Treatments for hyperemesis gravidarum and nausea and vomiting in pregnancy: a systematic review and economic assessment

Article history

The research reported in this issue of the journal was funded by the HTA programme as project number 12/152/01. The contractual start date was in September 2013. The draft report began editorial review in December 2014 and was accepted for publication in March 2015. 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 reviewers 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

Luke Vale is a member of the funding panel for the National Institute for Health Research (NIHR) Programme Grants for Applied Research and NIHR Health Technology Assessment. He is also a Director of the NIHR Research Design Service in the North East. Stephen C Robson is a panel member of NIHR Efficacy and Mechanism Evaluation. Catherine Nelson-Piercy is a co-developer of the Royal College of Obstetricians and Gynaecologists green-top guideline on management of nausea vomiting and hyperemesis gravidarum.

Copyright statement

© Queen’s Printer and Controller of HMSO 2016. This work was produced by O’Donnell et al. under the terms of a commissioning contract issued by the Secretary of State for Health. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) 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: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

Background

Nausea and vomiting in pregnancy (NVP) is one of the most common symptoms of pregnancy affecting 50–85% of all women during the first half of pregnancy. 1 Symptoms usually start between 6 and 8 weeks of gestation, rise to a peak before the end of the first trimester and, in the majority of women, resolve by 20 weeks. 2 Most women (65–70%) self-manage their symptoms with avoidance of dietary triggers and oral hydration. 2 However, in the remainder, symptoms are more severe and/or protracted, leading to physical and psychosocial sequelae. These can include reduced quality of life (QoL), lost work time and negative effects on relationships with family and friends. 3

The most severe form of NVP is referred to as hyperemesis gravidarum (HG), and is reported to affect 0.3–1.0% of pregnant women. 1 It is characterised by intractable vomiting, dehydration, ketosis, electrolyte imbalance, nutritional deficiencies and weight loss (usually defined as > 5% of pre-pregnancy weight). However, there is no widely accepted point at which NVP becomes HG. Likewise, the distinction between studies of women with NVP and HG is generally not possible as the degree of dehydration and weight loss prior to the intervention are rarely reported. Furthermore, although some studies report baseline symptom severity using a validated scale, this is insufficient to make a diagnosis of HG. For these reasons, study populations are seldom described as having HG, and are more frequently defined in terms of the severity of NVP. Therefore, for the purposes of this review, studies on interventions for both NVP and HG have been included.

Aetiology

The underlying pathophysiology of NVP/HG is poorly understood but is thought to involve a combination of biological, physiological, psychological and sociocultural factors. 4

Genetic factors increase the risk of occurrence: results from a Norwegian study which included over 500,000 women found that the risk of HG was 15.2% in the second pregnancy of women who had a previous history of HG compared with 0.7% in women who did not [odds ratio (OR) 26.4, 95% confidence interval (CI) 24.2 to 28.7]. 5 The risk of developing HG is also increased threefold in the daughters of women who suffered from HG (unadjusted OR 2.9, 95% CI 2.4 to 3.6). 6

Endocrine factors, especially higher levels of human chorionic gonadotropin, as is the case in multiple or molar pregnancies, have been associated with more severe forms of NVP/HG. A recent observational study found that free human chorionic gonadotropin and pappalysin-1 (also known as pregnancy-associated plasma protein A) were higher in women suffering from HG than in non-sufferers. 7

Gestational transient thyrotoxicosis, has been reported in 60% of women suffering from HG8 and thyroid-stimulating hormone levels are raised in women with HG. 9 Certain human chorionic gonadotropin subtypes can stimulate thyroid-stimulating hormone receptors and so contribute to the hyperthyroidism. The degree of hyperthyroidism has been found to correlate with the severity of NVP/HG. 10 Higher levels of oestrogen, progesterone and leptin, and lower levels of adrenocorticotrophic hormone and prolactin have also been associated with HG. 11

Delayed gastric emptying related to relaxation of smooth muscle during pregnancy may influence NVP symptoms. Furthermore, higher rates of Helicobacter pylori infection have been noted in women suffering from HG:11 in a meta-analysis of 25 studies investigating the association of H. pylori and HG, 14 studies demonstrated an increased risk of HG in infected women (with OR between 2.42 and 109.3), and 11 studies found no association. 12

Lack of a definitive physiological trigger for HG has, in the past, led to numerous psychosomatic and psychological theories such as resentment or ambivalence towards the pregnancy, immaturity, conversion disorder, symptom of hysteria, neurosis or depression. 13 It is now more commonly accepted that psychological afflictions are a consequence of the condition rather than a cause. 4,9

Impact on patients

Severe NVP causes emotional and psychological distress and can have a profound effect on a women’s QoL, behavioural and cognitive function, affecting work capacity, household activities and interaction with children. 14–16 Women with HG report feeling isolated, depressed and lonely, unable to cope with routine daily interactions or simple tasks. Two recent observational studies found higher incidences of depression, anxiety and stress in women diagnosed with HG compared with controls. 17,18 Following cessation of symptoms the depression, anxiety and stress scores took several weeks to resolve,18 not returning to control values until the third trimester. 17 However, in some women these psychological symptoms do not fully resolve and can result in post-traumatic stress disorder. 19

As a result women make greater use of health-care resources. Based on Hospital Episode Statistics data for England, there were nearly 26,000 admissions for NVP/HG in 2010–11 with an average length of stay of 2 days. 20 These NHS costs are likely to underestimate the full costs as women may purchase a variety of products over the counter, pay for alternative therapies, receive treatment in primary care settings or as a hospital outpatient, and may incur extra child care, living costs and lost earnings. In addition, the associated increased risk of cognitive, behavioural and emotional dysfunction in pregnancy18 may prompt the use of further services and resources.

In the absence of a definitive cause, management of NVP/HG tends to focus on the alleviation of symptoms and prevention of serious morbidity. Typically, women are admitted to hospital, prescribed intravenous (i.v.) fluid therapy and antiemetic medication, but there is little time spent dealing with their psychological, social and emotional needs or providing information and guidance about the condition. The result is that women can feel unsupported, dissatisfied with care and experience negative interpersonal interactions with health-care providers. 21

Finally, severe NVP/HG has implications for offspring. A recent systematic review and meta-analysis reported that women with HG were more likely to deliver preterm (OR 1.32, 95% CI 1.04 to 1.68) and to have a baby that was small for gestational age (OR 1.28, 95% CI 1.02 to 1.60), although there was no evidence of an association with congenital anomalies (pooled results from three studies: OR 1.17, 95% CI 0.68 to 2.03) or perinatal death (OR 0.92, 95% CI 0.61 to 1.41). 22 A large Swedish birth cohort reported that women with HG who had their first admission in the second trimester were at increased risk of preterm pre-eclampsia (OR 2.09, 95% CI 1.38 to 3.16), placental abruption (OR 3.07, 95% CI 1.88 to 5.00) and to have a baby that was small for gestational age (OR 1.39, 95% CI 1.06 to 1.83), suggesting an association between HG and placental-mediated disease. 23

Assessment and diagnosis

The diagnosis of NVP/HG is made after excluding differential diagnoses, including gastrointestinal disorders, urinary tract infection, metabolic and endocrine disorders, drugs, psychological disorders (such as eating disorders) and other pregnancy-associated conditions (in particular molar pregnancy). However, there is currently no widely accepted approach to measuring the severity of symptoms in women. The most commonly used tools for the assessment of NVP/HG severity are presented in Table 1, with actual examples of the tools provided in Appendix 1.

However, although the measurement of NVP/HG symptom severity is the main aim for women and practitioners, other wider outcomes are also relevant when assessing the broader effectiveness of interventions. Thus, key secondary outcomes in studies to date have included both measures related to maternal physical and psychosocial health, and fetal or neonatal outcomes (Table 2).

Current interventions for nausea and vomiting in pregnancy/hyperemesis gravidarum

For the purposes of this report, interventions are considered in three broad groups:

• First-line interventions, usually initiated by women before seeking medical care and hence tend to be used in less severe NVP.

• Second-line interventions, typically prescribed when a women presents to medical care. Initially this is likely to be a general practitioner (GP) in primary care but may involve referral of women with more severe symptoms for inpatient, outpatient or day case care in hospital.

• Third-line interventions, reserved for women in hospital with persistent or recurrent symptoms despite second-line therapies.

The relationship between these three intervention groups is described in Figure 1, with key individual interventions described in detail in the following sections.

FIGURE 1.

Treatments for NVP. a, Care may also involve urine ± blood tests, weight and maternal observations; and b, care will involve urine and blood tests, weight, maternal observations and pelvic ultrasound. i.m., intramuscular.

Current guidance and use of therapies within the NHS

Currently there are no national guidelines within the NHS pertaining to NVP/HG; however, the Royal College of Obstetricians and Gynaecologists are in the process of producing said guidelines which should be published in early 2016. Initially, GPs may try different antiemetics before referring women to hospital. Traditionally, secondary care would involve admission to either an antenatal or gynaecology ward for treatment with i.v. fluids, antiemetics and vitamin supplements. Oral intake would gradually be resumed followed by discharge back into the community. Resumption of symptoms would result in readmission and a repeat of previous care, possibly trying different antiemetics or a combination thereof.

Increasingly, more obstetric and gynaecology units are using ‘day case’ management as the first option for initial referrals. Care usually involves some form of rapid rehydration and treatment with an i.v. antiemetic, followed by discharge with oral antiemetics, ideally with advice, support and guidance regarding self-help measures. However, assessing symptom severity, as well as the packages of care, vary substantially and lack a strong evidence base.

When available, day case, outpatient management does result in fewer admissions to hospital. Consequently, women who are admitted tend to be suffering from more severe symptoms. These women are likely to have had repeated hospital attendances and to have tried a number of different combinations of interventions. This latter group of women are likely to experience persistent severe symptoms, weight loss, electrolyte imbalance and failure to cope. In some of these women corticosteroid therapy may be considered an appropriate option when more conventional options have failed. In rare circumstances where this proves unsuccessful, enteral or parenteral nutrition may be instigated and, as a last resort, some women will opt for termination of pregnancy.

Aims and objectives

This study aimed to:

• review systematically the evidence of the clinical effectiveness and cost-effectiveness of each treatment for NVP/HG

• determine which therapies are most likely to be cost-effective for implementation into the UK NHS

• identify and prioritise future research needs.

