Table 3.

Summary of included studies.

Author	Country	Participants	Aims	Research methods	Key findings	Weight of evidence A + B + C = D
Faull et al.3	United Kingdom	63 healthcare professionals working in one county: 22 GPs; 16 community nurses; 3 community pharmacists; 1 student Nurse; 4 community palliative care nurses; 17 community specialist nurses	To explore the issues that arise for practitioners working in the community, in relation to anticipatory prescribing for terminally ill patients who wish to die at home	Qualitative interviews and focus groups. Qualitative analysis	Participants valued the principle of anticipatory prescribing Decisions on when to prescribe more of an issue when non-cancer diagnosis It was uncommon to hear accounts of getting drugs in the home more than a day or two ahead of anticipated need Barriers to prescribing: potential drug waste, not knowing the patient well enough, concerns around prescriber accountability (especially opioids), situations where there may be drug misuse, and not knowing or trusting other professionals’ judgements Facilitators to prescribing: having known the patient for some time. Good communication between professionals	H H H – H
Wilson et al.19	United Kingdom	61 nurses working in two regions: 16 nursing home nurses; 27 community nurses; 18 community palliative care nurses 83 episodes of observations across 4 nursing homes and 4 community teams	To examine nurses’ decisions, aims, and concerns when using anticipatory medications	Ethnographic study using observations and qualitative interviews. Qualitative analysis	The aim expressed by nurses when using anticipatory medications was to ‘comfort’ and ‘settle’ dying patients and prevent admissions to hospital Nurses would only administer medication if symptoms that were both irreversible and due to entry into the dying phase, the patient consented (where possible) and was unable to take oral medication, and decisions were made independent of a patient’s relatives influence Nurses often worked in pairs to check prescriptions and aid decision-making Administering the medication raised a number of concerns: distinguish between pain and agitation so as to administer the most appropriate drug; not wanting to instigate administering drugs too soon; and balancing the risks of under-medicating against concerns about over-medicating and causing unwanted side effects Less experienced nurses expressed concerns about whether medications to control pain, particularly opioids, and symptoms hasten death. Concerns re ‘last injection’	H H H – H
Bowers and Redsell24	United Kingdom	11 nurses working in one county: 7 community palliative care nurses; 4 community nurses	To explore community nurses’ decision-making processes around the prescribing of anticipatory prescribing for people who are dying	Qualitative interviews. Qualitative analysis	Anticipatory medications represent a safety net and give nurses a sense of control in managing an individual’s last days of life symptom Management nurses felt that it was important to have medications to cover out-of-hour periods Nurses requested GPs prescribed drugs and negotiated with them over what drugs to prescribe facilitators to prescribing: keeping GPs up to date with the patient’s changing condition, good multidisciplinary communication, established relationship of mutual trust with GPs Barriers to prescribing: difficult to accurately predict when patients were likely to die; Some GPs worried that medications might be used inappropriately; Some GPs lacked up to date end-of-life drug knowledge and needed persuading to prescribe for all likely terminal symptoms	H M H – H
Rosenberg et al.22	Australia	18 family carers in one city	To examine the experiences of family caregivers supporting a dying person in the home setting, with particular regard to being supplied with an anticipatory prescribing kit	Qualitative interviews. Qualitative analysis	Patients are issued with anticipatory prescribing kits, and family carers are asked to administer injectable medications The introduction of the kit was viewed positively by most family carers Family carers found it reassuring that the kit improved accessibility should symptoms become difficult to control Some family carers were reluctant to give the medication and looked to nurses to administer drugs The expectation to administer medication was overwhelming and intimidating for some family carers	M H M – M
Finucane et al.38	United Kingdom	71 patients who died in eight nursing homes	To investigate the extent of anticipatory prescribing for residents who died in nursing homes in Lothian, Scotland	Retrospective notes review. Descriptive statistics	54% of residents who died in the nursing homes had a prescription for at least one anticipatory medicine 15% of residents had anticipatory prescriptions in place for all four common symptoms at the end of life There was great variation in anticipatory prescribing across the nursing homes: 100% of patients died with drugs prescribed in one nursing home compared with only 13% in another.	M M H – M
