The delivery of chemotherapy at home: an evidence synthesis

Population

Cancer patients receiving intravenous chemotherapy.

Interventions and comparators

Studies comparing intravenous chemotherapy in two (or more) of the following settings:

• home setting (includes nursing homes)

• community-based setting (e.g. GP practice, community clinic, community hospital or mobile units)

• hospital outpatient setting.

Within-setting comparisons were eligible if the study compared different organisational or management approaches.

Outcomes

Any of the following:

• safety

• patient quality of life

• preference

• satisfaction (including treatment compliance/adherence)

• social functioning

• clinical outcomes

• patient and carer opinions and experiences

• costs

• resource/organisational issues (including access).

The clinical outcomes of interest were self-rated health or measures of performance status.

Study designs

Any type of comparative design was eligible. To obtain information about patient quality of life, satisfaction, preferences and opinions, studies that reported results for only one eligible setting and qualitative research (any of the three settings) were considered, providing that they had a stated aim to evaluate one or more of these outcomes. Given the review focus on home and community settings, and the potential diversity and likely volume of these studies in an outpatient setting, we focused on studies of the home and community settings.

Full economic evaluations that compared two or more eligible settings and considered both costs and consequences (including cost-effectiveness, cost–utility or cost–benefit analyses) were eligible.

Clinical studies

Data were extracted on details of study methods, country and geographical region in which the study was conducted, whether it was single or multicentre, dates over which the study was conducted, patient characteristics, interventions, comparators where appropriate, all relevant outcome measures and results.

The quality of included comparative studies was assessed using criteria appropriate to the study design, adapted from published checklists. 12

Randomised controlled trials

Randomised controlled trials were assessed using the Cochrane risk of bias tool, which focuses on the domains shown to impact on the trial results in particular (selection, performance and detection biases and attrition). 13 The tool was modified to incorporate assessment of baseline imbalances when we evaluated selection bias. 14

Non-randomised comparative studies

Study quality evaluations were based on recently published papers detailing methodological issues and assessment of bias in non-randomised studies. 15–18 Confounding variables (variables other than the intervention being studied, which might affect study outcomes when groups are compared) are known to be a very important source of bias in non-randomised studies. 19 As there were many potentially important confounders for our review question, we focused our assessment on evaluation of the risk of bias due to confounding. This was done by answering the following questions:

1. How were the groups formed?

2. Were the effects of any confounders taken into consideration during the design and/or statistical analysis stages?

3. What methods were used to control for confounders?

4. Were data on the measured confounders recorded precisely enough?

5. Were any key confounders not controlled for?

The important confounders we considered were type of cancer, stage of cancer, type of chemotherapy, age, performance status [e.g. Karnofsky, Eastern Cooperative Oncology Group (ECOG) or Lansky scores], quality of life, treatment intent (curative or palliative) and distance from hospital. These confounders were chosen for their potential to affect outcomes such as quality of life and patient satisfaction. Some of these confounders are correlated.

Where these confounders were not measured, and not taken into account in the design or analysis of the study, the study results were deemed likely to be at a high risk of bias. Studies where such details were not clear were judged to have an unclear risk of bias. Given the non-randomised nature of the studies and the lack of assurance provided about the methods used, the implications of an unclear risk of bias judgement are similar to those of a high risk of bias judgement and the study results should not be interpreted as being reliable.

Where the answer to question 2 was ‘yes’, the details of which confounders were controlled for were recorded and the remaining questions were answered; the overall risk of bias judgement (from confounding) was then made based on the answers to questions 3, 4 and 5. Where the answer to question 2 was ‘no’ or ‘not reported’, a high or unclear risk of bias judgement was made and the remaining questions were not answered. An assessment of whether or not there was evidence that potential confounders did not actually result in confounding was also made when considering question 2.

For non-randomised studies there is evidence that confounding may not, on average, cause bias in the estimation of adverse effects. 17 We considered this during our assessments according to how likely a given adverse effect (as defined in individual studies) might be affected by confounding.

As cost and resource outcomes were extracted to inform the review’s economic modelling, formal synthesis and quality assessments were not routinely performed for these outcomes.

Non-comparative studies were not extracted or quality assessed, but they are listed for reference in Appendix 2.

Cost-effectiveness studies

Data extracted from economic evaluations included interventions compared study population; dates to which the data related; measures of effect; direct costs (medical and non-medical); currency used; utilities/measure of health benefit; and results and details of any decision modelling applied. The quality assessment of the economic evaluations was informed by use of the Drummond 36-point checklist. 20 The purpose of the review was to provide an overview of the current cost-effectiveness evidence base and help to inform the development of a de novo decision model. Any additional information that could aid development of a de novo decision model was also extracted.

Qualitative studies

Qualitative studies were assessed for methodological quality using criteria based on the work of Mays and Pope, among others. 21–23 As with the quantitative studies, the focus was on those domains which are expected to influence the reliability of the findings. Domains included transparency and documentation of the data collection and analysis processes, description and justification of sampling, validity appropriate to the method being used, reflexivity and clear distinction between data and interpretation.

The results sections from each included study [apart from one Doctor of Philosophy (Ph.D.)] thesis were extracted from portable document format files and entered into NVivo (QSR International, Warrington, UK) as text documents. 24 We used online translation for one paper in Danish. This unconventional approach appeared to generate reasonable approximations of the original meaning and avoided the expense and delay of professional translators; this was balanced against the possibility of losing some meaning in translation.

The Ph.D. thesis24 was read through with the other papers, but not extracted and coded until nearer the end of the process. The thesis was useful as a way of checking for gaps or absences in the data as a whole, but it was too dense and, in places, of less immediate relevance to warrant full extraction.

Clinical effectiveness data

Our detailed narrative synthesis explored the methodology and reported outcomes of included studies. Key study characteristics, patient outcomes and quality assessment were tabulated to provide clear summaries of the included studies. The clinical and statistical heterogeneity of the accumulated evidence was assessed. Differences between studies were discussed in the text and the potential impact of these differences on outcomes was explored. The results were interpreted in the context of the quality of the individual studies.

As anticipated, the available data were too heterogeneous for quantitative synthesis.

Cost-effectiveness data

The findings of the systematic review of full economic evaluations were summarised in a narrative synthesis.

Qualitative data

The qualitative studies were synthesised using meta-ethnography, an approach which searches each primary study and systematically extracts key findings and interpretations. 25,26 These data are then compared using a constant comparison method, which categorises key concepts to look for overlapping themes in order to enable linking of material. Each finding is subsequently assessed for similarity or difference to the other studies, and the goal is to develop, in an iterative manner, this reciprocal translation. Ultimately, new lines of argument can be developed which go beyond the data contained within the original studies.

The focus of the analysis was on themes and ideas relating directly to the provision of intravenous chemotherapy with particular reference to treatment location, rather than the experience of having cancer treatments per se.

Each results section was read closely on multiple occasions and coded line by line using participants’ words where possible. Initially, codes were tagged according to the setting (e.g. code ‘prefer to go home after treatment’ was linked to ‘outpatient setting’). Subsequent readings of the texts and resultant codes led to the reworking of the coding framework into key elements rather than distinguishing by treatment location. Codes were collapsed where possible and a process of diagramming used to explore links and interactions between the key lines of argument and categories.

Randomised trials

Results and synthesis of randomised clinical effectiveness trials

Non-randomised studies

Non-randomised study results

Three studies evaluated quality of life but their results did little to augment the RCT evidence: two studies were small, with a high risk that confounding would affect the reliability of their results; and one study was reported only as an abstract (making it difficult to interpret the results). 38,46,48 There were similar issues for the four studies of patient satisfaction. 37,39,40,42 One study included 435 patients but they were selected retrospectively, with no consideration made for confounding factors; and the study was in Canada, where the travel time and distances are different from those likely to be encountered in the UK. 37

Only one of the four studies which had a safety outcome yielded informative results;45 this Australian study reported that complications in the home setting were rare, although no comparative data were reported for the outpatient setting for this particular outcome. Two of the three other studies that looked at safety had very small sample sizes,40,49 and the other was only reported as an abstract. 48 Three studies that looked at qualitative patient experience are discussed in Results, Qualitative studies.

Only one comparative study looked at the issue of treatment compliance in any detail. 47 It was conducted in the USA and studied approximately 2800 follicular lymphoma patients receiving chemotherapy [with or without rituximab (Mabthera, Roche)]. The study concluded that patients treated in the outpatient setting tended to have longer times to treatment initiation and fewer cycles across all regimens, and were less likely to receive a compliant dosing schedule than patients treated in a community clinic setting. However, the reliability of these conclusions was unclear as the study was only reported as an abstract (the risk of bias due to confounding was unclear). The results for all the non-randomised studies are presented in Appendix 5.

Clinical results evidence summary

The included studies revealed inherent difficulties in conducting randomised trials of chemotherapy settings. Even trials that were designed appropriately to minimise avoidable biases faced problems not only of patient accrual but also of recruiting a population to enable an unbiased evaluation of the settings. These seemingly unavoidable selection biases might be expected to produce results that favour home (or community) settings. Even so, there was little evidence of clinically relevant differences between settings in terms of quality of life and clinical and psychological outcomes. The only potentially meaningful differences were seen for some patient satisfaction and preference outcomes. However, strength of preference was studied in only one trial, with preferences appearing not to be strong in around one-third of patients. The limited safety evidence available suggested there were no differences between settings.