Structure of the report

The following chapter (see Chapter 2) describes the methods employed for the systematic review and synthesis of evidence for interventions for HG and/or NVP. Chapter 3 provides an overview of the identified evidence, including the quality of the included studies, and a brief discussion of the issues that arose in attempting to synthesise the emergent data. Chapters 4–17 detail the findings for each individual intervention, focussing on the evidence for their effectiveness in terms of nausea, vomiting and retching. Chapter 18 presents the methods and results of the economic evaluation. Key issues considered likely to be important from the perspective of both patients and health-care practitioners are described in Chapters 19 and 20. The implications of the results of this review are discussed in depth in Chapter 21, with the final conclusions outlined in Chapter 22.

General methodology

The systematic review followed the approach suggested by the Evidence for Policy and Practice Information and Co-ordinating Centre at the Institute of Education, London. The review protocol was registered with PROSPERO, the International Prospective Register of Systematic Reviews,40 and it aimed to systematically appraise and summarise the evidence on available interventions for NVP/HG within the three broad groups described in Chapter 1:

• first-line interventions

• second-line interventions

• third-line inpatient interventions.

The review examined the evidence for these groups of interventions in relation to their clinical effectiveness and associated adverse events, and their cost-effectiveness.

Inclusion criteria

Search strategy

The search strategy was designed and executed by an experienced information specialist in collaboration with the rest of the research team. The original protocol stated that the search strategy would combine the two main conditions of pregnancy and NVP/HG with associated interventions and QoL outcomes. However, given both the extensive array of interventions used to address NVP/HG, and the relatively small available literature in this field, it was subsequently decided not to restrict the scope of the review by including key interventions or outcomes as search terms. Although such a strategy increases the number of papers to be reviewed; it minimises the risk of missing any relevant studies. The search was therefore structured around two core concepts: (1) nausea, vomiting and HG; and (2) pregnancy. Key words for both concepts were coupled with relevant medical subject heading (MeSH) and thesaurus terms. The search strategy was designed in MEDLINE and translated as appropriate to the other databases. All terms were truncated as appropriate and variant spellings were used. In order to reduce the number of studies returned, search filters for the relevant study types (RCTs or case series studies) were applied where possible. No time or language limit was set.

The full list of search terms for MEDLINE is presented in Table 3.

We searched the following electronic bibliographic databases on the dates described below, with update searches executed between the 11 and 16 September 2014 unless otherwise stated. As our initial scoping search highlighted the number of complementary medicine interventions for NVP/HG, the search was extended to include additional key databases of non-English-language studies44 (Latin American and Caribbean Health Sciences Literature and China National Knowledge Infrastructure).

• MEDLINE (Ovid) 1946–November 2013, searched 10 December 2013.

• EMBASE (Ovid) 1980–2013 week 50, searched 12 December 2013.

• Cochrane Central Register of Controlled Trials (Wiley) issue 11 2013, searched 16 December 2013.

• Cochrane Database of Systematic Reviews (Wiley) issue 11 2013, searched 16 December 2013.

• Database of Abstracts of Reviews of Effects (Wiley) issue 11 2013, searched 16 December 2013.

• Cumulative Index to Nursing and Allied Health Literature (EBSCOhost) 1981–November 2013, searched 17 December 2013.

• British Nursing Index (NHS Healthcare Databases) 1992–January 2014, searched 22 January 2014.

• PsycINFO (Ovid) 1806–December week 2 2013, searched 17 December 2013.

• Commonwealth Agricultural Bureaux (CAB) Abstracts (Ovid) 1910–2013 week 49, searched 18 December 2013.

• Latin American and Caribbean Health Sciences Literature (http://regional.bvsalud.org), searched 18 December 2013.

• Allied and Complementary Medicine Database (NHS Healthcare Databases) 1985–January 2014, searched 22 January 2014.

• Science Citation Index (Web of Knowledge) 1970–November 2013, searched 18 December 2013.

• Social Science Citation Index (Web of Knowledge) 1970–November 2013, searched 18 December 2013.

• Scopus, searched 8 January 2014.

• Conference Proceedings Index – Science (Web of Knowledge) 1990–November 2013, searched 18 December 2013.

• ClinicalTrials.gov searched 8 January 2014.

• NHS Economic Evaluation Database (Wiley) issue 11 2013, searched 16 December 2013.

• Health Economic Evaluations Database (Wiley), searched 30 January 2014.

• China National Knowledge Infrastructure (http://eng.cnki.net/grid2008/index.htm), searched 14 March 2014.

In addition, the reference lists of included papers and key relevant literature reviews identified during the search process were also examined for additional relevant studies, and Obstetric Medicine vol. 1(1) (September 2008) and vol. 7(2) (June 2014) were hand-searched. Furthermore, the following websites of relevant organisations were also searched in order to source as much unpublished literature as possible:

• Motherisk (URL: www.motherisk.org/women/drugs.jsp; accessed September 2014).

• American Congress of Obstetricians and Gynecologists (URL: www.acog.org/; accessed September 2014).

• Pregnancy Sickness Support (PSS) (URL: www.pregnancysicknesssupport.org.uk; accessed September 2014).

• National Institute for Health and Care Excellence Clinical Knowledge Summaries (URL: http://cks.nice.org.uk/nauseavomiting-in-pregnancy; accessed September 2014).

• Hyperemesis Education and Research (URL: www.helpher.org/health-professionals/treatments/index.php; accessed September 2014).

• UK Teratology Information Service (UKTIS) (URL: www.uktis.org/) including checking the references of the following relevant documents:

  • treatment of NVP (December 2013)

  • use of promethazine in pregnancy (October 2010)

  • use of vitamin B12 in pregnancy (September 2013)

  • use of pyridoxine (vitamin B6) in pregnancy (January 2011)

  • use of ginger in pregnancy (March 2013).

• European Medicines Agency (URL: www.ema.europa.eu/ema/; accessed September 2014).

• ProQuest Dissertations and Theses – UK & Ireland (URL: www.theses.com/; accessed September 2014).

• e-thesis online service (URL: http://ethos.bl.uk/; accessed September 2014).

• Trip (URL: www.tripdatabase.com/; accessed September 2014).

• System for Information on Grey Literature in Europe (URL: www.opengrey.eu/; accessed September 2014).

• Google Scholar (URL: https://scholar.google.co.uk; accessed September 2014).

First exclusion process

All records were imported into a bibliographic referencing software programme (EndNote v.X7; Thomson Reuters, CA, USA). Duplicate records were identified and deleted. The remaining references were assessed for relevancy by two independent investigators on the basis of the title and abstract (or title only if abstract not available). Papers were considered relevant to the systematic review if they met the inclusion criteria detailed in Inclusion criteria.

All of the titles and abstracts of all references were read by both investigators and classified as potentially eligible, not eligible or unclear within EndNote. Reconciliation of the resultant EndNote databases was conducted via Microsoft Excel 2013 (Microsoft Corporation, Redmond, WA, USA) and any discrepancies were discussed. In case of doubt, papers went through to the next stage of the exclusion process. Full-text copies of those papers identified as of potential relevance were obtained.

Second exclusion process

All full-text English-language papers obtained were assessed by two investigators independently and classified as relevant, not relevant or unclear on the basis of the same inclusion criteria. Any disagreements at this stage were resolved by discussion between the two researchers. Full-text papers published in languages other than English (German, French, Portuguese, Arabic, Chinese, Korean, Danish and Spanish) were assessed by native speakers of the relevant languages, working alongside one of the two investigators to ensure consistency and adequate compliance against the specified inclusion criteria. Tables of studies excluded at this stage were prepared, detailing reasons for exclusion.

Data extraction

Information from all papers identified as meeting the specified inclusion criteria was extracted using a structured data abstraction form. Key data extracted from eligible papers included:

1. study characteristics (bibliographic details, setting, intervention type, study population including definition of severity)

2. methodology and reporting

3. quantitative findings and conclusions.

For English-language papers, data extraction was carried out by one researcher and checked by another. For papers published in languages other than English, data extraction was carried out by a native speaker working alongside an investigator to identify and translate the relevant information. The data abstraction form for clinical effectiveness is presented as Appendix 2.

Risk of bias in included studies and quality assessment

The quality of the included studies was evaluated in accordance with the comprehensive approach advised by the Grading of Recommendations Assessment, Development and Evaluation (GRADE) Working Group. 45 The risk of bias of included RCTs was assessed with the Cochrane Collaboration’s tool46 (see Appendix 3 for full details). This included assessment of sequence generation; allocation concealment; blinding; selective reporting of outcomes; incomplete outcome data; and other possible sources of bias. For the incomplete outcome data item, we coded the satisfactory level of loss to follow-up for each outcome as ‘low risk of bias’, if fewer than 20% of patients were lost to follow-up and reasons for loss to follow-up were similar in all study arms. Disputes were resolved by discussion with another member of the review team. The risk of bias for case series studies was assessed using the component-based tool developed by the Effective Public Health Practice Project (EPHPP), Canada47 (see Appendix 4), which possesses a relatively high degree of inter-rater reliability in comparison to alternative tools. 48,49 For English-language papers, quality assessment was conducted independently by two investigators, with any disputes resolved by discussion. For papers published in languages other than English, quality assessment was carried out by a native speaker working alongside an investigator.

We had initially used an objective approach to make decisions about overall risk of bias; if at least one item was adjudged as being at high risk of bias then the trial was given a high risk of bias overall rating. Similarly, a trial that had no items scored as being at high risk of bias, but at least one was at an unclear risk of bias, then the trial was deemed to be at an unclear risk of bias. This meant that only trials that were scored at low risk of bias for all six risk of bias items could get an overall low risk judgement. Instead, we used a robust approach to assess overall risk of bias in a study, and which did not simply rely on a vote counting approach. This approach gave careful consideration to all six risk of bias item judgements, and the impact of unclear and high risk of bias in individual items, given the scope and context of the trial. For example, if a study had adequately addressed five of the six individual risk of bias items, but a placebo-controlled trial was unblinded in some capacity (either patients, personnel or both), then we scored this trial as being at high overall risk of bias, as this flaw in the conduct of such a trial could seriously impact on the results. However, in some trials blinding was not possible due to the type of interventions being compared so although the individual ‘blinding’ risk of bias item was scored as being at high risk of bias, this did not necessary mean the overall risk of bias in the study would follow. If other items were generally at low risk of bias, and we deemed blinding to be of little relevance, then we scored the study as being at low overall risk of bias. Justification for overall decisions has been provided where necessary.

In addition, two researchers independently assessed all included studies for the potential for imprecision, inconsistency and indirectness of results, using GRADE guidelines. 50–54 Summary of findings (SoF) tables were not presented due to the narrative nature of the review and the heterogeneity of the outcomes.

