future research or clinical practice

Using brief communication skills training to facilitate HCPs’ integration of the offer of antenatal screening into their pregnancy-confirming consultations is feasible and may be effective. Although active learning methods were perceived to be useful, it is important to bear in mind that individuals may have different learning style preferences. Further work investigating factors that affect training attendance is needed. There may be a role for computerised training programmes (e.g. BMJ Learning), with the integration of a scored online examination.

The training discussed in this chapter was carried out in all of the practices randomised to primary care parallel and primary care sequential prior to the trial.

Section I: trial methods

An abstract of the protocol for the trial is published on the Lancet website. 37 This, together with a copy of the full trial protocol, is included in Appendix 2.

Participants

Figure 3 shows a CONSORT flow diagram, illustrating the flow of participants through the SHIFT Trial.

FIGURE 3.

CONSORT diagram of flow of participants through SHIFT.

Practices

There were 123 practices in the two PCTs, and four did not agree to randomisation. This resulted in 119 eligible practices being invited to participate in the trial. There were 29 practices that participated in the trial, with two serving as pilot sites. Of the 27 practices that were randomised, two withdrew, and 25 completed the trial (Figure 3). It was not possible to obtain data from the practices that withdrew from the study, so an intention-to-treat analysis was not feasible. Data from the run-in phase are described in Chapter 2. In analyses for Chapter 4, we additionally excluded women who confirmed their pregnancies after 20 weeks’ gestation, giving 1390 women from the run-in phase as eligible to contribute to trial analyses.

Pregnant women

In the intervention phase, there were 2421 pregnancies identified in the 25 practices. Of these, 1708 pregnancies were eligible to be assessed for the primary outcome measure (Figure 3). There were 17 women with more than one pregnancy but only one woman was eligible for analysis in both pregnancies.

Cases excluded from the analysis:

• LMP was not known (n = 50).

• Carrier status was already known, i.e. recorded patient’s notes in primary care (n = 189).

• Termination of pregnancy for reasons other than fetal abnormality (n = 201).

• Miscarriage (n = 138).

• Pregnancy confirmation visit was after 20 weeks’ gestation (n = 248).

There were 87 women who were excluded on two or more criteria. There were 41 women who had a miscarriage after SCT testing and 28 women who had a termination of pregnancy after SCT testing and these were included.

Section II: primary outcome

Primary outcome measure: time of uptake and proportion screened

In both intervention groups, more women were screened before 70 days’ gestation than in the standard care group (Table 6, and Figures 4 and 5). Table 6a gives the primary outcome by family practice and group. Table 7 shows the cumulative proportion of women screened by gestational age and trial arm.

TABLE 6a - Screening outcomes by intervention group

In the standard care group, offer of test was ascertained for questionnaire respondents only.

In women who were screened before 182 days (26 weeks’ gestation).

Figures are frequencies (column per cent), except where indicated.

	Frequency (%)
	Standard care (441)	Primary care parallel (677)	Primary care sequential (590)
Women’s uptake of screening test < 70 days’ gestation	9 (2)	161 (24)	167 (28)
Women’s uptake of screening before 182 days’ gestation	324 (73)	571 (84)	481 (82)
Offer of screening test before 70 days’ gestation	3/90a(3)	321 (47)	281 (48)
Delay from pregnancy confirmation visit to screening uptake [days, mean (IQR)]b	60 (42 to 79)	35 (7 to 59)	31 (5 to 54)
Gestational age at screening uptakeb [days, mean (IQR)]	118 (101 to 134)	94 (66 to 118)	90 (61 to 113)

TABLE 6b - Primary outcome by family practice and group

Figures are number of subjects screened < 70 days’ gestation/number of eligible subjects at practice (row per cent).

Secondary care			Family practice parallel partner			Family practice sequential partner
Practice ID	Screened run-in (%)	Screened intervention (%)	Practice ID	Screened run-in (%)	Screened intervention (%)	Practice ID	Screened run-in (%)	Screened intervention (%)
A	2/37 (5)	1/33 (3)	I	2/33 (6)	7/33 (21)	Q	3/49 (6)	9/59 (15)
B	4/33 (12)	4/36 (11)	J	4/73 (5)	5/62 (8)	R	1/52 (2)	7/56 (13)
C	4/70 (6)	3/85 (4)	K	20/90 (22)	35/84 (42)	S	0/30 (0)	21/62 (34)
D	1/35 (3)	0/44 (0)	L	1/32 (3)	2/61 (3)	T	0/61 (0)	23/64 (36)
E	0/32 (0)	0/98 (0)	M	4/146 (3)	39/156 (25)	U	7/55 (13)	10/54 (19)
F	0/32 (0)	1/36 (3)	N	4/41 (10)	22/62 (35)	V	0/45 (0)	47/75 (63)
G	0/35 (0)	0/42(0)	O	0/57 (0)	25/108 (23)	W	0/37 (0)	29/69 (42)
H	1/62 (2)	0/67 (0)	P	4/122 (3)	26/111 (23)	X	0/64 (0)	13/87 (15)
						Y	1/67 (1)	8/64 (13)
	12/336 (4)	9/441 (2)		39/594 (7)	161/677 (24)		12/460 (3)	167/590 (28)

FIGURE 4.

Distributions for gestational age at first pregnancy-confirmation consultation (black bars) and at sickle cell and thalassaemia screening (white bars) by randomisation group. The proportions not screened by 26 weeks’ gestation were: standard care, 27%; primary care parallel, 16%; and primary care sequential 18%.

FIGURE 5.

Uptake of screening by gestational age (a) and by time since pregnancy-confirmation consultation in primary care (b).

TABLE 7 - Cumulative uptake of screening tests by gestational age and intervention group

Gestational age (completed weeks)	Uptake percentage (frequency)
	Standard care (441)	Primary care parallel (677)	Primary care sequential (590)
0–1	0	0.2 (1)	0
2–3	0	0	0.3 (2)
4–5	0.5 (2)	2.7 (18)	3.7 (22)
6–7	1.0 (4)	12.0 (81)	13.2 (78)
8–9	2.0 (9)	23.8 (161)	28.3 (167)
10–11	7.9 (35)	33.4 (226)	36.8 (217)
12–13	16.3 (72)	45.6 (309)	48.1 (284)
14–15	28.3 (125)	57.6 (390)	59.8 (353)
16–17	45.8 (202)	67.8 (459)	68.6 (405)
18–19	59.6 (263)	77.3 (523)	74.4 (439)
20–21	68.5 (302)	80.4 (544)	77.8 (459)
22–23	70.7 (312)	82.9 (561)	80.2 (473)
24–25	73.5 (324)	84.3 (571)	81.5 (481)
Not screened	117	106	109

After adjusting for age group, parity, family origin risk status, screening uptake in the run-in phase, partnership size and PCT, the increase in screening uptake was 16.5% (95% CI 7.1 to 25.8, p = 0.002) for primary care parallel, and 27.8% (95% CI 14.8 to 40.7, p < 0.001) for primary care sequential (Table 8). Adjustment for baseline screening performance had a larger impact on the estimates for the parallel testing group because screening uptake in the run-in phase was highest in this group. There was no difference in the proportion of women screened before 70 days’ gestation between the two primary care groups: offering parallel father testing neither facilitated nor impeded maternal SCT testing uptake (Table 8a).

TABLE 8a - Estimated effect of intervention on screening outcomes, cluster level analysis in primary care

Adjusted for Model 1 and proportion screened before 70 days’ gestation in run-in period.

Adjusted for Model 1 and uptake of screening before 182 days in the run-in period.

Adjusted for Model 1: age group, parity, proportion of ‘high-risk’ ethnic groups, PCT and number of GP partners at practice.

Adjusted for Model 1 and mean time interval from pregnancy confirmation to screening in the run-in period.

Adjusted for Model 1 and mean gestational age at screening in the run-in period.

Setting		Parallel (677)		Sequential (590)
		Estimate (95% CI)	p-value	Estimate (95% CI)	p-value
Uptake of screening test before 70 days’ gestation	Adjusted per cent differencea	16.5 (7.12 to 25.8)	0.002	27.8 (14.8 to 40.7)	<0.001
Uptake of screening before 182 days’ gestation	Adjusted per cent differenceb	3.80 (–7.93 to 15.54)	0.502	9.83 (–2.30 to 22.0)	0.105
Offer of screening test before 70 days’ gestation	Adjusted per cent differencec	39.2 (26.0 to 52.4)	< 0.001	44.2 (26.6 to 61.9)	< 0.001
Time from pregnancy confirmation to screening	Adjusted mean difference (days)d	–18.0 (–26.3 to –9.7)	< 0.001	–21.5 (–32.5 to –10.4)	0.001
Gestational age at screening	Adjusted mean difference (days)e	–15.5 (–26.4 to –4.63)	0.008	–21.8 (–34.8 to –8.80)	0.003

TABLE 8b - Estimated difference between primary care parallel and primary care sequential trial arms, cluster level analysis

Adjusted for Model 1 and proportion screened before 70 days’ gestation in run-in period.

Adjusted for Model 1 and uptake of screening before 182 days in the run-in period.

Adjusted for Model 1: age group, parity, proportion of ‘high-risk’ ethnic groups, PCT and number of GP partners at practice.

Adjusted for Model 1 and mean time interval from pregnancy confirmation to screening in the run-in period.

Adjusted for Model 1 and mean gestational age at screening in the run-in period.

		Difference between active intervention arms
		Estimate (95% CI)	p-value
Uptake of screening test before 70 days’ gestation	Adjusted per cent differencea	13.5 (–4.3 to 31.4)	0.119
Uptake of screening before 182 days’ gestation	Adjusted per cent differenceb	–6.08 (–21.6 to 9.48)	0.399
Offer of screening test before 70 days’ gestation	Adjusted per cent differencec	3.98 (–16.5 to 24.4)	0.674
Time from pregnancy confirmation to screening	Adjusted mean difference (days)d	–3.84 (–16.7 to 9.01)	0.516
Gestational age at screening	Adjusted mean difference (days)e	6.05 (–4.88 to 16.9)	0.242

Section III: secondary outcomes

Testing of maternal carriers and fathers

Testing of SCT carrier mothers

Among 2421 pregnancies, there were 160 women identified as SCT carriers. After excluding 17 women who met other criteria for ineligibility, there were 50 women whose carrier status was already known and 93 women who were newly identified as carriers in this trial (Figure 6). One woman, whose screening test was at 183 days’ gestation, and another woman, whose test did not originate in primary care, were included as newly identified carriers.

FIGURE 6.

Flow chart showing women identified as carriers.

The characteristics of carrier women are shown in Table 10. There were five women whose ethnicity was given as ‘British’, who were coded as ‘Northern European’. Women of African or Caribbean ethnicity accounted for more than half of carriers. Women of South and South East Asian ethnicity accounted for the next largest group, with 10 ‘Bangladeshi, five ‘Indian’ and five ‘Pakistani’ carriers. The distribution of carrier women by age and parity was generally similar to the overall sample.

TABLE 10 - Characteristics of women newly identified as carriers during the trial

	Frequency (%)
Ethnicity
North European	5 (4)
South and South East Asian	20 (14)
African or Caribbean	82 (57)
South European	2 (1)
Other	5 (4)
Mixed	6 (4)
Not known	23 (16)
Primiparous	77 (54)
Age group (years)
< 24	32 (22)
24–28	34 (24)
28–32	38 (27)
≥ 32	39 (27)

Table 11 shows the distribution of carrier mothers by intervention group. As the numbers of observations were small, p-values were estimated using exact logistic regression in stata version 10, assuming clustering by practice to be negligible. Carrier mothers were more likely to be tested before 70 days’ gestation when SCT testing was offered in primary care. In the primary care parallel group, 12/47 carriers (26%) were tested before 70 days’ gestation (p = 0.079), with 11/25 (44%) in the primary care sequential group (p = 0.005), compared with 1/21 (5%) in the standard care group. Combining the two primary care groups together, the relative odds of a maternal carrier being tested before 70 days’ gestation were 9.24 (1.31 to 405.7) p = 0.017.

TABLE 11 - Identification of SCT carrier women and father testing by trial arm

p-values were estimated by exact logistic regression because cell frequencies were small and were not adjusted for possible clustering.

There were 93 carriers identified but one was tested at 183 days gestation and one was Newham SCT screen only.

There were 84 fathers tested but only 74 of these had known dates of testing and four of these had dates of testing before LMP.

Based on cases with known date of testing and testing not before LMP.

There was one couple with status known date before LMP and this was omitted.

Figures are frequencies (per cent of column total except where shown).

	Standard care	Primary care parallel		Primary care sequential
		Frequency (%)	p-valuea	Frequency (%)	p-valuea
Eligible women	441	677		590
Couple carrier status known if mother carrier	8/21 (38)	18/47 (38)	1.000	12/25 (48)	0.708
Couple carrier status known < 77 days’ gestation if mother carrier	0/21 (0)	2/47 (4)	0.949	2/25 (8)	0.580
Mother identified carrier	21 (5)	47 (7)		25 (4)
Mother identified carrier, tested < 70 days’ gestation	1/21 (5)	12/47 (26)	0.079	11/25 (44)	0.005b
Fathers tested	13 (3)	51 (8)		16 (3)	–
Fathers tested < 77 days’ gestation	0 (0)	16 (2)	<0.001	4 (1)	0.214c,d
Mother identified carrier: father tested	9/21 (43)	19/47 (40)	1.000	11/25 (44)	1.000
Mother identified carrier: father tested < 77 days’ gestation	0/21 (0)	3/47 (6)	0.647	3/25 (12)	0.303c,d
Mother not carrier: father tested	4/420 (1)	32/630 (5)	–	5/565 (1)	–
Mother not carrier: father tested < 77 days’ gestation	0/420 (0)	13/630 (2)	0.003	1/565 (0)	1.000c,d
Couple carrier status known, all women	12 (3)	50 (7)	<0.001	17 (3)	1.000
Couple carrier status known < 77 days’ gestation, all women	0	13 (2)	0.003	3 (1)	0.374d

These analyses were performed using exact logistic regression because some cell frequencies were small and logistic regression may give biased results under this condition. The ICC for screening uptake before 70 days among carrier mothers only, by analysis of variance, was 0.02. This suggests that clustering should not be ignored. However, when ordinary logistic regression with robust variance estimates was implemented, the estimated p-value for the difference in screening uptake before 70 days in primary care parallel group, compared with standard care, was p = 0.044 and for the primary care sequential group, p = 0.003. Therefore, under these conditions, exact logistic regression appears to offer more conservative estimates.

Uptake of SCT screening tests by fathers

There were 84 cases with father testing results available, 74 of these also had data for date of testing; the test date was before the LMP in four cases which were excluded from analysis, leaving 80 father tests for analysis with 70 cases in which date of father testing known. The striking feature of the results, therefore, was the low uptake of father screening with 8% fathers tested in the primary care parallel testing group and 3% in both of the sequential testing groups. Based on these small numbers, fathers were more likely to be tested before 77 days’ gestation in the primary care parallel testing group as required by the protocol (Table 11).

Among fathers whose partners were carrier mothers, the uptake of father screening was higher than among all fathers, being 9/21 (44%) in standard care, 19/47 (40%) in primary care parallel testing and 11/25 (44%) in primary care sequential testing, with no overall difference among groups. Father testing before 77 days’ gestation, if the mother was a carrier, was performed for 0/21 (0%) in standard care, 3/47 (6%) in primary care parallel testing and 3/25 (12%) in primary care sequential (Table 11). If the mother was not a carrier then father testing was more frequent before 77 days’ gestation in the primary care parallel testing group, as required by the protocol, but only 13 (2%) of such fathers in this group were tested before 77 days’ gestation.

Test results in carriers identified in the trial

Table 12 shows the results of SCT testing in carrier women.

TABLE 12 - Test results for women identified either as carriers or having abnormal haemoglobins

Condition	Carriers identified in trial	Carrier status known	Total
AS	50	30	80
AC	15	7	22
AD	5	–	5
AE	4	2	6
SS	1	–	1
SC	1	–	1
EE	1	–	1
Alpha thalassaemia carrier	3	5	8
Beta thalassaemia carrier	9	5	14
Probable beta thalassaemia carrier	1	–	1
Alpha thalassaemia	2	1	3
Hereditary persistence of fetal haemoglobin	1	–	1
Total	93	50	143

Section IV: further analyses

A number of additional analyses were implemented to evaluate intervention effects in subgroups of participants and to evaluate possible interactions. The results of these analyses were generally negative and it was not considered important to present all the analyses in detail. The main findings are summarised below. Further details are available from the authors of the report.

Section V: discussion

Introduction

The objective of the economic evaluation was to model the cost-effectiveness of the three different patterns of screening, in part using data generated by the trial. The screening approaches investigated mirror the three trial arms: Group 1 (primary care parallel), Group Two (primary care sequential) and Group 3 (standard care sequential). The economic analysis sought to predict the costs associated with the three strategies and their outcomes in terms of uptake of earlier screening, and rates of downstream events such as PND, termination of affected pregnancies (TOP), affected births, and unexpected affected births.

There is no extensive literature on the cost-effectiveness of antenatal SCT carrier screening. A recent review is presented in Karnon et al. (2007)53 where four studies were identified: two small empirical studies that focus just on cost issues,54,55 and two model-based analyses. 9,56 The two modelling studies have adopted similar structures in terms of the pathways of the decision problem. However, different modelling frameworks have been applied: Zeuner et al. (1999)9 used a conventional decision tree model, and Gallivan et al. (2003)56 used a mathematical modelling approach. Further, the Gallivan et al. report is of a feasibility study, whereas the Zeuner et al. model has been more fully developed in that it has been validated by comparing outputs from the model against recorded UK data for key model predictions such as PND rates. The Zeuner et al. model also explicitly considers the ethnic mix of the screened population and the probability of both the mother and father carrying one of the six significant SCT traits, or of being a non-carrier. The model predicts outcomes from the antenatal process such as PND, TOP, affected births and unaffected births. On the basis of its appropriateness to the decision problem being addressed by the SHIFT study and its demonstrated model validity, the Zeuner et al. screening model has been used as the basis of the work undertaken in this project and is described more fully later in the chapter.

The cost-effectiveness model used in this work required estimates of the proportion of women screened by trial arm and gestational age. The first section of this chapter describes the analyses and results associated with deriving such estimates. The remainder of the chapter then describes the methods and results for the cost-effectiveness work, making use of the predicted proportions of women screened.

Methods to derive time-to-screening inputs for the CEA

Estimates of proportion of women screened by trial arm and gestational age based on the trial data were used as inputs to the cost-effectiveness analysis (CEA). These proportions were modelled from the time from LMP to screening date using a time-to-event framework. It was decided to use Bayesian–Markov chain Monte Carlo (MCMC) methods implemented in [sc]winbugs[/sc] 1.4.3. 57 MCMC draws samples from the joint posterior parameter distribution and not only has a simulation format compatible with probabilistic cost-effectiveness analysis,58 but also preserves the complex correlation structure between parameters, which is essential for appropriate propagation of uncertainty. This approach can readily take clustering, including clusters of different sizes, into account by including practice cluster as a ‘random effect’ within a hierarchical model, while at the same time providing a uniform analysis of proportion screened at any gestational age.

Based on preliminary analyses, and on the shape of the Kaplan–Meier survival curves of time to screening unadjusted for clustering (Figure 5), it was clear that conventional proportional hazards models did not fit the data because of the time dependency of the relative intervention effects. Based on these exploratory analyses, separate ‘baseline’ survival curves were fitted to each of the three trial groups, with covariates and cluster effects included on the assumption of proportional hazards.

Time from LMP to carrier testing was modelled as a Poisson counting process, one of several ways of parameterising Cox regression in a Bayesian framework. The model defines the hazard of an individual i with covariates Z_i in GP practice j in treatment group G with covariate vector Z_i at time t as:

where β is a vector of covariates, and γ_j are GP practice effects. The cumulative hazard is then defined as Λ_0ijG(t)=∫₀λ_0ijG(u)du.

In the counting-process formulation the data are in the form dN_i(t) = 1 if individual i ‘fails’ (is tested) at time t, and 0 otherwise, and Y_i(t) = 1 if i is under observation (at risk of being tested) at t, 0 otherwise. With time in finely divided intervals, the data likelihood is a step function at each observation point. We can closely approximate the continuous time process with observations occurring at discrete time points, and from here we use subscripts for t.

dN_it is Poisson distributed, with an expectation equal to the increment, dΛ_0t, in the cumulative hazard function at time t:

where the dΛ_0Gt are the step increments in the cumulative hazard for treatment group G. The regression coefficients β are given vague priors N (0, 1002), and the practice effects are assumed to be drawn from a common distribution of effects on the log-hazard scale, γj∼N(0,σc2).The between-practice precision 1/σc2 was given a vague Gamma (0.1, 0.1) prior. Finally, the increments d_{_0Gt} are given vague Gamma (0.1, 0.1) priors. The priors are designed throughout with the purpose of ensuring that all the posterior distributions would be dominated by the data rather than prior belief.

Note that the model specifies separate ‘baseline’ hazard functions for each treatment group. No proportional hazards assumption is being made regarding relative treatment effects. Covariate and GP practice ‘cluster’ effects are, however, based on proportional hazards assumptions.

The key output from the analysis is the cumulative proportion screened at time t, with treatment G, for woman in a ‘new’ GP practice having a practice effect γnew∼N(0,σc2) drawn from the estimated distribution of cluster effects, and with covariates set at their mean values Z¯ in the trial sample:

A related approach to variation due to cluster effects in policy models at local and national levels is suggested by Welton et al. 67 (in press). These ‘survival’ functions were used in the CEA. We also use the number of MCMC samples in which the proportion screened at time t in the primary care groups exceeds the proportion screened in the secondary care group, to construct a Bayesian ‘p-value’ for differences between treatment effects.

The covariates included in the analysis were: parity (0, > 0), age group (< 24 years, 24–28 years, 28–33 years, > 33 years), and ethnicity (North European, South or South East Asian, African or African Caribbean, South European, Other, Mixed or Not known). It was decided not to control for cluster level variables because there was no evidence that randomisation had failed to achieve a balanced allocation.

Convergence of the MCMC chain was assessed by setting different starting values. Convergence appeared satisfactory by 1000 iterations. In the analyses reported the first 3000 samples were discarded, and posterior summaries were based on the next 7000 samples.

Results from the Bayesian survival analysis

The posterior mean proportions screened in each group, adjusted for covariates and clustering, is shown in Figure 7. The results are shown in Table 14, along with posterior mean differences between the secondary care group and each of the primary care groups. The results show a close agreement between the crude proportion screened and the estimates derived from the [sc]winbugs[/sc] analysis. The proportion screened was considerably greater in the primary care groups. At 10 weeks’ gestational age, the per cent screened was 22.8% higher in the parallel testing group, and 26.3% higher in the sequential group. The difference was statistically significant at 6, 10, 14 and 18 weeks.

FIGURE 7.

Predicted proportion of women screened by group, adjusted for covariates and clustering.

There were covariate effects on the probability of being screened (Table 15). Older women were more likely to be screened earlier, but multiparous women were less likely. Women belonging to the ‘Not known’ ethnicity group were slightly more likely to be screened later.

There were extreme levels of variation between GP practices (Figure 8). The caterpillar plot shows that for each intervention arm, practice effects are randomly scattered around the baseline hazard. The between-cluster standard deviation on the log-hazard scale was 0.32, with 95% CI 0.22 to 0.47. On this basis, we would expect 95% of GP practices to have hazard ratios between 1.9 times greater and 1.9 times less than the median. Therefore, if the median level of screening by 10 weeks was 2.5%, 24.6% and 28.1% in the three groups as estimated in Table 15, we would expect screening rates to be between 0%, 8.8%, and 11.5% on each of the intervention, respectively, in a practice at the low-screening-rate extreme, and 13%, 40%, and 43%, respectively, in a practice at the high-screening-rate extreme.