The non-randomised studies added little to the randomised trial evidence (although community settings were more frequently studied). The main limitations were the small populations and the high risk that study results were biased due to confounding.

Study characteristics

Nine economic evaluations published between 1996 and 2013 met the criteria for inclusion. The key characteristics, methods and results for the nine studies are summarised in Table 4. Details of patient characteristics and treatment regimens can also be found in Table 4. All nine evaluations also met the inclusion criteria and have been assessed independently as comparative studies.

Three of the evaluations were conducted in Australia,31,32,45 two in the UK (England),4,29 two in France34,49 and one in each of Canada33 and Denmark. 38 Most studies assessed adult populations; two studies assessed paediatric populations. 33,38 Six evaluations were conducted alongside RCTs, two alongside non-randomised controlled studies,38,49 and one was conducted as part of a retrospective audit. 45 Most studies did not conduct a full incremental analysis i.e. to produce [incremental cost-effectiveness ratios (ICERs)], but instead reported cost and health outcomes separately. Costs and outcomes were generally assessed over a short time horizon (1 year or under).

All of the non-UK studies compared treatment delivered in the home with treatment delivered in a hospital outpatient setting. 31–34,38,45,49 The two UK studies included community settings: Pace et al. 29 compared treatment delivered in a community hospital setting with treatment delivered in a hospital outpatient setting; and OUTREACH4 compared home, GP surgery and hospital outpatient settings. None of the economic evaluations assessed the delivery of chemotherapy by mobile bus units. One study assessed the delivery of home chemotherapy by a third-party charity organisation. 34 In all other studies it was implied that home/community care was delivered by the health service.

As highlighted in the clinical study sections, most studies assessed mixed populations with various cancer types, including breast cancer, colon cancer, lung cancer, gastrointestinal cancer, lymphoma, pancreatic cancer and leukaemia. One study assessed only patients with acute lymphoblastic leukaemia. 33 An array of treatment regimens was used for a mixture of curative, supportive and palliative intent. The populations appeared to be heterogeneous in terms of disease severity.

The nine evaluations conducted alongside clinical studies recruited a total of 593 patients and 487 participants were evaluated in the concurrent economic evaluations. Most of these participants were evaluated in the two non-randomised studies; 179 patients were evaluated in the retrospective audit45 and 75 were evaluated in the controlled study. 38 The six evaluations made concurrently with RCTs were based on small data sets:4,29,31–34 the largest study (OUTREACH)4 had complete primary outcome data (EORTC QLQ-C30 Emotional Function subdomain) on only 57 participants across three treatment arms/settings. Chapter 6 contains an overview of recruitment and participation within the RCTs (see Limitations of the evidence and of the review). The number of participants informing the concurrent cost-effectiveness analysis is presented in Table 4, alongside other study characteristics and findings.

Economic quality assessment

The Drummond checklist was used to assess methods and reporting in the nine included economic evaluations. 20 Clinical studies undertaken alongside the economic evaluations were assessed for risk of bias as part of the quality evaluation of comparative studies (see Table 3 and Appendix 4). All of the evaluations suffered from limitations, as highlighted by the checklist. Many did not undertake appropriate data collection and/or sensitivity analyses. Seven of the nine studies reported no data on resource use outcomes, which significantly reduces the transparency and transferability of the results. Five studies reported disaggregated costs (costs for each of the included cost categories rather than a total cost only), but the usefulness of such costs in terms of informing potential UK NHS cost estimates is limited without resource use data. Reporting of methods used to derive cost and resource use outcomes was limited and this made the validity of the cost estimates unclear. Most studies failed to report the price year or any cost adjustments applied. None of the studies conducted an adequate analysis of the potential impact of uncertainty on the results; all three studies that included sensitivity analyses were limited in scope. Most studies failed to consider the generalisability of their results. Full results for the quality assessment can be found in Table 5.

Overall, the studies were deemed to be of low or uncertain quality due to small sample sizes; limited reporting on resource use, costs and methodology; and lack of robust sensitivity analyses. There is likely to be significant uncertainty around the cost-effectiveness results.

Economic study results

Economic evidence summary

Quality across the economic evaluations was variable, and overall should be considered poor. Biases in the clinical study, the level of reporting on resource use and associated unit costs, and the lack of health-related quality-of-life outcomes were major limitations. Poor reporting of resource use and the use of different perspectives across the different settings made the results difficult to compare. Several economic evaluations used patient preference as an outcome measure, rather than health-related quality-of-life outcomes, which are more widely used to inform decision-making. 29,32,34,45,54 High levels of uncertainty make it difficult to ascertain whether or not costs or outcomes differ between settings.

Overview of study characteristics and quality

Seventeen qualitative or mixed-methods studies were included in this review (published between 1984 and 2012). 4,24,29,30,32,39,41,51,55–64 The studies were conducted in Canada (four55–57,60), the USA (one59), the UK (nine4,24,29,30,39,51,62,63), Denmark (two41,61), Iceland (one64) and Australia (one32). UK studies were in England (six4,24,29,30,39,62) and one in each of Wales,51 Northern Ireland63 and Scotland. 58 Table 6 displays a summary of the included study characteristics. Appendix 8 gives full details of the extracted data. Table 7 provides summary quality assessment results. Stevens et al. 200655 and Stevens et al. 200456 are separate publications from the same study, where one paper reported on health-care professional views and the other focused on the experiences of the children and their parents; these two publications have been kept separate for clarity and, therefore, there were 18 sources of data contributing to the synthesis.

Participant views represented include chemotherapy patients (16 studies;4,24,29,30,32,39,41,51,55,57,59–64 two included children55,61); carers/partners or children’s parents (four studies24,39,55,61); and health-care professionals (five studies4,24,56,58,62). The study contexts varied, with half the studies taking place within a clinical study. Participants often spoke comparatively about their experiences based on current and prior treatment received or given.

Six papers contributed data on community provision of intravenous chemotherapy. 4,29,39,51,57,58 Eleven papers contributed data on the home setting4,24,30,32,41,55,56,59–61 and one paper was a theoretical discussion of a proposed service. 62 All 18 papers included discussion of outpatient settings and two studies focused exclusively on this. 63,64 The OUTREACH study contributed data on all three settings. 4

As discussed in the summary of the quantitative studies (see Chapter 6, Strengths of the review, Clinical effectiveness studies) study populations are likely to reflect a limited and biased perspective given the nature of the research. All papers dealing with community or home chemotherapy were based around new services, pilots or trials, which implies that the provision may be particularly high quality. It was rare to see any indications of problems with the new services. There was one comment about the chance of the community chemotherapy bus being cancelled but the patient noted that this never occurred. One Canadian study mentioned problems with drug deliveries. 55 There were no studies discussing new community cancer suites or outpatient chemotherapy suites; these facilities tend to offer integrated care in light and airy spaces designed for comfort. 65,66

Almost all patients expressed a preference for treatment location, with varying degrees of strength. The few patients who did not seem to mind about location were clear that the treatment was a necessary inconvenience regardless of the place it occurred.

It is worth noting that although Smith and Campbell58 was a discussion around a proposed community service, the views and comments from participants were essentially the same as those given by patients who had received community chemotherapy in other studies. This suggests that focus group work may be a useful preamble to setting up new services, as both patients and health professionals are likely to identify any potential barriers and benefits, enabling pre-emptive action.

Fifteen studies4,24,30,39,41,51,55–58,60–64 used interviews to gather data (one study used telephone interviews58), and two used open-ended questions on a questionnaire. 29,32 One study collected data through focus groups, open-ended questionnaires and telephone interviews. 24

Most papers did not clearly identify the author’s position, or appear to engage substantially with reflexivity. That is, they did not appear to consider the impact of the researcher on the data collection and analysis. Four papers at least partly addressed these issues including considering the potential impact of the author on the data collection and analyses process. 39,41,60,61 Other papers either did not clearly identify the researcher who collected and analysed the data or did not discuss the implications where the researcher was a cancer nurse. 30

The most common form of sampling was based on convenience; fewer studies used purposive sampling and three studies included all patients who participated in the trial or pilot. 29,30,32 Information on participation rates was scantly reported across the studies making it difficult to draw any general conclusions. In some studies, staff delivering care to study participants were responsible for recruitment or suggesting participation; this may have contributed some bias.

Study methodology was reported infrequently. Analyses were most commonly labelled as content analysis (five studies55,56,58,61,64) or thematic (four studies24,30,51,62). Two studies used framework,4,57 two used phenomenology,39,41 one used the constant comparative method60 and one used narrative analysis. 63 Three studies did not report their methods of analysis. 29,32,59

Validation and the consideration of validity can be a problematic topic within qualitative research. Ten out of 18 studies39,41,55–58,60,61,63,64 in this group reported using some form of validation. This included peer discussion, independent coding by more than one researcher and discussion of preliminary findings. Most studies provided clear quotations from the participants and emphasised the primacy of the interviewee’s voices. Thirteen out of 18 studies24,30,39,41,55–58,60–64 drew clear distinctions between collected data and researcher interpretation. Transferability and generalisability were discussed in some detail by nine studies.