Data synthesis

First, the range of interventions, populations and outcomes that have been studied were described. The direction and size of the reported effects from effectiveness studies were presented overall, as well as grouped according to population, intervention type, outcomes and study design. Results are summarised in tables. Groups of studies using similar definitions of severity were identified, based on the data extracted and expert opinion. A coding frame was developed for the different definitions used, which was checked by the second systematic reviewer. Two clinical specialists within the research team then grouped the studies into the coding frame. The grouping produced was compared and any discrepancies, including definitions that did not fit into the coding frame, were resolved by discussion.

Meta-analysis was considered to be inappropriate due to heterogeneity, as will be illustrated in the overview of included studies chapter (see Chapter 3). This judgement was made after consideration of interventions, trial populations, and especially the reporting and definitions of outcome measures and methods. We also explored whether data from different studies could be transformed on to a common scale (e.g. symptom severity might be recoded into number no longer experiencing severe symptoms) using imputation and subject to sensitivity analysis and methods, but this was not possible. Therefore it was not necessary for the team to investigate the validity of performing mixed-treatment (indirect) comparisons, using appropriate methods to compare interventions that have not been compared directly with each other. 55,56 Heterogeneity could not be assessed by visualisation of results or, where relevant in statistical terms, by the chi-squared test for homogeneity and the I² statistic as specified a priori. Instead, a narrative synthesis was conducted, with the effects split into numerous categories and classes of intervention comparison described in the following chapter (see Chapter 3, Interventions and comparators and Table 6). Trials with more than two randomised groups may appear in more than one category depending on the comparisons made.

Data on effectiveness, fetal outcomes and adverse events were tabulated and described narratively, including variation in the form, setting, study population and delivery of the interventions. Given the inconsistencies in the application of both published and author-defined severity scales, studies were recategorised by clinical experts on the review team according to whether the participants were predominantly suffering from mild, moderate or severe NVP/HG. The effects are generally presented in terms of whether or not there were statistically significant differences between randomised groups at the last time point at which outcomes were assessed. However, where possible, magnitude of effects was reported such as a mean difference or risk ratio with corresponding 95% CIs. All studies are included in the narrative synthesis, irrespective of their risk of bias but the weight of evidence is discussed accordingly. Where necessary, comments are made in the text to advise caution for serious methodological shortcomings as well as applying the GRADE approach50–54 in the overall assessment of the quality of the evidence, although SoF tables were not constructed (see Risk of bias in included studies and quality assessment).

Owing to limited available data, it was not possible to examine publication bias using funnel plots as specified in the original review protocol. We were also unable to conduct subgroup analyses to explore the variation with pre-determined factors (e.g. the setting in which the intervention was applied and the severity and duration of symptoms at baseline) or sensitivity analyses to explore the impact of study design, including variation in definitions of outcomes, on measures of effectiveness due to lack of suitable data.

Studies identified

A flow chart of the studies is shown in Figure 2. In total, 11,830 papers were identified from the combination of standard electronic databases (n = 11,659), specialist Chinese databases (n = 102) and various sources of grey literature (n = 69). Of these, 5152 duplicate papers were identified and deleted (5150 from the standard electronic databases, and two from the grey literature).

FIGURE 2.

Flow chart of clinical effectiveness literature.

The deletion of duplicate papers left 6678 individual papers for assessment. After screening titles and abstracts, 322 papers were identified as of potential relevance and full-text copies of 309 papers were obtained (with the remainder unobtainable). Of these, 96 were judged ineligible for the effectiveness review and immediately excluded (narrative overviews, systematic literature reviews or economic evaluations). After the second exclusion process, comprising more detailed reading of each full-text paper, a further 138 papers were judged not to meet the inclusion criteria of the review and were also excluded.

Key reasons for exclusion were duplicate paper already included; participant inclusion criteria for the identified study judged not relevant to our review; did not include any of the pre-specified outcomes; or ineligible study design (no comparator group).

As a result, 75 papers were identified for data extraction, from a total of 73 separate studies. A full list of included studies is provided as Appendix 5. A table of excluded studies detailing reasons for exclusion is provided as Appendix 6.

Quality of included studies

Randomised controlled trials

Overall risk of bias

The results of the quality assessment procedure for the 64 included RCTs (reported in 66 papers) are displayed in Figure 3 and Table 4. There was variation both in terms of the quality of the studies and the quality of the reporting. In a large number of papers, there was insufficient detail provided to permit clear judgement of risk of bias in a range of key areas. Overall, 33 RCTs were classed as having low within-study risk of bias, 11 RCTs were classed as having high within-study risk of bias, and the remainder (n = 20) were classed as unclear in this respect. The high proportion of studies at unclear risk of bias was due to poor reporting and a lack of detail, particularly in the methods section. There were also a number of publications in abstract form only. As an unclear judgement was often due to poor reporting rather than specific methodological concerns, it was not judged appropriate to categorise studies with those deemed at high risk of bias as a result of more serious methodological flaws. Our robust approach to the assessment of the overall risk of bias within individual studies is described in more detail in Chapter 2, Risk of bias in included studies and quality assessment. More detail is provided below to illustrate the range in quality in terms of each individual component of the Cochrane’s risk of bias tool. 46

FIGURE 3.

Risk of bias graph: review authors’ judgements about each risk of bias item presented as percentages across all included RCT studies.

TABLE 4 - Risk of bias summary: review authors’ judgements about each risk of bias item for included RCTs

✗, high risk of bias; ?, unclear risk of bias; ✓, low risk of bias; CONSORT, Consolidated Standards of Reporting Trials Consolidated Standards of Reporting Trials.

Study	Random sequence generation	Allocation concealment	Blinding of participants, personnel and outcome assessors	Incomplete outcome data	Selective outcome reporting	Other sources of bias	Within study risk of bias	Comments
Abas 201457	✓	✓	✓	✓	✓	?	Low	Study at low risk of bias
Adamczak 200758	?	?	✗	?	?	?	Unclear	Abstract form only, blinding of participants and personnel not possible
Babaei 201459	?	?	✓	✓	✓	?	Unclear	Double-blind RCT, but methods of sequence generation and allocation concealment not reported
Basirat 200960	✓	✓	✓	✓	✓	?	Low	Study at low risk of bias
Bayreuther 199461	✓	✓	✓	✓	✓	✓	Low	Study at low risk of bias
Belluomini 199462	✓	?	✓	✗	✓	✓	Low	Most women lost to follow-up and attrition between groups was similar
Biswas 201163	✓	✓	✓	✓	✓	?	Low	Study at low risk of bias
Bondok 200664	✓	✓	✓	✓	✓	✓	Low	Study at low risk of bias
Can Gurkan 200843	?	?	?	?	✓	?	Unclear	Study at unclear risk of bias
Capp 201465	?	?	?	✓	✓	?	Unclear	Study at unclear risk of bias
Carlsson 200066	✓	✓	✗	?	✓	?	Low	Not possible to blind personnel to acupuncture and acupuncture placebo
Chittumma 200767	✓	✓	✓	✓	✓	✓	Low	Study at low risk of bias
Diggory 196268	?	?	✗	✗	?	?	High	Study at high risk of bias
Ditto 199969	✓	?	?	?	?	?	Unclear	Study at unclear risk of bias
Ensiyeh 200970	✓	?	?	✓	✓	✓	Low	Study at low risk of bias
Erez 197171	?	?	✓	✓	✓	✓	Low	Labelled as double-blind crossover RCT and predates 1996 CONSORT statement
Eftekhari 201372	?	✗	✓	✓	✓	?	Unclear	Study at unclear risk of bias
Evans 199373	?	✓	?	?	?	?	Unclear	Study at unclear risk of bias
Fischer-Rasmussen 199174	?	?	✓	✓	✓	✓	Low	Labelled as double-blind crossover RCT and predates 1996 CONSORT statement
Ghahiri 201175	?	?	✗	✓	✓	?	Unclear	Study at unclear risk of bias
Ghani 201376	✓	✓	?	✓	✓	?	Low	Study at low risk of bias overall, but did not report control results
Haji Seid Javadi 201377	?	?	✗	?	?	?	High	Vitamin B6 tablets and ginger capsules given so trial unblinded
Heazell 200678	?	✓	✗	✓	✓	?	Unclear	Study at unclear risk of bias
Hsu 200379	?	?	✓	?	?	?	Unclear	Study at unclear risk of bias
Jamigorn 200780	✓	✓	✗	✓	✓	?	Low	Not possible to blind personnel to acupressure and acupressure placebo
Kashifard 201381	✓	✓	✓	✓	✓	✓	Low	Study at low risk of bias
Keating 200282	✓	✓	✓	✓	?	?	Low	Study at low risk of bias
Knight 200183	✓	✓	✓	✓	✓	✓	Low	Study at low risk of bias
Koren 201084	?	✓	✓	✓	✓	?	Low	Study at low risk of bias
Maina 201485	✓	✓	✓	✓	✓	?	Low	Study at low risk of bias
Maltepe 201386	✓	?	✓	✓	✓	?	Low	Study at low risk of bias
Mao 200987	✓	✗	✗	✓	✗	?	High	Study at high risk of bias
McParlin 200888	✓	✓	✗	✓	✓	✓	Low	Not possible to blind patients or personnel to interventions in this trial
Mohammadbeigi 201189	✓	✓	✗	✓	✓	?	High	Personnel member gave medicines to patients and conducted randomisation
Monias 195790	?	?	✓	?	✗	?	High	Unclear whether or not this was truly a RCT and outcomes selectively reported
Naeimi Rad 201291	✓	?	✗	✓	✓	?	Low	Not possible to blind personnel, but patients were blinded to sham devices
Narenji 201292	?	✗	✗	✓	✓	?	High	Study at high risk of bias
Nelson-Piercy 200193	✓	✓	✓	✓	✓	?	Low	Study at low risk of bias
Neri 200594	✓	?	?	✓	✗	?	High	Selective reporting of outcomes and just 2/6 core risk of bias items satisfied
Oliveira 201395	?	?	?	?	?	?	Unclear	Study at unclear risk of bias
Ozgoli 200996	?	?	✗	✓	✓	✓	Unclear	Unclear details on whether or not trial was truly randomised, personnel unblinded
Pasha 201297	?	?	?	✓	✓	?	Unclear	Study at unclear risk of bias
Pongrojpaw 200742	?	?	?	✓	?	✓	Unclear	Study at unclear risk of bias
Rosen 200398	✓	✓	?	✓	✓	?	Low	Not possible to blind personnel, but patients were blinded to sham devices
Saberi 201313	✓	?	✗	✓	✓	✓	Low	Trial of acupressure, ginger and no treatment so blinding not possible
Safari 199899	✓	✓	✓	✓	✓	?	Low	Study at low risk of bias
Sahakian 1991100	✓	?	✓	?	?	?	Unclear	Study at unclear risk of bias
Smith 2002101	✓	✓	✗	✓	✓	✓	Low	Not possible to blind personnel to sham control, but possible to blind patients
Smith 2004102	✓	✓	✓	✓	✓	✓	Low	Study at low risk of bias
Sripramote 2003103	✓	✓	✓	✓	✓	✓	Low	Study at low risk of bias
Steele 2001104	?	?	?	✓	✓	?	Unclear	Study at unclear risk of bias
Sullivan 1996105	?	✓	✓	✓	?	?	Unclear	If no change in nausea or emesis was observed after 48 hours of medication and hydration, patient was considered a treatment failure, but unclear if included in final analyses
Tan 2010106	✓	✓	✓	✓	✓	✓	Low	Study at low risk of bias
Tan 2009107	✓	✓	?	✓	✓	?	Unclear	Placebo used Tic Tac® (Ferrero UK Ltd, Greenford, UK) which could potentially lead to unblinding
Tan 2013108	✓	✓	✓	✓	✓	?	Low	Study at low risk of bias
Veciana 2001109	?	?	?	?	?	?	Unclear	Study at unclear risk of bias
Vutyavanich 199541	✓	?	✓	✓	✓	✓	Low	Study at low risk of bias
Vutyavanich 2001110	✓	✓	✓	✓	✓	✓	Low	Study at low risk of bias
Werntoft 2001111	✗	?	?	✗	?	?	High	Study at high risk of bias
Wibowo 2012112	✓	✓	✓	✓	✓	✓	Low	Study at low risk of bias
Willetts 2003113	✓	?	✓	✓	✗	?	High	Selective reporting of outcomes and just 3/6 core risk of bias items satisfied
Yost 2003114	✓	?	?	✓	✗	?	High	Selective reporting of outcomes and just 2/6 core risk of bias items satisfied
Zhang 2005115	✓	✗	✗	✓	✗	?	High	Study at high risk of bias
Ziaei 2004116	?	?	✓	✓	✓	?	Unclear	Study at unclear risk of bias