Perkins et al.5	United Kingdom	110 patients and 66 nurses and care staff in eleven nursing homes	To assess the impact of the Liverpool Care Pathway (LCP) on care in nursing homes and intensive care units	Mixed methods: retrospective case note review; 8 observations, linked with case note analysis; qualitative interviews with staff. Thematic analysis	Usually, when nursing home staff identified patients as being ‘weeks from death,’ they would request anticipatory prescribing Anticipatory prescribing was seen as a solution to problems with gaining timely medical input out of hours and avoidance of hospital admissions There was a strong emphasis in the nursing homes on being prepared for a patient’s death: anticipatory prescribing was viewed as essential Barriers to prescribing: GPs perceptions of the cost of wasted drugs; getting a timely review of the patient by the GP Most anticipatory medications went unused The administration of drugs often left nurses feeling uncomfortable, particularly if the patient died soon after their administration	H M M – M
Wilson and Seymour37	United Kingdom	72 healthcare professionals working in two regions: 61 nurses; 8 GPs; and 3 community pharmacists 83 episodes of observations	Aim not stated – reporting on a theme from a wider piece of research19	Ethnographic study using observations and qualitative interviews. Qualitative analysis	Nurses often initiated conversations with GPs about getting anticipatory prescribing in place. GPs were happy to take this advice Nursing participants reported that a small number of GPs were reluctant to prescribe anticipatory medications Barriers to prescribing: GPs did not regularly prescribe end-of-life drugs and lacked the confidence to do so without guidance, some nurses felt their expertise was not valued by GPs Facilitators to prescribing: trust, valuing each other’s knowledge and expertise, access to each other, and clarification of professional responsibilities comprise a central component of successful anticipatory prescribing	M H M – M
Brand et al.39	United Kingdom	12 healthcare professionals in one county: disciplines not stated	To explore the viewpoints of healthcare professionals involved in anticipatory prescribing in care homes	Qualitative interviews. Qualitative Thematic analysis analysis	Uncertainties surrounding when anticipatory prescribing should be initiated often results in residents not having drugs available until after symptoms appear Perception that anticipatory prescribing may reduce hospital admissions and provides symptom control Facilitators to prescribing: trusting relationships between professionals; good interdisciplinary communication	M M M – M
Brewerton et al.40	United Kingdom	150 patients accessing one community specialist palliative care service	To understand the current practice of anticipatory prescribing for patients referred to a community specialist palliative care service	Retrospective notes review. Descriptive statistics	63% had anticipatory prescribing. 55 of 100 patients with a cancer diagnosis had drugs in place verses 39 of 50 patients with a non-cancer diagnosis The median length of time from requesting anticipatory prescribing to death was 18 days 74 out of 97 patients who died in their preferred place of death had anticipatory prescribing	M M M – M
Griggs41	United Kingdom	17 community nurses within one county	To gain an insight into perceptions of a ‘good death’ among community nurses and to identify its central components	Qualitative interviews. Qualitative analysis	Nurses felt it was important to have drugs available ahead of need in homes Nurses were relied upon by GPs to recommend palliative drugs. Some Nurses did not like this responsibility Barriers to prescribing: perception that GPs reluctant to prescribe medications especially during out-of-hours periods	M M M – M
Israel et al.23	Australia	14 family caregivers in once city.	To investigate family caregivers perceptions of administering subcutaneous medications	Qualitative interviews. Qualitative analysis	All the family carers administered injectable anticipatory medications for at least 7 days Family carers felt they had no option but to give injections if their family member was to be cared for at home All placed a high value on the ability to contribute immediately to symptom control needs If symptoms were not controlled following injections, family carers felt disempowered and distressed Family carers expressed concern and uncertainty over timings of injections and feared causing medication overdose 2 family carers were concerned about the possibility of administering the ‘last injection’	M H L – M
Harris et al.35	United Kingdom	11 nurses from two different palliative care units and two head and neck wards: including 3 specialist palliative care nurses working in the community	To evaluate the utility of crisis medication in the management of terminal haemorrhage, through the experiences of nurses who have managed such events	Qualitative interviews. Interpretative phenomenological analysis	Participants’ experiences suggested that crisis medication had served little, if any, useful role in the management of terminal haemorrhage Challenging to know when to administer drugs and if events are reversible until it is too late	M H L – M