Synthesis

Overall, the data were grouped under three main lines of argument (Table 8). Two were relatively self-contained: one addressed issues around perceived cost and barriers to service use, while the second addressed satisfaction with treatment experience and provision.

The third line of argument was more substantial than the other two, and more clearly relates to specific issues of the setting for intravenous chemotherapy. The decisions made by patients, carers and family members were focused on maintaining normality in everyday life. Compromises were required to balance competing factors and patients’ location preferences reflected their individual situations.

The three main lines of argument are not entirely independent; where links were obvious these have been noted in the descriptive text that follows.

Qualitative summary

The synthesis incorporated data from 18 papers and over 450 participants including health professionals, patients (adults and children), parents, carers, siblings and partners. Most of the available data focused on experiences of outpatient and home settings for intravenous chemotherapy, with some data on community settings including mobile bus units. Overall, the quality of the included studies was moderate to good.

Maintaining normality was the important overarching theme which tied much of the data together. Across the data set, the importance of maintaining normality throughout a difficult illness/treatment was seen as key to being able to survive and look forward. Exactly what normal life constituted varied from patient to patient. Most patients were clearly making explicit trade-offs to maximise their resources (e.g. time, money or energy). Looking at the data set, it was rarely as simple as saying that one setting maintained normality and alleviated any safety-related concerns. Health professionals recognised and referred to this when talking about their patients, indicating a shared understanding.

Normality was more easily maintained when family life was minimally interrupted, the impact of cancer on daily life and family members was controllable, and patients were able to participate in activities of value. The time and energy consumed by chemotherapy underpinned much of this category: time spent travelling and waiting for treatment meant less time for normal life. The energy expended on treatment (including travel and waiting time) could leave patients unable to participate in important activities. Although treatment- and setting-related anxiety or side effects were mentioned, these seemed to be important because dealing with the anxiety, stress and nausea took up valuable time and energy.

While the outpatient settings were most often associated with increased confidence in the staff’s ability to deal with adverse reactions, there was some evidence that good, visible communication between an expert centre and an outreach location could ameliorate some of the safety concerns. Based on the available data, the time and energy consumed by outpatient treatment reduces overall quality of life and this is a sufficient driver for patients to prefer alternative treatment locations. These themes were particularly evident in the accounts from patients receiving palliative treatment and from parents of children with cancer.

Respondents and services provided

We sent 65 e-mails inviting stakeholders to participate and sent a reminder e-mail to non-respondents after 2 months. Respondents were encouraged to forward the survey to other contacts; this increased the number of responses at the expense of making the percentage response unclear.

In total, 26 people from 22 organisations provided usable survey responses to the NHS survey. Sixteen of the 22 organisations were in the north of England and six were in the south. Organisations included trusts, specific hospitals and one commissioning body. Respondents were in various roles including commissioners, pharmacists, cancer nurses, oncologists, regional managers, haematologists, directors and administrators. All survey respondents were in England.

Ten organisations provided chemotherapy at home or in the community: three delivered intravenous chemotherapy only in the home, three delivered only in a community setting and four delivered treatment in both settings.

Figure 6 provides a flow chart demonstrating the services provided by respondents.

FIGURE 6.

Flow chart of current NHS provision of home and community intravenous chemotherapy. a, Herceptin^®, Roche.

Aspects of the service and pharmacy use

Two of the seven organisations that provide chemotherapy at home indicated that the NHS delivered all aspects of the service, one failed to respond and four indicated that they used a private provider to deliver some or all of the service. Two of those using a private provider indicated that they used a hospital pharmacy for home chemotherapy; both of these deliver treatment in the home and community setting. The other two organisations which deliver in both settings were both NHS-provided services, including pharmacy.

Two of the three organisations providing treatment only in a community setting indicated that the NHS provided all aspects of the service, including hospital pharmacy, and one did not respond. Table 9 provides service provision details for settings offered, pharmacy use and training/recruitment requirements for services in the home or community.

Staff and training necessary for administration of chemotherapy at home or in the community

Staff involved in administration of chemotherapy in the home or the community included oncologists, nurses, haematologists and pharmacists. Five organisations responded that additional training and/or additional recruitment was required for their service, two indicated that no additional training was required and three did not respond. Nurses were the focus of additional training, which included training nurses to higher certification levels, training them on how to use mobile chemotherapy units (chemotherapy bus), and lone-worker training. Three of the five organisations that indicated additional training was required delivered in both settings (two NHS and one private provider). Two organisations delivered only in a community setting using a NHS service.

Three organisations indicated that hiring additional staff was required, while two indicated that it was not and five failed to respond. All of those that indicated that they would need to hire additional staff also indicated that additional training would be required; two of these were community NHS-delivered services and the other a NHS-delivered service across both settings. The two organisations that indicated no additional hiring also indicated no additional training requirements; both services had home services that were delivered by private providers.

The number of nurses involved in delivering intravenous chemotherapy in a home or community setting varied between organisations. Two organisations used one nurse at their community locations and three organisations used two nurses. Two organisations did not report how many nurses they used for community chemotherapy locations. The number of nurses who administered each individual treatment at home and community locations was fairly consistent. All five organisations that provided a response reported that one nurse was involved in each intravenous treatment.

The details of services provided, by whom and where, plus additional staff hiring and training requirements for each of the organisations, are presented in Table 9.

Eligibility for participation in home chemotherapy

Various eligibility policies were described for home chemotherapy. One of four organisations that provided home chemotherapy via a private provider reported an eligibility requirement of ‘a few’ cycles delivered in hospital; another reported that eligibility decisions were based on regimens; and two did not report eligibility requirements. There were few responses regarding eligibility requirements for chemotherapy; three organisations provided no response to eligibility requirements for chemotherapy at home and one reported that drug regimen suitability determined patient eligibility for chemotherapy at home.

Three of the seven organisations that provided chemotherapy at home reported that patients were referred to the service by consultants, and two indicated that consultants and specialist nurses could refer patients to the service. No other organisations provided referral details.

Three organisations provided estimates for the proportion of patients eligible for the home service and the proportion of patients accepting the service. There was a mix of proportions eligible and accepting, with some eligibility levels of < 5%. Both organisations that indicated < 5% eligibility also indicated that they used a home-care provider, with HaH named by one. Both organisations that had home services completely delivered by the hospital reported eligibility levels higher than those using a service delivered through a home-care provider (Table 10).

Eligibility for participation in community chemotherapy

In the community setting, patient eligibility criteria focused primarily on suitability of regimens and patient distance from their hospital. Only one organisation expressed eligibility limitations based on which cycle of chemotherapy was being administered; they reported that patients must be fit and have had two cycles in hospital. Two indicated that distance of patient travel was a factor in eligibility, but did not quantify what distances were acceptable. Three quantified the percentage of patients eligible for chemotherapy in the community: one reported that 60% of patients were eligible and 80% of eligible patients accepted; another reported that 10–15% were eligible and 10–15% accepted; and a third reported that 50% of patients were eligible, with 30% accepted (see Table 10).

Provision of chemotherapy services at home and in the community

Three organisations delivered chemotherapy in the community using mobile chemotherapy units. The other four organisations that delivered intravenous chemotherapy in a community setting used different locations: two in community hospitals, one in a satellite unit in a primary care centre and another used a room in a local hospice.

There were several similarities between home and community chemotherapy administration. Patients in both settings were cared for between chemotherapy treatments at their regular institutions and given access to standard 24-hour advice telephone lines, and patients were referred to both services by consultants and specialist nurses.

Organisations that indicated they did not provide chemotherapy at home or in the community

There were 12 organisations that did not provide intravenous chemotherapy at home or in the community; six indicated that they were interested in providing a service and three indicated that they may be interested in providing a service in the future. One organisation had indicated that they provide home chemotherapy, but only provided home trastuzumab. No organisations said that they would not consider providing a community or home service. Of the six organisations that said they would consider providing a service, one organisation was working on a proposal for a service, and another indicated that a service was not offered yet but was at an advanced stage of planning.

Barriers to service delivery

One of the aims of the survey was to identify barriers to service provision in the community and at home from those who provided services, and those who did not provide services. Common concerns existed in both groups. The full set of responses to these questions is available on request. Some commonly perceived barriers highlighted by responders were:

• costs of running a service

• value-added tax (VAT) savings, which were driving which drugs were offered at home rather than the suitability of the drugs for administration at home

• the fact that there might have been less expensive ways to deliver chemotherapy in the community than delivering at home, but current regulations did not allow or incentivise more efficient community delivery

• issues with consultant support for home services

• poor strategic planning

• broad geographical area that would be difficult to serve

• interest in delivering a service but lack of a commissioned service

• limited numbers of eligible chemotherapy regimens

• lack of nursing resources

• lack of suitably trained staff

• a need to convince patients to use the service.

There were several limitations to the survey: the sample was small; the information provided by respondents was generally not very detailed; the survey requested recollections and descriptions from providers rather than service data; and questions were not always interpreted as intended. However, the aim was to provide a picture of current provision and add clarity to the patient pathway, where possible. The survey highlights the wide variation in current provision.