Random sequence generation

The risk of bias arising from the method of generation of the allocation sequence was low in 39 of the included RCTs. 13,41,57,60,62–64,66,67,69,70,76,80–83,85–89,91,93,94,98–104,106–108,110,112–115 Methods employed included random number tables, computer-generated sequence generation61,63,64,81–83,88,93,94,98,99,106,108,112,114 and randomised block design. 62,67,80,87,89,91,101,102,107,113,115,117 One trial was classed as high risk because women were asked to draw an envelope from a box with the same appearance but with different contents. 111 It was categorised as unclear in the remaining 24 RCTs due to insufficient information provided by the authors to permit judgement either way. 42,43,58,59,65,68,71–75,77–79,84,90,92,95–97,104,105,109,116

Allocation concealment

Thirty studies employed allocation concealment methods judged to carry low risk of bias, such as the use of sequentially numbered sealed opaque envelopes containing allocation assignment. 57,60,61,63,64,66,67,73,76,78,80–85,88,89,93,98,99,102,103,105–108,112,118 Thirty studies did not provide sufficient information to allow a judgement of low or high risk and were therefore classed as unclear. 13,41–43,58,59,62,65,68–71,74,75,77,79,90,91,94–97,100,104,109,111,113,114,116,117 The remaining four RCTs were judged as having high risk of allocation concealment bias. 72,87,92,115 For example, one study stated that patients were randomly divided into two groups by those involved in the study,72 or the nature of the intervention being tested meant it was not possible to conceal allocation. 87,92,115

Blinding of participants, personnel and outcome assessors

Of the included RCT studies, 32 were judged to have low risk of bias in relation to the blinding of participants and other personnel involved in the trial,41,57,59–64,67,71,72,74,79,81–85,90,93,99,100,102,103,105,106,108,110,112,113,116,117 generally through the provision of medication in identical formats for both active and placebo. Sixteen studies were judged to have high risk of bias in this respect, for example due to clear differences in either the appearance, dosage rates or mode of delivery between intervention and placebo comparator, or as a result of evidence that the research staff involved were aware of allocation status. 13,58,66,68,75,77,78,80,87–89,91,92,96,101,115 In some instances, however, despite lack of blinding, the nature of the intervention meant that this was not relevant; for example, in McParlin and colleagues88 where blinding of participants and staff was not possible as the packages of care delivered to the intervention and control groups varied in content. However, it is important to highlight that although it might not have been possible to blind patients or clinicians, outcome assessors and analysts handling the resultant data may nevertheless have been blinded. The remaining 16 studies did not provide sufficient information to permit a judgement of low or high bias, often due to imprecise, poor reporting, and were thus classed as unclear. 42,43,65,69,70,73,76,94,95,97,98,104,107,109,111,114

Incomplete outcome data

Most studies (n = 50) were judged as carrying low risk of bias in relation to this component. 13,41,42,57,59–61,63–65,67,70–72,74–76,78,80–85,87–89,91–94,96–99,102–108,110,112–118 Although published protocols were rarely available, all data for the primary outcomes pre-specified in the paper were reported for all randomised participants, or rates of drop-out were either sufficiently low (< 20%), or proportionately comparable between groups, so that it was not considered likely to result in a clinically relevant bias. Three studies displayed a high risk of bias in this regard, all as a result of high numbers of participant drop-outs. 62,68,111 The remainder (11 studies in total) were judged as unclear due to lack of sufficient information. 43,58,66,69,73,77,79,90,95,100,109

Selective outcome reporting

Six studies were judged as having high risk of bias in terms of selective outcome reporting, due to either not reporting data for pre-specified outcomes, or for reporting data in the results that were not pre-specified in either the original study protocol or methods section. 87,90,94,113–115 Forty-five studies were classed as having low risk of bias, with all outcomes specified and subsequently reported. 13,41,43,57,59–67,70–72,74–76,78,80,81,83–85,88,89,91–93,96–99,101–104,106–108,110,112,116,117 Risk of bias was judged as unclear for the final 13 studies. 42,58,68,69,73,77,79,82,95,100,105,109,111

Other sources of bias

Twenty of the included RCT studies were judged as having low risk of bias in this area. 13,41,42,61,62,64,67,70,71,74,81,83,88,96,101–103,106,110,112 However, a substantial number (n = 44) were classed as unclear, due to lack of sufficient information in the paper to permit detailed assessment of whether or not an important risk of bias existed, or due to insufficient rationale or evidence that an identified problem had introduced serious levels of bias to the study. 43,57–60,65,66,68,69,72,73,75–80,82,84,85,87,89–95,97–100,104,105,107–109,111,113–117,119 For example, in one paper,76 lack of reporting of full results for the control group resulted in an unclear judgement in this area.

Interventions and comparators

The included studies were grouped into the three broad groups of interventions outlined in Chapter 1: patient-initiated first-line interventions; clinician-prescribed second-line interventions; and clinician-prescribed third-line interventions. It should be noted that, for patient-initiated first-line interventions, the only studies identified that could be classified as lifestyle interventions were those which trialled ginger preparations and/or vitamin B6. No studies of dietary- or hypnotherapy-based interventions were identified. However, studies of a number of novel therapies not covered by our original review protocol were identified, namely the use of aromatherapy, transdermal clonidine and gabapentin. The studies comprising the evidence base for each group of interventions are detailed in Table 6. Note that all studies are two-arm RCTs unless otherwise stated.

In addition, the network plot (Figure 4) shows the range of interventions from all comparative studies included in the review. Individual interventions have been grouped where appropriate.

FIGURE 4.

Network plot of range of interventions and comparisons for NVP/HG. Size of node is proportional to frequency of intervention and width of line to frequency of comparisons between two interventions. Plot does not include one pre-emptive trial,117 outpatient care trial127 and two four-arm trials. 68,101

The size of the nodes in the network plot is proportional to the frequency of the intervention in the review, and the width of the lines indicates the frequency of the comparisons made between two interventions. These nodes and lines, however, do not represent the weight of evidence in the review as this would also be influenced by sample size and the precision of estimates, as well as other factors. The plot did not include a trial on pre-emptive treatment of doxylamine/pyridoxine combination, outpatient versus inpatient care117,127 or two four-arm trials,68,101 which would have over-reported the number of comparisons in the network plot. These interventions included dietary instructions only, or together with either placebo, antihistamines or antihistamine/vitamin B6 combination in one trial68 and traditional acupuncture, P6 acupuncture, placebo or no acupuncture in another trial. 101 Ginger, vitamin B6, antihistamines, acupressure, metoclopramide, corticosteroids, doxylamine/pyridoxine combination and the serotonin antagonist ondansetron are more widely reported than other interventions, but there is also information on interventions such as acupuncture, nerve stimulation therapy and aromatherapy oils which have been considered as treatments for NVP/HG. Evidence on the effects of interventions such as Chinese herbal medicine, dextrose saline, transdermal clonidine and diazepam is very limited and in most cases is reported in single trials. As expected, placebo interventions are most widely reported as comparators, and so this has the biggest node on the network plot (emphasised by the square node). The most commonly reported treatment comparisons are ginger capsules versus placebo; acupressure versus placebo; ginger capsules versus vitamin B6 capsules; corticosteroids versus ‘treatment as usual’; metoclopramide versus ondansetron; and acupuncture versus nocebo (nocebo is an inert intervention that creates comparable side effects/harmful effects in a patient, as opposed a placebo, which is an inert substance that creates either a beneficial response or no response in a patient).

Participants and symptom severity

In addition to substantial variation in terms of the range of interventions and comparators evident within the literature, it is also important to highlight the heterogeneity of symptom severity found among patient populations.

It was initially intended that as part of this review only studies that recruited women with severe NVP or HG would be included. However, assessment of symptom severity varied within and across studies, and it was not possible to easily place every participant population into categories. We therefore attempted to categorise the symptom severity of participants for each study, using the description of severity in the inclusion criteria and, if available, any severity score given at baseline. These two items of information were assessed by two independent assessors (CMP and SCR) to assign severity as mild, moderate, severe or unclear. Agreement was reached for all but one study, which was classified as unclear.

This classification was then used in each results chapter to describe symptoms and outcomes in terms of severity.