Harris et al.36	United Kingdom	8 nurses working in palliative care or head and neck setting	To explore nurse’s experiences of the role of crisis medication in the management of terminal haemorrhage in patients with advanced cancer	Qualitative interviews. Thematic analysis	Terminal haemorrhage is a rapid event and there is often no time for crisis medication to be given or take effect. Determining whether to give crisis medication is challenging and raises anxiety Nurses feel reassured to have medication prescribed even if it may not be used or has time to take effect	M H L – M
Kemp et al.42	United Kingdom	Patients registered with 12 GP surgeries in one county	To evaluate the prevalence and impact of anticipatory prescribing on home death/utilisation of healthcare in the last month of life	Retrospective case note review. Statistical analysis	Anticipatory prescribing was in place for 16% of predictable deaths in a 1-year period: levels of usage varied widely between GP surgeries Patients living at home were less likely to have drugs prescribed than those in care homes The use of anticipatory prescribing was associated with an increased chance in home death – however, a causal association was not demonstrated Anticipatory prescribing use was also associated with decreased risk of hospitalisation in last month of life, and increased GP contact in both care home and community residents	M M M – M
Owen et al.43	United Kingdom	550 patients who died in 19 nursing homes	Review of care since the GP surgery–based MDT took over medical and pharmacological care of the nursing homes	Retrospective notes review. Statistical analysis	Anticipatory prescribing frequency varied across the nursing homes: 3 nursing homes had it in place for 62% of deaths, and 3 nursing homes had it in place for only 28% of deaths Less than a third of patients who were prescribed drugs had them administered Midazolam and morphine were the most commonly used medications There was a clear correlation (r2 = 0.64) between the proportion of patients prescribed anticipatory medications and the proportion of patients dying at the home instead of in hospital There was no correlation between administration of anticipatory medications and place of death	M M M – M
Wilson et al.20	United Kingdom	575 nurses working in two regions: 231 nursing home nurses; 151 palliative care nurses; 193 district nurses	To gain insight into the roles and experiences of a wide range of community nurses in end-of-life medication decisions	Staff survey. Descriptive statistics. Thematic analysis of free-text comments	Responses suggest anticipatory prescribing is a widespread practice Where patients’ age categories were reported (n = 412), 63.8% (n = 263) were said to be aged 70 or over A primary cause of death was provided for 434 patient cases and in 79.3% of these, cancer was reported by nurses as the registered cause of death Decision to prescribe often dictated by the nurses rather than the GP Facilitators to prescribing: nurses reported working well with GPs and perceived that they had good access to the medications needed; 79.2% of nurses reported that they ‘infrequently or never’ found doctors reluctant to prescribe anticipatory medication Barriers to prescribing: anticipatory prescriptions being incorrectly written up by doctors; 8.6% of nurses said they ‘always or frequently’ experienced significant difficulties in obtaining the anticipatory drugs Nurses reported that the anticipatory medications successfully controlled those symptoms they were intended to relieve in 89.6% of the patient cases they recalled Midazolam was the drug most commonly reported to have been used in the last month of the patient’s life Nurses felt they were responsible for assessing the patient’s response to drugs	M M M – M
Addicott44	United Kingdom	11 healthcare professionals working in two surgeries: 8 GPs; 1 practice nurse; 2 community nurses	To identify challenges and examples of good practice in providing good-quality end-of-life care in general practice	Case study using qualitative interviews. Qualitative analysis	GPs happy to prescribe anticipatory drugs to cover out of hours periods Prescribing considered a significant responsibility as accountable for use/misuse Concerns around large amounts of medication left in home without supervision	L H L – M
Amass and Allen6	United Kingdom	23 patients in the community across one region	To evaluate an anticipatory medication pilot	Audit of care. Descriptive statistics	23 anticipatory prescribing kits issued and 16 (70%) were used The intervention was well received by nurses, patients, and carers None of the 16 patients required admission to a hospital or hospice for end-of-life symptom control The net cost of wasted medicines was about £10 per patient	L M M – M