Patient choice

The initial distribution of patients to different settings within the model was determined by patient choice assumptions derived from OUTREACH. 4 As highlighted in the systematic review of clinical effectiveness, patients may participate in a trial because they want the new treatment; others may elect not to participate to ensure that they are not exposed to unwanted treatment. In OUTREACH, patients who wanted outpatient hospital chemotherapy or did not want to receive treatment in a particular setting could simply opt not to participate in the trial. The proportion of patients in the model choosing each setting was, therefore, adjusted for recruitment bias, based on methodology from King et al. 31

Although OUTREACH reported preference data, it did not utilise a crossover design and patients were asked whether or not they would prefer to continue treatment in their allocated setting. These patients experienced only one setting during the trial and so it is unclear if their decision was informed by past experience of chemotherapy in the hospital outpatient setting or if patients were naive to chemotherapy. Owing to this, it was unclear in which setting any given patient would choose to have treatment had they experienced multiple settings.

To adjust for this limitation of the OUTREACH data we assumed that the patient population had had weak preferences as in King et al. 31 Assuming indifference to treatment settings may be considered conservative as we cannot know the motivation of patients who chose to participate in the trial; they might have participated because of preference for treatment settings other than the outpatient default. Patients who chose not to participate in the trial owing to a stated preference were by definition not indifferent to the setting and so we needed to adjust for this potential bias. Our adjustments took into account patients’ expressed reasons for preferring one setting over another as described in Chapter 3 (see Qualitative studies) and the analysis in King et al. ,31 where patients who had declined to participate in the trial due to strong preferences were reintegrated into the data to adjust preferences; this resulted in 48% of patients preferring chemotherapy at home, as outlined earlier (see Chapter 3, Clinical effectiveness studies and Satisfaction and preferences). 31,54

In OUTREACH, 33 patients were randomised to home, 32 were randomised to community (GP practice) and 32 were randomised to outpatient. Fifty-three patients declined to take part in the trial: 35 wanted an outpatient setting; 16 did not want chemotherapy in a GP practice; and two preferred chemotherapy at home. Following the King et al. 31 methodology, these 53 patients excluded owing to preference were added to those who chose to participate in the OUTREACH trial with the assumption that patients with a stated preference for outpatient would receive outpatient and those with a preference against the GP or home setting would be split evenly between the alternative settings. As the results from the King et al. 31 study suggested that overall patient preference may not be particularly strong, we assumed patients who stated no preference to be indifferent to treatment setting allocation.

After adjustments, the resultant probabilities suggested that 27.3% of patients would choose to have their chemotherapy at home, 22.0% would choose to have it in the community and 50.7% would choose the hospital outpatient setting. In a probabilistic sensitivity analysis, a Dirichlet distribution was used to reflect the uncertainty surrounding choice. Table 11 shows the transition probabilities between states used within the model and the mortality rate applied to all states.

At this point, including patient choice does not affect OUTREACH results because any benefits and costs in a setting are divided by the number of people in that setting and so they are cancelled out. However, if throughput is introduced to the model, and this throughput determines fixed costs and potential savings due to economic efficiency, then patient choice will impact the results of the model.

The model allows for the possibility that patients may die based on Office for National Statistics mortality statistics for 20 common cancers. 73 Although there were no mortalities recorded in the OUTREACH trial, it is possible that, in practice, some patients receiving chemotherapy will die over a 12-week or longer period. The mortality rate was assumed to be the same for each setting and so had no effect on the model outcomes. The mortality parameter was included to allow for longer model lengths and to improve the face validity of the model.

Patient quality of life

The OUTREACH trial reported the number of QALYs gained over the 12-week length of the trial. Utility values were derived using the EQ-5D questionnaire as completed by the patients in the trial. Questionnaires were completed at baseline and 4, 8 and 12 weeks; only outcomes at 12 weeks were analysed. The model uses the 12-week QALYs derived from OUTREACH. These data are treated as utility scores, assuming constant utility over the time period, implying that any QALY gains are equivalent to multiplying the utility score at 12 weeks by 12/52. The QALYs in OUTREACH were reported without baseline values and so the base-case model has no adjustment for baseline values. Table 12 presents the utility values in the model derived from OUTREACH. The uncertainty in the underlying utility data was represented using beta distributions in probabilistic sensitivity analysis. The utility value for the death state was assumed to be the customary value of zero.

Costs

The original OUTREACH publication reported total costs with no details of what these comprised. We obtained additional cost category data from the authors.

Total costs were divided into the following categories: inpatient, outpatient, day hospital, A&E visits, community care, medication (excluding cancer drugs) and nurse contact. Follow-up data provided by the authors were reported as average costs per patient for each category in each setting, with no reported resource use. However, included data suggested that A&E visits, as a cost category, cost on average £3–4 per patient. Clarification was sought from the authors whether the value they reported was the average number of visits, as their table appeared to indicate, or the total number of visits to A&E for each treatment arm, but no additional information was provided. Therefore, we have assumed that this is an error and that the number represented total visits to A&E per arm, rather than the average costs of those visits. In order to produce the average cost of each emergency room visit, we multiplied the NHS reference cost for A&E visits by the number supplied from the OUTREACH trial and divided this by the number of participants in each arm. The cost of one A&E visit without any follow-up visits according to NHS reference costs was £122 in 2011–12. 74

Cost data provided on request from the OUTREACH trial did not contain SDs or standard errors; aggregate costs reported in the published study provided SDs for total costs in each study arm. We assumed that the proportion of the SD to the mean for cost categories would be similar to the proportion of the SD to the mean for total costs. The proportions of the SD of costs to mean costs are shown in Table 13. We assumed that the proportion of the SD to costs for all arms would be the average of the three, that is 75%.

In a probabilistic sensitivity analysis, gamma distributions were used to represent costs, as the skew of the gamma distribution and bounding between zero and positive infinity make it a good choice for representing costs. Table 14 incorporates all updated costs from the OUTREACH trial inclusive of our assumptions modifying A&E costs and inflating prices using 2011–12 Personal and Social Services Research Unit price indices for medical services. 75 The cost of an A&E visit was derived from 2011–12 NHS reference costs;74 all other costs were derived from the OUTREACH trial. 4

Adjusting for baseline imbalances in utility scores

Information provided from the OUTREACH trial indicated that the authors had adjusted QALY gains for baseline imbalances. Because no baseline values were reported and only a difference in difference analysis was provided, this sensitivity analysis assumes that the difference in baseline values is maintained in the difference in final QALYs gained. To maintain consistency, these values were converted to utility scores for use in the model. Table 18 shows the utility adjustments made in the OUTREACH data.

Using these adjusted values in the model changes the ICER of the chemotherapy settings. In the base case, chemotherapy administered in a community setting was the most effective option at a threshold of £20,000. In this sensitivity analysis, chemotherapy in the community is both more costly and less effective than administration in the outpatient setting. The ICER for outpatient chemotherapy compared with chemotherapy in the home setting also improves.

This analysis emphasises the uncertainty surrounding the utility values used in the model. Given the small sample size in the OUTREACH trial and the suggestion of baseline differences, the utility data cannot be considered reliable. It is likely that the quality of life of patients depends on their disease, treatment status (palliative, supportive or curative), age, and other factors that the OUTREACH data may not have captured. Clarification was sought from the authors on whether or not the data were adjusted for patient characteristics other than baseline EQ-5D scores, but none was received. Given the small patient numbers, it is unlikely that any such adjustment would have been plausible or would reduce the uncertainty. Further research is necessary to confirm or refute differences in patient quality of life between settings.

In scenario 2, making the mean cost of community care in the home setting equal to the mean cost of community care in the community (GP) setting, and the mean nurse diary costs for the community (GP) setting equal to those in the home setting, resulted in outpatient care being extendedly dominated; this means that a hypothetical combination of home and community chemotherapy services would be more cost-effective than outpatient chemotherapy on its own. In scenario 2, chemotherapy in the community setting is the most cost-effective option with an ICER that is lower than the base-case analysis. Similarly, scenario 3 makes home more costly than outpatient and it remains less effective. In scenario 3, home is dominated and community is the most cost-effective option.

Other areas of interest for exploration

The perspective of all our modelling analyses is the NHS and Personal and Social Services perspective (only costs to the NHS and Personal and Social Services are included). Other perspectives could be analysed and may produce different results.

Some economic evaluations included in our systematic review included patient or carer costs: patient travel costs;29 parent travel costs, payments for physician/care provider visits, medication, babysitting, lost productivity and government transfer payments;33 and patient travel costs. 34 These are valid perspectives, but their relevance to NHS decision-makers may be limited even though they are interesting.

The Centre for Health Economics, University of York, published a contingent valuation study of patient time. 77 This study asked patients to value different usages of their time such as time to start of treatment (between scheduling an appointment and receiving treatment), travel time, time waiting at a health facility for treatment and time receiving treatment.

In our model, time to start of treatment and time receiving treatment are expected to be independent of the setting and so only values related to travel time and waiting time are reported. The Centre for Health Economics study measured Dutch patients’ valuations of travel and waiting time for patients in three treatment areas (radiotherapy, orthopaedics and rehabilitation) and found that they valued each hour of travel time at £11.54 and each hour of waiting time at £34.76. 77 Chemotherapy patients were keenly aware of how time spent travelling to hospital and waiting for treatment had negative impacts on their time and energy. Including valuations of patient time could have significant impacts on the cost-effectiveness of treatment settings for chemotherapy. However, our systematic review identified only limited information on time spent travelling to and waiting in hospitals in the UK.