Outcome measures

Finally, and linked to the issues discussed above, the identified literature in this field was also characterised by the range of symptom severity scales employed from study to study to assess intervention outcomes. Out of the 73 included studies (reported in 75 papers), only 23 used validated NVP/HG assessment scales such as PUQE (10 studies), RINVR (11 studies) or the McGill Nausea Questionnaire (one study). Thirty-one studies assessed nausea and/or vomiting severity using a 10-point VAS. Twenty-one studies employed either a study-specific, non-validated author-defined assessment scale (including, for example, numbers of episodes of vomiting combined with the use of a Likert scale to assess subjective feelings of symptom severity among participants), or used the various proxy measures of symptom severity outlined in our protocol [e.g. percentage weight loss, length of hospital stay, or hospital (re-)admission episodes]. Table 7 illustrates the primary symptom severity outcome measures employed by each included study.

Additional sources of outcome data on medications

The UKTIS is currently commissioned by Public Health England to provide advice to UK health professionals on the fetal effects of therapeutic, poisoning and chemical exposures in pregnancy, and to conduct surveillance of known and emerging teratogens. The UKTIS database currently contains a record of just under 60,000 enquiries dating back to 1978, of which 320 relate to use of specific drugs in the treatment of HG (period of enquiry 18 June 1978 to 18 March 2014). Surveillance data collected by the UKTIS are reviewed periodically and published in UKTIS monographs through the National Poisons Information Service database (www.TOXBASE.org). Data collected by the UKTIS in relation to medications for NVP/HG, including specific monograph data on ginger, vitamin B6, vitamin B12, promethazine and olanzapine, are provided in Table 43, Appendix 7 for information.

Meta-analysis of included randomised controlled trials

As highlighted in the previous sections, there was wide variation across studies. Specifically, there was considerable heterogeneity between interventions within each of the categories of comparisons, and in terms of how interventions were administered/delivered. The measurement of outcomes also differed substantially between trials reporting the same comparisons, so in most cases the trials were not directly comparable. In a meta-analysis it is important not to combine outcomes that are too diverse; even if it had been possible to extract data for a meta-analysis, such an analysis is likely to produced misleading results due to the considerable heterogeneity between studies. 46 Furthermore, many of these trials were extremely poorly reported and their conduct was often uncertain. In summary, clinical and methodological variations between studies were considerable, and the intervention effect was likely to be affected by the factors that varied across studies. Consequently, we have not conducted a meta-analysis of findings from the RCTs.

Structure of individual results chapters

The following chapters present more detailed findings from the evidence review for each individual intervention. As already indicated, given it was not possible to meta-analyse the data from individual studies for any group of interventions and comparators, the results are summarised in narrative form. The narrative content of each chapter focuses on the findings from the included studies in terms of their reported effectiveness for addressing our primary outcomes of interest, that is, the key symptoms associated with HG/NVP. Thus, where available, effectiveness is reported in terms of the validated overall HG/NVP assessment scales (PUQE, RINVR or McGill Nausea Questionnaire). Otherwise, the effectiveness of interventions is reported in relation to their impact on the three key symptoms: nausea, vomiting and retching. Data illustrating significant results in relation to these key symptoms are detailed in the narrative text; otherwise, results are described as not significance or not clear. Data for case series studies are not included in the narrative but available in the accompanying results tables for information. Additional secondary outcome data reported by included studies (see Table 2 for a full list) are presented in Appendix 8.

Introduction

Ginger was used as an intervention to treat HG, NVP or various forms of pregnancy sickness in a total of 16 RCTs. 13,42,60,63,67,70,74,77,82,89,92,96,102,103,110,113 Heterogeneity was observed in relation to the clinical setting and patient populations in which the studies were conducted, as well as the interventions, comparators and outcomes reported in each trial. As previously described (see Chapter 3, Meta-analysis of included randomised controlled trials) given the differences between trials in patient populations, settings, interventions and, in particular, the heterogeneous nature of the reported outcomes across trials, we did not attempt to perform meta-analyses and have thus reported a narrative summary only for each intervention and comparator set, with additional outcome data presented in Table 8.

Of the included trials, half (n = 8) described the women participants as experiencing symptoms categorised at the mild end of the NVP severity scale. 60,63,67,70,82,89,110,113 Six were classed as including participants with mild–moderate symptom severity. 13,74,92,96,102,103 The information provided by the remainder was insufficient to categorise objectively severity, although the authors describe the women as experiencing pregnancy-related nausea and vomiting. 42,77 Of the 16 studies, 4 were judged to carry a high risk of bias,77,89,92,113 10 low risk,13,60,63,67,70,74,82,102,103,110 and the remainder, unclear due to lack of information. 42,96

Ginger capsules versus placebo capsules

Five trials74,89,96,110,113 compared ginger and placebo capsules for the treatment of NVP. The trial by Mohammadbeigi and colleagues89 was a three-arm RCT which compared ginger, metoclopramide and placebo. 89 Only one trial74 explicitly reported the effects of ginger in the treatment of HG (which they define as vomiting occurring during pregnancy, presenting prior to the 20th week and of sufficient severity to require hospital admission). However, it is important to note that this deviates from the standard definition of HG (see Chapter 1, Background). Three of these trials were judged to carry a low risk of bias, two high risk due to issues regarding blinding of personnel administering the treatment89 or selective outcome reporting113 and the risk of bias in the remaining trial was unclear. 96

Ginger syrup versus placebo syrup

The trial of Keating and Chez82 reported on the effect of ginger syrup versus non-ginger-containing placebo syrup on nausea and vomiting in early pregnancy. This trial was judged as carrying a low risk of bias.

Ginger biscuit versus placebo biscuit

We identified one trial,60 judged at low risk of bias, which reported on the effect of ginger biscuits on nausea and vomiting in early pregnancy.

Ginger versus vitamin B6

Six trials compared ginger with vitamin B6 capsules for the treatment of NVP in pregnancy. 67,70,77,92,102,103 Most of these trials were judged as being at a low risk of bias. 67,70,102,103 However, risk of bias in the trial of Haji Seid Javari and colleagues77 was judged as unclear, and Nanreji and colleagues92 was judged as being at a high risk of bias due to concerns with regard to the blinding process and allocation concealment.

Ginger capsules versus acupressure

The trial of Saberi and colleagues13 reported a randomised comparison of the effectiveness of ginger capsules versus acupressure to relieve NVP. A control arm was also included where no intervention was performed during the 7 days of the trial. Analysis was performed on all participants. This study was judged as carrying low risk of bias.

Ginger versus doxylamine–pyridoxine

Biswas and colleagues63 reported a randomised, comparison of ginger extract tablets versus doxinate (doxylamine 10 mg plus pyridoxine 10 mg) in the treatment of nausea and vomiting during pregnancy. Patients were blinded to the treatment, whereas investigators were not. However, overall, this study was judged to have a low risk of bias.

Ginger versus antihistamine (dimenhydrinate) capsules

Pongrojpaw and colleagues42 examined the effect of ginger versus antihistamine (dimenhydrinate) capsules in the treatment of NVP. Risk of bias was unclear due to lack of information provided in this abstract.

Ginger versus metoclopramide

The 2011 trial of Mohammadbeigi and colleagues89 randomised women to ginger, metoclopramide or placebo capsules for the treatment of pregnancy sickness. Details are reported in Ginger capsules versus placebo capsules concerning ginger versus placebo. In summary, both were found to be were efficacious but there was no evidence that one treatment was superior to the other. Furthermore, as detailed above, the trial was judged as carrying a high risk of bias due to concerns over lack of blinding of trial personnel.

Summary

• The evidence available for ginger was predominantly at low risk of bias or the risk of bias was unclear, with four exceptions. 77,89,92,113

• Ginger is one of the most widely reported interventions but the evidence is limited and generalisability relates to women with less severe nausea and vomiting symptoms receiving ginger treatment.

• The quality of the evidence for ginger versus placebo is low and was downgraded due to clinical heterogeneity and imprecision. Further research is very likely to have an important impact on our confidence in the estimate of effect and may change the estimate. There was a sparseness of data in the other comparisons, so they were further downgraded to very low quality of evidence and we are very uncertain about the estimates.

• The identified studies comparing ginger preparations with placebo generally reported evidence of an improvement over a range of symptoms with the use of ginger.

• When consideration is restricted to those studies at low risk of bias, the results are less clear cut. Ginger looks promising in reducing symptoms, but the findings are not conclusive.

• For the comparison of ginger preparations with acupressure, ginger again looks promising, but the evidence is very limited in both quantity and quality.

• For the comparison of ginger versus vitamin B6 there are some higher-quality studies, but in general there was little evidence of a difference in the severity of symptoms between groups. There were few data for the comparisons of ginger against doxylamine–pyridoxine or antihistamine or metoclopramide, and minimal evidence suggesting any difference between groups.

• Overall, there is a suggestion that ginger might be better than placebo in reducing the severity of symptoms, but more larger, better-quality studies are required for this and all other comparators.

Introduction

Acupuncture or acupressure was used as an intervention for NVP in a total of 18 RCTs and one case series study. The majority (n = 13) of studies compared acupressure, acupuncture or nerve stimulation against placebo or sham treatment. However, heterogeneity was observed in relation to the clinical setting and patient populations in which the studies were conducted, as well as the interventions, comparators and outcomes reported in each trial. As previously described (see Chapter 3, Meta-analysis of included randomised controlled trials), given the differences between trials in patient populations, settings, interventions and, in particular, the heterogeneous nature of the reported outcomes across trials, we did not attempt to perform meta-analyses. We have thus reported a narrative summary only for each intervention and comparator set, with additional outcome data presented in Table 9.

Of the identified trials, eight were judged as being low risk of bias,61,62,66,80,83,91,98,101 with a further six unclear due to lack of sufficient information on the research process. 43,73,78,79,104,109 Four were judged as carrying high risk of bias on the basis of incomplete outcome data and weak randomisation process,115 lack of blinding,87 or selective outcome reporting. 94,111,115 We also identified a case series which evaluated the effect of P6 acupressure by Markose and colleagues,124 which was categorised as weak based on the EPHPP tool.