FIGURE 8.

Caterpillar plots showing the scatter of practice effects (95% CI) around the baseline hazard.

Methods for the CEA

A cost-effectiveness framework has been used, encompassing both estimation and prediction of the costs and effects associated with the three models of care. The focus is the prenatal period, so only costs and outcome measures associated with screening and prenatal care have been included. The trial has been used as the data source for events up to the screening outcome, with data from other sources used for events post screening, such as PND and termination of pregnancy.

A key outcome in the economic analysis is the proportion of women screened before 70 days’ gestation, based on the trial results. Given that the trial found no gains associated with primary care screening in terms of ‘informed choice’ (as described in Chapter 6), this outcome was not included in the economic analysis. Using the model-based analysis, the economic evaluation has extrapolated beyond the trial to consider the longer-term process indicators, such as uptake of PND and TOP, and outcomes of affected live births and unexpected affected live births.

An incremental analysis is adopted, the increment being the difference in costs and effects between trial arms. Secondary sequential is assumed to be ‘standard practice’ and hence the default comparator. The alternative screening policies are compared in terms of the following incremental cost-effectiveness ratios (ICERs):

• additional cost per extra woman screened before 70 days’ gestation

• additional cost per extra PND undertaken

• additional cost per extra unexpected affected live birth prevented.

Perspective, scope and time horizon

The perspective of the study indicates the breadth of coverage in terms of cost issues and is particularly important when the costs and savings might be experienced by different sectors of the economy, for example the health sector, social services and families. If the use of testing in primary care leads to earlier and greater uptake of screening, there will be cost implications for the health-care sector and for the screened women. In addition, parallel testing of partners in primary care will have cost implications. This is because the partner is being tested before the result of the woman’s test is known and the woman may not be a carrier. Therefore, the economic evaluation has adopted a broad perspective and considered costs falling both on the NHS and on service users. The base-case analysis adopted a health-sector perspective, with service-user costs additionally considered as part of the sensitivity analysis.

The scope of a study determines how extensive is the consideration of the consequences resulting from the programmes being evaluated. The time horizon of the analysis covers pregnancies to their conclusion, i.e. birth, termination or other pregnancy loss unrelated to screening. It does not include the potential effects of the screening cycle on future pregnancies.

The model – an overview

The trial work provides new data on whether the screening alternatives are associated with differences in three outcomes: the time from LMP to screening; the proportion of women screened by 70 days; and the proportion of women making informed choices about screening. The modelling component of the project allows extrapolation beyond these observed outcomes to predict longer-term antenatal screening outcomes and comprehensive cost estimates.

The project has made use of an existing model, that reported by Zeuner et al. (1999)9 (originally programmed in sas), and the further development of that model undertaken by Karnon in his work for the National Screening Committee (programmed in excel). In the model, the screening process pathways are depicted for antenatal populations described by ethnic composition, interethnic unions, the frequency of six significant SCT carrier states and the non-carrier state, and the mendelian recessive inheritance patterns. This information allows calculation of the number of homozygous, heterozygous and unaffected fetuses with their corresponding genotypes, expected each year for any given ethnic composition in the antenatal population. The second function of the model is to predict costs and effects of a screening process for an antenatal population defined by ethnic mix.

In the model runs for this project, the screening process pathways are depicted for the antenatal population observed in the trial. The model structure is based on the chronological sequence of steps during the screening process, as described in Figure 9. Minor modifications have been made to the model to allow for parallel screening, whereby the declared father is offered screening in advance of the result of the woman’s test.

FIGURE 9.

Screening flow diagram.

In line with the screening approaches evaluated in the trial, a universal screening strategy is modelled such that all women (with the exception of those who present too late) are eligible for screening. Not all women in the trial underwent screening and this is captured as a model parameter. While some women may have declined the offer, it is likely that others were simply not offered screening. Some of the key model assumptions are listed below.

• In the case where a woman is identified as a carrier and the father is not tested, the woman is automatically offered PND.

• The model allows for the scenario where the declared partner is not the biological father.

• Any woman presenting after the end of the second trimester (26 weeks’ gestation) is too late to receive screening.

• The scenario of spontaneously miscarriage of a pregnancy is allowed for.

• The testing process is not assumed to be perfect so the model allows test results for women and partners to be false negative or false positive.

• The acceptance rates for PND and TOP are time dependent, so earlier screening can lead to higher rates.

• Prenatal diagnosis can induce miscarriage.

Thus, we have extended the Zeuner model to allow comparison of the screening approaches compared in this trial. In addition, the model has been modified in order to allow for time dependency of post-screening decisions such as PND and TOP, and also it has been adapted to allow probabilistic analyses to be undertaken.

Cost data

The screening programme costs considered in this work are categorised as: invitation/screening, counselling through the screening and diagnosis process, laboratory testing, PND and TOP. Given the scope of the analysis, we are interested in the prenatal costs incurred by the screening programmes up to and including TOP. Costs associated with antenatal care and averted lifetime treatment costs are, therefore, not considered.

As part of the empirical work, key resource use data have been collected to estimate the short-term costs associated with the alternative approaches to screening and diagnosis of women and their partners. Data were collected prospectively on patient-specific resource use, including short-term events, such as screening offer; counselling carrier women, partners and couples; and diagnostic testing and subsequent procedures. These anonymous data were extracted from routine medical records of trial participants.

Data were also collected on the additional time, both for consultations and administration, associated with primary care screening, and the personnel involved. Such data came from routine practice records and from interviews with clinical personnel. Data on length of consultation were estimated for pregnant women offered testing in primary care and those offered testing in secondary care, as part of a midwifery booking appointment.

Using a combination of objective and self-report data collected as part of the trial, the estimated additional length of the consultation as a result of introducing the screening test was estimated to be between 0.5 and 3 minutes, with a further 0.5 minutes if the father was present. For the base-case analysis, the assumption was made that, in the absence of a father, the additional length of this consultation was 3 minutes.

The estimated cost of a consultation in primary care, for either the sequential screening arm or the parallel screening arm varies, depending on whether the consultation is carried out by a GP or a nurse. Data on who conducted the consultation were recorded as part of the study and so the costs were calculated based on a weighted average for the unit cost associated with the test offer. It was noted in the trial data that the majority of individuals were not tested on the same day as the offer but returned to receive the test from a phlebotomist. This use of the phlebotomist increased the cost by approximately £1.74, in addition to the cost of the test itself, estimated to be £4.00. The unit cost estimates and the main cost inputs into the model are given in Table 16.

The unit costs associated with laboratory tests were accessed from laboratory pricing schedules. Information on other unit costs or prices (e.g. relating to PND and TOP) were collected from relevant routine sources [e.g. Department of Health Reference Cost schedules59 and Curtis and Netten (2006)60 for unit cost sources for community services] and hospital finance departments in order that an overall cost per screened woman could be calculated. All costs are reported in 2006 prices.

In the acceptability component of the study, a sample of women and their partners in all arms of the trial were asked to complete a patient cost questionnaire, which aimed to record the private costs of undergoing screening (e.g. travel costs, time off work, lost income and child-care costs). Private costs for post-testing procedures, such as PND and TOP, were not collected from participants in this study for both practical and ethical reasons. Information on time taken to deliver counselling sessions was obtained through interviews with key service personnel. Other information required to estimate private costs for other procedures (such as travel modality and travel time, length of TOP procedure, etc.) were based on assumptions.

Table 16 reports the unit cost estimates used in the economic analysis.

Other model parameters

Tables 17 and 18 report the model parameters describing the antenatal population studied and the screening pathways modelled. All parameters in the original Zeuner et al. (1999)9 model were reviewed and a judgement made concerning the updating of the parameter value. Where new data were available from this trial, they were used. Where new data could be accessed or found in published sources for other parameters, these values were also updated. Tables 17 and 18 detail the sources for all main parameters, with updates indicated.

For the time-to-screening variable, the model used the screening time predictions given by the winbugs analysis, where adjustment was introduced for clustering, parity, ethnicity, age, etc. (see Table 20). In line with this analysis, the cost-effectiveness work also used 3000 iterations as a burn-in phase and the next 7000 iterations to derive model inputs.

Base-case and sensitivity analyses, and presentation of results

A base-case analysis is first presented, representing the population of women seen in the trial and using the parameter values as defined in Tables 16–18. The published Zeuner model was entirely deterministic so part of the project was to undertake a model development process to allow probabilistic analyses in order that parameter uncertainty might be explored.

Given the highly complex model employed in this research, for pragmatic reasons it was necessary to adopt a partial approach to the investigation of parameter uncertainty. Thus, the probabilistic analysis has focussed exclusively on parameters populated by new data from the SHIFT study. This approach serves to underplay the full level of uncertainty associated with the results and the reader needs to take this caveat into account when reviewing the analysis results. The following parameters have been varied:

• ethnic mix of the screened population

• proportion of women not screened

• probability that father is tested when the woman is a carrier

• proportions of women screened by 6, 10, 14, 18 and 22 weeks.

For the ethnic mix of the population, the sampling was from a Dirichlet distribution for the vector of probabilities for the complete set of proportions. For the not screened rates and the probability of father testing, beta distributions were used. When different values were used for different options, independent samplings were employed. For the time to screening data, the probabilistic model was populated by the iterations from the winbugs survival analysis. The probabilistic analyses were based on 7000 replications of the model.

The base-case cost-effectiveness results have been derived from the probabilistic model analyses and so outputs represent mean values. Results have been presented using ICERs and scatters of points on the cost-effectiveness plane reflecting uncertainty in the results. The scatters contain 7000 points, each point representing one of 7000 estimates of incremental costs and effects generated by each model run. (Note: 7000 iterations were available from the winbugs analysis and so consistency was maintained in the probabilistic component of the CEA. This is a much larger number of samples than is commonly seen in probabilistic sensitivity analyses.)

Sensitivity analysis allows exploration of the robustness of the base-case results to plausible variations in key assumptions. Two sensitivity analyses have been undertaken. First, the perspective has been broadened beyond the health sector in order to include, additionally, costs falling on service users. Second, the uptake of PND was varied because the levels of uptake seen in the trial, albeit based on small numbers, were considerably lower than the levels assumed in the base-case run of the model. Given the length of time to run the model probabilistically, these sensitivity analyses were undertaken using deterministic model runs.

Results

Base-case analysis

The base-case results from the probabilistic analyses are reported in Table 19. For convenience and ease of interpretation, all figures are expressed as a rate per 10,000 pregnancies. Where results derive from the probabilistic analyses, 95% credible intervals are additionally reported, which simply represent the range (from 2.5% to 97.5%) of the distribution of outputs from the probabilistic model runs. Results from baseline probabilistic and deterministic analyses are very similar and so deterministic analysis results are reported for sensitivity analyses.

TABLE 19 - Base-case results – model predictions per 10,000 pregnancies (point estimates and 95% credible intervals)

Note: If credible intervals for the same measure under different strategies overlap, this does not necessarily mean that there is substantial uncertainty as to which strategy has the larger value. It is possible that there is considerable uncertainty as to the baseline value, but that one strategy consistently gives a higher value than the other. For example, there is substantial overlap between the credible intervals for costs of primary care sequential and control, but the credible interval for the difference between them is (£13,837 to £47,342), which does not cross zero.

	Standard sequential	Primary care parallel	Primary care sequential
Health sector costs	£133,469 (£109,906 to £153,720)	£185,265 (£155,382 to £206,757)	£164,086 (£136,635 to £186,850)
Women screened by 70 days	264 (92 to 580)	2556 (1276 to 4444)	2887 (1509 to 4930)
Blood samples with positive result (woman)	737 (645 to 816)	780 (685 to 852)	772 (677 to 851)
Blood samples with positive result (partner)	61 (34 to 89)	56 (36 to 76)	75 (49 to 102)
PNDs undertaken	50 (40 to 61)	57 (48 to 66)	50 (41 to 60)
PND-induced miscarriages	0.75 (0.60 to 0.92)	0.86 (0.72 to 0.98)	0.75 (0.61 to 0.91)
PND result (any positive)	4.68 (3.65 to 5.67)	4.84 (3.96 to 5.76)	5.62 (4.50 to 6.80)
PND result (SCD positive)	2.45 (1.93 to 2.94)	2.79 (2.21 to 3.36)	3.01 (2.34 to 3.72)
TOPs undertaken	3.82 (2.94 to 4.68)	3.88 (3.17 to 4.61)	4.56 (3.63 to 5.53)
Total births	8595.52 (8594.77 to 8596.31)	8595.44 (8594.76 to 8596.10)	8594.89 (8594.02 to 8595.73)
Affected births	27.83 (25.77 to 30.00)	27.83 (25.79 to 29.96)	27.17 (25.07 to 29.38)
Unaffected births	8567.70 (8565.58 to 8569.71)	8567.60 (8565.46 to 8659.63)	8567.72 (8565.59 to 8569.75)

The transfer of screening into a primary care setting is associated with an increase in costs to the NHS of between £31,000 and £52,000 per 10,000 pregnancies, depending on the screening approach considered. However, this increased expenditure is matched by a large increase in effectiveness measured by the number of women screened by 70 days’ gestation. Whilst the increase in screening rates by 70 days is dramatic, it still leaves the vast majority of women (i.e. over 70%) not screened by the 10-week deadline. The predictions from the model for process measures and outcomes not collected in the trial, such as PND rates, TOP rates and affected births, are also reported in Table 19. Ultimately, this process improvement in screening leads to a prediction that, on the basis of the assumptions used in constructing and populating this model, the number of PNDs, TOPs and affected births is not likely to be very different under the alternative screening scenarios.

The results for effectiveness are put alongside the cost results in Table 20, with results reported as incremental cost-effectiveness ratios, comparing both primary care strategies to the sequential secondary care model. The results indicate a cost-effectiveness ratio of £12 per additional woman screened by 70 days, for the move away from the Secondary Care service to a primary care sequential programme. Given the very small and uncertain differences between strategies in terms of other measures such as PND, TOP and affected births, ICERs for these outcomes are not reported.

TABLE 20 - Base-case ICERs

D, dominated (i.e. primary care parallel is associated with both a higher cost and poorer outcomes).

Rate per 10,000 women screened.

Cost per additional woman screened by 70 days.

	Increase in costa	Increase in number of women screened by 70 daysa	ICER (woman screened 70 days)b
Primary care sequential vs standard sequential	£30,617	2623	£12
Primary care parallel vs standard sequential	£51,796	2292	£23
Primary care parallel vs primary care sequential	£21,179	–331	D

The more costly primary care option is parallel testing but this appears not to be associated with more women being screened by 70 days and so this option, at least in terms of the short-term screening ICER (i.e. cost per woman screened by 70 days), is dominated by sequential screening in primary care. That is, in primary care, sequential screening has both a lower cost and better outcome than parallel screening.

The base-case results from the probabilistic analyses are also reported using cost-effectiveness planes given as Figures 10 and 11. A total of 7000 points is plotted on each scatter diagram, representing the mean incremental cost and effectiveness estimates from each of the model runs. The analyses are based on women screened by 70 days and the scatter is entirely within the north-east quadrant of the CE plane, indicating that the primary care policy appears very likely to be associated with both a higher cost and a larger number of women screened. The central points in the scatters indicate incremental costs of £52,000 (primary care parallel versus standard sequential) and £31,000 (primary care sequential versus standard sequential), in line with the cost figures reported in Table 20.

FIGURE 10.

Cost-effectiveness plane scatter diagrams.

FIGURE 11.

Cost-effectiveness plane scatter diagrams.

Discussion

These evaluation results indicate that the primary-care-based screening observed in the trial was associated with some improvement in screening coverage and the CEA indicates that the NHS would have to pay, on average, £12 to screen one additional pregnant woman before the 10-week threshold (using the sequential screening process). This additional cost is primarily driven by the extra time required by the GP or other HCP conducting the screening. Whilst one might be disappointed that higher levels of early screening uptake were not achieved, the cost picture suggests a rather modest sum is required in order to provide women and couples with the opportunity to make informed decisions, subsequently on further investigations and actions in relation to the pregnancy. Thus, we feel that our results lend support for a policy of primary care sequential screening in populations similar to those observed in the trial. The attractiveness of the cost-effectiveness ratios is despite the fact that over 70% of women were not screened by 10 weeks in the primary care arms of the trial. The longer-term process and outcome measures modelled as part of this work highlight the need for some caution as the suggestion is that the improved uptake of screening through primary care intervention might not be associated with increased uptake of PND or TOP, and thus may have little affect on the overall numbers of affected births.

Strengths and weaknesses of the cost-effectiveness and modelling work

• An important caveat relating to the cost-effectiveness results concerns the paucity of data on the additional costs associated with the screening process. As part of the trial, data were collected on consultation times but these data were partial and not straightforward to analyse and interpret. Therefore, it should be recognised that if primary care screening could be undertaken without extending the consultation by the 3 minutes estimated in our analyses then, unsurprisingly, the cost-effectiveness results for such screening would look considerably more attractive.

• This work built on an existing published model, and having this as a starting point has both advantages and drawbacks. On the plus side, the model used in this work had been developed over several years, and fully reflects the complexity of the clinical context and the screening pathways. Thus, modelling some of the detail of the clinical condition was already in place and so the focus of our work was adaptation rather than modelling from scratch. Further, the model had been through an extensive validation process with comparison made of model predictions and observed screening outcomes. On the negative side, the process of familiarising ourselves with such a highly complex model was not an inconsiderable task, especially given that the model was originally built to address a different policy question.

• One of the main data inputs for the model is the time to screen by trial arm. An important strength of the work is that it is based on new data from a large randomised controlled trial. Further, these model inputs have been analysed such that they both adjust for the cluster nature of the trial and for various covariates using the winbugs-based analyses of the time-to-screen data.

• The modelling work has, where possible, based its inputs on new published evidence for model parameters. The project, therefore, serves to re-emphasise the realities of important gaps in knowledge relating to this clinical area. For example, data on uptake of PND and TOP in this context is very sparse, particularly when issues of variation by ethnic group and timing of screening are considered.

• Probabilistic analyses allow the importance of parameter uncertainty to be explored. The scatter diagrams reported in this chapter give an indication of the uncertainty that exists in the mean estimates of cost and effect. The work reported here represents a partial investigation of the extent of parameter uncertainty because of the exclusive focus on those parameters where we had new data from the SHIFT Trial. Thus, the uncertainty demonstrated reflects that seen in the trial alone and is an underestimate of the full parameter uncertainty as the values on other parameters are not known with certainty.

Some notes on choice of survival analysis methods and differences between methods

The choice of methods and comparisons between methods on the primary trial outcome deserve some comment. The GEE approach and the cluster-based method used as a sensitivity analysis produced estimates of the proportion screened, and differences between interventions in proportion screened that were comparable to the Bayesian MCMC analysis (Table 14) used in the CEA. The confidence intervals for the Bayesian analysis are a little wider. This is a reflection of the fact that the CEA results apply not to a cluster with an ‘average’ screening rate, but to a ‘new’ practice drawn at random from the estimated population at practices, of which the trial practices were a sample. Practices were so variable that the predicted absolute difference in per cent screened by 10 weeks between primary care and secondary care settings, while averaging about 20% (Table 14) might be anywhere between 10% and 30%.

With the increasing trend towards probabilistic methods, there has been a move in recent years in health technology assessment either to estimate treatment effects and their variances and then enter these as parameters in the CEA, or, to integrate the efficacy and CEAs in a single unified analysis. This latter approach, developed at Duke University61,62 in the 1990s, has been advocated by a number of authorities,63–66 and is now commonly seen in technology assessments undertaken for NICE. Analyses based on Bayesian posterior estimation using MCMC are ideal for this purpose, because the simulation format links seamlessly to probabilistic decision modelling. This approach is particularly useful for cluster randomised trials, where the clustering can be accommodated in a Bayesian hierarchical model within the CEA. 67,68

Of course, a further advantage of a single integrated analysis is that it provides a coherent basis for inference and decision-making. While, in this case, treatment differences were very clear, and all forms of analysis would lead to the same conclusions, this will not always be the case.

In the SHIFT Trial, the form of data analysis that was specified in advance was the GEE analysis. However, the analytic requirements of the CEA were not fully considered at that stage. As a result, we have had to undertake, and present, two separate analyses, on the basis that the GEE analyses must be presented as it was specified in advance, but GEE was not capable of being extended to meet the requirements of the CEA.

Future research

The separation between disciplines is giving rise to a series of related anomalies that need to be urgently addressed in methodological research. For example, standard methods for powering trials are based on efficacy alone, even when a CEA is to be run alongside the trial, as was the case in SHIFT. Sample size considerations, and data analysis methods, should take account of both the need for sound statistical inference and cost-effectiveness.

Informed choice

As with other screening programmes, a central aim of the NHS SC&T Screening Programme is to offer couples the opportunity to make an informed choice about participation in the screening programme. Research on the factors associated with informed choice for people making health-care decisions has tended to assess one dimension only, most often knowledge about a procedure. It is now widely acknowledged that informed choice is more complex than this and involves several dimensions. A consensus is emerging that informed choices have two core characteristics: first, they reflect an individual’s values, and second, they are made in the context of good knowledge. Building on this, we have developed, to our knowledge, the first operational definition of an informed choice: ‘a decision based on relevant knowledge, consistent with the decision-maker’s values and behaviourally implemented’.

Thus, an informed choice to accept testing is one based on good knowledge, where those with positive attitudes towards undergoing the test have it; an informed choice to decline testing is one based on good knowledge where those with negative attitudes towards undergoing the test do not have it. Choices based on poor knowledge, or which are inconsistent with values, are classified as uninformed.

Lack of informed choice in the context of screening for SCT

Many factors have been identified as important in limiting the opportunity for couples to make informed choices in the context of antenatal screening for SCT. They include:

• failure to offer screening or screening offered too late in pregnancy

• couples lack of knowledge about the screening test

• screening process takes too long

• diagnostic testing offered too late in pregnancy.

Reducing delays in the screening process

Offering screening early in pregnancy has the potential to maximise both the uptake of tests and the range of reproductive choices that are available subsequently. 15,16 The earliest stage at which antenatal screening can be offered practically is at the pregnancy-confirmation consultation in primary care. However, some characteristics of GPs’ primary care consultations may impede the facilitation of informed choices. The duration and context of the primary-care pregnancy-confirmation consultation is perhaps less conducive to facilitating informed choices than the booking appointment conducted by a midwife. GPs have less time than midwives for their first consultation with a pregnant woman. Further, women booking with a midwife may have had their pregnancies confirmed several weeks earlier, and are therefore more likely to be primed to receive, retain and process information regarding antenatal care in a way that women confirming their pregnancies are less likely to be.

Emotional and cognitive impact of antenatal carrier detection for SCT

Antenatal SCT screening is now routinely offered in high-prevalence areas in England as part of the NHS SC&T Screening programme. One potential adverse effect of such screening is the anxiety that may accompany carrier identification. To date, only one study has described women’s emotional responses to being identified as SCT carriers during pregnancy. 90 The authors concluded that screening was acceptable to women and undue anxiety was not generated. No measures appear to have been used to substantiate this observation. Six studies have formally assessed emotional responses to the antenatal detection of carriers for cystic fibrosis. Five of the six studies used standardised measures. Four reported increased levels of anxiety in women identified as carriers before learning that their partners were not carriers. 27,51,91,92 We are unaware of any studies that have assessed emotional outcomes in carrier couples or in carrier women who learn that their partners are also carriers. A 3-year follow-up study of general and antenatal populations found no increase in anxiety 3 years post testing, although carriers were less positive than non-carriers in their feelings about their test results. 93

The present study used standardised measures of anxiety in a cohort of women participating in the SHIFT Trial (see Chapter 4) to describe the emotional and cognitive impact of identifying women as carriers of sickle cell or thalassaemia during pregnancy.