Several respondents in our survey of provision mentioned that drugs delivered at home were exempt from VAT (current rate 20%) and that these savings were used to finance home chemotherapy services. This view was reflected by the advisory group. Private providers have used VAT savings in trying to market their home services to NHS providers. 69 HM Revenue & Customs has indicated that medicines delivered in the home may be eligible for a zero rating for VAT where requirements are met regarding drug use and provision. 68

Considering that cost differences between the settings were small, reducing drug costs by 20% could alter cost-effectiveness decisions on home provision of specific chemotherapeutic drugs. Drug costs were not included in our model as they were not included in the OUTREACH data. However, it is likely that drug costs make up a significant proportion of the total cost of treatment. This will be particularly true in the case of newer chemotherapy drugs where the cost of the drug may represent most of the total treatment cost. In these instances the impact of VAT exemptions may have an impact on where these drugs are prepared and delivered.

Some of the issues outlined will not be relevant across the whole NHS, but local service configurations may make them worthy of further investigation should data become available.

Search strategy

1. (cancer$ or neoplas$ or tumor$ or tumour$ or malignan$ or oncolog$ or carcinoma$).ti,ab. (13,694)

2. chemotherapy.ti,ab. (1083)

3. systemic therapy.ti,ab. (30)

4. intravenous drug therapy.ti,ab. (0)

5. adjuvant therapy.ti,ab. (76)

6. or/2-5 (1167)

7. ((service$ or therapy or treatment$) adj6 (home or community or outreach or out-reach or ambulatory or domicil$)).ti,ab. (2940)

8. (hospital at home or hospital in the home or own home$ or home care or homecare or closer to home).ti,ab. (1167)

9. 7 or 8 (3782)

10. 1 and 6 and 9 (26)

11. (chemotherapy adj6 (home or community or outreach or out-reach or ambulatory or domicil$)).ti,ab. (21)

12. (chemotherapy adj6 service$).ti,ab. (8)

13. (chemotherapy adj6 (general practitioner$ or family practitioner$ or family doctor$ or family physician$ or primary care physician$)).ti,ab. (1)

14. (self-infusion adj6 home).ti,ab. (0)

15. home infusion.ti,ab. (8)

16. or/11-15 (34)

17. 1 and 16 (31)

18. 10 or 17 (44)

Key:

$ = truncation

.ti,ab. = terms in either title or abstract fields

adj6 = terms within six words of each other (any order).

Search strategy

S1 TI,AB(cancer* or neoplas* or tumor* or tumour* or malignan* or oncolog* or carcinoma*) 12003

S2 TI,AB(chemotherapy) (1121)

S3 TI,AB(“systemic therapy”) (15)

S4 TI,AB(“intravenous drug therapy”) (6)

S5 TI,AB(“adjuvant therapy”) (39)

S6 S2 or s3 or s4 or s5 (1167)

S7 TI,AB((service* or therapy or treatment*) NEAR/6 (home or community or outreach or “out-reach” or ambulatory or domicil*)) (2719)

S8 TI,AB(“hospital at home” or “hospital in the home” or “own home*” or “home care” or homecare or “closer to home”) (1084)

S9 S7 or s8 (3566)

S10 s1 and s6 and s9 (22)

S11 TI,AB(chemotherapy NEAR/6 (home or community or outreach or “out-reach” or ambulatory or domicil*)) (56)

S12 TI,AB(chemotherapy NEAR/6 service*) (37)

S13 TI,AB(chemotherapy NEAR/6 (“general practitioner*” or “family practitioner*” or “family doctor*” or “primary care physician*”)) (0)

S14 TI,AB(“self-infusion” NEAR/6 home) (0)

S15 TI,AB(“home infusion”) (2)

S16 s11 or s12 or s13 or s14 or s15 (82)

S17 s1 and s16 (61)

S18 s10 or s17 (65)

Key:

TI,AB = terms in either title or abstract fields

* = truncation

NEAR/6 = terms within six words of each other (any order)

“ ” = phrase search.

Update

Date range searched: 1982 to March 2013.

Date searched: 29 October 2013.

Records found: 1069.

Search strategy

S1 (MH “Neoplasms+”) (162,398)

S2 cancer* or neoplas* or tumor* or tumour* or malignan* or oncolog* or carcinoma* (201,467)

S3 (MH “Oncologic Nursing”) OR (MH “Pediatric Oncology Nursing”) (10,675)

S4 S1 OR S2 OR S3 (216,493)

S5 (MH “Chemotherapy, Adjuvant”) OR (MH “Chemotherapy, Cancer”) (11,453)

S6 (MH “Administration, Intravenous”) OR (MH “Infusions, Intravenous”) (6862)

S7 chemotherapy (21,389)

S8 “systemic therapy” (967)

S9 “intravenous drug therapy” (45)

S10 “adjuvant therapy” (1311)

S11 S5 OR S6 OR S7 OR S8 OR S9 OR S10 (29,231)

S12 (MH “Home Health Care”) (13,640)

S13 (MH “Home Nursing, Professional”) (6375)

S14 (MH “Community Health Nursing”) OR (MH “Community Health Services”) OR (MH “Community Health Centers”) (29,922)

S15 (MH “Outpatients”) (28,597)

S16 (MH “Ambulatory Care”) (5611)

S17 (MH “Ambulatory Care Facilities”) OR (MH “Ambulatory Care Nursing”) (3924)

S18 (MH “Family Practice”) (9893)

S19 (MH “Physicians, Family”) (7696)

S20 (service* or therapy or treatment*) N6 (home or community or outreach or “out-reach” or ambulatory or domicil*) (32,720)

S21 “hospital at home” or “hospital in the home” or “own home*” or “home care” or homecare or “closer to home” (14,801)

S22 S12 OR S13 OR S14 OR S15 OR S16 OR S17 OR S18 OR S19 OR S20 OR S21 (120,856)

S23 S4 AND S11 AND S22 (745)

S24 (MH “Home Intravenous Therapy”) (1197)

S25 chemotherapy N6 (home or community or outreach or “out-reach” or ambulatory or domicil*) (214)

S26 chemotherapy N6 service* (77)

S27 chemotherapy N6 (“general practitioner*” or “family practitioner*” or “family doctor*” or “primary care physician*”) (4)

S28 “self-infusion” N6 home  (3)

S29 “home infusion” (263)

S30 S24 OR S25 OR S26 OR S27 OR S28 OR S29 (1476)

S31 S4 and S30 (320)

S32 S23 OR S31 (884)

Key:

MH = indexing term (MeSH heading)

* = truncation

“ ” = phrase search

N6 = terms within six words of each other (any order)

Search strategy

chemotherapy AND home 154 - none relevant

“ambulatory chemotherapy” (0)

“chemotherapy in the community” (0)

“domiciliary chemotherapy” (0)

“home intravenous therapy” (0)

“self-infusion at home” (0)

“home infusion” 5 - none relevant (0)

“outreach chemotherapy” (0)

“out-reach chemotherapy” (0)

chemotherapy AND “general practice” (9) - none relevant

chemotherapy AND “primary care” (76) - none relevant

Key:

“ ” = phrase search.

Update

Date searched: 29 October 2013.

Cochrane Database of Systematic Reviews, Issue 10 of 12, October 2013.

Records found: 3.

Search strategy

#1 MeSH descriptor: [Neoplasms] explode all trees

#2 cancer* or neoplas* or tumor* or tumour* or malignan* or oncolog* or carcinoma*:ti,ab,kw (Word variations have been searched)

#3 MeSH descriptor: [Oncologic Nursing] this term only

#4 #1 or #2 or #3

#5 MeSH descriptor: [Drug Therapy] this term only

#6 MeSH descriptor: [Antineoplastic Combined Chemotherapy Protocols] this term only

#7 MeSH descriptor: [Chemotherapy, Adjuvant] this term only

#8 MeSH descriptor: [Consolidation Chemotherapy] this term only

#9 MeSH descriptor: [Maintenance Chemotherapy] this term only

#10 MeSH descriptor: [Administration, Intravenous] this term only

#11 MeSH descriptor: [Infusions, Intravenous] this term only

#12 chemotherapy or “systemic therapy” or “intravenous drug therapy” or “adjuvant therapy”:ti,ab,kw (Word variations have been searched)

#13 #5 or #6 or #7 or #8 or #9 or #10 or #11 or #12

#14 MeSH descriptor: [Home Care Services] this term only

#15 MeSH descriptor: [Home Care Services, Hospital-Based] this term only

#16 MeSH descriptor: [Outpatients] this term only

#17 MeSH descriptor: [Ambulatory Care] this term only

#18 MeSH descriptor: [Ambulatory Care Facilities] this term only

#19 MeSH descriptor: [Outpatient Clinics, Hospital] this term only

#20 MeSH descriptor: [Community Health Services] this term only

#21 MeSH descriptor: [Community Health Nursing] this term only

#22 MeSH descriptor: [Community Health Centers] this term only

#23 MeSH descriptor: [General Practitioners] this term only

#24 MeSH descriptor: [Physicians, Family] this term only

#25 MeSH descriptor: [Physicians, Primary Care] this term only

#26 MeSH descriptor: [General Practice] this term only

#27 MeSH descriptor: [Family Practice] this term only

#28 service* near/6 (home or community or outreach or “out-reach” or ambulatory or domicil*):ti,ab,kw (Word variations have been searched)