The severity of symptoms for patient participants in the studies comprising this group of interventions varied substantially. Three studies were classed as focussing on women with mild NVP,61,62,104 seven with mild to moderate,43,80,83,91,98,101,124 two with moderate severity,78,115 four with moderate to severe,66,87,94,111 and a final three for which severity was not possible to classify. 73,79,109

Acupressure versus placebo

Eight trials compared acupressure with placebo or sham treatment. 43,61,62,78,79,91,104,111 Seven trials43,61,62,78,79,104,111 involved the comparison of acupressure administered to the P6 acupoint, which lies between the tendons of the palmaris longus and flexor carpi radialis muscles, four centimetres proximal to the wrist crease. 129 Six of these trials examined the application of acupressure to point P6 by the use of pressure bands worn by study participants;43,61,78,79,104,111 one comprised self-administered acupressure at the P6 acupoint. 62 A final acupressure versus placebo trial compared acupressure self-administered with the two KID21 acupoints, which are located above the navel. 91 The placebo comparison intervention came in a two formats. The vast majority of studies (7/8) compared the active acupressure intervention with the application of acupressure to a sham acupoint (i.e. at which there is no known acupressure point or meridian pathway). 43,61,62,78,79,91,111 One final study compared the use of pressure bands with the wearing of placebo pressure bands which did not have a pressure button. 104 Three of these trials were judged as having low risk of bias61,62,91 and four were judged to be unclear. 43,78,79,104 Werntoff and Dykes111 was judged as having high risk of bias on the basis of incomplete outcome data and weak randomisation process.

Acupressure versus vitamin B6

One trial (Jamigorn and Phupong80), judged as having a low risk of bias, compared the use of acupressure administered via a pressure band at acupoint P6 against vitamin B6 taken in tablet form.

Acupressure (case series)

Markose and colleagues124 was a case series study that evaluated the effect of P6 acupressure on women with mild to moderate nausea following an initial control phase of 3 days without treatment (categorised as weak in terms of quality).

Nerve stimulation versus placebo

Three studies compared the effect of a nerve stimulation device administered at pressure point P6 (as described in Acupressure versus placebo) against an identical inactive unit. 73,98,109 Of these trials, two provided insufficient data to permit a clear judgement of bias,73,109 and the other trial was judged as having a low risk of bias. 98 Given the differences between trials in patient populations, settings, interventions and in particular the heterogeneous nature of the reported outcomes across trials, we did not attempt to perform meta-analyses and have outlined individual trial results in a narrative.

Acupuncture versus placebo

The trials of Carlsson and colleagues66 and Knight and colleagues83 both compared acupuncture at various treatment points with placebo (sham) treatment. Smith and colleagues101 was a four-arm RCT comparing a combination of various treatment points (group 1); with comparators including single point acupuncture to P6 (group 2); sham acupuncture to ineffective points (group 3); and a control comprising a standard information sheet and telephone support (group 4). Smith and colleagues101 and Knight and colleagues83 were both judged as carrying a low risk of bias; however, Carlsson and colleagues66 was judged as high due to lack of blinding.

Acupuncture versus metoclopramide

The trial of Neri and colleagues94 compared acupuncture to acupoints PC6, Central Venter 12 and Stomach 36 plus the wearing of an acupressure device at the PC6 point against the administration of a metoclopramide infusion at hospital plus oral supplementation with a vitamin B12 complex. The trial was judged as carrying a high risk of bias due to selective outcome reporting.

Acupuncture versus Chinese herbal medicine versus barbiturates

Two studies compared acupuncture against both Chinese herbal medicine and Western medicine. In the Mao and Liang trial,87 all women were given i.v. fluids for rehydration purposes and to correct electrolyte imbalance. The following interventions were then compared: (1) traditional Chinese acupuncture at acupoints BL11, ST37, PC6 and SP4; (2) a traditional herbal remedy; and (3) phenobarbital. In the Zhang 2005 trial,115 traditional Chinese acupuncture was compared with a traditional herbal drink and i.v. fluids plus phenobarbitol. Both trials were found to have a high risk of bias due to inadequate blinding and/or selective outcome reporting.

Summary

• The quality of evidence available for this group of interventions varied considerably, from low (eight studies61,62,66,80,83,91,98,101) to high (five studies87,94,111,115,124) risk of bias.

• Comparisons with placebo were equivocal: two studies (both where participants had mild symptoms and at low risk of bias)61,62 found no evidence of a difference or did not report the outcome for nausea and vomiting symptoms respectively.

• Acupressure at the KID21 acupoint reduced both nausea and vomiting compared with placebo.

• No evidence of a difference between acupressure and vitamin B6. 80

• Two studies73,98 of nerve stimulation at the P6 acupressure point reported better outcomes versus a non-stimulating device [one study = participants with mild–moderate severity with low risk of bias,98 one study = unclear severity and risk of bias73]. A third study found no evidence of a difference (unclear severity and risk of bias). 109

• Three studies66,83,101 compared acupuncture with placebo. There was no evidence of an effect on vomiting but one study (mild–moderate symptoms, low risk of bias)101 reported reduced nausea and retching in both traditional and P6 acupuncture groups.

• No evidence of a difference between acupuncture compared with metoclopramide, but symptoms improved from baseline in both groups (one study: high risk of bias, severity assessed as moderate–severe94).

• Comparisons of traditional Chinese acupuncture and herbal medicine with Western medicine were at high risk of bias and impossible to emulate within the UK NHS setting.

• Overall conclusions are limited by the quality of included studies. Many participants appear to have improved symptoms from baseline in all treatment arms.

• There is some suggestion that acupressure may have some beneficial effect in women with mild symptoms of NVP.

• Acupressure at the KID21 point appears to improve symptoms of nausea and vomiting compared with acupressure to a false point, but data are limited.

• The evidence for nerve stimulation is mixed.

• Acupuncture may reduce symptoms of nausea and retching in women with mild–moderate symptoms, but data are limited and inconclusive.

• More larger, better-quality studies are required for these interventions.

Introduction

Aromatherapy was used as an intervention to treat nausea and/or vomiting in two pilot RCTs. 76,97 Heterogeneity was observed in relation to the clinical setting and patient populations in which the studies were conducted, as well as the interventions, comparators and outcomes reported in each trial. As previously described (see Chapter 3, Meta-analysis of included randomised controlled trials) given these, we did not attempt to perform meta-analyses, and have thus reported a narrative summary only for each intervention and comparator set.

One trial compared aromatherapy with placebo in pre-natal wards97 whereas the other trial compared aromatherapy with routine antenatal care. 76 Both trials were at unclear risk of bias: mainly due to poor reporting in the trial of Pasha and colleagues;97 and for uncertainty surrounding whether the trial was truly randomised or was a quasi-RCT in Ghani and Ibrahim. 76 The women were described as experiencing symptoms at the mild to moderate end of the severity spectrum at baseline in both trials (Table 10).

Aromatherapy versus no aromatherapy

One trial97 compared mint oil aromatherapy against placebo for the treatment of mild NVP. 97 Another trial compared mixed essential oils (lavender and peppermint) and routine antenatal care, where no treatment oils were given, in women with mild to moderate nausea and/or vomiting (i.e. requiring antiemetics but not hospitalisation).

Summary

• The evidence from the two trials76,97 available for aromatherapy was predominantly at an unclear risk of bias.

• The identified studies reported evidence of an improvement in symptoms over time, but there was no evidence of a difference compared with placebo or routine antenatal care.

• Overall, few data are available with no evidence of an effect.

• More larger, better-quality studies are required for to make any conclusions about this intervention.

Introduction

Five trials compared the effectiveness of pyridoxine (vitamin B6) for the treatment of NVP. 41,59,100,107,112 Two trials compared doses of vitamin B6 against placebo tablets,41,100 and they were at low41 and unclear100 risk of bias, respectively. Tan and colleagues107 examined the effect of vitamin B6 and metoclopramide combination versus metoclopramide alone and was at an unclear risk of bias, mainly due to the use of a dubious placebo (Tic Tac^®, Ferrero UK Ltd, Greenford, UK). The trial of Wibowo and colleagues112 compared high and low doses of vitamin B6, and was at low risk of bias. Babaei and Foghaha59 compared the effectiveness of vitamin B6 against dimenhydrinate in the treatment of NVP and was at an unclear risk of bias.

In all five studies,41,59,100,107,112 women were described as experiencing mild to moderate symptoms at baseline (Table 11). However, as previously described (see Chapter 3, Meta-analysis of included randomised controlled trials), given the differences between trials in patient populations, settings, interventions and, in particular, the heterogeneous nature of the reported outcomes across trials, we did not attempt to perform meta-analyses. Thus we reported a narrative summary only for each intervention and comparator set.

Vitamin B6 versus placebo

Vitamin B6 and metoclopramide combination versus metoclopramide alone

The trial of Tan and colleagues107 used the 10-point VAS to assess median nausea scores in women with severe nausea and vomiting during pregnancy with clinical features. The trial also reported the mean number of daily vomiting episodes. 107

Vitamin B6 versus serotonin antagonist

The trial of Babaei and Foghaha59 compared effectiveness of vitamin B6 against dimenhydrinate in the treatment of NVP in a double-blind RCT which was adjudged as being at unclear risk of bias.

Summary

• The evidence available for vitamin B6 was predominantly at low risk of bias or the risk of bias was unclear.

• Participants in the five studies41,59,100,107,112 had symptoms categorised as mild to moderate at baseline.

• Comparisons of vitamin B6 preparations with placebo generally reported evidence of a reduction in symptoms of nausea, especially in women with more severe symptoms, and vomiting.

• Higher doses of vitamin B6 preparations resulted in a greater improvement in NVP symptoms.

• There was no evidence to suggest that vitamin B6 and metoclopramide as a combination treatment had an advantage over metoclopramide alone.

• Overall, there is a suggestion that vitamin B6 might be better than placebo in reducing the severity of symptoms especially at higher doses, but more studies are required using a range of comparators.

Introduction

Four trials65,84,95,117 assessed the effectiveness of vitamin B6 in combination with doxylamine for the treatment of NVP in comparison with a placebo (Koren and colleagues84), ondansetron (Capp and colleagues65 and Oliveira and colleagues95) or when administered pre-emptively versus following symptom onset (Maltepe and Koren117). In the Koren and colleagues84 and Maltepe and Koren117 studies this took the form of Diclectin (delayed-release doxylamine/pyridoxine combination), while Oliveira and colleagues95 and Capp and colleagues65 gave pyridoxine and doxylamine as separate preparations.