Antenatal SCT screening in primary care – acceptability for women

These analyses examine the feasibility of implementing SCT screening in primary care in terms of its acceptability to women. There are many factors that determine the acceptability of antenatal SCT screening, including prior knowledge about SCT, the screening process and sociocultural background. This study uses women’s perspectives to explore the acceptability of being offered antenatal SCT screening in primary care.

Background

There is little evidence about the best ways to successfully implement screening programmes in primary care. Policy and practice often assume an unproblematic approach to implementing new ways of working. Dissemination and implementation strategies, however, have costs that may outweigh the benefits of the new technology. Traditional models of implementation, which assume health-care providers and managers have the resources, skills and motivation to introduce new practices in their working environment, are often flawed because they neglect the barriers to introducing new practices. 110 Barriers to implementation exist at many levels, including the individual practitioner, the clinical team, the practice setting and wider organisational factors. 111,112 Implementation research has tended to focus on the role of individual health-care practitioners,113 although even in a setting such as general practice, where clinical autonomy and discretion is relatively strong, there is evidence that practitioners can equally be a product of social and organisational circumstances in which they work. 114

The evidence concerning the most effective way of implementing SCT screening among pregnant women is sparse, relying on speculation rather than empirical insight. Implementing screening in primary care is seen as especially difficult,30,115 with the need to set up new practice systems serving as a major barrier to implementation. Other barriers include a lack of training and resources and enthusiasm among GPs to play a strong role. 22 Many GPs had little interest in screening, particularly given the many other pressures they had to deal with. It was also noted that very few GPs had much interest in SCT, compared with diabetes or general health promotion. Further, Qureshi et al. (2006)116 found that GPs were more confident in providing prenatal genetic advice for CF carriers than thalassaemia carriers. These GPs also demonstrated poor awareness of the importance of rapid referral to diagnostic services. Several studies have investigated the negative attitudes of GPs and other health professionals towards prenatal testing in primary care, while also raising patient concerns about the quality of their care. 117,118 Further, patients from minority ethnic groups can be viewed by some professionals as a burden in primary care, creating another barrier to successful implementation. 119 Other barriers to good quality care include poor communication and a lack of collaboration between primary and secondary care. 115 In addition, HCPs also believed that there was little knowledge of screening among affected communities. 21,120

In summary, there is limited evidence about the best ways to successfully implement screening programme in primary care. The current study builds on the available evidence by examining the feasibility of implementing screening in primary care. It explores the perspectives of GPs working in the general practices where screening was offered antenatally as part of the SHIFT Trial.

Methods and procedure

Semi-structured face-to-face interviews were conducted with 34 participating GPs. The content of the interview varied according to roles, but covered the acceptability and feasibility of offering antenatal SCT screening in primary care. Informants were identified from amongst SHIFT-participating practices using a purposive sampling frame. A nominated person consented on behalf of the whole practice for two participants to complete tape-recorded interviews.

All interviews were carried out or supervised by an experienced qualitative, primary-care researcher in each of the clinics and recorded with informants’ consent. Each interview lasted approximately 30 minutes. The interviews were open ended and a semi-structured interview schedule was used to aid the researcher. This was developed during the pilot phase of the study and modified as data collection proceeded. The topic guide began with the general themes about the organisation of care of pregnant women within practices, followed by specific themes about the experience of offering SCT screening, and, finally, the obstacles and enabling factors to offering the test.

Data collection and analysis was an iterative process. Informal analysis began during the interview process to allow insights gained from one interview could inform subsequent interviews. As a result, the researcher was able to clarify and further probe issues that are being elicited giving further insight into the area under investigation. The researcher began formal analysis of the interview data after the transcriptions process was completed. The data were analysed thematically across all groups, drawing on grounded theory using the method of constant comparison; common themes and ideas were identified and then categorised. The data were organised into defined categories and their meaning abstracted. nvivo (2006)121 software package was used to organise the transcripts. A provisional, inductive coding frame was derived from the early stage of the analysis and modified as new themes emerged. This was used to assign codes to the transcribed data. To increase reliability, some interviews were double-coded by others in the research team. All interviews were conducted by VT.

Results

Organisational barriers

Professional barriers

Perceived patient barriers

Organisational facilitators

Professional facilitators

Perceived patient facilitators

Discussion

Overall, GPs included in the study were enthusiastic about offering antenatal SCT screening. They believed that screening would improve the health care they offered their patients. In particular, they perceived that screening early in pregnancy would provide important additional options for pregnant women. They did, however, perceive a number of barriers as well as facilitators.

The main organisational barrier described was inflexible appointment systems. GPs believed that there was insufficient time to offer screening in primary care. Time, or the lack of it, is a common theme in GPs’ accounts of their working activities. 122 The majority of GPs felt that while they had provided women with a choice, the lack of time was a major barrier to facilitating informed choice. Another major barrier to implementation, and one closely related to a lack of time in the context of patient–practitioner consultation was language. Patients who did not have English as a first language were perceived as challenging, often burdensome,123 and less likely to be offered screening in a way that would facilitate informed choices. Poor communication between doctor and patient has been attributed to language differences, and associated with dissatisfaction with health services. A study of immigrant women in Australia found that their lack of English and the providers’ inability to use, or interest in using, their language generates feelings of frustration. 124 In the current study, the absence of a shared language was more problematic, particularly as GPs suggested that the use of interpreters was not feasible due to a lack of this resource in primary care. This finding highlights the need to develop systems, including evaluation, that facilitate the use of interpreters in primary care to ensure equity of access for people who do not speak English. This, however, is a longstanding problem. 101

Practitioners face a number of factors militating against women making informed choices in relation to antenatal screening. These include the offer of a test being perceived as a recommendation to have it, and a lack of time to present choices fully. 125 Although most informants in our study expressed the view that patients were given a choice as to whether to undergo screening, informed choice was perceived as difficult to implement for many reasons, including time constraints, language barriers and patient reliance on GP recommendation. However, the quality of decisions made by women in the three trial groups did not vary, suggesting that the decision quality was not lower if the test was offered by GPs.

The current study supports the findings of other studies that suggest the implementation of antenatal screening for haemoglobinopathies in general practice may be difficult. 26,30,115 A low priority, for example, may be placed on activities that are not part of new performance and quality targets in the UK GP contract. More broadly, introducing antenatal screening in primary care involves more than the offer of a test. It occurs within a historical and social context, in which roles and responsibilities are constantly being negotiated and re-negotiated among the key stakeholders, such as patients, midwives, GPs, strategic health authorities, hospitals and the Department of Health, as they support, implement and challenge normative values and assumptions about health-care provision, of which screening for genetic conditions is part. 126 The implementation process has been shown to be shaped by the ability of stakeholders to impose their different interests and agenda. 127 In the context of this study, GPs beliefs and attitudes and their working practices acted as both facilitators and barriers to implementation.

General practitioners perceived the need for good communication at all levels. In addition, GPs in our study expressed the view that implementation of SCT screening in primary care may be facilitated when factors such as practice cohesion and GP training are addressed. For example, GPs suggested that smaller practices, where GPs exercised more control, would be better equipped to accommodate SCT screening than would larger practices where good communication systems are not in place. Often GPs from larger practices, who regularly employed locums, felt that SCT screening in primary care was not feasible.

Women’s generally positive attitudes towards the general care offered by their GP was perceived by GPs as facilitating the introduction of SCT screening in primary care. Trust (or lack of it) is fundamental in understanding this. 123 GPs linked trust with the woman’s moral responsibility to her child. That is, GPs perceived that women underwent screening because they wanted a healthy baby, feeling obliged to comply with medical advice, advice which most women trusted implicitly. Women who do not comply with the authorial knowledge regarding the value of technology in childbearing are ultimately held responsible for the health of their baby. 105 This has implications for informed choice. That is, women’s trust in doctors can undermine the concept of autonomy embedded in informed choice. An alternative viewpoint is that women who act in line with the values of significant others, such as GPs, can exercise an informed choice if the individual ‘chooses’ to act in line with the attitudes of significant others. 128

General practitioners had positive attitudes towards antenatal SCT screening, as they believed in the benefits of early diagnosis in facilitating timely choice for pregnant women, having seen the effect that sickle cell disease had on their patients. They expressed the importance of educating patients about these conditions, particularly those at risk.

Strengths and weaknesses

Qualitative methods were used in this study to provide a deeper understanding of GPs’ subjective experiences. This extends work by Thomas and colleagues (2005)26 by providing an in-depth analysis of GP’s perceptions about implementing antenatal SCT in primary care. The findings documented in this study are based on self-reports of GPs based in high-prevalence areas, who were taking part in a trial. Such GPs may more be open to the idea of antenatal SCT screening than other GPs, and may not be representative of GPs as a whole. Observational data of the encounters between GPs and women and their partners may provide a different account. Our results present perceptions and experiences, and some of these may be based on generalisations or, perhaps, even myths. The aim of this paper is not to interrogate these in any great detail or determine whether they are ‘true’ or not. However, what this paper does attempt to do is to explore how their meaning is expressed in the context of service delivery.

Conclusion

Overall, GPs were positive about offering antenatal SCT screening. They identified a number of barriers and facilitators, at organisational, professional and patient levels, which could usefully be considered in the implementation of such screening within primary care.

Main findings

In areas with high prevalence, offering antenatal SCT screening as part of the pregnancy-confirmation visit in primary care significantly increases the proportion of women screened before 10 weeks (70 days), from 2% in standard care to between 16% and 27% in primary care. It is important to note that the majority of women remain unscreened at this gestational age.

There are two likely causes for this:

• First, nearly 50% of women were not offered screening before 70 days This could be due to late confirmation of pregnancy or GPs failing to offer screening at the pregnancy confirmation visit. GPs attributed their failure to offer screening to women who attended primary care by 10 weeks to lack of time, language barriers and lack of training.

• Second, there were substantial delays between the offer of screening and the screening test These delays may be due to women taking time to make their decision about whether to be screened, or they may represent systemic organisational problems that led to delays in the availability of blood tests, for example the phlebotomy services in primary care did not always allow same-day testing.

Fewer than 5% of fathers underwent carrier testing. Of the fathers offered parallel testing in primary care, less than 3% were tested by 77 days’ gestation. While this was significantly higher than the numbers of father tested using sequential father testing, the overall number is very low. This raises questions about the use of parallel testing in this context.

Overall, 17 women knew their couple carrier status by 77 days. Again, this number was significantly higher for those offered parallel father testing in primary care, but the numbers are still low. The reasons for this are unclear, but, for carrier women, the failure to screen fathers means that a couple’s carrier status cannot be determined.

Offering screening in primary care requires additional resources. The transfer of screening into a primary-care setting is associated with an increase in costs to the NHS of between £31,000 and £52,000 per 10,000 pregnancies, depending on the screening approach considered. The cost-effectiveness analysis results indicate that the NHS would have to pay, on average, £12 to screen one additional pregnant woman before the 10-week threshold (using the primary-care sequential screening process). An important caveat relating to the cost-effectiveness results concerns the paucity of data on the additional costs associated with the screening process. As part of the trial, data were collected on consultation times but these data were partial and not straightforward to analyse and interpret. Therefore, it should be recognised that if primary-care screening could be undertaken without extending the consultation by the 3 minutes estimated in our analyses then, unsurprisingly, the cost-effectiveness results for such screening would look considerably more attractive.

Whether this is an efficient use of resources will depend upon the values attached to early screening and the consequences that arise from its implementation. It is unclear how much women value early screening for SCT carrier status. The proportion of women screened by 26 weeks was similar across the trial arms: an early offer of screening did not result in more women undergoing testing. This suggests that women may not perceive early screening to be more valuable that later screening. Studies of early screening for Down syndrome have demonstrated that women value the safety and detection rates of screening more highly than screening early in pregnancy, in contrast with HCPs who value earlier tests more highly. 129 Note that these findings may not be applicable in this context as detection rate does not vary by gestation. They do, however, highlight the importance of ascertaining the values of users and not assuming that these will be shared by those providing the services. Alternatively, similar uptake rates when the test is offered later in pregnancy make reflect women’s lack of understanding of the potential benefits of earlier testing. For example, they may not be aware of the possibility of PND or termination of pregnancy.

The principal value of early testing is that it provides carrier couples with the option of prenatal diagnostic testing in the early stages of pregnancy and, for those found to have an affected pregnancy, the option of a termination at an early stage of pregnancy. The evidence regarding the strength of value attached to earlier terminations is weak. There is evidence to show that the proportion of women undergoing PND following the diagnosis of sickle cell disorder or thalassaemia is lower the later in pregnancy the PND is made. 15,16 However, these are observational data that may be subject to ascertainment bias (those more ambivalent about PND or TOP may present later in pregnancy for testing). The data from the current population-based trial suggested relatively low interest in prenatal testing, regardless of gestational age. Of the 14 carrier couples detected, none went on to have PND. Two women opted for PND based on maternal carrier status only. These are very small numbers but they raise an important question that deserves systematic study: how much do couples value PND for SCT, and to what extent does this vary with the gestational age at which tests are offered, carrier status is discovered, and the condition identified?

Although the data on sickle cell disorder and thalassaemia were aggregated in this trial, there may be important differences, between and within the two conditions, which influence the behaviour of at-risk couples, for example the clinical effects, treatment options and prognoses. In addition, the demographic and ethnic characteristics of the at-risk populations for the two conditions are different: sickle cell disease primarily affects African Caribbean couples, while thalassaemia primarily affects Mediterranean, Middle Eastern and Asian couples. These differences need to be better understood.

National Screening Committee criteria

The National Screening Committee (NSC) in England uses a set of criteria by which to judge whether a screening test should become a programme. Antenatal SCT screening was deemed to meet these criteria as part of the decision to develop and implement a national programme. The SHIFT Trial informs three of the 22 principles (see Appendix 5, NSC criteria) guiding the implementation of any screening programme, which concern the acceptability of the programme, the balance between the benefits and harms from screening, and the opportunity costs of screening:

• 1. Principle 14 There should be evidence that the complete screening programme (test, diagnostic procedures and treatment) is clinically, socially and ethically acceptable to health professionals and the public.

  1. – GPs were enthusiastic about offering antenatal SCT screening. They believed that screening would improve the health care they offered to their patients. In particular, they perceived that screening early in pregnancy would provide important additional options for pregnant women. Women also had positive attitudes towards being offered SCT screening in primary care at the pregnancy-confirmation visit.

• 2. Principle 15 The benefit from the screening programme should outweigh the physical and psychological harm (caused by the test, diagnostic procedures and treatment).

  1. –The results of the current study suggest that antenatal SCT carrier detection does not appear to impact negatively on the emotional well-being of pregnant women. Follow-up of larger numbers and over a longer time frame is required.

• 3. Principle 16 The opportunity cost of the screening programme (including testing, diagnosis and treatment, administration, training and quality assurance) should be economically balanced in relation to expenditure on medical care as a whole (i.e. value for money).

  1. – The key policy and value-for-money question is whether, for areas of high prevalence, where universal testing is currently being applied, the NHS should be willing to pay, on average, £12 to screen one additional pregnant woman before the 10-week threshold. This research suggests that important gains in early screening rates can be achieved, using the primary-care sequential screening process, at a rather modest cost. Whilst the early screen potentially gives key information to facilitate informed decisions on subsequent investigations and actions in relation to the pregnancy, the true value associated with providing women and couples with such early information, however, is unclear at this stage.

Recommendations for research

The following recommendations are equally weighted.

1. The principal value of early testing is that it provides carrier couples with the option of prenatal diagnostic testing in the early stages of pregnancy, and, for those found to have an affected pregnancy, the option of a termination at an early stage of pregnancy. The evidence regarding the strength of value attached to earlier terminations is weak. It would be useful to determine the impact of gestational age at screening on uptake of prenatal diagnostic testing, and reproductive decisions following the detection of affected pregnancies.

2. The values attached by individuals and society to having information about SCT carrier status early in pregnancy. From this trial, it is hard to ascertain the reasons why women did not have screening. It would be useful to explore their reasons, and to determine if their decision was an informed choice.

3. Low uptake of fathers could threaten feasibility of early screening. Without father testing, there is no early knowledge of couple carrier status, and reproductive choices are not facilitated. Limited test uptake may be explained by high levels of social and material deprivation in the trial area or if biological fathers are not registered at the same practice as the mothers. Research needs to identify the factors limiting the uptake of testing by fathers in order to determine how they can be addressed.

4. Exploring other models of care to increasing the proportion of women screened by 10 weeks’ gestation. Possible mechanisms include exploring further the role of midwifery, the use of Quality Outcomes framework in primary care or increasing the role of community pharmacists.

5. The results of the current study suggest that antenatal SCT carrier detection may not impact negatively on the emotional well-being of pregnant women. Longer-term follow-up on larger numbers of carrier couples is needed to estimate more precisely the extent and nature of the emotional impact on them and whether there are variations according to risk to a particular type of SCT disorder. Due to the small sample, the findings in this trial should be generalised with caution.

6. An important caveat relating to the cost-effectiveness results concerns the paucity of data on the additional costs associated with the screening process. The cost-effectiveness results are highly sensitive to the value of this parameter and further data collection to assess the true impact on consultation length of adding screening in primary care would be helpful.

Ethics committee

Ethics committe approval was granted for all of the studies reported in this monograph.

Contribution of authors

The trial was designed by Theresa Marteau, Elizabeth Dormandy, Tony Ades and Martin Gulliford with additional contributions from all other authors.

The analyses for Chapters 2, 4, 6 and 7 were all led by Martin Gulliford. The analyses for Chapter 5 were led by Stirling Bryan and Tony Ades. The qualitative analyses for Chapters 7 and 8 were led by Michael Calnan and Karl Atkin. The drafting of the report was led by Elizabeth Dormandy and Theresa Marteau with contributions from all other authors.

All authors read the final report and gave their approval.

Disclaimers

The views expressed in this publication are those of the authors and not necessarily those of the HTA programme or the Department of Health.

Abstract

• Aim To formulate hypotheses about effective ways of recruiting and retaining practices to clinical trials based on a case study.

• Design Case study of practice recruitment and retention to a trial of delivering antenatal sickle cell and thalassaemia screening.

• Setting Two UK Primary Care Trusts with 123 practices, with a high incidence of SCT and high levels of social deprivation.

• Outcome measures Number of practices recruited to and completing a trial.

• Methods Practices were invited to take part in the trial using a Research Information Sheet for Practices. Invitations were sent to all practice managers, GPs, practice nurses and nurse practitioners. Expenses of approximately £3000 per practice were available. Practices and the research team signed Research Activity Agreements, detailing a payment schedule based on deliverables. Semistructured interviews were completed with 20 GPs who participated in the trial.

• Results Four practices did not agree to randomisation and were excluded. Of 119 eligible practices, 29 expressed an interest in participation. Two practices withdrew from the trial; 27 practices participated; two hosted pilot studies and 25 completed the trial, giving a retention rate of 93% (27/29). The 27 participating practices did not differ from non-participating practices in list size, number of GPs, social deprivation or minority ethnic group composition of the practice population.

• Conclusion Three factors appeared important in recruiting practices: research topic, invitation method and interest in research. Three factors appeared important in retaining practices: good communication, easy data collection methods and payment upon meeting pre-agreed targets. The effectiveness of these factors at facilitating recruitment and retention requires assessment in experimental studies.

Keywords: primary health care; clinical trials; recruitment; retention

How this fits in

The focus of research into chronic conditions will shift from hospital to primary care as these conditions are increasingly managed in primary care. There is limited evidence, however, regarding the factors that facilitate recruitment and retention of general practices in clinical trials. A case study describing recruitment and retention of 25 practices into one clinical trial is used to generate hypotheses about effective methods.

Introduction

Research activity in primary care will increase as the number of chronic conditions managed in primary care continues to rise. For example, the recent ‘Roadmap’ published by the Royal College of General Practitioners asserts ‘virtually all health problems could be dealt with in primary care’. 1 It is therefore pertinent to consider the factors that facilitate research in primary care.

Recruiting and retaining general practitioners to participate in trials is challenging. 2–4 A recent review identified the lack of evidence about factors associated with recruitment to research of HCPs in primary care. 5 A range of interventions, including the use of printed educational materials, financial incentives, reminders, computer prompts and trial organisation, were described, but it was noted that there was an absence of evidence regarding their effectiveness. Experience from seven dyspepsia trials indicated that organisational strategies, such as the use of experienced researchers, methods of identifying eligible patients, GP workload and simplicity of patient eligibility criteria, may be more effective than more specific strategies at increasing recruitment in primary care. 5 In addition, concerns have been raised that the most successful recruitment strategies may lead to biased results. For example, the ‘physicians recruiting physicians’ strategy leads to acceptable recruitment rates but may not include representative practices. 6

One example of a potentially effective approach to practice recruitment that is built on an intimate knowledge of research in primary care involves the use of Research Information Sheets for Practices. 7 These provide a template for the generation of clear and succinct information about a trial or research study. Recipients report these provide a good basis for practices to decide whether to participate in a study. 7 There has, however, been no formal evaluation of their effectiveness in recruiting general practitioners and practices to research.

While there are several published case studies describing the recruitment of practices to trials, there is limited evidence on the strategies most likely to achieve retention of representative primary care practices to trials. 8 In a US study, 22% (46 out of 210) of practices dropped out of a trial of cancer prevention in primary care within 5 months of agreeing to take part in the trial. 9 There is a lack of evidence of factors associated with successful retention of practices in primary care trials. Failure to retain practices may also reflect a failure to recruit participants, and there is some evidence on failure to recruit participants in primary care. 10–12 Prout and colleagues13 interviewed nine GPs and one practice nurse after they had recruited varying numbers of children to a trial. Good trial organisation, simple documentation and trial procedures were reported as facilitating recruitment. While this evidence points to features associated with successful recruitment of participants by GPs there is limited evidence or guidance on strategies to successfully retain practices in trials.

Research networks of practices have been set up in the UK and elsewhere to facilitate the recruitment of practices to clinical trials and other research studies and to encourage primary care workers to undertake their research. The first primary care research network in the UK was the MRC General Practice Research Framework (MRC-GPRF), set up in 1973 for an MRC-funded trial of treatment for mild hypertension. In 1986 this was developed into a national research resource. The Framework now provides research access to 10% of general practices in the UK, although these practices are not representative of UK general practices as a whole. 14 The Primary Care Research Network (PCRN) was set up in March 2007 to increase the number of patients recruited or involved in research in primary care. 15 Neither network, however, currently provides published guidance on recruitment and retention of primary care practices.

The aim of the current paper is to present a detailed case study of the methods used to recruit and retain general practices in a clinical trial conducted in primary care, to form the basis for formulating hypotheses about effective ways of recruiting and retaining general practices to clinical trials.