#29 therapy near/6 (home or community or outreach or “out-reach” or ambulatory or domicil*):ti,ab,kw (Word variations have been searched)

#30 treatment* near/6 (home or community or outreach or “out-reach” or ambulatory or domicil*):ti,ab,kw (Word variations have been searched)

#31 “hospital at home” or “hospital in the home” or “own home*” or “home care” or homecare or “closer to home”:ti,ab,kw (Word variations have been searched)

#32 #14 or #15 or #16 or #17 or #18 or #19 or #20 or #21 or #22 or #23 or #24 or #25 or #26 or #27 or #28 or #29 or #30 or #31

#33 #4 and #13 and #32

#34 MeSH descriptor: [Home Infusion Therapy] this term only

#35 chemotherapy near/6 (home or community or outreach or “out-reach” or ambulatory or domicil*):ti,ab,kw (Word variations have been searched)

#36 chemotherapy near/6 service*:ti,ab,kw (Word variations have been searched)

#37 chemotherapy near/6 (“general practitioner*” or “family practitioner*” or “family doctor*” or “family physician*” or “primary care physician*”):ti,ab,kw (Word variations have been searched)

#38 “self-infusion” near/6 home:ti,ab,kw (Word variations have been searched)

#39 “home infusion”:ti,ab,kw (Word variations have been searched)

#40 #34 or #35 or #36 or #37 or #38 or #39

#41 #4 and #40

#42 #33 or #41

Key:

MeSH descriptor = indexing term (MeSH heading)

* = truncation

“ “ = phrase search

:ti,ab,kw = terms in title, abstract or keyword fields

near/6 = terms within six words of each other (any order).

Update

Date searched: 29 October 2013.

• Database of Abstracts of Reviews of Effects: Issue 3 of 4, July 2013.

  • Records found: 9.

• Health Technology Assessment Database: Issue 3 of 4, July 2013.

  • Records found: 2.

• NHS Economic Evaluation Database: Issue 3 of 4, July 2013.

  • Records found: 67.

• Cochrane Central Register of Controlled Trials: Issue 9 of 12, September 2013.

  • Records found: 161.

• Cochrane Methodology Register: Issue 3 of 4, July 2012.

  • Records found: 9.

Search strategy

#1 MeSH descriptor: [Neoplasms] explode all trees

#2 cancer* or neoplas* or tumor* or tumour* or malignan* or oncolog* or carcinoma* (Word variations have been searched)

#3 MeSH descriptor: [Oncologic Nursing] this term only

#4 #1 or #2 or #3

#5 MeSH descriptor: [Drug Therapy] this term only

#6 MeSH descriptor: [Antineoplastic Combined Chemotherapy Protocols] this term only

#7 MeSH descriptor: [Chemotherapy, Adjuvant] this term only

#8 MeSH descriptor: [Consolidation Chemotherapy] this term only

#9 MeSH descriptor: [Maintenance Chemotherapy] this term only

#10 MeSH descriptor: [Administration, Intravenous] this term only

#11 MeSH descriptor: [Infusions, Intravenous] this term only

#12 chemotherapy or “systemic therapy” or “intravenous drug therapy” or “adjuvant therapy” (Word variations have been searched)

#13 #5 or #6 or #7 or #8 or #9 or #10 or #11 or #12

#14 MeSH descriptor: [Home Care Services] this term only

#15 MeSH descriptor: [Home Care Services, Hospital-Based] this term only

#16 MeSH descriptor: [Outpatients] this term only

#17 MeSH descriptor: [Ambulatory Care] this term only

#18 MeSH descriptor: [Ambulatory Care Facilities] this term only

#19 MeSH descriptor: [Outpatient Clinics, Hospital] this term only

#20 MeSH descriptor: [Community Health Services] this term only

#21 MeSH descriptor: [Community Health Nursing] this term only

#22 MeSH descriptor: [Community Health Centers] this term only

#23 MeSH descriptor: [General Practitioners] this term only

#24 MeSH descriptor: [Physicians, Family] this term only

#25 MeSH descriptor: [Physicians, Primary Care] this term only

#26 MeSH descriptor: [General Practice] this term only

#27 MeSH descriptor: [Family Practice] this term only

#28 service* near/6 (home or community or outreach or “out-reach” or ambulatory or domicil*) (Word variations have been searched)

#29 therapy near/6 (home or community or outreach or “out-reach” or ambulatory or domicil*) (Word variations have been searched)

#30 treatment* near/6 (home or community or outreach or “out-reach” or ambulatory or domicil*) (Word variations have been searched)

#31 “hospital at home” or “hospital in the home” or “own home*” or “home care” or homecare or “closer to home” (Word variations have been searched)

#32 #14 or #15 or #16 or #17 or #18 or #19 or #20 or #21 or #22 or #23 or #24 or #25 or #26 or #27 or #28 or #29 or #30 or #31

#33 #4 and #13 and #32

#34 MeSH descriptor: [Home Infusion Therapy] this term only

#35 chemotherapy near/6 (home or community or outreach or “out-reach” or ambulatory or domicil*) (Word variations have been searched)

#36 chemotherapy near/6 service* (Word variations have been searched)

#37 chemotherapy near/6 (“general practitioner*” or “family practitioner*” or “family doctor*” or “family physician*” or “primary care physician*”) (Word variations have been searched)

#38 “self-infusion” near/6 home (Word variations have been searched)

#39 “home infusion” (Word variations have been searched)

#40 #34 or #35 or #36 or #37 or #38 or #39

#41 #4 and #40

#42 #33 or #41

Key:

MeSH descriptor = indexing term (MeSH heading)

* = truncation

“ ” = phrase search

near/6 = terms within six words of each other (any order).

Update

Date range: 1990 to October 2013.

Date searched: 29 October 2013.

Records found: 141.

Search strategy

# 1 266,359 Topic=(cancer* or neoplas* or tumor* or tumour* or malignan* or oncolog* or carcinoma*)

# 2 27,630 Topic=(chemotherapy)

# 3 625 Topic=(“systemic therapy”)

# 4 3 Topic=(“intravenous drug therapy”)

# 5 1960 Topic=(“adjuvant therapy”)

# 6 29,227 #5 OR #4 OR #3 OR #2

# 7 5659 Topic=((service* or therapy or treatment*) NEAR/6 (home or community or outreach or out-reach or ambulatory or domicil*))

# 8 1213 Topic=(“hospital at home” or “hospital in the home” or “own home*” or “home care” or homecare or “closer to home”)

# 9 6636 #8 OR #7

# 10 61 #9 AND #6 AND #1

# 11 119 Topic=(chemotherapy NEAR/6 (home or community or outreach or “out-reach” or ambulatory or domicil*))

# 12 13 Topic=(chemotherapy NEAR/6 service*)

# 13 2 Topic=(chemotherapy NEAR/6 (“general practitioner” or “family practitioner*” or “family doctor*” or “family physician*” or “primary care physician*”))

# 14 2 Topic=(“self-infusion” NEAR/6 home)

# 15 11 Topic=(“home infusion”)

# 16 146 #15 OR #14 OR #13 OR #12 OR #11

# 17 91 #16 AND #1

# 18 135 #17 OR #10

Key:

TS= topic tag; searches terms in Title, Abstract, Author Keywords and Keywords Plus fields

* = truncation

“ ” = phrase search

NEAR/6 = terms within six words of each other (any order).

Search strategy

chemotherapy AND home 254 - none relevant

‘ambulatory chemotherapy’ 0

‘chemotherapy in the community’ 1 - none relevant

‘domiciliary chemotherapy’ 0

‘home intravenous therapy’ 0

‘self-infusion at home’ 0

‘home infusion’ 6 - none relevant

‘outreach chemotherapy’ 0

‘out-reach chemotherapy’ 0

chemotherapy AND ‘general practice’ 6 - none relevant

chemotherapy AND ‘primary care’ 66 - none relevant

Key:

‘ ’ = phrase search

Search strategy

S1 47073 (CANCER? OR NEOPLAS? OR TUMOR OR TUMORS OR TUMOUR OR TUMOURS OR MALIGNANT OR MALIGNANCY OR ONCOLOGY OR CARCINOMA?)

S2 3395 CHEMOTHERAPY

S3 54 SYSTEMIC(W)THERAPY

S4 2 INTRAVENOUS(W)DRUG(W)THERAPY

S5 119 ADJUVANT(W)THERAPY

S6 3523 S2 OR S3 OR S4 OR S5

S7 9903 (SERVICE? OR THERAPY OR TREATMENT?)(6N)(HOME OR COMMUNITY - OR OUTREACH OR OUT(W)REACH OR AMBULATORY OR DOMICIL?)

S8 1689 HOSPITAL(W)AT(W)HOME OR HOSPITAL(W)IN(W)THE(W)HOME OR OWN(- W)HOME? OR HOME(W)CARE OR HOMECARE OR CLOSER(W)TO(W)HOME

S9 11189 S7 OR S8

S10 11 S1 AND S6 AND S9

S11 18 CHEMOTHERAPY(6N)(HOME OR COMMUNITY OR OUTREACH OR OUT(W)REACH OR AMBULATORY OR DOMICIL?)