Ashton et al.33	United Kingdom	13 care staff working in four care homes and one NHS mental health ward	To assess the effects of the Gold Standards Framework and LPC on the experience of staff	Qualitative focus group. Analysis not stated	Staff acknowledged the difficulties for GPs in anticipatory prescribing, particularly relating to: pain management, the experience of the GP and their understanding of advanced dementia, the reluctance to prescribe diamorphine Staff felt these difficulties would resolve as the GP developed a trusting relationship with them	L H L – M
Ashton et al.34	United Kingdom	200 healthcare professionals working in four care homes and one NHS mental health ward	To determine the effects of introducing Gold Standards Framework and LCP from the perspectives of staff involved in the care of older people with dementia	Case study using mixed methods: interviews, focus groups, survey of staff. Analysis not stated	Anticipatory prescribing was viewed as a key element in the management of pain and other distressing symptoms	L H L – M
Bullen et al.45	Australia	8 community palliative care nurses. 43 community palliative care services	To conduct a survey of a local service to examine views on medication management before and after the implementation of an anticipatory prescribing kit and to conduct a nationwide prevalence survey examining the use of anticipatory prescribing kits	Quantitative single-arm intervention study with pre- and post-questionnaires in a community specialist palliative care service. Nationwide prevalence survey of the use of anticipatory prescribing kits in Australia	88% of nurses reported the implementation of the anticipatory prescribing kits had improved patient outcomes The administration of anticipatory medications was highly variable and usually occurred when the patient entered a deteriorating or terminal phase of care In the majority of instances where kits were used, the medications were perceived to have met patients’ needs The majority of services surveyed reported they did not use anticipatory prescribing kits Most participants from services who did not utilise anticipatory prescribing kits believed that they could improve patient care Having access to the low-cost kits was perceived to help avoid unnecessary crisis hospital admissions	L M M – M
Harris and Nobel46	United Kingdom	152 community, hospice and hospital palliative care teams across the United Kingdom	To explore current practice in the management of terminal haemorrhage by palliative care teams in the United Kingdom	Survey with open and closed questions. Descriptive statistics	Midazolam was the most commonly used crisis medication although there is a large variation in the dose of this and other drugs used Unclear role of crisis medication, as patients often die before they can be given or take effect	M M L – M
Kinley et al.47	United Kingdom	319 residents who died in 38 nursing homes taking part in an end-of-life programme	To identify the prescribing practice for symptom control in the last month of life for residents dying in nursing homes	Retrospective notes review. Descriptive statistics	37% of residents had anticipatory prescribing in place at the time of death	M M L – M
Lawton et al.48	United Kingdom	58 community nursing teams in one county	To audit staff awareness of an anticipatory prescribing scheme	Audit of practice. Descriptive statistics. Grouping of free-text comments received	The majority of patients issued drugs were diagnosed with a malignancy (n = 43) Difficulty in predicting right time to prescribe anticipatory medications Having prescribed medication available in the home was perceived as reassuring for families Barriers to prescribing: patient not wanting drugs in home; professionals not thinking about anticipatory prescribing; GPs declining to consider anticipatory prescribing The costs of prescriptions were estimated to be to £22.12 per patient A significant amount of medicines went unused, but 77% of boxes issued had at least one drug used	L M M – M
Wowchuk et al.49	Canada	457 patients in one region	To evaluate the use of a anticipatory prescribing kit	Service evaluation based on complete data collection forms from accessed anticipatory prescribing kits. Statistical analysis	Pilot project issuing 457 patients with anticipatory prescribing kits over a 5-year period Majority of patients in pilot had cancer (8.5% non-malignant) 293 kits were both placed in patients’ homes and accessed The mean survival from the time the kit was open until the time the patient died was 4.54 days Home death rate much higher in those participating in the pilot medication kit scheme compared to the home death rate for the overall programme 79%–88% home death rate for those who used the kit; 60% home death rate for those who had the kit placed but did not use it; 25%–29% home death rate for those not in the pilot	L M M – M
Dale et al.13	United Kingdom	995 surgeries in England and Northern Ireland: those returning baseline and follow up questionnaires	To identify factors associated with the extent of change in processes that occurred in practices in the year following adoption of the Gold Standards Framework	Quantitative uncontrolled observational cohort study with pre-post questionnaire. Statistical analysis	48.9% of surgeries had a procedure for anticipatory prescribing at baseline 82.3% of surgeries had a procedure for anticipatory prescribing a year later	M L L – L