Methods

Results

Of the 123 practices in the two PCTs, four practices were ineligible because they did not agree to be randomised. Two practices indicated they did not agree to randomisation before expressing an interest in trial participation, and two practices indicated they did not agree to randomisation after expressing an interest in trial participation (Figure 1). Twenty-nine of the 119 eligible practices expressed an interest in participating in the trial, giving a recruitment rate of 24% [29/119, 95% confidence interval (CI) 17 to 33]. We did not, however, continue the recruitment process in the remaining 90 practices as we had exceeded the target number of practices (n = 24). We have not speculated on how many additional practices would have participated in the trial if we had continued an active recruitment policy. Therefore, the recruitment rate of general practices to the trial is at least 24%.

FIGURE 1.

Flow diagram of practices in the trial

Twenty-seven practices participated in and completed the trial, two as pilot sites and 25 as trial sites (Figure 1). Thus, we retained 93% (95% CI 77 to 99) of practices in the trial. Two practices withdrew because of work pressures, one prior to the run-in data collection phase starting, and one after training was completed but before the intervention phase started.

Practices that completed the intervention phase did not differ from the other practices in the two PCTs with respect to number of GPs (p = 0.64); the list size per GP (p = 0.99); the Townsend score (p = 0.69) or the resident per cent of minority ethnic groups (p = 0.80) (Table 1).

Trial recruitment and retention

Qualitative analysis of interviews identified two major themes that seemed important in GPs’ decisions to take part in the trial (Box 1): perceived importance or the research topic and a general interest in research. In addition, one GP cited practical reasons for participation, while other GPs were not sure why they took part and one speculated that it might be for financial reasons. The two practices that withdrew from the study cited time pressures as the reason. One practice withdrew before starting the trial citing general pressures in primary care at the time. The second practice withdrew after the training session because offering the test in primary care was perceived as too time consuming.

The research team synthesised the above findings into a list of factors that are potentially important in recruitment and retention (Box 2).

Discussion

Conclusion

Three factors appeared important in recruiting practices: the research topic, invitation method and interest in research. Three factors appeared important in retaining practices: good communication between practice and researchers, easy data collection methods and payment upon meeting pre-agreed targets. The effectiveness of these factors in facilitating recruitment and retention requires assessment in experimental studies.

Acknowledgements

This study is funded by the UK Department of Health through its Health Technology Assessment programme, grant number 03/02/03 – ‘Antenatal screening for haemoglobinopathies in primary care: a cluster randomised trial to inform a simulation model. The Screening for Haemoglobinopathies in First Trimester (SHIFT) trial.’ The opinions and conclusions expressed here are those of the authors and do not necessarily reflect those of the UK NHS or the Department of Health. We are very grateful to the general practices who participated in this study.

Competing interests

The authors have no competing interest to disclose.

Ethical approval

Ethical approval was granted to collect anonymised data on all pregnancies reported in participating practices (05/Q0501/36).

Appendix 1 – Research Information Sheet for Practices (RISP) and invitation letter

SHIFT (Screening for Haemoglobinopathies In First Trimester) Trial

Purpose

• I.  To examine the feasibility, acceptability and effectiveness of offering antenatal screening for sickle cell and thalassaemia in one of three ways:

  1. – Group 1 In primary care, when women first report their pregnancies with partners offered testing at the same time.

  2. – Group 2 In primary care, when women first report their pregnancies with partners offered testing later and only if women are identified as carriers.

  3. – Group 3 In community-based secondary care, when women are booked by midwives, with partners offered testing later and only if women are identified as carriers.

  4. Practices will be randomised to offer screening in one way only.

• II.  To model the cost-effectiveness of these three methods of offering screening.

Context

Many women are not making informed choices about antenatal screening for sickle cell and thalassaemia because they are not offered the test or they are offered the test too late in pregnancy. For example, in a recent audit in Newham in 2002, 59 couples were at risk of having an affected baby, but only nine had been screened by 11 weeks’ gestation.

Practice involvement (details overleaf)

The practice will:

1. offer antenatal screening for sickle cell and thalassaemia to all pregnant women when the pregnancy is first confirmed, and

2. provide anonymous data on the number of pregnancies and time of offer and uptake of antenatal screening for sickle cell and thalassaemia.

Ethical committee approval

Ethical and R&D approval for the Trial has been obtained.

Period of data collection

Six months, plus retrospective data on number of pregnancies and time of screening for 6 months prior to data collection period.

Suggested start date in this practice

March 2006.

Practice costs are reimbursed

Yes, approximately £3000. This covers administrative costs of providing anonymised data, costs of offering the screening test to pregnant women and locum costs for attending a training session.

Your project contacts:

Name: Elizabeth Dormandy, SHIFT Trial Manager

Phone: 020 0000 0000

Email: elizabeth.dormandy@kcl.ac.uk

Lead researcher: Professor Theresa Marteau, Professor of Health Psychology

Host institution: King’s College London

Funder: NHS R&D Health Technology Assessment programme

Details of practice involvement

What the researcher will do

• Provide nationally approved training in offering antenatal screening for sickle cell and thalassaemia and information for pregnant women and their partners about the screening process.

• Seek consent from women to take part in the evaluation of the trial.

• Conduct trial evaluations.

• Maintain contact with local midwives, counsellors, laboratory staff and obstetricians.

To undertake the research the researcher will request access to the following personnel, records and/or practice facilities

• Anonymised data on the number of pregnancies and the offer and time of screening for all pregnant women for a 12-month period (6 months prior to data collection and during the data collection period).

• Contact details of pregnant women who agree to be contacted by a researcher.

What the practice personnel will be asked to do

• Agree to randomisation to one of three patterns of care.

• Every GP to attend a training session in (1) offering antenatal screening for sickle cell and thalassaemia (2) the trial protocol (approximately 2 hours, provided at the practice).

• Provide data on the number of pregnancies and the offer and time of screening uptake for all pregnant women for a 12-month period.

• Offer screening to all eligible women for 6 months of the data collection period. Eligible women are those confirming pregnancy, aged 18 and over, wanting to proceed with the pregnancy and whose carrier status is not documented in primary CARE records. Practices randomised to Group 1 to offer screening to partners of eligible women.

• Seek permission for a researcher to contact pregnant women. One GP from each practice to participate in an interview about the experiences of offering screening in primary care.

Local sickle cell and thalassaemia counsellors will provide the support pathway for any carriers that are identified through SHIFT.

How consent and confidentiality will be handled

• The proposed research will be conducted in accord with the Research Governance Framework, COREC, LREC and PCT approvals.

• A researcher will only approach women who agree to be contacted. Women will be informed about the risks and benefits of study participation before consent to participate in the evaluation is sought.

Practice feedback

This will be provided by personal contacts with Elizabeth Dormandy and trial updates via SHIFT website. Feedback of results prior to presentation to the scientific community will be offered to practices either as a seminar held at the practice or a written report in the summer of 2007.

We thank you for your help and interest.

Please contact Elizabeth Dormandy if you would like any further information or clarification (email elizabeth.dormandy@kcl.ac.uk or tel. 020 0000 0000).

This Research Information Sheet for Practices was initiated by the General Practice & Primary Care Research Unit, Cambridge University.

Version 1.3: October 2005

email: elizabeth.dormandy@kcl.ac.uk

5th November 2004

Dear Dr

SHIFT – Screening for Haemoglobinopathies In the First Trimester

We are writing to invite your practice to participate in a NHS R&D-funded study that aims to find the most effective way of offering antenatal screening for sickle cell and thalassaemia as early as possible. The study is being conducted in Lambeth and Newham and has the full backing of both PCTs.

Changing demographics means that the number of affected pregnancies affected by sickle cell and thalassaemia is increasing: these are now the most common genetic condition in England and Wales. This is particularly evident in Lambeth and Newham, where those at high risk for these conditions form a majority of the population. To date, screening services have served these populations poorly. For example, in a recent audit in Lambeth, Lewisham and Southwark fewer than half of the mothers of affected babies were offered PND. As well as being of practical clinical relevance, our study complements recent guidelines from NICE, the NSF for Children and Maternity Services, and the National Screening Programme for Sickle Cell and Thalassaemia.

We realise that you have many calls on your time and therefore propose funds to cover the costs. This has been estimated by local GPs to be about £3000 per practice. We will offer training to all practices to help them in the offer of screening and the practicalities of data collection. Benefits of participation include responding to a well-documented local need, as well as taking a lead locally and nationally and helping to shape future health policy. Details of what participation involves are included in the enclosed practice information sheet. Please note that data collection for the study will start in December 2005.

Elizabeth Dormandy, the Trial Manager, will be contacting your practice manager to discuss this in the next 2 weeks. If, in the interim, you have any questions do not hesitate to contact her.

Yours sincerely

On behalf of the SHIFT Trial Team

1. Lakhani M, Baker M, Field S. The future direction of general practice: a roadmap. London: Royal College of General Practitioners; 2007.
2. Yallop JJ, McAvoy BR, Croucher JL, Tonkin A, Piterman L. Primary health care research: essential but disadvantaged. Med J Aust 2006;185:118-20.
3. Huibers MJH, Bleijenberg G, Beurskens AJHM, Kant IJ, Knottnernus JA, van der Windt DAWM, et al. An alternative trail design to overcome validity and recruitment problems in primary care research. Fam Pract 2004;21:213-18.
4. Wilson S, Delaney BC, Roalfe A, Roberts L, Redman V, Wearn AM, et al. Randomised controlled trials in primary care: case study. BMJ 2000;321:24-7.
5. Foy R, Parry J, Duggan A, Delaney BC, Wilson S, Lewin-van den Broek NT, et al. How evidence based are recruitment strategies to randomized controlled trials in primary care? Experience from seven studies. Fam Pract 2003;20:83-92.
6. Ellis SD, Bertoni AG, Bonds DE, Clinch C, Balasubramanyam A, Blackwell C, et al. Value of recruitment strategies used in a primary care practice based trial. Contemp Clin Trials 2007;28:258-67.
7. Bateman H. A Research Information Sheet of Practices (RISP): a tool to facilitate research participation. Fam Pract 2002;19:691-7.
8. Hoddinott P, Britten J, Harrild K, Godden DJ. Recruitment issues when primary care population clusters are used in randomised controlled trials: climbing mountains or pushing boulders uphill. Contemp Clin Trials 2007;28:232-41.
9. Shelton BJ, Wofford JL, Gosselink CA, McClatchey MW, Brekke K, Conry C, et al. Recruitment and retention of physicians for primary care research. J Community Health 2002;27:79-8.
10. Fransen GAJ, van Marrewijk CJ, Mujakovic S, Muris JWM, Laheij RJF, Numans ME, et al. Pragmatic trials in primary care. Methodological challenges and solutions demonstrated by the DIAMOND-study. BMC Med Res Methodol 2007;7.
11. Farrin A, Russell I, Torgerson D, Underwood M. Differential recruitment in a cluster randomized trial in primary care: the experience of the UK Back pain, Exercise, Active management and Manipulation (UK BEAM) feasibility study. Clin Trials 2005;2:119-24.
12. Hetherton J, Matheson A, Robson M. Recruitment by GPs during consultations in a primary care randomized controlled trial comparing computerized psychological therapy with clinical psychology and routine care: problems and possible solutions. Primary Health Care Res 2004;5:5-10.
13. Prout H, Butler C, Kinnersley P, Robling M, Hood K, Tudor-Jones R. A qualitative evaluation of implementing a randomized controlled trial in general practice. Fam Pract 2003;20:675-81.
14. MRC General Practice Research Framework n.d. URL: www.gprf.mrc.ac.uk (accessed 5 March 2008).
15. Primary Care Research Network n.d. URL: www.ukcrn.org.uk/index/networks/primarycare (accessed 5 March 2008).
16. Marteau TM, . Protocol 06PRT 921: Antenatal Screening for Haemoglobinopathies in Primary Care: A Cluster Randomised Trial to Inform a Simulation Model. SHIFT (Screening for Haemoglobinopathies in First Trimester) Trial n.d. URL: www.thelancet.com/journals/lancet/misc/protocol/06PRT-921 (accessed 5 March 2008).
17. Dormandy E, Gulliford M, Reid EP, Brown K, Marteau TM. Delay between pregnancy confirmation and sickle cell thalassaemia screening: a population-based cohort study. Br J Gen Pract 2008;58:154-9.
18. Neighbourhood Statistics n.d. URL: http://neighbourhood.statistics.gov.uk/dissemination/LeadTableView.do?a=7%26b=276767%26c=newham%26d=13%26e=10%26g=343478%26i=1001x1003x1004%26o=35%26m=0%26r=1%26s=1201083025343%26enc=1%26dsFamilyId=724 (accessed 23 January 2007).
19. NHS Sickle Cell and Thalassaemia Programme . Information for Community Practitioner Leaflet n.d. URL: http://sct.screening.nhs.uk/Documents/CommPracInfo.pdf (accessed 5 March 2008).
20. NHS Sickle Cell and Thalassaemia Screening Programme (2006): Standards for Linked Antenatal and Newborn Screening Programme. UK National Screening Committee; n.d.
21. Silagy CA, Carson NE. Factors affecting the level of interest and activity in primary care research among general practitioners. Fam Pract 1989;6:173-6.
22. Rendell JM, Merritt RK, Geddes JR. Incentives and disincentives to participation by clinicians in randomised controlled trials (review). Cochrane Database Syst Rev 2007;2:1-19.
23. Dormandy E, Brown K, Reid EP, Marteau TM. Towards socially inclusive research: an evaluation of telephone questionnaire administration in a multilingual population. BMC Med Res Methodol 2008;8.

Abstract

Design

SHIFT is a partial factorial cluster-randomised trial. General (family) practices will be randomised to one of the three groups above, and pregnant women will be offered screening according to the group to which the practice at which they are registered has been randomised. There will be 24 general practices from two primary care trusts in London: Newham and Lambeth.

Target populations

• All pregnant women receiving antenatal care in the study settings, wanting to proceed with the pregnancy, and whose carrier status is not documented in their primary care records.

• Fathers of babies of eligible women randomised to Group 1.

• Health professionals providing antenatal care to eligible parents.

Outcomes

Primary outcomes The primary effectiveness outcome measures are the gestational age that antenatal screening for sickle cell and thalassaemia is done and the gestational age that such screening is offered.

Secondary outcomes Effectiveness will also be assessed by quantitative measures of informed choice. Feasibility and acceptability will be assessed by semistructured interviews with general practitioners and women. Acceptability for women will also be assessed by quantitative measures that examine the emotional response to carrier testing and concern about results. Costs to women and the health service will be assessed by collecting data on resource use for observed short-term events, including screening and counselling carrier couples.

Simulation model

Outcome data will be used to inform a new simulation model, developed from an existing and published sickle cell and thalassaemia screening model. The simulation model will be used to predict longer-term antenatal screening outcomes, such as miscarriage, after prenatal testing. The model will be used to explore the incremental costs and effectiveness of the alternative screening approaches.

Sample size

The primary outcome will be collected on 792 women, which gives sufficient power to detect an increase from 30% to 50% of women offered screening by 10 weeks’ gestation in primary versus community-based secondary care, with 90% power at the 5% level of significance, assuming an intraclass correlation coefficient of 0.03. Rates of informed choice will also be determined for these women.

Analysis plan

We shall compare the relative odds of a woman achieving three outcomes (uptake of screening, uptake of screening by 10 weeks, and the making of an informed choice) in Groups 1 and 2 versus Group 3, and the relative odds of women knowing their baby’s father’s carrier status as well as their own by 11 weeks in Groups 1 versus 2. We shall adjust for confounding variables, including age, parity, ethnicity, education, and other indices of socioeconomic status. We will use regression methods for clustered data to estimate odds ratios and their confidence intervals after adjusting for confounders and clustering by care provider.

Sponsor

NHS R&D Health Technology Assessment programme.

Date trial started: January 2006

Expected end date: October 2007

Expected reporting date: December 2007

SHIFT Trial protocol

Version 3.0 (without appendices) 23 February 2007. Changes to protocol listed at end of this appendix.

SHIFT protocol

Research objectives

Objective I

To assess and compare the feasibility, acceptability and effectiveness of three ways of offering universal antenatal sickle cell and thalassaemia (SCT) screening:

• Group 1: In primary care, when women first report their pregnancies with parallel father testing (i.e. the baby’s father is offered screening at the same time as the pregnant woman).

• Group 2: In primary care, when women first report their pregnancies with sequential father testing (i.e. the baby’s father is only offered screening if the pregnant woman is found to be a carrier).

• Group 3: In community-based secondary care, when women are booked by midwives with sequential father testing.

Objective II

To model the cost-effectiveness of these three patterns of care, using the data collected as part of achieving Objective I.

Review of existing research

The problem

Many at-risk pregnant women and the fathers of their baby are either not being offered SCT carrier tests or the process of screening and diagnostic testing is taking too long to allow couples the full ranges of reproductive options. These include prenatal diagnostic testing for carrier couples, and, where an affected pregnancy is identified, the offer of a termination. Given the recent statutory requirements to reduce disparities between ethnic and social groups in the provision of health care,1 it is germane to consider how antenatal screening for SCT can be organised to minimise the failure to offer screening early in pregnancy, the basic premise upon which the policy of universal screening is based. 2

Screening policy

Screening policy options for SCT in the UK were reviewed in two systematic reviews commissioned by the Health Technology Assessment (HTA) programme. 2,3 The results of these reviews informed the proposed introduction of a universal antenatal screening policy in areas of high prevalence by 2004. 4 Having introduced this policy, it is now important to determine how it is most effectively translated into practice to ensure not only that all at-risk pregnant women are offered screening, but also that this is done sufficiently early in pregnancy to allow women and babies’ fathers the full range of reproductive options. It is the aim of the proposed research to assess the feasibility, acceptability and effectiveness in achieving earlier testing by offering screening in primary care at the time a woman first reports her pregnancy to her general practitioner.

Current patterns of care

Currently, screening is offered to pregnant women in a variety of settings, most often when they attend for a booking appointment with a midwife. The location of such appointments varies across the country: they may take place in hospital, in the woman’s home or in a community-based clinic. The community-based clinic (which we have designated ‘community-based secondary care’) appears to be the most common pattern of screening in the UK, especially in Greater London, where about half of the UK minority ethnic population live. 5,6 We have therefore used this pattern of screening as the comparison against which screening in primary care will be judged.

Informed choice

As with other screening programmes, a central aim of the NHS Sickle Cell and Thalassaemia (SC&T) Screening Programme is to offer couples the opportunity to make an informed choice about participation in the screening programme. 4,7 Research on the factors associated with informed choice for people making health-care decisions has tended to assess one dimension only, most often knowledge about a procedure. 8,9 It is now widely acknowledged that informed choice is more complex than this and involves several dimensions. 10,11 A consensus is emerging that informed choices have two core characteristics: first, they reflect an individual’s values, and second, they are made in the context of good knowledge. 7,11–14 Building on this, we have developed, to our knowledge, the first operational definition of an informed choice: ‘A decision based on relevant knowledge, consistent with the decision-maker’s values and behaviourally implemented’. 15

Thus, an informed choice to accept testing is one based on good knowledge, where those with positive attitudes towards undergoing the test have it; an informed choice to decline testing is one based on good knowledge, where those with negative attitudes towards undergoing the test do not have it. Choices based on poor knowledge or which are inconsistent with values are classified as uninformed.

Lack of informed choice in the context of screening for SCT

Many factors have been identified as important in limiting the opportunity for couples to make informed choices in the context of antenatal screening for SCT. 16

Failure to offer screening

This can occur when women are incorrectly judged as having low risks of being carriers and hence are not offered the test. Non-northern European ethnicity has been used as a poor proxy measure of risk status for SCT, with misclassification occurring at a rate as high as 20%. 17 This reflects the practice of some health professionals offering screening on the basis of skin colour or family name. 18,19

Failure to offer screening also follows erroneous assumptions by health professionals about women’s attitudes towards undergoing the screening tests. For example, some health professionals assume that Asian women will not terminate pregnancies, and as a result do not offer screening tests to them. 20 In fact, existing evidence suggests that ethnicity is a very poor predictor of women’s attitudes towards undergoing antenatal screening. 21,22

The NHS SC&T Screening Programme has sought to address the failure to offer women screening by making it national policy for screening to be offered to all pregnant women in areas of high prevalence (the universal screening policy), facilitating this by instigating nationally co-ordinated education and training of health professionals.

Couples lack of knowledge about the screening test

Making an informed choice requires potential participants of screening programmes to have good knowledge about the screening programme. Several surveys show poor knowledge of SCT and of antenatal screening for these conditions. 23–25

The NHS SC&T Screening Programme has commissioned the development of information materials for screening participants (including those who do not speak or read English and those with low levels of literacy) and health professionals. These are being evaluated to help achieve high levels of knowledge in couples making choices about screening.

Diagnostic testing is offered too late in pregnancy

The variation in uptake of prenatal diagnostic tests is associated with the time in pregnancy that the screening test is offered. For example, 73% of British Pakistani women accepted prenatal diagnosis for thalassaemia when it was offered in the first trimester compared with 39% when offered in the second trimester. 26 This may be a particular problem for women whose pregnancies are at risk for sickle cell disease, with one retrospective study reporting that the mean gestation at which women who were carriers for sickle cell were first seen for counselling was 15 weeks, compared with 12.3 weeks for women who were carriers for thalassaemia. 27

There are several reasons why delays occur in couple’s learning about their carrier status. Some laboratories produce screening test results within 48 hours,28 whereas others take 5 days or more. 5 In addition, once results are known, it can take anything from 6–34 days for women to receive an appointment for counselling about diagnostic testing options. 29 Thus, the speed at which laboratories process samples and report screening results can limit couples’ opportunities to make informed choices.

The NHS SC&T Screening Programme has set standards for the time taken for laboratories to report a result (three working days4) and for the time interval between informing women of their carrier status and offering an appointment for a consultation to discuss diagnostic testing options (5 working days).

Reducing delays in the screening process: the proposed study

We propose to assess two ways of delivering antenatal screening for SCT that have the potential to reduce delays in the screening process: (1) offering screening at the time when women first report their pregnancies in primary care and (2) offering testing to the baby’s father at the same time as pregnant women are offered screening.

Time in pregnancy when screening is offered

Antenatal care is most often initiated when a woman reports her pregnancy to her GP, who then refers the pregnant woman to a community midwife to ‘book’ her for antenatal and maternity care. Screening is usually offered at this midwifery booking appointment, which commonly occurs between two and four weeks later, with some women not having this until 15 weeks’ gestation,30 many weeks after reporting their pregnancies to their GPs. The delay between reporting a pregnancy and seeing a midwife can be greater for women from ethnic minorities. General practitioners tend to book South Asian women later than other women, despite these women reporting their pregnancies at similar gestations to other women. 31,32 The extent and possible causes of these delays are not known.

Offering antenatal screening for SCT when women first report their pregnancies to their GPs has the potential to reduce this delay, a solution recognised in the recent NICE guidelines on maternity care33 and National Service Framework on Maternal and Child Health. 34 Such an approach to offering carrier testing in primary care for another recessively inherited condition, cystic fibrosis, was successfully conducted by one of our team in Manchester. 35 The extent to which this approach is feasible, acceptable and effective in the different cultural context of SCT is unknown. While offering screening at the earliest opportunity, i.e. as women report their pregnancies, has high face validity as a way of ensuring a timely offer and uptake of screening, offering screening at this time may be less effective at achieving informed choice than an offer made by a midwife as part of a longer booking appointment. We shall evaluate this and other outcomes in the proposed study.