One of the trials95 was only available in abstract form and risk of bias was unclear, whereas the trial of Koren and colleagues84 was at a low risk of bias. Another trial117 examined pre-emptive treatment with Diclectin and this trial appeared to be at low risk of bias but some items were unclear. Capp and colleagues65 was at unclear risk of bias due to lack of provided details across a number of measures. 65 The reporting of severity of symptoms was not possible for the Maltepe and Koren117 pre-emptive study, poor for the Oliveira and colleagues95 study and unclear for the Capp and colleagues65 study, but in the Koren and colleagues84 study, severity appeared to range from mild to moderate. As previously described (see Chapter 3, Meta-analysis of included randomised controlled trials), given the differences between trials in patient populations, settings, interventions and, in particular, the heterogeneous nature of the reported outcomes across trials, we did not attempt to perform meta-analyses, and have thus reported a narrative summary only for each intervention and comparator set.

We also identified a non-randomised prospective observational study (Ashkenazi-Hoffnung and colleagues36) which compared a combination regimen of doxylamine and pyridoxine versus metoclopramide only (control group). This study is prone to extreme selection bias and there was a noticeable difference at baseline: moderate to severe symptoms were present in 97% of the intervention group women versus 69% of control group women (p < 0.01). A summary of study conduct and results is depicted in Table 12 for completeness.

Doxylamine/pyridoxine versus placebo

One double-blind, multicentre placebo-controlled trial84 compared treatment with Diclectin, a combination preparation of doxylamine succinate (10 mg) and pyridoxine hydrochloride (10 mg), to placebo in women with symptoms of NVP and a PUQE score ≥ 7. Women in the trial had not previously responded to conservative management.

Doxylamine/pyridoxine versus ondansetron

The double-blind RCT of Oliveira and colleagues95 was in abstract form only and randomised women in the first trimester of pregnancy requesting treatment for NVP to either one tablet of ondansetron (4 mg) plus a second (placebo) tablet or one tablet of pyridoxine (25 mg) plus one tablet of doxylamine (12.5 mg) administered every 8 hours for 5 days. 95 All study medications were identical in appearance. Capp and colleagues65 also compared 4-mg ondansetron plus a placebo tablet administered every 8 hours for 5 days against 25-mg pyridoxine plus 12.5-mg doxylamine. 65

Pre-emptive doxylamine/pyridoxine

One RCT117 compared pre-emptive treatment with Diclectin before the onset of symptoms of NVP, versus when the symptoms first began in patients at high risk for recurrence of severe NVP.

Summary

• The evidence available for pyridoxine/doxylamine combinations is varied, but two trials appeared to be at low risk of bias84,117 with the other two having an unclear risk of bias profile. 65,95

• The quality of the evidence is low and was downgraded due to clinical heterogeneity and sparseness of data in most comparisons. Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimates.

• Diclectin appears to be more effective at relieving symptoms of NVP than placebo.

• Ondansetron appears more effective at reducing nausea than pyridoxine plus doxylamine but there was no difference for vomiting.

• Limited data from a single small study36 showed no difference in efficacy between pyridoxine plus vitamin B6 versus metoclopramide but was subject to selection bias.

• Pre-emptive treatment with Diclectin appears to result in a reduced risk of moderate/severe NVP compared with treatment initiation once symptoms begin.

• Further larger, well-conducted trials are required to test the effectiveness of Diclectin or pyridoxine/doxylamine in combination compared with other treatment options.

Introduction

Antihistamines were used as an intervention to treat nausea and/or vomiting in three RCTs. 68,71,90 Heterogeneity was observed in relation to the clinical setting, and the patient populations in which the studies were conducted, but most notably there were differences in interventions, comparators and outcomes reported in each trial. As previously described (see Chapter 3, Meta-analysis of included randomised controlled trials), given these differences, we did not attempt to perform meta-analyses and have thus reported a narrative summary only for each intervention and comparator set.

The three trials had varying risk of bias profiles, with just one of the trials appearing to be at low risk of bias,71 and the remaining two at high risk of bias due to concerns with regard to blinding and incomplete outcome reporting,68 and selective outcome reporting. 90 The women were described as experiencing mild symptoms at baseline in all three trials (Table 13).

Antihistamines versus placebo

Two trials71,90 compared antihistamines and placebo: one trial71 in women with mild recurrent nausea and who had vomited at least three times per week over the previous 2 weeks; and the other trial90 in women complaining of nausea and/or vomiting. Hydroxyzine hydrochloride (25 mg) capsules were given twice daily for 3 weeks in the trial of Erez and colleagues,71 and cyclizine pyridoxine (dose not reported) tablets twice daily for 2 weeks in the Monias90 trial.

Antihistamines alone or in combination with Vitamin B6 versus control

One four-arm trial68 compared antihistamines (meclizine hydrochloride) with and without vitamin B6 (pyridoxine) with a placebo group and a no treatment group in women experiencing mild nausea or vomiting. In the intervention arms, meclizine hydrochloride was given at doses of 25 mg and 50 mg in the morning and evening, respectively. In the combination group, pyridoxine was given at 50 mg in the morning and 100 mg in the evening.

Summary

• Of the three studies available for antihistamines, two were at high risk of bias68,90 while one was at low risk. 71 The quality of the evidence was very low and was downgraded due to the reporting of author-defined scales to measure symptom severity, and other concerns about the overall risk of bias in included RCTs. We are very uncertain about the effectiveness estimates in terms of validation and estimates may change in future trials.

• Participants in all three studies68,71,90 had mild symptoms so generalisability is restricted.

• Use of antihistamines resulted in an improvement over a range of symptoms.

• The addition of vitamin B6 does not appear to improve effectiveness.

• Evidence is limited in both quantity and quality.

• Antihistamines appear to be better than placebo in reducing the severity of symptoms, but better-quality, large studies are required for this and all other comparators.

Introduction

Dopamine antagonists were used as an intervention to treat HG in the trial of Tan and colleagues. 106 The trial was a head-to-head comparison of two interventions (promethazine vs. metoclopramide) and was at low risk of bias. We will focus on the results from this trial here. We also identified a non-randomised prospective study122 which examined two study groups (67 patients treated with i.v. droperidol 1 mg/hour plus diphenhydramine 25–50 mg every 6 hours; and 34 patients treated with i.v. droperidol 0.5 mg/hour plus diphenhydramine 50 mg every 6 hours; and a historical control of 54 patients receiving conventional antiemetic treatment). The study groups were then gradually weaned onto oral hydroxyzine and metoclopramide. This study is prone to extreme selection bias and other biases, and classified as ‘weak’ according to the EPHPP quality assessment tool. A summary of study conduct and results is depicted in Table 14, with safety data reported in Appendix 8.

Promethazine versus metoclopramide

The trial of Tan and colleagues106 randomised women in early pregnancy (gestation of ≤ 16 weeks) with clinical HG to either 25-mg promethazine (n = 76) or 10-mg metoclopramide (n = 73), administered intravenously every 8 hours for 24 hours.

Summary

• Limited data suggest that promethazine is as effective as metoclopramide in reducing the symptoms of NVP.

• The quality of evidence was very low and was downgraded due to sparseness of data and imprecision, although the one trial106 that reported a comparison on promethazine and metoclopramide was at low risk of bias.

• We are very uncertain about the estimates of effectiveness. Further, well-conducted studies are required to compare the effectiveness of metoclopramide against other comparators.

Introduction

A total of five trials57,72,75,81,105 and one cohort study121 compared the serotonin antagonist (ondansetron) against a range of alternatives for the treatment of women experiencing various severities of nausea in pregnancy. Of these, one study focussed on the safety of ondansetron versus the usual treatment regimen,121 with symptom severity not specified. Three trials tested ondansetron against metoclopramide, with participants’ symptoms classified as mild to moderate in two trials,75,81 and severe in one trial. 57 The remaining two trials compared ondansetron with antihistamines,72,105 with women classed as experiencing symptoms at the moderate to severe end of the spectrum.

Two of the identified trials were found to have a low risk of bias. 57,81 The trial of Sullivan and colleagues105 was classed as unclear due to insufficient information, with the remaining two found to carry a high risk of bias. 72,75 The included cohort study of Einarson and colleagues121 was classed as ‘weak’ according to the EPHPP quality assessment tool. As previously described (see Chapter 3, Meta-analysis of included randomised controlled trials), given the differences between trials in patient populations, settings, interventions and, in particular, the heterogeneous nature of the reported outcomes across trials, we did not attempt to perform meta-analyses and have thus reported a narrative summary only for each intervention and comparator set. A summary of study conduct and results is depicted in Table 15, with safety data reported in Appendix 8.

Ondansetron versus usual treatment

Einarson and colleagues121 examined ondansetron versus the usual treatment regimen. 121 As this involved a telephone survey of women already taking ondansetron, however, no direct measures of symptom relief were assessed. The outcomes of interest related to the incidence of fetal abnormalities in the surveyed patient population. These data are provided in Appendix 8, with additional UKTIS data on the intervention detailed in Appendix 7.

Ondansetron versus metoclopramide

Ondansetron versus antihistamines

Summary

• The evidence available for serotonin receptor antagonists, specifically ondansetron, was predominantly at high or unclear risk of bias with only one at low risk. 81

• Evidence for the comparison of ondansetron with metoclopramide was mixed, with both drugs improving symptoms.

• Low risk of bias studies found ondansetron more effective at reducing symptoms of vomiting compared with metoclopramide after 4 days.

• Both ondansetron and antihistamine improved symptoms with no evidence of a difference between them.

• Overall, ondansetron does reduce the severity of symptoms, but more larger, better-quality studies are required to assess benefit over other comparators.

Introduction

Two studies compared the effectiveness of i.v. fluids for the treatment of women with HG. 69,108 One trial by Tan and colleagues108 tested different compositions of i.v. solution (dextrose plus saline vs. saline only)108 and was judged as low risk of bias. The other study compared the use of diazepam against i.v. fluids containing vitamins69 but was judged as carrying an unclear risk of bias due to lack of sufficient information in a number of areas. Both papers described the trial participants as suffering from HG; however, we classified severity as either moderate,69 or moderate to severe108 based on the participant data provided. As previously described (see Chapter 3, Meta-analysis of included randomised controlled trials), given the differences between trials in patient populations, settings, interventions and, in particular, the heterogeneous nature of the reported outcomes across trials, we did not attempt to perform meta-analyses and have thus reported a narrative summary only for each intervention and comparator set. A summary of study conduct and results is depicted in Table 16, with safety data reported in Appendix 8.

Dextrose saline versus saline only

Intravenous fluids versus intravenous fluids plus diazepam

Summary

• The evidence available for i.v. fluids was at low108 or unclear69 risk of bias.

• i.v. fluid improves reported symptoms. Dextrose saline may be more effective at improving nausea over time for those with moderate nausea.

• Diazepam appears to be more effective than i.v. fluids alone at reducing nausea on day 2 but there was no evidence post treatment for those with moderate/severe nausea.