Hardy et al.50	Australia	20 patients in one nursing home (as part of a study looking at four hospitals, three hospices and one nursing home)	To evaluate the care of patients who died in institutes in Queensland	Retrospective notes review. Descriptive statistics	Few Patients in the nursing home were prescribed drugs in anticipation of symptoms (no numbers given)	L M L – L
Healy et al.14	Australia	76 family carer questionnaires. Focus groups with 26 nurses	To evaluate of the effectiveness of an education package that supports laycarers of home-based palliative patients to manage breakthrough subcutaneous medications used for symptom control	Mixed methods: single-arm intervention study with two post-intervention questionnaires for family carers. Focus groups with nurses	In Australia laycarers, mostly family members may be required to administer subcutaneous medications Family carers found the package was useful and enabled them to deal confidently with symptoms arising in the home-based palliative patient Nurses were uncertain in when to train family carers in the patient’s trajectory Contention between nurses on if it is safe or appropriate for family carers to give drugs	L M L – L
Jamal et al.51	United Kingdom	GPs and community nurses (numbers not stated) working in one county	To evaluate the awareness of network guidelines along with the prescribing and usage ratios of anticipatory prescribing kits	Service evaluation. Descriptive statistics	90% of GPs responding indicated that they had prescribed anticipatory prescribing kits 69% of GP’s stated prescribing was influenced by access to anticipatory prescribing information, and 75% stated that levels of confidence impacted on decision-making 55% of GPs respondents indicated that prescribing was influenced by concerns about misuse of drugs 41% of GPs indicated that cost was a factor The recommended network guidelines for 2–3 days’ supply of anticipatory medications costs £30.26 per patient	L M L – L
Lawton et al.52	United Kingdom	181 after death reviews with home staff in 56 nursing homes and 25 care homes	To describe factors that promote a ‘good death’ in care homes	Qualitative interviews. Qualitative analysis.	Nursing home staff felt having anticipatory medications in place gave reassurance to residents, staff, and relatives	L M L – L
Lee et al.53	United Kingdom	5 informal carers in one county	To audit the feasibility of the policy and practice of informal caregivers administering subcutaneous medication	Audit of care. Reporting on informal carers comments	Informal carers gave injectable anticipatory medications, with nurse support and training All informal carers stated that, if required, they would administer subcutaneous injections again to a family member	L M L – L
Mathews and Finch54	United Kingdom	10 patients in one nursing home. Reflective group with nursing staff (number not stated)	To evaluate the impact of implementing the LPC in a nursing home	Audit of patient notes and a reflective group discussion with nurses on implementing the LPC. Analysis methods not stated	GPs prescribe anticipatory medications and nursing home staff judge when to administer drugs Barriers to prescribing: nurses reported GPs reluctant to prescribe diamorphine to opioid naive patients Facilitators to prescribing: GPs familiarity with anticipatory prescribing practice Some nurses worried about administering injectable opioids and felt uneasy when a patient died within hours of an injection	L M L – L
O’Loghlen and Baines55	United Kingdom	295 service evaluation forms from 83 GPs surgeries in one county	To evaluate an anticipatory prescribing scheme	Service evaluation. Descriptive statistics	Perception that the scheme offered peace of mind for patients and relatives The information gathered from the completed forms suggested that 121 admissions to hospital or hospice were prevented	M L L – L
Lee and Headland56	United Kingdom	2 patients in one county	To report on the feasibility of relatives giving subcutaneous injections	Descriptive case reports from a nurse perspective. Description of care received	Reports on two cases where family carers gave injectable anticipatory medication following training Accounts that the family carers felt this was acceptable and helped with providing effective symptom control at home	L L L – L

Care home: a community residence without trained nurse on site; nursing home: a community residence with trained nurses on site; GP: family doctor; H: high; M: medium; L: low.

Quality of the evidence was assessed using Gough’s Weight of Evidence framework:³⁰ (A) coherence and integrity of the evidence in its own terms; (B) appropriateness of the study design in answering the review questions; (C) relevance of the evidence for answering the review questions; and (D) overall assessment of the quality and relevance of the study, derived by combining judgements (A), (B), and (C).