The proposed study will address the two main problems identified in two recent pilot studies offering antenatal screening for SCT in primary care. 28,36 The first problem concerned low levels of knowledge in couples. 28 The authors called for better information for those offered screening and training of health-care professionals (HCPs), all of which will be incorporated in our proposed study. The second main problem concerned the failure to provide primary care practices with the resources needed to provide the screening service. 36 All of the HCPs included in our proposed study will receive training in collaboration with PEGASUS (Professional Education for Genetic Assessment and Screening), the nationally co-ordinated health professional training programme. Participating practices will receive NHS costs to cover the offer of SCT screening according to the study protocol, research costs for providing anonymised data on the number of pregnancies, time and uptake of testing, and costs to provide locum cover when GPs are trained in both offering antenatal SCT screening and study protocol.

Stage in the screening process when fathers are offered screening

Even if women are offered screening early in pregnancy, much time may elapse in offering testing to the baby’s father, which delays couples knowing their joint carrier status. Sequential testing is the usual pattern of care offered in SCT testing. In one study male partners were offered screening up to 4 weeks after their female carrier partners. 30 In cystic fibrosis screening, father testing has been conducted in one of two ways: parallel father testing, in which samples are taken and tested from both parents at the same time, and sequential testing, in which blood is taken from the father only after the woman has been tested and found to be a carrier. 37

While parallel father testing has the potential to provide couples with information on their carrier status earlier than sequential testing, it is not known how feasible, acceptable or effective it is in the population at highest risk of SCT. The project is being implemented in deprived inner city areas, where the population is characterised by high mobility, low income and low education. This may lead to difficulties if few women attend with their baby’s father, if fathers are not registered with the same GP or if fathers are not resident in the UK. 17 Estimates of partners attending for sickle cell testing after women have been identified as carriers vary between 63% and 81%. 2,27 Many of these characteristics pose problems for sequential as well as parallel father testing, including the suspicion and fear of stigmatisation that some men feel regarding carrier testing. 38 To ascertain whether parallel father testing is a viable model, we propose to evaluate how feasible, acceptable and effective it is in bringing forward the time at which couples who want screening can know their joint carrier status. We propose to evaluate this method of offering father testing in primary care only, as this is the most likely pattern of care to result in earlier knowledge of couple carrier status.

Future patterns of care

It is likely that as the NHS SC&T Screening Programme becomes established and the community becomes familiar with it, so the ways in which screening will be conducted will change. Thus, the experience of one of us (Dr Jane Logan) is that as a community served by a general practice routinely incorporates carrier screening for SCT into care, so carrier testing is more likely to take place before pregnancy. There will, however, always be a need to run several models of care of which the early offer of screening in pregnancy will be an important one. 39 It is unclear whether this is best conducted by a GP, a practice nurse or a midwife. Discussions with members of primary care teams suggest that the pattern we propose is, given the way care is currently organised, the most feasible way of delivering an early offer of screening.

The proposed study

Research methods

For the study to inform national policy it requires a robust design with appropriate outcome measures, as well as data collection methods that will minimise the biases that arise from missing data.

Study design

A partial factor cluster randomised controlled trial, with general practice as the unit of randomisation, will be used to evaluate first, screening offered in primary care at the time women report their pregnancies, and, second, father testing offered at this same time. It has a partial factorial design as we have chosen to evaluate parallel father testing only in primary care, to determine the extent to which this approach to father testing can build upon offering testing to women in primary care to achieve the earliest possible time at which couples can know their carrier status.

Outcome measures

We have chosen the timely conduct of screening for SCT as the main outcome measure. This is a necessary first outcome in ensuring that carrier couples learn of their carrier status in sufficient time to select, if they so choose, prenatal diagnostic testing and, if appropriate, termination of affected pregnancies. It is also the outcome measure that allows for robust appropriately powered studies to be designed within a realistic time frame. By using test uptake as the primary end point for the proposed studies we do not consider uptake per se to be a desirable outcome of the offer of screening: uptake in the absence of informed choice is, in the view of the Human Genetics Commission and similar other bodies, an undesirable outcome. Therefore, we propose to assess the time of offer of screening and the extent to which choices in the patterns of care being studied are informed. To our knowledge we propose the first systematic assessment of the extent to which fathers’ choices are informed. 15 Other outcomes, such as the offer of prenatal diagnostic testing, miscarriage rates, and the numbers of affected and unaffected births require data collection from the whole of England for more than 1 year to have sufficient power to examine the effectiveness of different patterns of care, given the estimated annual number of affected conceptions in England (28–60 for thalassaemia and 133–238 for sickle cell2,3). The NHS SC&T Screening Programme will be able to assess these outcomes when a nationally co-ordinated IT system of data collection is in place. In the interim, we propose to collect data regarding the performance of different patterns of care in the initial stages of the screening process, while modelling outcomes for the latter stages, in which significant outcomes are rarer.

Minimising missing data

We aim to minimise this potential bias of missing data in two ways:

1. Maximising study participation across all ethnic groups Consumers will be actively involved at all stages of planning the study, including data collection, so that the cultural sensitivities of the local consumers are addressed. This will ensure, as far as possible, that a representative sample and response rate is achieved. All study materials will be available in translation, in writing and on audiotape. In addition, we aim to recruit researchers who can speak one or more of the main Asian and African languages spoken by women invited to participate in the study.

2. Obtaining the main outcome measure on all pregnant women The main outcome measure, the time of offer and uptake of screening, will be obtained using anonymised data from the participating general practices for all pregnant women reporting their pregnancies during the 6-month period of data collection. We do not anticipate any selection bias with this method as records from all eligible women will be included.

Research setting

The study will take place in two PCTs, Lambeth and Newham, covering areas with high prevalence of Black African, Black Caribbean, Indian, Pakistani and Bangladeshi and mixed-heritage women. This sample reflects ethnic and cultural diversity, as well as a variety of socioeconomic backgrounds. The sample is likely to detect an equal mix of couples at risk from SCT. The estimated ethnic composition in the two areas, based on 2001 Census data, is given in Table 1,40 along with the main languages spoken apart from English.

TABLE 1 - Estimated ethnic composition from study sites

	Lambeth	Newham
Black Caribbean	11.8	7.3
Black African	17.5	13.1
Black other	6.7	1.1
Indian	1.4	12.1
Pakistani	0.7	8.5
Bangladeshi	1.0	8.8
Other	9.6	9.7
Northern European	51.3	39.4
Main languages spoken (listed alphabetically)	Eritrean French, Hindi, Lingala, Portuguese, Spanish and Urdu	Bengali, French, Lingala, Lunganda, Punjabi and Urdu

Earlier research into women’s experiences of thalassaemia screening has failed due to communication failure between HCPs. 19 We have sought to address this problem at an early stage in the research process by establishing collaborative links with the key stakeholders involved in the provision of antenatal screening for SCT. These include representatives from the two PCTs, as well as the sickle cell and thalassaemia centres, midwifery and obstetric leads, and haematology laboratories in the two PCTS.

We have identified 33 practices (19 from Lambeth PCT and 14 from Newham PCT) that have expressed an interest in participating in SHIFT. We feel this positive response reflects the recognised importance of the problem and the desire of many primary care professionals to participate in an adequately resourced research study. Participating practices will receive high-quality training to provide a service they perceive to be of great value to their patients.

Training health-care professionals

The NHS SC&T Screening Programme has commissioned a training programme, PEGASUS, led by Professor Joe Kai, to develop and implement training for HCPs working in primary and secondary care. We shall work with Professor Kai to implement this in the context of the trial (see appendix I).

Objective I

To assess the feasibility, acceptability and effectiveness of offering universal antenatal SCT screening in three ways:

• Group 1 – in primary care, when women first report their pregnancies with parallel father testing (i.e. the baby’s father is offered screening at the same time as the pregnant woman) Women will be offered screening for SCT when they first report their pregnancies to their GPs, the first opportunity to offer SCT screening in pregnancy. A verbal explanation of the test will be supplemented by written information, prepared for the NHS SC&T Screening Programme (see appendix II). Women wishing to be tested at that time will have blood taken using the usual procedure for phlebotomy at the practice. Those wanting more time to decide will be invited to return within 1 week. Fathers of babies of women receiving care from general practices in this study group will be offered parallel father testing, i.e. fathers are offered testing at the same time as the woman. If the father is present when a woman first reports her pregnancy to the GP, the test will be offered by the GP at this time. If the father does not attend this visit, the woman will be invited to offer the opportunity for her baby’s father to be tested, using a take-home pack (see appendix III). The pack contains information about the test, details of several local places (primary care, local hospital, SCT centre) where this can be undergone, and a completed request form that identifies both parents and provides the address of the father’s GP for reporting results. Fathers’ samples will be analysed upon receipt in the laboratory, in parallel with women’s samples. Follow-up of women identified as carriers will be led by local SCT counsellors, in line with the NHS SC&T Screening Programme protocol.

• Group 2 – in primary care, when women first report their pregnancies with sequential father testing (i.e. the baby’s father is only offered screening if the pregnant woman is found to be a carrier) Women will be offered screening for SCT when they first report their pregnancies to their GPs, as in Group 1. Fathers of babies of women receiving care from general practices in this study group will be offered sequential testing, i.e. offered testing only if the woman is found to be a carrier for sickle cell or thalassaemia. The offer of testing to fathers of babies of women identified as carriers will be made by local SCT counsellors, in keeping with the NHS SC&T Screening Programme protocol.

• Group 3 – in community-based secondary care, when women are booked by midwives with sequential father testing When women first report their pregnancies to their GPs, they are referred to a community midwife to be ‘booked’ for antenatal and maternity care. The midwife offers antenatal SCT screening at this booking visit, which can be conducted at the woman’s home, in a community-based clinic or at a hospital. The fathers of babies of women receiving care in practices in this study group will be offered sequential testing, as described above. This is currently standard care in Lambeth and Newham PCTs.

Design

The design is a partial factorial cluster randomised controlled trial in which 24 general practices consenting to participate in the study will be allocated to one of three groups (Figure 1).

FIGURE 1.

Study design.

Outcome measures

Effectiveness

In addition to cost-effectiveness (see Objective II), three other aspects of effectiveness will be assessed:

1. Time of uptake of screening (primary outcome) These data will be collected in two ways to ensure completeness of data collection:

   1. GP databases – from these, GPs will provide anonymised data on the number of pregnant women who are tested for SCT and the time in pregnancy when the woman was tested. These data will be provided for the 6 months before the study commences, as well as the 6 months’ period of data collection in each practice. As these data are anonymised we do not need to obtain individual patient consent or apply for exemption under Section 60 of the Health and Social Care Act (2001). 41

   2. Women participating in the quantitative and qualitative components of the study will be asked to give permission for information on testing to be taken from their laboratory and maternity notes.

2. Time of offer of screening This will be collected in the same way as time of uptake of screening.

3. Time at which women know there babies fathers’ carrier status The ideal outcome is the proportion of carrier couples that know their carrier status by 11 weeks. To power the study using this outcome measure would require UK-wide data. For the study to be feasible we have powered it using the proportion of women who know the carrier status of their babies’ fathers by 11 weeks.

Informed choice about screening

This will be measured for participants who accept or decline screening, using our standardised measure, comprising a multiple-choice measure of knowledge about the screening test and a measure of attitudes towards undergoing the screening test. 15,42 Choices are classified as ‘informed’ in individuals with high levels of knowledge (above the mid-point) and who act in line with their attitudes, i.e. those with negative attitudes do not have the test, while those with positive attitudes do have the test. We have good evidence of its validity in assessing informed choice in the context of antenatal screening. 15,42–45

Acceptability

This will be assessed in two ways:

• Women’s and the babies’ fathers’ emotional responses to carrier testing This will be assessed using two questionnaire-based measures:

  1. – An adjective checklist, found to discriminate between carriers and non-carriers of autosomal conditions. 46

  2. – A two-item measure assessing concern about test results, a measure that we have found sensitive when assessing women’s responses to other screening tests. 47

• Semi-structured face-to-face interviews from a sample of women and men, to ascertain their experiences of:

  1. – the offer of antenatal screening for SCT

  2. – deciding whether to accept the offer

  3. – for those accepting the offer, their experiences of the process of screening

  4. – for those declining the offer, their experiences of this.

Participants will be asked to consent to a possible interview at a place of their choice after the testing process. Approximately 20 women and, if consenting, their babies’ fathers will be asked to complete an interview. A semistructured interview schedule and questionnaire will be used to elicit women’s and fathers’ experiences, as listed in the aims above (see appendix IV, Interview schedule). For people who do not speak English these interviews will be conducted in a language of their choice, using either research staff or interpretation services. The interviews will be tape recorded and transcribed. The analysis of these interviews will draw on grounded theory using the method of constant comparison. The software package atlas-ti will be used to organise the transcripts. A provisional, inductive coding frame will be derived from the earlier stages of the analysis but will be modified as new themes emerge. This will be used to assign codes to the transcribed data. The data in each category will be summarised and themes identified and developed. Simultaneous sampling and analysis will continue until all categories are saturated and no new information is forthcoming. This component of the study will be led by Professor Michael Calnan and Dr Karl Atkin.

Feasibility

This will be assessed using semistructured face-to-face interviews with participating GPs, midwives, practice nurses and counsellors. The interviews will focus upon identifying obstacles and ways of removing these.

Participants will be identified using a purposive sampling frame of key informants (GPs and midwives) from the participating practices including, if possible, at least one informant from each practice. Interviews will be tape recorded and carried out or supervised by an experienced qualitative, primary care researcher (Professor Michael Calnan). The themes explored in the interview will depend on the role of the informant, but the general themes will involve their experiences of offering the test; the obstacles and enabling factors, and how any obstacles were or could have been overcome; and descriptions of the training and the extent to which they felt it met their needs. The analysis of these interviews will draw on grounded theory using the method of constant comparison as described above. For details of the interview schedule, see appendix V.

Other measures

Ethnicity Women will be asked to provide information on their ethnicity and that of their babies’ fathers, with question wording from the Health Survey of England:

• Socioeconomic status Women will be asked to complete a simple three-item measure based on educational qualifications, car ownership and home ownership to assess socioeconomic status. 48

• Parity Details will be collected from the women.

• Knowledge of sickle cell and thalassaemia carrier status Knowledge at the time of reporting pregnancies or booking appointment will be recorded.

• Reproductive decisions Information will also be collected on any choices that carrier women and carrier couples make regarding diagnostic testing and pregnancy termination and their satisfaction with the decision.

• Direct access to midwifery care Midwives offering antenatal care to women in the trial will be asked to record the number of women who access midwifery care without a referral from their GP. For these women, midwives will be asked to record the offer of screening and time of screening, and ask if a researcher can contact the women.

The involvement of practice nurses in offering antenatal SCT screening will be recorded for Group 1 and 2 practices. See Appendices 6 and 7, respectively, for details of the Time One and Two questionnaires.

Study samples

Health-care professionals

The study sample will comprise general practitioners and midwives providing antenatal care to pregnant women in the 24 participating general practices in two PCTs.

• Women Eligible pregnant women attending participating practices who wish to continue with the pregnancy, are less than 19 weeks 6 days’ gestation at their first visit to the GP in this pregnancy, and who have no written record of SCT carrier status in primary care. All participants offered the test aged 18 or over are eligible to complete the questionnaire.

• Fathers Those eligible for the study are fathers of babies of women in Group 1.

Proposed sample size

We estimate that we require data on the main dependent variable for 792 women attending 24 general practices (33 women per practice) (see Figure 1). This gives sufficient power to detect a difference of 20% in the proportion of women undergoing screening by 10 weeks’ gestation in different settings. In a sample of pregnant women in South Thames, 30% were screened by 10 week32 and the proportion is predicted to be 50% when screening is offered in primary care. We have assumed 90% power at the 5% level of significance and an intraclass correlation coefficient (ICC) of 0.03. This ICC estimate is based on data from 31 studies in primary care, in which 75% of ICCs were less than 0.032. 49 In order to detect a difference between 30% and 50% under simple random allocation, 134 participants per group would be required. To allow for cluster randomisation this must be appropriately inflated (eq2 of Campbell50), giving a requirement for 33 participants per practice and with eight practices per group there will be 264 women per group in total. This gives at least the same power (> 90%) to detect an difference in the proportion of women knowing their babies’ fathers’ carrier status from an estimated 10% to 30%, predicted to occur with parallel father testing (Group 1) compared with sequential father testing (Group 2). Adjustment of analyses for sociodemographic confounders and baseline screening performance should enhance the precision of estimated differences in screening outcomes between settings. A recruitment rate of 33 eligible women in a 6-month period is feasible, given the reported 100–150 pregnancies per year in the general practices wishing to participate in the study. Even if we lost one practice from each group, leaving seven, then the required recruitment would be feasible in the study time period (42 women from each of the remaining practices). Questionnaire based data will be collected from 232 women in each of the three study groups (including fathers of babies of 232 women in Group 1, the only group in which fathers are routinely offered the test), which allows an estimate of the proportions making an informed choice in each group within 10%, with 95% confidence, allowing for clustering.

Procedure

All general practices in the two PCTs have been contacted and invited to participate in the study. Eligible practices are those that have a computerised database with means of providing anonymised data on the number of pregnancies, time of offer of screening, and time of uptake for those women who accept testing and who agree to be randomised. Twenty-four of these will be randomised to the trial arms, and stratified by PCT and partnership size. Training in collaboration with PEGASUS will be arranged for all participating health professionals in these practices, as described above. The research team will train these health professionals in the research protocol and maintain close links with the participating practices to ensure fidelity to the protocol (see page 15 for Quality Assurance procedures). Practices in Group Three will be offered training after the recruitment period. Data collection will last for 7 months in each general practice, and commencement of data collection will be staggered at a rate of three practices per month for the first 8 months of the data collection phase.

General practitioners will offer SCT screening using the usual procedures for consent for antenatal blood tests, using information materials developed by the National programme. Midwives in the participating practices will be trained not to duplicate the offer of testing to women who have been offered testing by their GPs.

The GP will ask women if the trial manager can contact them to provide information about the research in order for them to decide if they would like to participate. For those willing to be contacted, a researcher will contact women by telephone, in the first instance, to explain the study and seek consent to participate, allowing time for consideration and questions. The information provided about the study will be based on the patient information sheet (see appendix VIII). This leaflet is described in more detail in the section ‘Informing women about the risks and benefits of participation in the study’ on Page 18 of the protocol. Those wishing to participate will be asked to sign three copies of a consent form (see appendix IX). More detail on this process is given in the section ‘Informed consent to participate in the study evaluation’ on Page 19 of the protocol. Data from women will be collected in three ways to minimise data loss: by telephone; face to face; and by post. Fathers of babies of women in Group 1 will be asked if a researcher can contact them about the study if they attend with women for this first appointment. Those not attending will be informed via the women, who will be asked to give the babies’ fathers information about the study, including a reply-paid postcard indicating their willingness to participate in the study (see appendix X).

All participants will be asked to complete a Time One questionnaire when they consent to participate in the study; a subgroup of 120 participants will be asked to complete a Time Two questionnaire at about 20 weeks’ gestation and a subgroup of 20 women will be asked to complete interviews at about 20 weeks.

Participation rate

As described in this protocol we plan to obtain the main outcome data using anonymised data from participating practices. We do not anticipate an unrepresentative participation rate as records from all eligible women will be included.

A secondary outcome measure is informed choice. A sample size of 232 participants per group, allowing for clustering, is required to detect a 10% difference in rates of informed choice, that is 29 women per practice. In a 6-month period it is estimated that 50–75 pregnant women are seen in each practice. In a recently completed cluster randomised controlled trial of the method of conducting antenatal Down syndrome screening, we approached 1292 women: 1156 (89%) agreed to take part and 982 (79%) completed questionnaires. 42 This study took place in a health action zone, i.e. an area with high levels of social deprivation. Thus there are data to suggest that we are likely to obtain an adequate and representative response rate for the secondary outcome measure.

Pilot study

We propose to run a pilot study at a general practice in Newham and Lambeth PCTs, prior to data collection, to finalise the protocol without influencing health professionals’ behaviour in the study sites.

Proposed analysis

We shall compare the relative odds of a woman achieving three outcomes (uptake of screening; uptake of screening by 10 weeks; and the making of an informed choice) for two patterns of care (Groups 2 and 3), and the relative odds of women knowing their baby’s father’s carrier status as well as their own by 11 weeks (Groups 1 and 2). We shall adjust for confounding variables such as age, parity, ethnicity, education and other indices of socioeconomic status. We will use regression methods for clustered data to estimate odds ratios and their confidence intervals after adjusting for confounders and clustering by care provider. Initially, we propose to use the method of generalised estimating equations. However, since regression methods for clustered binary data can sometimes be associated with problems of non-convergence or biased estimates, we will also investigate whether conclusions are sensitive to the method of analysis by comparing these analyses with results obtained using random effects logistic regression, or ordinary logistic regression with robust variance estimates. 51 Where anonymised baseline data are available, these will be collected for 6 months in each practice prior to practices receiving training in the new screening protocol. Where retrospective anonymised baseline data are not available, protocols for collecting them will be instigated as practices are recruited to the study, and baseline data collection will take place in the training period prior to recruitment of women. Comparison of baseline data across practices will allow us to evaluate the degree of imbalance in screening outcomes that may have resulted from randomisation of a small number of practices. We will adjust analyses for baseline observations using analysis of covariance as described by Ukoumunne and Thompson. 52 The cluster-specific proportions at baseline will be used for adjustment, as different pregnancies will be evaluated in the two study periods. Secondary analyses will be performed to explore demographic predictors of uptake (including ethnicity, socioeconomic status and age) in relation to the models of care. Rates of informed choice and emotional responses to screening will also be compared across the three study groups.

Quality assurance and quality control

Fidelity with the clinical protocol in which training was given, i.e. the universal offer of screening, will be assessed by comparing two sets of numbers, that should tally: (1) GP computer records for a 2-week period for the numbers of pregnant women recorded as having been offered screening when the pregnancy is first reported and (2) practice records of the number of women reporting pregnancies during the same period of time. Fidelity with the study protocol will be assessed by comparing the number of patient information sheets given out with the GP and practice records during the same period of time. Any discrepancies will be discussed with the participating practices and efforts made to reduce these.

Objective II

To model the cost-effectiveness of the different patterns of care, using the data generated as part of achieving Objective I.

Data for the cost-effectiveness study

If the use of early testing in primary care is shown to be associated with improvements in the screening process (e.g. increases in the proportion of pregnant women undergoing screening by 10 weeks’ gestation) compared with those offered screening in community-based secondary care then it is likely that there will be cost implications for both the health-care sector and for the screened women. For example, given that there is a relatively high rate of miscarriage in pregnancies in the first trimester, some women who will undergo early testing in the primary care arm of the trial would not have been tested if the testing were delayed until the booking because a miscarriage is experienced. In addition, parallel testing of fathers in primary care compared with the sequential testing of fathers, if accepted, will have cost implications. Again this is because of the increased risk of miscarriage in the first trimester and because the father is being tested before the results of the woman’s test is known, and it is likely that the woman may not be a carrier. Therefore, the economic evaluation will take a broad perspective and consider costs falling both on the NHS and on patients.

As part of the empirical work, key resource use data will be collected to estimate the short-term costs associated with the alternative approaches to screening and diagnosis of women and their baby’s father. We will prospectively collect data on training costs and patient-specific resource use. These will include short-term events, such as screening offer, counselling carrier women, fathers and couples, diagnostic testing, and subsequent procedures (e.g. termination of pregnancy) and counselling.

During the semistructured face-to-face interviews with GPs and midwives, information will be collected on their estimates of additional time, both for consultations and administration. These data will be calibrated with data collected in a simple time-and-motion study carried out on a sample of both the primary care practices and the community-based secondary care services. In particular, data on length of consultation time will be recorded for pregnant women who are offered testing as part of a GP consultation, and those offered testing in secondary care as part of a midwifery booking appointment. The additional administrative time devoted to clerical activities as a result of screening in these two contexts will also be collected.