• Overall, i.v. fluids help correct dehydration and improve symptoms, dextrose saline may be more effective at reducing nausea than normal saline. (The lower concentration of sodium in dextrose saline may exacerbate any pre-existing hyponatraemia. High doses/concentrations of dextrose solutions may increase the risk of Wernicke’s encephalopathy, but concentrations in dextrose saline are unlikely to provoke this response.)

• Future studies are required which focus on interventions given alongside rehydration therapy.

Introduction

One trial by Maina and colleagues85 studied the efficacy of a 5-mg transdermal clonidine patch in the treatment of HG, in a double-blind, placebo-controlled, crossover RCT. Patients were classified at the severe end of the HG spectrum, with the trial judged as having a low risk of bias. A summary of study conduct and results is depicted in Table 17, with safety data reported in Appendix 8.

Transdermal clonidine versus placebo patch

Summary

• Evidence from one study85 suggests that the use of transdermal clonidine patches may be effective for the treatment of severe HG, but more and larger studies are required to compare effectiveness against comparators.

Introduction

One study compared the effectiveness of either midwife-led outpatient management versus standard inpatient care,88 while another examined the feasibility of day case management,120 for the treatment of moderate to severe HG/NVP. The trial of McParlin and colleagues88 was judged as having a low risk of bias, while a 2007 case series study by Alalade and colleagues120 was classed as of weak quality. A summary of study conduct and results is depicted in Table 18, with safety data reported in Appendix 8.

Outpatient management versus standard inpatient care

Day case management

The feasibility and efficacy of day case management of HG was assessed by Alalade and colleagues120 using rates of patient discharges within a 24-hour period. These case series data are reported in Appendix 8.

Summary

• The evidence available for day case management was at low risk88 and high risk120 of bias.

• The identified studies indicate that day case management of women with moderate to severe symptoms is feasible and acceptable to women.

• The results indicate that day case management is as effective at improving severity scores as inpatient management for some women.

• More, larger studies are required to provide definitive results and women’s views.

Introduction

Corticosteroids were tested against a variety of alternative treatments in seven individual studies. 58,64,93,99,114,116,125 Two trials compared the effectiveness of corticosteroids against placebo tablets in patient populations categorised as experiencing moderate to severe symptoms. 93,114 Three trials compared corticosteroids with either promethazine99,116 or phenergan suppositories58 in patients classed as experiencing moderate, unclear or moderate to severe level symptoms respectively. One trial tested the effectiveness of steroids against metoclopramide for women with moderate to severe HG. 116 As previously described (see Chapter 3, Meta-analysis of included randomised controlled trials), given the differences between trials in patient populations, settings, interventions and, in particular, the heterogeneous nature of the reported outcomes across trials, we did not attempt to perform meta-analyses and have thus reported a narrative summary only for each intervention and comparator set.

Of these trials, three were classed as carrying a low risk of bias64,93,99 and two were unclear due to lack of information. 58,116 One study was judged to have a high risk of bias due to lack of blinding and unclear outcome reporting. 114 A final case series study examined the effect of corticosteroids in comparison with the usual treatment regimen in women with severe HG. 125 A summary of study conduct and results is depicted in Table 19, with safety data reported in Appendix 8.

Corticosteroids versus placebo

Corticosteroids versus promethazine (Phenergan)

Corticosteroids versus metoclopramide

Corticosteroids against usual treatment

Summary

• Evidence available for corticosteroids was predominantly at low risk of bias (three studies64,93,99), or the risk of bias was unclear/high (three studies58,114,116).

• Steroids versus placebo had a trend towards improved symptoms, but results were not statistically significant.

• For steroids versus promethazine, there was no evidence of a difference in improvement by 1 week.

• Steroids were more effective at reducing vomiting episodes than Phenergan suppositories or metoclopramide.

• Overall, treatment with corticosteroids reduces the severity of symptoms, but more and larger studies are required to compare effectiveness against comparators.

Introduction

Two case series of nasogastric123 and jejunostomy126 feeding for women experiencing HG at the severe end of the spectrum were found. Both studies were judged as weak in quality. A summary of study conduct and results is depicted in Table 20, with safety data reported in Appendix 8.

Nasogastric enteral feeding

Hsu and colleagues123 reported their experience of treating HG with nasogastric enteral feeding in a case study with seven women conducted at the Genesee Hospital, Rochester, New York. Enteral feeding was administered via an 8-Fr Dobbhoff nasogastric tube (Ross Products Division, Abbott Laboratories, Columbus, OH), delivering Jevity or Osmolite incrementally to meet daily caloric requirements (initial rate of 25 ml/hour, increasing as tolerated by 25 ml an hour per day). Patients were discharged once stabilised. Six women continued enteral feedings at home with nasogastric feedings discontinuing when nutritional needs were being met orally. Key outcomes of interest were the period of time (in days) after which patients were discharged and the mean duration of feedings (see Appendix 8).

Jejunostomy feeding

A further case series study by Saha and colleagues126 looked at the feasibility and efficacy of feeding via jejunostomy in five women with HG compared with standard therapy at the Women and Infants Hospital, Providence, Rhode Island. Participants all displayed persistent severe nausea and vomiting plus one of the following: weight loss of > 5% of pre-pregnancy weight; ketonuria; multiple emergency room visits for dehydration; and/or inability to tolerate oral intake, despite i.v. hydration, i.v. ondansetron, i.v. ranitidine or pantoprazole, and i.v. metoclopramide. The J-tubes were placed between 12 and 26 weeks’ gestation (median 14 weeks) for a mean duration of 19 weeks (range 8–28 weeks). Four J-tubes remained in place until delivery. One tube was removed at 34 weeks at the patient’s request because of emotional distress and one tube fell out at 30 weeks’ gestation and was not replaced. They reported that all patients had continued nausea and vomiting over the study period which required continued standard therapy in addition to J-tube feedings. Outcomes measured by the study included use of additional medication and adverse events (see Appendix 8).

Summary

• Enteral feeding is an effective but extreme method of supporting women suffering from very severe symptoms as a last resort.

Introduction

No controlled trials of gabapentin use in the treatment of HG were identified. The only available data were derived from one uncontrolled pilot study by Guttuso and colleagues39 which enrolled seven consecutive women of 20 weeks’ gestation with HG to examine gabapentin therapy. This study was classed as ‘weak’ by the EPHPP quality assessment tool.

Summary

• We identified only one very small pilot trial which examined gabapentin therapy in women with HG.

• More research is needed, including monitoring of safety, in light of the reported cases of congenital anomalies among the seven exposed infants.

Introduction

This chapter has two aims: (i) to review systematically the published evidence relating to the cost-effectiveness of interventions used in the treatment of NVP/HG; and (ii) to assess the relative cost-effectiveness of such interventions from a NHS and Personal Social Services perspective. To assess cost-effectiveness, an economic model was developed in the first instance. However, due to the lack of evidence required to populate the economic model, an alternative method of economic evaluation was used. The methods and findings of the systematic review are presented first, followed by those of the economic evaluation.

Systematic review of economic evaluations

Results

From the main systematic review, 11 papers (no duplicates) were identified. None of the records were deemed relevant to this economic review and, as such, no hard copies were obtained for scrutiny against the inclusion criteria for the review. A flow diagram presenting the process of selecting studies can be found in Figure 5.

FIGURE 5.

Flow diagram showing study selection for economic evaluations review.

Economic modelling

Methods

Outline of model

We initially proposed to develop an economic model to estimate costs, long-term effects and relative cost-effectiveness of the alternative interventions for NVP and HG from the perspective of the UK NHS and Personal Social Services. The model was to describe the pathways of individuals who have different severities of symptoms and have treatment initiated in different sectors of the health service (primary care, hospital outpatients and hospital inpatients). It was to cover the period of initial intervention and the costs and consequences of any subsequent outcomes including further interventions. Some of the effects of NVP and HG are short term; however, there may be some persisting impact on the mother and longer-term effects on the child. It was the intention to model the cumulative costs and quality-adjusted life-years for the mother and longer-term effects on the child (reported either in natural units or, if data allowed, quality-adjusted life-years). An outline structure of the core pathways for this proposed model is presented in Figure 6. The full model is available from the authors.

FIGURE 6.

Structure of the decision model.

The economic model was built in TreeAge Pro^® 2014 (TreeAge Software, Inc., Williamstown, MA, USA) to estimate the relative cost-effectiveness of the pharmacological and non-pharmacological interventions used to treat NVP/HG. The populations considered were women experiencing mild, moderate and severe nausea, vomiting and/or retching in pregnancy.

The decision tree structure can be described as follows. Depending on the severity of symptoms, patients began treatment on first-, second- or third-line treatment. The categorisation of interventions is described in detail in Chapter 2 but can be distinguished loosely as patient-initiated treatments (first line), clinician-prescribed antiemetics (second line) and clinician-prescribed corticosteroids (third line). The categorisation of these treatments as first line, second line and third line is not medically recognised and there may be discrepancies in how clinicians describe particular interventions. However, such a hierarchy allows us to compare the relative cost-effectiveness of categories of treatments rather than the multiple individual treatments that exist, while allowing us to think through the sequence of treatments that are likely to be relevant to patients at the different levels of severity of symptoms.

Beginning at ‘first-line treatment’, symptoms may worsen or not worsen. If symptoms worsened, patients moved to a ‘new first-line treatment’, ‘second-line treatment’ or ‘third-line treatment’. From each of these points, symptoms may again worsen or not worsen. If symptoms worsened, patients moved to the next hierarchical treatment strategy. This sequence was followed until patients reached ‘third-line treatment’, at which point symptoms may worsen or not worsen. If symptoms worsened, patients moved to a ‘new third-line treatment’ or ‘came off treatment’, at which point symptoms could either ‘stay the same’ or ‘resolve’ (terminal nodes). Beginning at the same starting point (‘first-line treatment’), if symptoms did not worsen, symptoms could ‘stay the same’ or ‘resolve’ (terminal node). If ‘symptoms stayed the same’, the pathways described at the same point as symptoms worsening were replicated.

For those patients beginning at ‘second-line’ and third-line’ treatment, the subsequent care pathways were broadly similar to those previously described. However, the model assumes that those patients starting at second- and third-line treatments are unable to progress to less severe treatment strategies. Therefore, the care pathways are progressively shorter as we move through the hierarchy of treatments. The decision model allows one to assess costs and effects across the three alternative treatment strategies and to assess the relative cost-effectiveness of interventions used to treat NVP/HG.

Economic evaluation