S12 3 CHEMOTHERAPY(6N)SERVICE?

S13 0 CHEMOTHERAPY(6N)(GENERAL(W)PRACTITIONER? OR FAMILY(W)PRACTITIONER? OR FAMILY(W)DOCTOR? OR PRIMARY(W)CARE(W)PHYSICIAN?)

S14 0 SELF(W)INFUSION(6N)HOME

S15 6 HOME(W)INFUSION

S16 27 S11 OR S12 OR S13 OR S14 OR S15

S17 16 S1 AND S16

S18 24 S10 OR S17

Key:

? = truncation

(W) = terms adjacent to each other (same order)

(6N) = terms within 6 words of each other (any order).

Search strategy

1. (cancer$ or neoplas$ or tumor$ or tumour$ or malignan$ or oncolog$ or carcinoma$).ti,ab. (1014)

2. chemotherapy.ti,ab. (50)

3. systemic therapy.ti,ab. (3)

4. intravenous drug therapy.ti,ab. (0)

5. adjuvant therapy.ti,ab. (2)

6. or/2-5 (54)

7. ((service$ or therapy or treatment$) adj6 (home or community or outreach or out-reach or ambulatory or domicil$)).ti,ab. (893)

8. (hospital at home or hospital in the home or own home$ or home care or homecare or closer to home).ti,ab. (345)

9. 7 or 8 (1175)

10. 1 and 6 and 9 (1)

11. (chemotherapy adj6 (home or community or outreach or out-reach or ambulatory or domicil$)).ti,ab. (1)

12. (chemotherapy adj6 service$).ti,ab. (0)

13. (chemotherapy adj6 (general practitioner$ or family practitioner$ or family doctor$ or family physician$ or primary care physician$)).ti,ab. (0)

14. (self-infusion adj6 home).ti,ab. (0)

15. home infusion.ti,ab. (1)

16. or/11-15 (2)

17. 1 and 16 (1)

18. 10 or 17 (1)

Key:

$ = truncation

.ti,ab. = terms in either title or abstract fields

adj6 = terms within six words of each other (any order).

Update

Date range: 1974 to 28 October 2013.

Date searched: 29 October 2013.

Records found: 2940.

Search strategy

1. exp neoplasm/ (3,205,629)

2. (cancer$ or neoplas$ or tumor$ or tumour$ or malignan$ or oncolog$ or carcinoma$).ti,ab. (2,642,872)

3. exp oncology nursing/ (5654)

4. or/1-3 (3,707,658)

5. chemotherapy/ or adjuvant chemotherapy/ or cancer chemotherapy/ or combination chemotherapy/ or consolidation chemotherapy/ or induction chemotherapy/ or maintenance chemotherapy/ or multimodal chemotherapy/ (252,289)

6. antineoplastic agent/ (211,787)

7. chemotherapy.ti,ab. (314,754)

8. systemic therapy.ti,ab. (8828)

9. intravenous drug therapy.ti,ab. (53)

10. adjuvant therapy.ti,ab. (20,039)

11. or/5-10 (580,292)

12. home care/ (45,243)

13. *outpatient/ (6050)

14. *ambulatory care/ or *ambulatory care nursing/ (11,954)

15. *outpatient department/ (11,572)

16. community care/ or community health nursing/ (68,672)

17. general practitioner/ or general practice/ (115,882)

18. ((service$ or therapy or treatment$) adj6 (home or community or outreach or out-reach or ambulatory or domicil$)).ti,ab. (54,756)

19. (hospital at home or hospital in the home or own home$ or home care or homecare or closer to home).ti,ab. (18,147)

20. or/12-19 (289,243)

21. 4 and 11 and 20 (2138)

22. home intravenous therapy/ (17)

23. (chemotherapy adj6 (home or community or outreach or out-reach or ambulatory or domicil$)).ti,ab. (1000)

24. (chemotherapy adj6 service$).ti,ab. (254)

25. (chemotherapy adj6 (general practitioner$ or family practitioner$ or family doctor$ or family physician$ or primary care physician$)).ti,ab. (28)

26. (self-infusion adj6 home).ti,ab. (27)

27. home infusion.ti,ab. (350)

28. or/22-27 (1624)

29. 4 and 28 (954)

30. 21 or 29 (2733)

31. animals/ or nonhumans/ (1,821,160)

32. humans/ (14178823)

33. 31 not (31 and 32) (1,363,257)

34. 30 not 33 (2719)

Key:

/ = indexing term (EMTREE heading)

* = focussed EMTREE heading

exp = exploded EMTREE heading

$ = truncation

.ti,ab. = terms in either title or abstract fields

adj6 = terms within six words of each other (any order).

Search strategy

(“home infusion therapy” AND chemotherapy AND (cancer OR neoplasms)) filetype:pdf = 21,000 hits

(“home chemotherapy” AND (cancer OR neoplasms)) filetype:pdf = 1160 hits

(“chemotherapy at home” AND (cancer OR neoplasms)) filetype:pdf = 4360 hits

(“outreach chemotherapy” AND (cancer OR neoplasms)) filetype:pdf = 177 hits

(“ambulatory chemotherapy” AND (cancer OR neoplasms)) filetype:pdf = 1190 hits

(“hospital at home” AND chemotherapy AND (cancer OR neoplasms)) filetype:pdf = 27,000 hits

(“closer to home” AND chemotherapy AND (cancer OR neoplasms)) filetype:pdf = 22,800 hits

Search strategy

1. (cancer$ or neoplas$ or tumor$ or tumour$ or malignan$ or oncolog$ or carcinoma$).ti,ab. (12,171)

2. chemotherapy.ti,ab. (542)

3. systemic therapy.ti,ab. (9)

4. intravenous drug therapy.ti,ab. (1)

5. adjuvant therapy.ti,ab. (35)

6. or/2-5 (579)

7. ((service$ or therapy or treatment$) adj6 (home or community or outreach or out-reach or ambulatory or domicil$)).ti,ab. (9283)

8. (hospital at home or hospital in the home or own home$ or home care or homecare or closer to home).ti,ab. (2469)

9. 7 or 8 (10,981)

10. 1 and 6 and 9 (24)

11. (chemotherapy adj6 (home or community or outreach or out-reach or ambulatory or domicil$)).ti,ab. (30)

12. (chemotherapy adj6 service$).ti,ab. (37)

13. (chemotherapy adj6 (general practitioner$ or family practitioner$ or family doctor$ or family physician$ or primary care physician$)).ti,ab. (0)

14. (self-infusion adj6 home).ti,ab. (0)

15. home infusion.ti,ab. (0)

16. or/11-15 (55)

17. 1 and 16 (35)

18. 10 or 17 (44)

Key:

$ = truncation

.ti,ab. = terms in either title or abstract fields

adj6 = terms within six words of each other (any order).

Search strategy

S1 113738 (CANCER? OR NEOPLAS? OR TUMOR OR TUMORS OR TUMOUR OR TUMOURS OR MALIGNANT OR MALIGNANCY OR ONCOLOGY OR CARCINOMA?)

S2 10350 CHEMOTHERAPY

S3 97 SYSTEMIC(W)THERAPY

S4 0 INTRAVENOUS(W)DRUG(W)THERAPY

S5 408 ADJUVANT(W)THERAPY

S6 10815 S2 OR S3 OR S4 OR S5

S7 2387 (SERVICE? OR THERAPY OR TREATMENT?)(6N)(HOME OR COMMUNITY - OR OUTREACH OR OUT(W)REACH OR AMBULATORY OR DOMICIL?)

S8 656 HOSPITAL(W)AT(W)HOME OR HOSPITAL(W)IN(W)THE(W)HOME OR OWN(- W)HOME? OR HOME(W)CARE OR HOMECARE OR CLOSER(W)TO(W)HOME

S9 2938 S7 OR S8

S10 12 S1 AND S6 AND S9

S11 24 CHEMOTHERAPY(6N)(HOME OR COMMUNITY OR OUTREACH OR OUT(W)REACH OR AMBULATORY OR DOMICIL?)

S12 3 CHEMOTHERAPY(6N)SERVICE?

S13 0 CHEMOTHERAPY(6N)(GENERAL(W)PRACTITIONER? OR FAMILY(W)PRACTITIONER? OR FAMILY(W)DOCTOR? OR PRIMARY(W)CARE(W)PHYSICIAN?)

S14 2 SELF(W)INFUSION(6N)HOME

S15 12 HOME(W)INFUSION

S16 40 S11 OR S12 OR S13 OR S14 OR S15

S17 18 S1 AND S16

S18 25 S10 OR S17

Key:

? = truncation

(W) = terms adjacent to each other (same order)

(6N) = terms within 6 words of each other (any order).