In addition, data on additional resource use and costs, if any, in the laboratory, as a result of the screening, will be collected as part of the study. The unit costs associated with these tests will be accessed from laboratory pricing schedules. Information on other unit costs or prices will then be required to attach to each resource item in order that an overall cost per patient can be calculated. Such data will be collected from relevant routine sources, (e.g. Personal Social Services Research Unit53) and hospital finance departments.

In the acceptability component of the study, a sample of women and their partners will be asked to complete a patient cost questionnaire, which will record the private costs to patients of undergoing screening (e.g. travel costs, time off work, lost income and child-care costs).

Extrapolating beyond observed outcomes

The empirical work, described above, will indicate whether the screening alternatives being compared in the trial are associated with differences in three key outcomes: the proportion of women undergoing screening by 10 weeks’ gestation; the proportion of women who undergo screening whose baby’s father’s carrier status is known by 11 weeks; and the proportion of women making informed choices about screening. The main purpose of the modelling component of this project is to allow for extrapolation beyond these observed outcomes, i.e. the use of a modelling framework provides the opportunity to predict longer-term antenatal screening outcomes based on the study results of testing. Such longer-term outcomes include: the proportion of all affected fetuses detected prenatally, the proportion of live births, the proportion of terminations of affected pregnancies, the proportion of miscarriages induced by prenatal diagnostic testing, full screening costs and costs associated with long-term rare events.

As a starting point we will make use of two existing models: the simulation model reported by Zeuner et al. 7 (programmed in sas), and the further development of that model, which has been undertaken by Jon Karnon in his work for the National Screening Committee (programmed in excel). [Note that Tony Ades, a coapplicant on this bid, constructed the original model, and Jon Karnon is a member of the research team.] In both models the screening process pathways were depicted for antenatal populations described by ethnic composition, interethnic unions, the frequency of six significant haemoglobinopathy carrier states and the non-carrier state, and the mendelian recessive inheritance patterns. This information allows calculation of the number of homozygous, heterozygous and unaffected fetuses, with their corresponding genotypes, expected each year for any given ethnic composition in the antenatal population. The second function of the model is to put each of these subgroups of an antenatal population through an antenatal and neonatal screening process. Flow diagrams are used to describe the chronological sequence of steps during the screening process.

This research project will develop these models further and use them as a framework in order to allow data from the new empirical work to be incorporated. The trial will produce data on key variables within the cost-effectiveness model, which will inform probability distributions for these parameters. For other model parameters, for which trial data cannot be used to inform probability distributions, these will be described using published data where available.

The new analyses will explore the incremental cost and effectiveness of the alternative screening approaches. The alternative screening policies can be compared in terms of a number of possible incremental cost-effectiveness ratios (ICERs). The most straightforward is the additional costs incurred per additional unwanted affected live birth prevented (‘affected live birth prevented’ ICER). However, the Zeuner et al. model7 divides mothers with affected fetuses into two groups: those to whom reproductive choice was offered (‘choice offered’) and those to whom choice was denied (‘choice not offered’). This allows alternative policies to be compared in terms of an ICER expressed as the additional costs incurred per additional choice offered (‘choice’ ICER). This can be extended, using data from this trial, to explore the additional costs incurred per additional informed choice offered (‘informed choice’ ICER), where choices regarding screening are classified as ‘informed’ if they are based on good knowledge and if they are in line with a woman’s values. In keeping with the guiding principles of the NHS SC&T Screening Programme, the latter is our preferred ICER in this research.

The results of these economic analyses will also be presented using cost-effectiveness acceptability curves to reflect sampling variation and uncertainties in the appropriate threshold cost-effectiveness value. We shall also use both simple and probabilistic sensitivity analyses to explore the robustness of these results to plausible variations in key assumptions and variations in the analytical methods used, and to consider the broader issue of the generalisability of the results. For example, if the screening approaches explored in this research appear cost-effective on the basis of the data collected and analyses undertaken as part of this study, we shall explore the feasibility, costs and cost-effectiveness of their use routinely in other centres and other settings. This will allow the model to be run to explore not only the cost-effectiveness of universal in high prevalence areas, but also in medium- and low-prevalence areas.

Using qualitative data and the modelling framework to explore additional policy questions

The qualitative data to be collected as part of the overall project will identify further possible changes in the organisation and delivery aspects of the screening programmes. Clearly, it will not be possible to explore empirically these additional issues within this research project. However, we will make use of the model framework to predict the longer-term impact of such further policy changes and their overall cost-effectiveness. The level of uncertainty in these further analyses will, by definition, be greater. Uncertainties will be explored using a similar approach to that described above.

Research process and management

Ethical arrangements for the conduct of studies to achieve project objectives

The proposed research will be conducted in accord with the Research Governance Framework, involving COREC and research governance approvals.

Informing women about the risks and benefits of participation in the study

Potential participants will be informed about the risks and benefits of study participation, in writing and face-to-face, using materials that have been piloted for their comprehensibility and approved by the relevant ethics committees. For women who do not speak English the study will be discussed in a language of their choice using research staff, interpreters or the services of ‘Language Line’, a telephone translating service. We are working with the London IDEAS Genetics Knowledge Park to develop the materials in languages appropriate to the study populations. A copy of the information sheet is shown in appendix VIII. It follows COREC guidelines on patient information sheets. The written information first outlines the evidence regarding the current failure to offer these tests sufficiently early in pregnancy to allow full reproductive options. Second, it outlines the aim of the study. Third, it indicates that participating in the study may not benefit individual women but should benefit women in the future, and, fourth, that participation will require the research team to examine specific aspects of their records, namely uptake and time of testing. The written information also includes who is being invited to take part, what taking part involves, how the study is funded, the name and telephone number of the trial manager. This information emphasises that participants are free to withdraw at any time and without obligation. Information provided verbally about the study will be based on this written information. As well as COREC guidelines, development of study materials has been guided by MRC guidelines for good clinical practice,54 liaison with our consumer representatives, services users and the research team. In addition we have worked with the NHS SC&T Screening Programme to ensure that the materials informing women about the study and the screening programme reflect information provided by the National Programme.

Informed consent to participate in the study evaluation

In keeping with the practice of cluster trials in which the health-care organisation is the unit of randomisation, consent to receive the pattern of care provided will not be sought. Consent will be sought for women to participate in the evaluation of this care. Women will be asked by the HCP offering screening if a researcher can contact them to inform them about the study and invite them to participate in the qualitative and quantitative evaluation. A researcher will only approach women who agree to be contacted. Written and verbal information about the study in appropriate languages will be provided before consent is sought. After time for consideration and questions, women wishing to participate will be asked to sign three consent forms: one to be retained by the participant, one to be kept in the woman’s medical records and one by the researcher (see appendix IX). This discussion will take place in a language and format that is easily accessible to the woman. The research team has considerable experience of negotiating and gaining consent from minority ethnic populations and understands the importance of offering choice in terms of the linguistic and cultural background of the interviewer.

In keeping with MRC guidelines, consent to participate in the research evaluation will not be sought from women aged under 18. Women who do not speak or read English will be informed about the study using an interpreter with appropriate study materials. The trial documentation will be stored for 5 years following completion of the trial. Data will be stored in accordance with MRC guidelines.

Independent supervision of trials

The study will be supervised by two committees that are independent of the research team and King’s College London, as outlined in the MRC guidelines. 54

Trial Steering Committee

Nola Ishmael OBE, recently retired, as a Nursing Officer in the Department of Health, chairs this committee. Three other people, who are independent of the research team and King’s College London, are committee members: Dr Allison Streetly, National Co-ordinator of the NHS SC&T Screening Programme; Ms Audrey Braithwaite Kelly, a service user; and Dr Alison Hill, Department Health Advisor on Genetics. The remit of the committee is overall supervision of the trial. In particular, it will monitor progress against the milestones set in the study protocol, adherence to the protocol as assessed by the quality assurance and quality control measures set out in the protocol, and consideration of new information and safety of participants in the study. It is anticipated that this committee will meet twice a year during the study lifetime. Theresa Marteau, as principal investigator, will attend these meetings and a representative of the funding body, HTA, will also be invited to attend.

Data Monitoring and Ethics Committee

This will be chaired by Professor Mahesh (‘Max’) Parmar, Professor of Medical Statistics and Epidemiology and Head of the Cancer Division of the MRC Clinical Trials Unit. Dr Simon Griffin, Department of Primary Care, University of Cambridge, and Dr Lyn Chitty, Institute of Child Health, University of London, are members of the committee. The Data Monitoring and Ethics Committee will be independent of the research team, Trial Steering Committee, HTA and King’s College London, and reports to the Trial Steering Committee. The remit is to monitor study data and make recommendations about stopping the trial. Meetings of this committee are called and organised by Theresa Marteau, principal investigator, but she will attend meetings only if specifically asked to do so by the Chair.

Study management

Theresa Marteau and the trial manager Elizabeth Dormandy will review the study on a weekly basis. ED will liaise with research assistants on a daily basis. These reviews will form the basis of monthly research team discussions and two monthly reports submitted to the Trial Steering Committee. A core group of the research team will meet or speak as part of a telephone conference once a month.

Research team

Professor Tony Ades, University of Bristol

Professor Elizabeth Anionwu, Thames Valley University

Dr Karl Atkin, University of Leeds

Professor Stirling Bryan, University of Birmingham

Professor Mike Calnan, University of Bristol

Mrs Verna Davis, Manchester PCT

Dr Moira Dick, Lambeth PCT

Dr Elizabeth Dormandy, King’s College London

Professor Gene Feder, Bart’s and Royal London Hospitals

Dr Martin Gulliford, Kings College London

Dr Hilary Harris, Brooklands Medical Centre

Dr Tracy Johnston, St Mary’s Hospital Manchester

Mrs Patricia Jones, University College London Hospitals

Dr Jon Karnon, University of Sheffield

Dr Fred Kavalier, Guy’s Hospital

Dr Jane Logan, Mawbey Group Practice

Professor Theresa Marteau (Principal Investigator), King’s College London

Ms Tracy Roberts, University of Birmingham

Dr Barbara Wild, King’s College Hospital

Shift research staff

• King’s College London, Department of Psychology:

  1. – project manager, 3 years, full-time, (Elizabeth Dormandy)

  2. – research assistant, 2 years, full-time (to be appointed)

  3. – research assistant, 2 years, part-time (to be appointed)

  4. – clerical assistant, 2 years, part-time (Hazel Showell)

• King’s College London, Department of Public Health:

  1. –research assistant, 6 months, part-time (to be appointed)

• Birmingham University

  1. – research assistant, 2 years, part-time (to be appointed).

Expertise of the research team

Our research team brings a wealth of local, national and internationally recognised clinical, management and academic expertise, all of which are needed to succeed in conducting a multicentred project in primary- and community-based secondary care, delivering services to a multi-ethnic population. The team brings extensive clinical expertise in primary care, including providing genetic services to people from ethnic minorities (JL, FK, HH, GF) as well as research on providing genetic services in this context (HH, EA, KA, TM), clinical expertise in SCT (MD, VD), providing laboratory services (BW) and obstetric/maternity care at secondary level (TJ, PJ), training health professionals to provide genetic services (EA, VD), facilitating informed choices in antenatal contexts (EA, KA, ED, TM), and experience in negotiating fieldwork with the various minority ethnic populations that are part of the study population (KA, EA).

Our team also brings considerable relevant methodological expertise in economic modelling (SB, TR, TA, JK), the design and analysis of cluster randomised trials (MG), barriers to HCP implementing change (MC), the measurement of informed choice (ED, TM), and qualitative methods (MC, KA). A further strength of our team is that we can build upon the extensive modelling conducted for the HTA by three of our team (TA, EA, JK), which formed the basis for the universal antenatal screening policy of the NHS SC&T Screening Programme. Our team has close links with the NHS SC&T Screening Programme, with five co-applicants being members of the steering committee for this programme (EA, MD, TJ, JL, TM). Finally, the principal investigator has considerable expertise in successfully delivering large multicentre studies in primary and secondary care.

A small group of consumers have advised and will continue to advise the research team, comprising women and men who have had recent experiences of maternity services, and who are from populations at high and low risk of SCT. They have been recruited through contacts from the SCT centre in Camden, from the Sickle Cell Society and from the National Childbirth Trust. In addition, two workshops will be convened with representatives of the Sickle Cell Society and the UK Thalassaemia Society before the study commences to verify that, within the commissioning brief, we are addressing issues relevant to service users.

Funding body

NHS R&D National Coordinating Centre for Health Technology Assessment: £598,569 from October 2004 to September 2007.

Abstract

Background

Missing data may bias the results of clinical trials and other studies, with low response rates compromising the validity of the findings. 1,2 Acceptable questionnaire response rates are considered to be in the range of 60–70%, with response rates of over 70% described as very good. 3 Self-administered questionnaires sent and returned by post offer a cost-effective way to obtain data from a large number of participants, but often has low response rates. A recent systematic review of methods to increase response rates to questionnaires identified a number of strategies, such as shortening questionnaires, repeat mailing of questionnaires and telephone reminders. 4 Intensive reminders by telephone and post were the most effective, improving response rates by an average of 24%. Financial incentives have also been used to increase response rates. 5 For example a recent descriptive study reported that an unconditional payment of £5 increased the response rate from 78% to 88%. 6

Some studies comparing telephone and postal response rates from patients have found that response rates from questionnaires administered by telephone are higher,7 while others have found no difference. 8 Studies comparing responses from general practitioners have also shown varying results. One study reported a lower response rate for postal surveys than for telephone methods,9 while another found a higher response rate for postal surveys than for telephone methods. 10 Hocking and colleagues suggested that the lower telephone response rate in their study was because practice receptionists blocked telephone access to the general practitioners who were the target of the survey. A consensus is emerging that a combination of direct contact and postal methods leads to higher response rates to questionnaires than the use of postal methods alone. 8,11

Two areas of uncertainty exist about the use of telephone compared with postal administration of questionnaires. First, research studies not only need a good response rate, but also representative responses from different socioeconomic, demographic and clinical groups. It is not known whether postal and telephone methods of administering questionnaires affect responses differently from different socioeconomic, demographic or clinical groups. Second, responses to questionnaire items administered by post or telephone need to be equivalent. Only two randomised studies have examined this. One found a difference in responses to the same question, while the other did not. 8,12

The observational study described here took place within SHIFT (Screening for Haemoglobinopathies in the First Trimester), a cluster randomised controlled trial assessing the feasibility, effectiveness and acceptability of offering antenatal sickle cell and thalassaemia (SCT) screening in primary care (ISRCTN00677850). 13 The trial was set in two inner city UK PCTs with a large number of people from minority ethnic groups and with high levels of material and social deprivation. The questionnaires had been designed to assess informed choice about antenatal SCT screening in a population with low levels of literacy. 14

At the start of the SHIFT Trial, participants were posted questionnaires (in one of 12 languages, selected according to information provided at trial consent) to complete at home and return by post. Initial monitoring indicated that this method gave an unacceptably low response rate. Based on a literature search we developed a second strategy that was more likely to achieve a high response rate in a multilingual population, namely offering participants the opportunity to complete a questionnaire over the telephone with a researcher and an interpreter if necessary. Participants offered the second strategy were also offered postal completion. This paper compares (1) the response rates of offering participants a choice of telephone and postal completion methods with those of offering postal completion only; (2) the responses obtained from the method selected, i.e. postal or telephone completion; and (3) the financial costs to the research team of offering postal and telephone administration of questionnaires in order that reliable estimates can be included in future research cost estimates.

Methods

Results

The response rate for women offered postal completion only was 26% compared with 67% for women offered a choice of telephone or postal completion [41% difference, 95% confidence interval of difference (CI diff) 30 to 52] (Table 1). 19 Women who spoke English had a higher response rate to a choice of completion methods compared to women who did not speak English (71% versus 56%, 95% CI diff 7 to 23). There were insufficient women who did not speak English and were only sent questionnaires by post to make a valid comparison with women who spoke English and were only sent questionnaires by post, i.e. between English and non-English speakers who were not offered a choice of completion methods.

The preferences and response rates of the subgroup of women offered a choice of completion methods are shown in Table 2. Among the women offered a choice of completion methods, 58% chose telephone completion (416/714 95% CI 55 to 62). The response rate for women choosing telephone completion was 98% compared with 23% for women choosing postal completion (75% difference, 95% CI diff 70 to 80). Eight women, having opted for telephone completion, did not complete the questionnaire. Questionnaires were sent in the post to these women. Of these eight, none was completed or returned. Women who did and did not speak English had similarly low response rates to postal questionnaires (23% versus 23%, 95% CI diff –10 to 10). While both groups had very high response rates to telephone completion, women who spoke English had slightly higher rates (99% versus 94%, 95% CI diff 0.7 to 11).

More women were identified as having positive attitudes towards undergoing antenatal SCT screening using telephone questionnaires than using postal questionnaires (96% versus 87%, 95% CI diff 0.006 to 15). There were no differences in the proportion of women identified with good knowledge about the test using telephone or postal questionnaires (31% versus 37% 95% CI diff –6 to 17) (Figure 1).

FIGURE 1.

Identification of knowledge and attitude from telephone and postal questionnaires.

Telephone questionnaires were completed in 20 languages other than English. Postal questionnaires were completed in nine languages other than English.

Discussion

The results of this study suggest that offering a choice of telephone or postal completion methods can result in a dramatic increase in response rates compared with postal completion alone. This study was not randomised trial and there could be other explanations for the observed effect. The intervention was complex. It included offering a choice, and offering a choice of two methods that differed in several ways. While we cannot exclude the possibility that the mere offer of a choice was responsible for the effect, the pattern of results suggests that it is the opportunity to complete questionnaires by telephone that is crucial. Telephone completion compared with postal completion allows for ready translation to more languages and provides social support to complete the questionnaire. Perhaps most importantly it removes the reading obstacle to questionnaire completion, thereby allowing the estimated 20–25% of the UK population who are functionally illiterate20 to participate in the research process.

The increased response rate obtained by administering questionnaires by telephone raises two questions: first, did the increase in response rate vary by social group, and, second, were responses obtained by telephone different to those obtained by post? Regarding differential response rates across social groups, the response from those offered a choice in this study was higher for English speakers than for non-English speakers. The lack of data from non responders means, however, that we are not able to assess how representative the responders are of the telephone or postal groups overall.

The data do not identify a reason why non-English speakers were less likely to opt for completing the questionnaire by telephone than were English speakers. It may reflect a failure of the trial to engage non-English speakers in the research process. Alternatively, it may reflect cultural differences in, for example, willingness to talk to strangers over the telephone. Further work is required to understand why non-English-speaking women were less likely to choose to complete questionnaires by telephone than were English-speaking women. Understanding this may allow telephone administration to be offered in ways that increase acceptability, and hence, use for non-English speakers and thereby increase response rates further.

Comparing responses from the two modalities revealed differences in assessed attitudes: women completing the questionnaire over the telephone were more likely to be classified as having positive attitudes towards undergoing antenatal SCT screening than women completing the postal questionnaire. There were no differences in knowledge by completion modality. One explanation is that the difference may be due to the way the questions were asked or social desirability. Alternatively, it may be because the study was not randomised. That is women who opted to complete the questionnaire over the telephone had more positive attitudes towards undergoing the test than women who completed the questionnaire by post.

There are financial costs involved in offering women a choice of telephone or postal questionnaire completion methods. These include the researcher time to administer the questionnaire over the telephone, as well as the cost of using telephone interpreters. The use of telephone interpreters did not negate the need (or cost) for translating written materials. Research budgets should include funding to cover all these costs.

There are potential problems associated with administering questionnaires over the telephone. The person administering the questionnaire needs to be trained to ensure that the potential participant does not feel under any pressure to participate in the evaluation or to complete the questionnaire. Training is also required to ensure that questions are not asked in a leading way, i.e. in a way likely to guide respondents to answer in a particular way.

For women who opted to complete the questionnaire by telephone, the use of interpreters did not appear to pose any problems. The use of telephone interpreters allows greater flexibility than using written translations because the languages required do not require specifying in advance. However, the use of telephone interpreters does not allow for the use of quality control procedures such as back translation that are available with written translations.

There has been some debate about the use of translation services within the NHS, which cost in the region of £55 million per annum. 21 It has been argued that the use of such services may compound rather than ameliorate the health problems of non-English speakers by reducing the need for them to learn English and thereby implicitly encouraging non-English speakers to remain outside of the dominant culture. 22 Others have argued that the lack of translation services results in poorer health care for non-English speakers. 23 Whilst this debate is likely to continue within the context of service provision, it is important to acknowledge that this debate is not applicable in research settings. A prerequisite for reliable trials is that trial outcomes are obtained for all participants and are representative of the population in general. 2 They therefore need to include people who do as well as those who do not speak English.

Conclusion

Studies requiring data to be collected by questionnaire may obtain higher response rates from both English and non-English speakers when a choice of telephone or postal administration and, where necessary, an interpreter, is offered compared with offering postal administration only. This approach will, however, incur additional research costs and uncertainty remains about the equivalence of responses obtained in the two methods.

Competing interests

The authors declare that they have no competing interests.

Authors’ contributions

ED participated in the design of the study and performed the statistical analysis and drafted the manuscript; KB participated in the design of the study and conducted the telephone interviews; ER participated in the design of the study and helped with drafting the manuscript; TMM conceived of the study, participated in its design and co-ordination and helped to draft the manuscript. All authors read and approved the final manuscript.

Acknowledgements

This study is funded by the UK Department of Health through its Health Technology Assessment programme, grant number 03/02/03 – ‘Antenatal screening for haemoglobinopathies in primary care: a cluster randomised trial to inform a simulation model. The Screening for Haemoglobinopathies in First Trimester (SHIFT) trial.’ The opinions and conclusions expressed here are those of the authors and do not necessarily reflect those of the UK NHS or the Department of Health. We are very grateful to the general practices, pregnant women and interpreters who participated in this study.

References

Ideally, all of the following criteria should be met before screening for a condition is initiated.

Health Technology Assessment reports published to date

1. Home parenteral nutrition: a systematic review.

   By Richards DM, Deeks JJ, Sheldon TA, Shaffer JL.

2. Diagnosis, management and screening of early localised prostate cancer.

   A review by Selley S, Donovan J, Faulkner A, Coast J, Gillatt D.

3. The diagnosis, management, treatment and costs of prostate cancer in England and Wales.

   A review by Chamberlain J, Melia J, Moss S, Brown J.

4. Screening for fragile X syndrome.

   A review by Murray J, Cuckle H, Taylor G, Hewison J.

5. A review of near patient testing in primary care.

   By Hobbs FDR, Delaney BC, Fitzmaurice DA, Wilson S, Hyde CJ, Thorpe GH, et al.

6. Systematic review of outpatient services for chronic pain control.

   By McQuay HJ, Moore RA, Eccleston C, Morley S, de C Williams AC.

7. Neonatal screening for inborn errors of metabolism: cost, yield and outcome.

   A review by Pollitt RJ, Green A, McCabe CJ, Booth A, Cooper NJ, Leonard JV, et al.

8. Preschool vision screening.

   A review by Snowdon SK, Stewart-Brown SL.

9. Implications of socio-cultural contexts for the ethics of clinical trials.

   A review by Ashcroft RE, Chadwick DW, Clark SRL, Edwards RHT, Frith L, Hutton JL.

10. A critical review of the role of neonatal hearing screening in the detection of congenital hearing impairment.

    By Davis A, Bamford J, Wilson I, Ramkalawan T, Forshaw M, Wright S.