Search strategy

1. (cancer$ or neoplas$ or tumor$ or tumour$ or malignan$ or oncolog$ or carcinoma$).ti,ab. (63,025)

2. chemotherapy.ti,ab. (2100)

3. systemic therapy.ti,ab. (40)

4. intravenous drug therapy.ti,ab. (0)

5. adjuvant therapy.ti,ab. (54)

6. or/2-5 (2176)

7. ((service$ or therapy or treatment$) adj6 (home or community or outreach or out-reach or ambulatory or domicil$)).ti,ab. (5829)

8. (hospital at home or hospital in the home or own home$ or home care or homecare or closer to home).ti,ab. (1339)

9. 7 or 8 (6986)

10. 1 and 6 and 9 (3)

11. (chemotherapy adj6 (home or community or outreach or out-reach or ambulatory or domicil$)).ti,ab. (4)

12. (chemotherapy adj6 service$).ti,ab. (2)

13. (chemotherapy adj6 (general practitioner$ or family practitioner$ or family doctor$ or family physician$ or primary care physician$)).ti,ab. (0)

14. (self-infusion adj6 home).ti,ab. (0)

15. home infusion.ti,ab. (4)

16. or/11-15 (9)

17. 1 and 16 (3)

18. 10 or 17 (4)

Key:

$ = truncation

.ti,ab. = terms in either title or abstract fields

adj6 = terms within six words of each other (any order).

Search strategy

chemotherapy AND home 52 - 19 relevant

‘ambulatory chemotherapy’ 2

‘chemotherapy in the community’ 0

‘domiciliary chemotherapy’ 1

‘home intravenous therapy’ 3 - 2 relevant

‘self-infusion at home’ 0

‘home infusion’ 13 - 5 relevant

‘outreach chemotherapy’ 0

‘out-reach chemotherapy’ 0

chemotherapy AND ‘general practice’ 3 - 2 relevant

chemotherapy AND ‘primary care’ 12 - 1 relevant

Key:

‘ ’ = phrase search.

Search strategy

1. (cancer$ or neoplas$ or tumor$ or tumour$ or malignan$ or oncolog$ or carcinoma$).ti,ab. (46,619)

2. chemotherapy.ti,ab. (3347)

3. systemic therapy.ti,ab. (369)

4. intravenous drug therapy.ti,ab. (0)

5. adjuvant therapy.ti,ab. (246)

6. or/2-5 (3906)

7. ((service$ or therapy or treatment$) adj6 (home or community or outreach or out-reach or ambulatory or domicil$)).ti,ab. (24,037)

8. (hospital at home or hospital in the home or own home$ or home care or homecare or closer to home).ti,ab. (4691)

9. 7 or 8 (27,793)

10. 1 and 6 and 9 (46)

11. (chemotherapy adj6 (home or community or outreach or out-reach or ambulatory or domicil$)).ti,ab. (42)

12. (chemotherapy adj6 service$).ti,ab. (17)

13. (chemotherapy adj6 (general practitioner$ or family practitioner$ or family doctor$ or family physician$ or primary care physician$)).ti,ab. (2)

14. (self-infusion adj6 home).ti,ab. (1)

15. home infusion.ti,ab. (6)

16. or/11-15 (61)

17. 1 and 16 (44)

18. 10 or 17 (72)

Key:

$ = truncation

.ti,ab. = terms in either title or abstract fields

adj6 = terms within six words of each other (any order).

Update

Date range: all dates to 29 October 2012.

Date searched: 29 October 2013.

Records found: 1007.

Search strategy

#1 Search neoplasms[MeSH Terms] 2,415,714

#2 Search ((((((cancer*[Title/Abstract]) OR neoplas*[Title/Abstract]) OR tumor*[Title/Abstract]) OR tumour*[Title/Abstract]) OR malignan*[Title/Abstract]) OR oncolog*[Title/Abstract]) OR carcinoma*[Title/Abstract] 2,032,899

#3 Search oncologic nursing[mh:noexp] 6072

#4 Search ((#1) OR #2) OR #3 2,923,865

#5 Search drug therapy[mh:noexp] 33,188

#6 Search antineoplastic combined chemotherapy protocols[mh:noexp] 96,554

#7 Search ((chemotherapy, adjuvant[mh:noexp]) OR consolidation chemotherapy[mh:noexp]) OR maintenance chemotherapy[mh:noexp] 27,399

#8 Search infusion, intravenous[mh:noexp] 45,706

#9 Search chemotherapy[Title/Abstract] 225,942

#10 Search “systemic therapy”[Title/Abstract] 5969

#11 Search “intravenous drug therapy”[Title/Abstract] 40

#12 Search “adjuvant therapy”[Title/Abstract] 14,837

#13 Search (((((((#5) OR #6) OR #7) OR #8) OR #9) OR #10) OR #11) OR #12 359,727

#14 Search (home care services[mh:noexp]) OR home care services, hospital based[mh:noexp] 27,991

#15 Search outpatients[majr:noexp] 2120

#16 Search ambulatory care[majr:noexp] 14,543

#17 Search (ambulatory care facilities[majr:noexp]) OR outpatient clinics, hospital[majr:noexp] 13,413

#18 Search ((community health services[mh:noexp]) OR community health nursing[mh:noexp]) OR community health centers[mh:noexp] 47,623

#19 Search ((general practitioners[mh:noexp]) OR physicians, family[mh:noexp]) OR physicians, primary care[mh:noexp] 16,302

#20 Search (general practice[mh:noexp]) OR family practice[mh:noexp] 60,996

#21 Search (((((((((((((((((“home service*”[Title/Abstract]) OR “home therapy”[Title/Abstract]) OR “home treatment*”[Title/Abstract]) OR “community service”[Title/Abstract]) OR “community therapy”[Title/Abstract]) OR “community treatment”[Title/Abstract]) OR “outreach service*”[Title/Abstract]) OR “outreach therapy”[Title/Abstract]) OR “outreach treatment*”[Title/Abstract]) OR “out-reach service*”[Title/Abstract]) OR “out-reach therapy”[Title/Abstract]) OR “out-reach treatment*”[Title/Abstract]) OR “ambulatory service*”[Title/Abstract]) OR “ambulatory therapy”[Title/Abstract]) OR “ambulatory treatment*”[Title/Abstract]) OR “domicil* service*”[Title/Abstract]) OR “domicil* therapy”[Title/Abstract]) OR “domicil* treatment*”[Title/Abstract] 5808

#22 Search (((((“hospital at home”[Title/Abstract]) OR “hospital in the home”[Title/Abstract]) OR “own home*”[Title/Abstract]) OR “home care”[Title/Abstract]) OR homecare[Title/Abstract]) OR “closer to home”[Title/Abstract] 14,748

#23 Search ((((((((#14) OR #15) OR #16) OR #17) OR #18) OR #19) OR #20) OR #21) OR #22 180,529

#24 Search ((#4) AND #13) AND #23 845

#25 Search home infusion therapy[mh:noexp] 580

#26 Search (((((“home chemotherapy”[Title/Abstract]) OR “community chemotherapy”[Title/Abstract]) OR “outreach chemotherapy”[Title/Abstract]) OR “out-reach chemotherapy”[Title/Abstract]) OR “ambulatory chemotherapy”[Title/Abstract]) OR “domicil* chemotherapy”[Title/Abstract] 171

#27 Search “chemotherapy service*”[Title/Abstract] 11

#28 Search ((((“general practitioner chemotherapy”[Title/Abstract]) OR “family practitioner chemotherapy”[Title/Abstract]) OR “family doctor chemotherapy”[Title/Abstract]) OR “family physician chemotherapy”[Title/Abstract]) OR “primary care physician chemotherapy”[Title/Abstract] 0

#29 Search “self-infusion at home”[Title/Abstract] 0

#30 Search “home infusion”[Title/Abstract] 259

#31 Search (((((#25) OR #26) OR #27) OR #28) OR #29) OR #30 890

#32 Search (#4) AND #31 238

#33 Search (#24) OR #32 975

Key:

[MeSH Terms] = indexing term (MeSH heading)

[mh:noexp] = non-exploded MeSH heading

[Title/Abstract] = terms in either title or abstract fields

“ ” = phrase search

* = truncation.

Search strategy

1. (cancer$ or neoplas$ or tumor$ or tumour$ or malignan$ or oncolog$ or carcinoma$).ti,ab. (1854)

2. chemotherapy.ti,ab. (42)

3. systemic therapy.ti,ab. (58)

4. intravenous drug therapy.ti,ab. (0)

5. adjuvant therapy.ti,ab. (0)

6. or/2-5 (100)

7. ((service$ or therapy or treatment$) adj6 (home or community or outreach or out-reach or ambulatory or domicil$)).ti,ab. (10,540)

8. (hospital at home or hospital in the home or own home$ or home care or homecare or closer to home).ti,ab. (4276)

9. 7 or 8 (13,575)

10. 1 and 6 and 9 (0)

11. (chemotherapy adj6 (home or community or outreach or out-reach or ambulatory or domicil$)).ti,ab. (2)

12. (chemotherapy adj6 service$).ti,ab. (3)

13. (chemotherapy adj6 (general practitioner$ or family practitioner$ or family doctor$ or family physician$ or primary care physician$)).ti,ab. (0)

14. (self-infusion adj6 home).ti,ab. (0)

15. home infusion.ti,ab. (0)

16. or/11-15 (4)

17. 1 and 16 (3)

18. 10 or 17 (3)

Key:

$ = truncation

.ti,ab. = terms in either title or abstract fields

adj6 = terms within six words of each other (any order).