11. Newborn screening for inborn errors of metabolism: a systematic review.

    By Seymour CA, Thomason MJ, Chalmers RA, Addison GM, Bain MD, Cockburn F, et al.

12. Routine preoperative testing: a systematic review of the evidence.

    By Munro J, Booth A, Nicholl J.

13. Systematic review of the effectiveness of laxatives in the elderly.

    By Petticrew M, Watt I, Sheldon T.

14. When and how to assess fast-changing technologies: a comparative study of medical applications of four generic technologies.

    A review by Mowatt G, Bower DJ, Brebner JA, Cairns JA, Grant AM, McKee L.

1. Antenatal screening for Down’s syndrome.

   A review by Wald NJ, Kennard A, Hackshaw A, McGuire A.

2. Screening for ovarian cancer: a systematic review.

   By Bell R, Petticrew M, Luengo S, Sheldon TA.

3. Consensus development methods, and their use in clinical guideline development.

   A review by Murphy MK, Black NA, Lamping DL, McKee CM, Sanderson CFB, Askham J, et al.

4. A cost–utility analysis of interferon beta for multiple sclerosis.

   By Parkin D, McNamee P, Jacoby A, Miller P, Thomas S, Bates D.

5. Effectiveness and efficiency of methods of dialysis therapy for end-stage renal disease: systematic reviews.

   By MacLeod A, Grant A, Donaldson C, Khan I, Campbell M, Daly C, et al.

6. Effectiveness of hip prostheses in primary total hip replacement: a critical review of evidence and an economic model.

   By Faulkner A, Kennedy LG, Baxter K, Donovan J, Wilkinson M, Bevan G.

7. Antimicrobial prophylaxis in colorectal surgery: a systematic review of randomised controlled trials.

   By Song F, Glenny AM.

8. Bone marrow and peripheral blood stem cell transplantation for malignancy.

   A review by Johnson PWM, Simnett SJ, Sweetenham JW, Morgan GJ, Stewart LA.

9. Screening for speech and language delay: a systematic review of the literature.

   By Law J, Boyle J, Harris F, Harkness A, Nye C.

10. Resource allocation for chronic stable angina: a systematic review of effectiveness, costs and cost-effectiveness of alternative interventions.

    By Sculpher MJ, Petticrew M, Kelland JL, Elliott RA, Holdright DR, Buxton MJ.

11. Detection, adherence and control of hypertension for the prevention of stroke: a systematic review.

    By Ebrahim S.

12. Postoperative analgesia and vomiting, with special reference to day-case surgery: a systematic review.

    By McQuay HJ, Moore RA.

13. Choosing between randomised and nonrandomised studies: a systematic review.

    By Britton A, McKee M, Black N, McPherson K, Sanderson C, Bain C.

14. Evaluating patient-based outcome measures for use in clinical trials.

    A review by Fitzpatrick R, Davey C, Buxton MJ, Jones DR.

15. Ethical issues in the design and conduct of randomised controlled trials.

    A review by Edwards SJL, Lilford RJ, Braunholtz DA, Jackson JC, Hewison J, Thornton J.

16. Qualitative research methods in health technology assessment: a review of the literature.

    By Murphy E, Dingwall R, Greatbatch D, Parker S, Watson P.

17. The costs and benefits of paramedic skills in pre-hospital trauma care.

    By Nicholl J, Hughes S, Dixon S, Turner J, Yates D.

18. Systematic review of endoscopic ultrasound in gastro-oesophageal cancer.

    By Harris KM, Kelly S, Berry E, Hutton J, Roderick P, Cullingworth J, et al.

19. Systematic reviews of trials and other studies.

    By Sutton AJ, Abrams KR, Jones DR, Sheldon TA, Song F.

20. Primary total hip replacement surgery: a systematic review of outcomes and modelling of cost-effectiveness associated with different prostheses.

    A review by Fitzpatrick R, Shortall E, Sculpher M, Murray D, Morris R, Lodge M, et al.

1. Informed decision making: an annotated bibliography and systematic review.

   By Bekker H, Thornton JG, Airey CM, Connelly JB, Hewison J, Robinson MB, et al.

2. Handling uncertainty when performing economic evaluation of healthcare interventions.

   A review by Briggs AH, Gray AM.

3. The role of expectancies in the placebo effect and their use in the delivery of health care: a systematic review.

   By Crow R, Gage H, Hampson S, Hart J, Kimber A, Thomas H.

4. A randomised controlled trial of different approaches to universal antenatal HIV testing: uptake and acceptability. Annex: Antenatal HIV testing – assessment of a routine voluntary approach.

   By Simpson WM, Johnstone FD, Boyd FM, Goldberg DJ, Hart GJ, Gormley SM, et al.

5. Methods for evaluating area-wide and organisation-based interventions in health and health care: a systematic review.

   By Ukoumunne OC, Gulliford MC, Chinn S, Sterne JAC, Burney PGJ.

6. Assessing the costs of healthcare technologies in clinical trials.

   A review by Johnston K, Buxton MJ, Jones DR, Fitzpatrick R.

7. Cooperatives and their primary care emergency centres: organisation and impact.

   By Hallam L, Henthorne K.

8. Screening for cystic fibrosis.

   A review by Murray J, Cuckle H, Taylor G, Littlewood J, Hewison J.

9. A review of the use of health status measures in economic evaluation.

   By Brazier J, Deverill M, Green C, Harper R, Booth A.

10. Methods for the analysis of quality-of-life and survival data in health technology assessment.

    A review by Billingham LJ, Abrams KR, Jones DR.

11. Antenatal and neonatal haemoglobinopathy screening in the UK: review and economic analysis.

    By Zeuner D, Ades AE, Karnon J, Brown J, Dezateux C, Anionwu EN.

12. Assessing the quality of reports of randomised trials: implications for the conduct of meta-analyses.

    A review by Moher D, Cook DJ, Jadad AR, Tugwell P, Moher M, Jones A, et al.

13. ‘Early warning systems’ for identifying new healthcare technologies.

    By Robert G, Stevens A, Gabbay J.

14. A systematic review of the role of human papillomavirus testing within a cervical screening programme.

    By Cuzick J, Sasieni P, Davies P, Adams J, Normand C, Frater A, et al.

15. Near patient testing in diabetes clinics: appraising the costs and outcomes.

    By Grieve R, Beech R, Vincent J, Mazurkiewicz J.

16. Positron emission tomography: establishing priorities for health technology assessment.

    A review by Robert G, Milne R.

17. The debridement of chronic wounds: a systematic review.

    By Bradley M, Cullum N, Sheldon T.

18. Systematic reviews of wound care management: (2) Dressings and topical agents used in the healing of chronic wounds.

    By Bradley M, Cullum N, Nelson EA, Petticrew M, Sheldon T, Torgerson D.

19. A systematic literature review of spiral and electron beam computed tomography: with particular reference to clinical applications in hepatic lesions, pulmonary embolus and coronary artery disease.

    By Berry E, Kelly S, Hutton J, Harris KM, Roderick P, Boyce JC, et al.

20. What role for statins? A review and economic model.

    By Ebrahim S, Davey Smith G, McCabe C, Payne N, Pickin M, Sheldon TA, et al.

21. Factors that limit the quality, number and progress of randomised controlled trials.

    A review by Prescott RJ, Counsell CE, Gillespie WJ, Grant AM, Russell IT, Kiauka S, et al.

22. Antimicrobial prophylaxis in total hip replacement: a systematic review.

    By Glenny AM, Song F.

23. Health promoting schools and health promotion in schools: two systematic reviews.

    By Lister-Sharp D, Chapman S, Stewart-Brown S, Sowden A.

24. Economic evaluation of a primary care-based education programme for patients with osteoarthritis of the knee.

    A review by Lord J, Victor C, Littlejohns P, Ross FM, Axford JS.

1. The estimation of marginal time preference in a UK-wide sample (TEMPUS) project.

   A review by Cairns JA, van der Pol MM.

2. Geriatric rehabilitation following fractures in older people: a systematic review.

   By Cameron I, Crotty M, Currie C, Finnegan T, Gillespie L, Gillespie W, et al.

3. Screening for sickle cell disease and thalassaemia: a systematic review with supplementary research.

   By Davies SC, Cronin E, Gill M, Greengross P, Hickman M, Normand C.

4. Community provision of hearing aids and related audiology services.

   A review by Reeves DJ, Alborz A, Hickson FS, Bamford JM.

5. False-negative results in screening programmes: systematic review of impact and implications.

   By Petticrew MP, Sowden AJ, Lister-Sharp D, Wright K.

6. Costs and benefits of community postnatal support workers: a randomised controlled trial.

   By Morrell CJ, Spiby H, Stewart P, Walters S, Morgan A.

7. Implantable contraceptives (subdermal implants and hormonally impregnated intrauterine systems) versus other forms of reversible contraceptives: two systematic reviews to assess relative effectiveness, acceptability, tolerability and cost-effectiveness.

   By French RS, Cowan FM, Mansour DJA, Morris S, Procter T, Hughes D, et al.

8. An introduction to statistical methods for health technology assessment.

   A review by White SJ, Ashby D, Brown PJ.

9. Disease-modifying drugs for multiple sclerosis: a rapid and systematic review.

   By Clegg A, Bryant J, Milne R.

10. Publication and related biases.

    A review by Song F, Eastwood AJ, Gilbody S, Duley L, Sutton AJ.

11. Cost and outcome implications of the organisation of vascular services.

    By Michaels J, Brazier J, Palfreyman S, Shackley P, Slack R.

12. Monitoring blood glucose control in diabetes mellitus: a systematic review.

    By Coster S, Gulliford MC, Seed PT, Powrie JK, Swaminathan R.

13. The effectiveness of domiciliary health visiting: a systematic review of international studies and a selective review of the British literature.

    By Elkan R, Kendrick D, Hewitt M, Robinson JJA, Tolley K, Blair M, et al.

14. The determinants of screening uptake and interventions for increasing uptake: a systematic review.

    By Jepson R, Clegg A, Forbes C, Lewis R, Sowden A, Kleijnen J.

15. The effectiveness and cost-effectiveness of prophylactic removal of wisdom teeth.

    A rapid review by Song F, O’Meara S, Wilson P, Golder S, Kleijnen J.

16. Ultrasound screening in pregnancy: a systematic review of the clinical effectiveness, cost-effectiveness and women’s views.

    By Bricker L, Garcia J, Henderson J, Mugford M, Neilson J, Roberts T, et al.

17. A rapid and systematic review of the effectiveness and cost-effectiveness of the taxanes used in the treatment of advanced breast and ovarian cancer.

    By Lister-Sharp D, McDonagh MS, Khan KS, Kleijnen J.

18. Liquid-based cytology in cervical screening: a rapid and systematic review.

    By Payne N, Chilcott J, McGoogan E.

19. Randomised controlled trial of non-directive counselling, cognitive–behaviour therapy and usual general practitioner care in the management of depression as well as mixed anxiety and depression in primary care.

    By King M, Sibbald B, Ward E, Bower P, Lloyd M, Gabbay M, et al.

20. Routine referral for radiography of patients presenting with low back pain: is patients’ outcome influenced by GPs’ referral for plain radiography?

    By Kerry S, Hilton S, Patel S, Dundas D, Rink E, Lord J.

21. Systematic reviews of wound care management: (3) antimicrobial agents for chronic wounds; (4) diabetic foot ulceration.

    By O’Meara S, Cullum N, Majid M, Sheldon T.

22. Using routine data to complement and enhance the results of randomised controlled trials.

    By Lewsey JD, Leyland AH, Murray GD, Boddy FA.

23. Coronary artery stents in the treatment of ischaemic heart disease: a rapid and systematic review.

    By Meads C, Cummins C, Jolly K, Stevens A, Burls A, Hyde C.

24. Outcome measures for adult critical care: a systematic review.

    By Hayes JA, Black NA, Jenkinson C, Young JD, Rowan KM, Daly K, et al.

25. A systematic review to evaluate the effectiveness of interventions to promote the initiation of breastfeeding.

    By Fairbank L, O’Meara S, Renfrew MJ, Woolridge M, Sowden AJ, Lister-Sharp D.

26. Implantable cardioverter defibrillators: arrhythmias. A rapid and systematic review.

    By Parkes J, Bryant J, Milne R.

27. Treatments for fatigue in multiple sclerosis: a rapid and systematic review.

    By Brañas P, Jordan R, Fry-Smith A, Burls A, Hyde C.

28. Early asthma prophylaxis, natural history, skeletal development and economy (EASE): a pilot randomised controlled trial.

    By Baxter-Jones ADG, Helms PJ, Russell G, Grant A, Ross S, Cairns JA, et al.

29. Screening for hypercholesterolaemia versus case finding for familial hypercholesterolaemia: a systematic review and cost-effectiveness analysis.

    By Marks D, Wonderling D, Thorogood M, Lambert H, Humphries SE, Neil HAW.

30. A rapid and systematic review of the clinical effectiveness and cost-effectiveness of glycoprotein IIb/IIIa antagonists in the medical management of unstable angina.

    By McDonagh MS, Bachmann LM, Golder S, Kleijnen J, ter Riet G.

31. A randomised controlled trial of prehospital intravenous fluid replacement therapy in serious trauma.

    By Turner J, Nicholl J, Webber L, Cox H, Dixon S, Yates D.

32. Intrathecal pumps for giving opioids in chronic pain: a systematic review.

    By Williams JE, Louw G, Towlerton G.

33. Combination therapy (interferon alfa and ribavirin) in the treatment of chronic hepatitis C: a rapid and systematic review.

    By Shepherd J, Waugh N, Hewitson P.

34. A systematic review of comparisons of effect sizes derived from randomised and non-randomised studies.

    By MacLehose RR, Reeves BC, Harvey IM, Sheldon TA, Russell IT, Black AMS.

35. Intravascular ultrasound-guided interventions in coronary artery disease: a systematic literature review, with decision-analytic modelling, of outcomes and cost-effectiveness.

    By Berry E, Kelly S, Hutton J, Lindsay HSJ, Blaxill JM, Evans JA, et al.

36. A randomised controlled trial to evaluate the effectiveness and cost-effectiveness of counselling patients with chronic depression.

    By Simpson S, Corney R, Fitzgerald P, Beecham J.

37. Systematic review of treatments for atopic eczema.

    By Hoare C, Li Wan Po A, Williams H.

38. Bayesian methods in health technology assessment: a review.

    By Spiegelhalter DJ, Myles JP, Jones DR, Abrams KR.

39. The management of dyspepsia: a systematic review.

    By Delaney B, Moayyedi P, Deeks J, Innes M, Soo S, Barton P, et al.

40. A systematic review of treatments for severe psoriasis.

    By Griffiths CEM, Clark CM, Chalmers RJG, Li Wan Po A, Williams HC.

1. Clinical and cost-effectiveness of donepezil, rivastigmine and galantamine for Alzheimer’s disease: a rapid and systematic review.

   By Clegg A, Bryant J, Nicholson T, McIntyre L, De Broe S, Gerard K, et al.

2. The clinical effectiveness and cost-effectiveness of riluzole for motor neurone disease: a rapid and systematic review.

   By Stewart A, Sandercock J, Bryan S, Hyde C, Barton PM, Fry-Smith A, et al.

3. Equity and the economic evaluation of healthcare.

   By Sassi F, Archard L, Le Grand J.

4. Quality-of-life measures in chronic diseases of childhood.

   By Eiser C, Morse R.

5. Eliciting public preferences for healthcare: a systematic review of techniques.

   By Ryan M, Scott DA, Reeves C, Bate A, van Teijlingen ER, Russell EM, et al.

6. General health status measures for people with cognitive impairment: learning disability and acquired brain injury.

   By Riemsma RP, Forbes CA, Glanville JM, Eastwood AJ, Kleijnen J.

7. An assessment of screening strategies for fragile X syndrome in the UK.

   By Pembrey ME, Barnicoat AJ, Carmichael B, Bobrow M, Turner G.

8. Issues in methodological research: perspectives from researchers and commissioners.

   By Lilford RJ, Richardson A, Stevens A, Fitzpatrick R, Edwards S, Rock F, et al.

9. Systematic reviews of wound care management: (5) beds; (6) compression; (7) laser therapy, therapeutic ultrasound, electrotherapy and electromagnetic therapy.

   By Cullum N, Nelson EA, Flemming K, Sheldon T.

10. Effects of educational and psychosocial interventions for adolescents with diabetes mellitus: a systematic review.

    By Hampson SE, Skinner TC, Hart J, Storey L, Gage H, Foxcroft D, et al.

11. Effectiveness of autologous chondrocyte transplantation for hyaline cartilage defects in knees: a rapid and systematic review.

    By Jobanputra P, Parry D, Fry-Smith A, Burls A.

12. Statistical assessment of the learning curves of health technologies.

    By Ramsay CR, Grant AM, Wallace SA, Garthwaite PH, Monk AF, Russell IT.

13. The effectiveness and cost-effectiveness of temozolomide for the treatment of recurrent malignant glioma: a rapid and systematic review.

    By Dinnes J, Cave C, Huang S, Major K, Milne R.

14. A rapid and systematic review of the clinical effectiveness and cost-effectiveness of debriding agents in treating surgical wounds healing by secondary intention.

    By Lewis R, Whiting P, ter Riet G, O’Meara S, Glanville J.

15. Home treatment for mental health problems: a systematic review.

    By Burns T, Knapp M, Catty J, Healey A, Henderson J, Watt H, et al.

16. How to develop cost-conscious guidelines.

    By Eccles M, Mason J.

17. The role of specialist nurses in multiple sclerosis: a rapid and systematic review.

    By De Broe S, Christopher F, Waugh N.

18. A rapid and systematic review of the clinical effectiveness and cost-effectiveness of orlistat in the management of obesity.

    By O’Meara S, Riemsma R, Shirran L, Mather L, ter Riet G.

19. The clinical effectiveness and cost-effectiveness of pioglitazone for type 2 diabetes mellitus: a rapid and systematic review.

    By Chilcott J, Wight J, Lloyd Jones M, Tappenden P.

20. Extended scope of nursing practice: a multicentre randomised controlled trial of appropriately trained nurses and preregistration house officers in preoperative assessment in elective general surgery.

    By Kinley H, Czoski-Murray C, George S, McCabe C, Primrose J, Reilly C, et al.

21. Systematic reviews of the effectiveness of day care for people with severe mental disorders: (1) Acute day hospital versus admission; (2) Vocational rehabilitation; (3) Day hospital versus outpatient care.

    By Marshall M, Crowther R, Almaraz- Serrano A, Creed F, Sledge W, Kluiter H, et al.

22. The measurement and monitoring of surgical adverse events.

    By Bruce J, Russell EM, Mollison J, Krukowski ZH.

23. Action research: a systematic review and guidance for assessment.

    By Waterman H, Tillen D, Dickson R, de Koning K.

24. A rapid and systematic review of the clinical effectiveness and cost-effectiveness of gemcitabine for the treatment of pancreatic cancer.

    By Ward S, Morris E, Bansback N, Calvert N, Crellin A, Forman D, et al.

25. A rapid and systematic review of the evidence for the clinical effectiveness and cost-effectiveness of irinotecan, oxaliplatin and raltitrexed for the treatment of advanced colorectal cancer.

    By Lloyd Jones M, Hummel S, Bansback N, Orr B, Seymour M.

26. Comparison of the effectiveness of inhaler devices in asthma and chronic obstructive airways disease: a systematic review of the literature.

    By Brocklebank D, Ram F, Wright J, Barry P, Cates C, Davies L, et al.

27. The cost-effectiveness of magnetic resonance imaging for investigation of the knee joint.

    By Bryan S, Weatherburn G, Bungay H, Hatrick C, Salas C, Parry D, et al.

28. A rapid and systematic review of the clinical effectiveness and cost-effectiveness of topotecan for ovarian cancer.

    By Forbes C, Shirran L, Bagnall A-M, Duffy S, ter Riet G.

29. Superseded by a report published in a later volume.

30. The role of radiography in primary care patients with low back pain of at least 6 weeks duration: a randomised (unblinded) controlled trial.

    By Kendrick D, Fielding K, Bentley E, Miller P, Kerslake R, Pringle M.

31. Design and use of questionnaires: a review of best practice applicable to surveys of health service staff and patients.

    By McColl E, Jacoby A, Thomas L, Soutter J, Bamford C, Steen N, et al.

32. A rapid and systematic review of the clinical effectiveness and cost-effectiveness of paclitaxel, docetaxel, gemcitabine and vinorelbine in non-small-cell lung cancer.

    By Clegg A, Scott DA, Sidhu M, Hewitson P, Waugh N.

33. Subgroup analyses in randomised controlled trials: quantifying the risks of false-positives and false-negatives.

    By Brookes ST, Whitley E, Peters TJ, Mulheran PA, Egger M, Davey Smith G.

34. Depot antipsychotic medication in the treatment of patients with schizophrenia: (1) Meta-review; (2) Patient and nurse attitudes.

    By David AS, Adams C.

35. A systematic review of controlled trials of the effectiveness and cost-effectiveness of brief psychological treatments for depression.

    By Churchill R, Hunot V, Corney R, Knapp M, McGuire H, Tylee A, et al.

36. Cost analysis of child health surveillance.

    By Sanderson D, Wright D, Acton C, Duree D.

1. A study of the methods used to select review criteria for clinical audit.

   By Hearnshaw H, Harker R, Cheater F, Baker R, Grimshaw G.

2. Fludarabine as second-line therapy for B cell chronic lymphocytic leukaemia: a technology assessment.

   By Hyde C, Wake B, Bryan S, Barton P, Fry-Smith A, Davenport C, et al.

3. Rituximab as third-line treatment for refractory or recurrent Stage III or IV follicular non-Hodgkin’s lymphoma: a systematic review and economic evaluation.

   By Wake B, Hyde C, Bryan S, Barton P, Song F, Fry-Smith A, et al.

4. A systematic review of discharge arrangements for older people.

   By Parker SG, Peet SM, McPherson A, Cannaby AM, Baker R, Wilson A, et al.

5. The clinical effectiveness and cost-effectiveness of inhaler devices used in the routine management of chronic asthma in older children: a systematic review and economic evaluation.

   By Peters J, Stevenson M, Beverley C, Lim J, Smith S.

6. The clinical effectiveness and cost-effectiveness of sibutramine in the management of obesity: a technology assessment.

   By O’Meara S, Riemsma R, Shirran L, Mather L, ter Riet G.

7. The cost-effectiveness of magnetic resonance angiography for carotid artery stenosis and peripheral vascular disease: a systematic review.

   By Berry E, Kelly S, Westwood ME, Davies LM, Gough MJ, Bamford JM, et al.

8. Promoting physical activity in South Asian Muslim women through ‘exercise on prescription’.

   By Carroll B, Ali N, Azam N.

9. Zanamivir for the treatment of influenza in adults: a systematic review and economic evaluation.

   By Burls A, Clark W, Stewart T, Preston C, Bryan S, Jefferson T, et al.

10. A review of the natural history and epidemiology of multiple sclerosis: implications for resource allocation and health economic models.

    By Richards RG, Sampson FC, Beard SM, Tappenden P.

11. Screening for gestational diabetes: a systematic review and economic evaluation.

    By Scott DA, Loveman E, McIntyre L, Waugh N.

12. The clinical effectiveness and cost-effectiveness of surgery for people with morbid obesity: a systematic review and economic evaluation.

    By Clegg AJ, Colquitt J, Sidhu MK, Royle P, Loveman E, Walker A.

13. The clinical effectiveness of trastuzumab for breast cancer: a systematic review.

    By Lewis R, Bagnall A-M, Forbes C, Shirran E, Duffy S, Kleijnen J, et al.

14. The clinical effectiveness and cost-effectiveness of vinorelbine for breast cancer: a systematic review and economic evaluation.

    By Lewis R, Bagnall A-M, King S, Woolacott N, Forbes C, Shirran L, et al.

15. A systematic review of the effectiveness and cost-effectiveness of metal-on-metal hip resurfacing arthroplasty for treatment of hip disease.

    By Vale L, Wyness L, McCormack K, McKenzie L, Brazzelli M, Stearns SC.

16. The clinical effectiveness and cost-effectiveness of bupropion and nicotine replacement therapy for smoking cessation: a systematic review and economic evaluation.

    By Woolacott NF, Jones L, Forbes CA, Mather LC, Sowden AJ, Song FJ, et al.

17. A systematic review of effectiveness and economic evaluation of new drug treatments for juvenile idiopathic arthritis: etanercept.

    By Cummins C, Connock M, Fry-Smith A, Burls A.

18. Clinical effectiveness and cost-effectiveness of growth hormone in children: a systematic review and economic evaluation.

    By Bryant J, Cave C, Mihaylova B, Chase D, McIntyre L, Gerard K, et al.

19. Clinical effectiveness and cost-effectiveness of growth hormone in adults in relation to impact on quality of life: a systematic review and economic evaluation.

    By Bryant J, Loveman E, Chase D, Mihaylova B, Cave C, Gerard K, et al.

20. Clinical medication review by a pharmacist of patients on repeat prescriptions in general practice: a randomised controlled trial.

    By Zermansky AG, Petty DR, Raynor DK, Lowe CJ, Freementle N, Vail A.

21. The effectiveness of infliximab and etanercept for the treatment of rheumatoid arthritis: a systematic review and economic evaluation.

    By Jobanputra P, Barton P, Bryan S, Burls A.

22. A systematic review and economic evaluation of computerised cognitive behaviour therapy for depression and anxiety.

    By Kaltenthaler E, Shackley P, Stevens K, Beverley C, Parry G, Chilcott J.

23. A systematic review and economic evaluation of pegylated liposomal doxorubicin hydrochloride for ovarian cancer.

    By Forbes C, Wilby J, Richardson G, Sculpher M, Mather L, Riemsma R.

24. A systematic review of the effectiveness of interventions based on a stages-of-change approach to promote individual behaviour change.

    By Riemsma RP, Pattenden J, Bridle C, Sowden AJ, Mather L, Watt IS, et al.

25. A systematic review update of the clinical effectiveness and cost-effectiveness of glycoprotein IIb/IIIa antagonists.

    By Robinson M, Ginnelly L, Sculpher M, Jones L, Riemsma R, Palmer S, et al.

26. A systematic review of the effectiveness, cost-effectiveness and barriers to implementation of thrombolytic and neuroprotective therapy for acute ischaemic stroke in the NHS.

    By Sandercock P, Berge E, Dennis M, Forbes J, Hand P, Kwan J, et al.

27. A randomised controlled crossover trial of nurse practitioner versus doctor-led outpatient care in a bronchiectasis clinic.

    By Caine N, Sharples LD, Hollingworth W, French J, Keogan M, Exley A, et al.

28. Clinical effectiveness and cost – consequences of selective serotonin reuptake inhibitors in the treatment of sex offenders.

    By Adi Y, Ashcroft D, Browne K, Beech A, Fry-Smith A, Hyde C.

29. Treatment of established osteoporosis: a systematic review and cost–utility analysis.

    By Kanis JA, Brazier JE, Stevenson M, Calvert NW, Lloyd Jones M.

30. Which anaesthetic agents are cost-effective in day surgery? Literature review, national survey of practice and randomised controlled trial.

    By Elliott RA Payne K, Moore JK, Davies LM, Harper NJN, St Leger AS, et al.

31. Screening for hepatitis C among injecting drug users and in genitourinary medicine clinics: systematic reviews of effectiveness, modelling study and national survey of current practice.

    By Stein K, Dalziel K, Walker A, McIntyre L, Jenkins B, Horne J, et al.

32. The measurement of satisfaction with healthcare: implications for practice from a systematic review of the literature.

    By Crow R, Gage H, Hampson S, Hart J, Kimber A, Storey L, et al.

33. The effectiveness and cost-effectiveness of imatinib in chronic myeloid leukaemia: a systematic review.

    By Garside R, Round A, Dalziel K, Stein K, Royle R.

34. A comparative study of hypertonic saline, daily and alternate-day rhDNase in children with cystic fibrosis.

    By Suri R, Wallis C, Bush A, Thompson S, Normand C, Flather M, et al.

35. A systematic review of the costs and effectiveness of different models of paediatric home care.

    By Parker G, Bhakta P, Lovett CA, Paisley S, Olsen R, Turner D, et al.

1. How important are comprehensive literature searches and the assessment of trial quality in systematic reviews? Empirical study.

   By Egger M, Jüni P, Bartlett C, Holenstein F, Sterne J.

2. Systematic review of the effectiveness and cost-effectiveness, and economic evaluation, of home versus hospital or satellite unit haemodialysis for people with end-stage renal failure.

   By Mowatt G, Vale L, Perez J, Wyness L, Fraser C, MacLeod A, et al.

3. Systematic review and economic evaluation of the effectiveness of infliximab for the treatment of Crohn’s disease.

   By Clark W, Raftery J, Barton P, Song F, Fry-Smith A, Burls A.

4. A review of the clinical effectiveness and cost-effectiveness of routine anti-D prophylaxis for pregnant women who are rhesus negative.

   By Chilcott J, Lloyd Jones M, Wight J, Forman K, Wray J, Beverley C, et al.

5. Systematic review and evaluation of the use of tumour markers in paediatric oncology: Ewing’s sarcoma and neuroblastoma.

   By Riley RD, Burchill SA, Abrams KR, Heney D, Lambert PC, Jones DR, et al.

6. The cost-effectiveness of screening for Helicobacter pylori to reduce mortality and morbidity from gastric cancer and peptic ulcer disease: a discrete-event simulation model.

   By Roderick P, Davies R, Raftery J, Crabbe D, Pearce R, Bhandari P, et al.

7. The clinical effectiveness and cost-effectiveness of routine dental checks: a systematic review and economic evaluation.

   By Davenport C, Elley K, Salas C, Taylor-Weetman CL, Fry-Smith A, Bryan S, et al.

8. A multicentre randomised controlled trial assessing the costs and benefits of using structured information and analysis of women’s preferences in the management of menorrhagia.

   By Kennedy ADM, Sculpher MJ, Coulter A, Dwyer N, Rees M, Horsley S, et al.

9. Clinical effectiveness and cost–utility of photodynamic therapy for wet age-related macular degeneration: a systematic review and economic evaluation.

   By Meads C, Salas C, Roberts T, Moore D, Fry-Smith A, Hyde C.

10. Evaluation of molecular tests for prenatal diagnosis of chromosome abnormalities.

    By Grimshaw GM, Szczepura A, Hultén M, MacDonald F, Nevin NC, Sutton F, et al.

11. First and second trimester antenatal screening for Down’s syndrome: the results of the Serum, Urine and Ultrasound Screening Study (SURUSS).

    By Wald NJ, Rodeck C, Hackshaw AK, Walters J, Chitty L, Mackinson AM.

12. The effectiveness and cost-effectiveness of ultrasound locating devices for central venous access: a systematic review and economic evaluation.

    By Calvert N, Hind D, McWilliams RG, Thomas SM, Beverley C, Davidson A.

13. A systematic review of atypical antipsychotics in schizophrenia.

    By Bagnall A-M, Jones L, Lewis R, Ginnelly L, Glanville J, Torgerson D, et al.

14. Prostate Testing for Cancer and Treatment (ProtecT) feasibility study.

    By Donovan J, Hamdy F, Neal D, Peters T, Oliver S, Brindle L, et al.

15. Early thrombolysis for the treatment of acute myocardial infarction: a systematic review and economic evaluation.

    By Boland A, Dundar Y, Bagust A, Haycox A, Hill R, Mujica Mota R, et al.

16. Screening for fragile X syndrome: a literature review and modelling.

    By Song FJ, Barton P, Sleightholme V, Yao GL, Fry-Smith A.

17. Systematic review of endoscopic sinus surgery for nasal polyps.

    By Dalziel K, Stein K, Round A, Garside R, Royle P.

18. Towards efficient guidelines: how to monitor guideline use in primary care.

    By Hutchinson A, McIntosh A, Cox S, Gilbert C.

19. Effectiveness and cost-effectiveness of acute hospital-based spinal cord injuries services: systematic review.

    By Bagnall A-M, Jones L, Richardson G, Duffy S, Riemsma R.

20. Prioritisation of health technology assessment. The PATHS model: methods and case studies.

    By Townsend J, Buxton M, Harper G.

21. Systematic review of the clinical effectiveness and cost-effectiveness of tension-free vaginal tape for treatment of urinary stress incontinence.

    By Cody J, Wyness L, Wallace S, Glazener C, Kilonzo M, Stearns S, et al.

22. The clinical and cost-effectiveness of patient education models for diabetes: a systematic review and economic evaluation.

    By Loveman E, Cave C, Green C, Royle P, Dunn N, Waugh N.

23. The role of modelling in prioritising and planning clinical trials.

    By Chilcott J, Brennan A, Booth A, Karnon J, Tappenden P.

24. Cost–benefit evaluation of routine influenza immunisation in people 65–74 years of age.

    By Allsup S, Gosney M, Haycox A, Regan M.

25. The clinical and cost-effectiveness of pulsatile machine perfusion versus cold storage of kidneys for transplantation retrieved from heart-beating and non-heart-beating donors.

    By Wight J, Chilcott J, Holmes M, Brewer N.

26. Can randomised trials rely on existing electronic data? A feasibility study to explore the value of routine data in health technology assessment.

    By Williams JG, Cheung WY, Cohen DR, Hutchings HA, Longo MF, Russell IT.

27. Evaluating non-randomised intervention studies.

    By Deeks JJ, Dinnes J, D’Amico R, Sowden AJ, Sakarovitch C, Song F, et al.

28. A randomised controlled trial to assess the impact of a package comprising a patient-orientated, evidence-based self- help guidebook and patient-centred consultations on disease management and satisfaction in inflammatory bowel disease.

    By Kennedy A, Nelson E, Reeves D, Richardson G, Roberts C, Robinson A, et al.

29. The effectiveness of diagnostic tests for the assessment of shoulder pain due to soft tissue disorders: a systematic review.

    By Dinnes J, Loveman E, McIntyre L, Waugh N.

30. The value of digital imaging in diabetic retinopathy.

    By Sharp PF, Olson J, Strachan F, Hipwell J, Ludbrook A, O’Donnell M, et al.

31. Lowering blood pressure to prevent myocardial infarction and stroke: a new preventive strategy.

    By Law M, Wald N, Morris J.

32. Clinical and cost-effectiveness of capecitabine and tegafur with uracil for the treatment of metastatic colorectal cancer: systematic review and economic evaluation.

    By Ward S, Kaltenthaler E, Cowan J, Brewer N.

33. Clinical and cost-effectiveness of new and emerging technologies for early localised prostate cancer: a systematic review.

    By Hummel S, Paisley S, Morgan A, Currie E, Brewer N.

34. Literature searching for clinical and cost-effectiveness studies used in health technology assessment reports carried out for the National Institute for Clinical Excellence appraisal system.

    By Royle P, Waugh N.

35. Systematic review and economic decision modelling for the prevention and treatment of influenza A and B.

    By Turner D, Wailoo A, Nicholson K, Cooper N, Sutton A, Abrams K.

36. A randomised controlled trial to evaluate the clinical and cost-effectiveness of Hickman line insertions in adult cancer patients by nurses.

    By Boland A, Haycox A, Bagust A, Fitzsimmons L.

37. Redesigning postnatal care: a randomised controlled trial of protocol-based midwifery-led care focused on individual women’s physical and psychological health needs.

    By MacArthur C, Winter HR, Bick DE, Lilford RJ, Lancashire RJ, Knowles H, et al.

38. Estimating implied rates of discount in healthcare decision-making.

    By West RR, McNabb R, Thompson AGH, Sheldon TA, Grimley Evans J.

39. Systematic review of isolation policies in the hospital management of methicillin-resistant Staphylococcus aureus: a review of the literature with epidemiological and economic modelling.

    By Cooper BS, Stone SP, Kibbler CC, Cookson BD, Roberts JA, Medley GF, et al.

40. Treatments for spasticity and pain in multiple sclerosis: a systematic review.

    By Beard S, Hunn A, Wight J.

41. The inclusion of reports of randomised trials published in languages other than English in systematic reviews.

    By Moher D, Pham B, Lawson ML, Klassen TP.

42. The impact of screening on future health-promoting behaviours and health beliefs: a systematic review.

    By Bankhead CR, Brett J, Bukach C, Webster P, Stewart-Brown S, Munafo M, et al.

1. What is the best imaging strategy for acute stroke?

   By Wardlaw JM, Keir SL, Seymour J, Lewis S, Sandercock PAG, Dennis MS, et al.

2. Systematic review and modelling of the investigation of acute and chronic chest pain presenting in primary care.

   By Mant J, McManus RJ, Oakes RAL, Delaney BC, Barton PM, Deeks JJ, et al.

3. The effectiveness and cost-effectiveness of microwave and thermal balloon endometrial ablation for heavy menstrual bleeding: a systematic review and economic modelling.

   By Garside R, Stein K, Wyatt K, Round A, Price A.

4. A systematic review of the role of bisphosphonates in metastatic disease.

   By Ross JR, Saunders Y, Edmonds PM, Patel S, Wonderling D, Normand C, et al.

5. Systematic review of the clinical effectiveness and cost-effectiveness of capecitabine (Xeloda^®) for locally advanced and/or metastatic breast cancer.

   By Jones L, Hawkins N, Westwood M, Wright K, Richardson G, Riemsma R.

6. Effectiveness and efficiency of guideline dissemination and implementation strategies.

   By Grimshaw JM, Thomas RE, MacLennan G, Fraser C, Ramsay CR, Vale L, et al.

7. Clinical effectiveness and costs of the Sugarbaker procedure for the treatment of pseudomyxoma peritonei.

   By Bryant J, Clegg AJ, Sidhu MK, Brodin H, Royle P, Davidson P.

8. Psychological treatment for insomnia in the regulation of long-term hypnotic drug use.

   By Morgan K, Dixon S, Mathers N, Thompson J, Tomeny M.

9. Improving the evaluation of therapeutic interventions in multiple sclerosis: development of a patient-based measure of outcome.

   By Hobart JC, Riazi A, Lamping DL, Fitzpatrick R, Thompson AJ.

10. A systematic review and economic evaluation of magnetic resonance cholangiopancreatography compared with diagnostic endoscopic retrograde cholangiopancreatography.

    By Kaltenthaler E, Bravo Vergel Y, Chilcott J, Thomas S, Blakeborough T, Walters SJ, et al.

11. The use of modelling to evaluate new drugs for patients with a chronic condition: the case of antibodies against tumour necrosis factor in rheumatoid arthritis.

    By Barton P, Jobanputra P, Wilson J, Bryan S, Burls A.

12. Clinical effectiveness and cost-effectiveness of neonatal screening for inborn errors of metabolism using tandem mass spectrometry: a systematic review.

    By Pandor A, Eastham J, Beverley C, Chilcott J, Paisley S.

13. Clinical effectiveness and cost-effectiveness of pioglitazone and rosiglitazone in the treatment of type 2 diabetes: a systematic review and economic evaluation.

    By Czoski-Murray C, Warren E, Chilcott J, Beverley C, Psyllaki MA, Cowan J.

14. Routine examination of the newborn: the EMREN study. Evaluation of an extension of the midwife role including a randomised controlled trial of appropriately trained midwives and paediatric senior house officers.

    By Townsend J, Wolke D, Hayes J, Davé S, Rogers C, Bloomfield L, et al.

15. Involving consumers in research and development agenda setting for the NHS: developing an evidence-based approach.

    By Oliver S, Clarke-Jones L, Rees R, Milne R, Buchanan P, Gabbay J, et al.

16. A multi-centre randomised controlled trial of minimally invasive direct coronary bypass grafting versus percutaneous transluminal coronary angioplasty with stenting for proximal stenosis of the left anterior descending coronary artery.

    By Reeves BC, Angelini GD, Bryan AJ, Taylor FC, Cripps T, Spyt TJ, et al.

17. Does early magnetic resonance imaging influence management or improve outcome in patients referred to secondary care with low back pain? A pragmatic randomised controlled trial.

    By Gilbert FJ, Grant AM, Gillan MGC, Vale L, Scott NW, Campbell MK, et al.

18. The clinical and cost-effectiveness of anakinra for the treatment of rheumatoid arthritis in adults: a systematic review and economic analysis.

    By Clark W, Jobanputra P, Barton P, Burls A.

19. A rapid and systematic review and economic evaluation of the clinical and cost-effectiveness of newer drugs for treatment of mania associated with bipolar affective disorder.

    By Bridle C, Palmer S, Bagnall A-M, Darba J, Duffy S, Sculpher M, et al.

20. Liquid-based cytology in cervical screening: an updated rapid and systematic review and economic analysis.

    By Karnon J, Peters J, Platt J, Chilcott J, McGoogan E, Brewer N.

21. Systematic review of the long-term effects and economic consequences of treatments for obesity and implications for health improvement.

    By Avenell A, Broom J, Brown TJ, Poobalan A, Aucott L, Stearns SC, et al.

22. Autoantibody testing in children with newly diagnosed type 1 diabetes mellitus.

    By Dretzke J, Cummins C, Sandercock J, Fry-Smith A, Barrett T, Burls A.

23. Clinical effectiveness and cost-effectiveness of prehospital intravenous fluids in trauma patients.

    By Dretzke J, Sandercock J, Bayliss S, Burls A.

24. Newer hypnotic drugs for the short-term management of insomnia: a systematic review and economic evaluation.

    By Dündar Y, Boland A, Strobl J, Dodd S, Haycox A, Bagust A, et al.

25. Development and validation of methods for assessing the quality of diagnostic accuracy studies.

    By Whiting P, Rutjes AWS, Dinnes J, Reitsma JB, Bossuyt PMM, Kleijnen J.

26. EVALUATE hysterectomy trial: a multicentre randomised trial comparing abdominal, vaginal and laparoscopic methods of hysterectomy.

    By Garry R, Fountain J, Brown J, Manca A, Mason S, Sculpher M, et al.

27. Methods for expected value of information analysis in complex health economic models: developments on the health economics of interferon-β and glatiramer acetate for multiple sclerosis.

    By Tappenden P, Chilcott JB, Eggington S, Oakley J, McCabe C.

28. Effectiveness and cost-effectiveness of imatinib for first-line treatment of chronic myeloid leukaemia in chronic phase: a systematic review and economic analysis.

    By Dalziel K, Round A, Stein K, Garside R, Price A.

29. VenUS I: a randomised controlled trial of two types of bandage for treating venous leg ulcers.

    By Iglesias C, Nelson EA, Cullum NA, Torgerson DJ, on behalf of the VenUS Team.

30. Systematic review of the effectiveness and cost-effectiveness, and economic evaluation, of myocardial perfusion scintigraphy for the diagnosis and management of angina and myocardial infarction.

    By Mowatt G, Vale L, Brazzelli M, Hernandez R, Murray A, Scott N, et al.

31. A pilot study on the use of decision theory and value of information analysis as part of the NHS Health Technology Assessment programme.

    By Claxton K, Ginnelly L, Sculpher M, Philips Z, Palmer S.

32. The Social Support and Family Health Study: a randomised controlled trial and economic evaluation of two alternative forms of postnatal support for mothers living in disadvantaged inner-city areas.

    By Wiggins M, Oakley A, Roberts I, Turner H, Rajan L, Austerberry H, et al.

33. Psychosocial aspects of genetic screening of pregnant women and newborns: a systematic review.

    By Green JM, Hewison J, Bekker HL, Bryant, Cuckle HS.

34. Evaluation of abnormal uterine bleeding: comparison of three outpatient procedures within cohorts defined by age and menopausal status.

    By Critchley HOD, Warner P, Lee AJ, Brechin S, Guise J, Graham B.

35. Coronary artery stents: a rapid systematic review and economic evaluation.

    By Hill R, Bagust A, Bakhai A, Dickson R, Dündar Y, Haycox A, et al.

36. Review of guidelines for good practice in decision-analytic modelling in health technology assessment.

    By Philips Z, Ginnelly L, Sculpher M, Claxton K, Golder S, Riemsma R, et al.

37. Rituximab (MabThera^®) for aggressive non-Hodgkin’s lymphoma: systematic review and economic evaluation.

    By Knight C, Hind D, Brewer N, Abbott V.

38. Clinical effectiveness and cost-effectiveness of clopidogrel and modified-release dipyridamole in the secondary prevention of occlusive vascular events: a systematic review and economic evaluation.

    By Jones L, Griffin S, Palmer S, Main C, Orton V, Sculpher M, et al.

39. Pegylated interferon α-2a and -2b in combination with ribavirin in the treatment of chronic hepatitis C: a systematic review and economic evaluation.

    By Shepherd J, Brodin H, Cave C, Waugh N, Price A, Gabbay J.

40. Clopidogrel used in combination with aspirin compared with aspirin alone in the treatment of non-ST-segment- elevation acute coronary syndromes: a systematic review and economic evaluation.

    By Main C, Palmer S, Griffin S, Jones L, Orton V, Sculpher M, et al.

41. Provision, uptake and cost of cardiac rehabilitation programmes: improving services to under-represented groups.

    By Beswick AD, Rees K, Griebsch I, Taylor FC, Burke M, West RR, et al.

42. Involving South Asian patients in clinical trials.

    By Hussain-Gambles M, Leese B, Atkin K, Brown J, Mason S, Tovey P.

43. Clinical and cost-effectiveness of continuous subcutaneous insulin infusion for diabetes.

    By Colquitt JL, Green C, Sidhu MK, Hartwell D, Waugh N.

44. Identification and assessment of ongoing trials in health technology assessment reviews.

    By Song FJ, Fry-Smith A, Davenport C, Bayliss S, Adi Y, Wilson JS, et al.

45. Systematic review and economic evaluation of a long-acting insulin analogue, insulin glargine

    By Warren E, Weatherley-Jones E, Chilcott J, Beverley C.

46. Supplementation of a home-based exercise programme with a class-based programme for people with osteoarthritis of the knees: a randomised controlled trial and health economic analysis.

    By McCarthy CJ, Mills PM, Pullen R, Richardson G, Hawkins N, Roberts CR, et al.

47. Clinical and cost-effectiveness of once-daily versus more frequent use of same potency topical corticosteroids for atopic eczema: a systematic review and economic evaluation.

    By Green C, Colquitt JL, Kirby J, Davidson P, Payne E.

48. Acupuncture of chronic headache disorders in primary care: randomised controlled trial and economic analysis.

    By Vickers AJ, Rees RW, Zollman CE, McCarney R, Smith CM, Ellis N, et al.

49. Generalisability in economic evaluation studies in healthcare: a review and case studies.

    By Sculpher MJ, Pang FS, Manca A, Drummond MF, Golder S, Urdahl H, et al.

50. Virtual outreach: a randomised controlled trial and economic evaluation of joint teleconferenced medical consultations.

    By Wallace P, Barber J, Clayton W, Currell R, Fleming K, Garner P, et al.