Appendix 3

Articles included in the review

First author	Citation	Sample	Aim	Research methods used	Key Findings	Weight of Evidence^*
Black22	J Gerontol Soc Work 2004;43:13 1–46	29 social workers in 6 hospitals in upstate New York, US.	To examine social workers’ advance directive communication with hospitalised elderly patients	Questionnaire	Social workers play an active role in advance directive discussions They reported the time they spent was inadequate	A) Medium B) Low C) Low Overall: Low
Black27	Home Health Care Serv Q 2007;26:41–58	27 case managers across Florida, US	To explore case managers perceptions of facilitators and barriers to advance care planning practices.	Focus groups	15/27 case managers indicated clients willingly discussed future care plans Time constraints of case management. Lack of available immediate family. Families were sometimes “in denial”, “unrealistic” “dysfunctional” 16/27 said their proficiency in ACP was an issue. Many said dementia limited communication	A) Medium B) Low C) Low Overall: Low
Bradley10	J Am Geriatr Soc 1998;46:12 35–41.	600 residents admitted between 1990 and 1994 to 6 nursing homes in Connecticut, US	To measure the frequency nursing home residents discuss with clinicians their wishes for future treatment and to assess the influence of the PSDA.	Review of nursing home medical records	28.5% of residents had at least one discussion of future treatment wishes documented This had increased since the implementation of the PSDA from 20.3% in 1990 to 36.7% in 1994).	A) Medium B) Low C) Medium Overall: Medium
Carrese21	Brit Med J 2002;324:1 25–7	20 chronically ill housebound patients over 75 from Baltimore US	To understand how elderly patients think about and approach future illness and the end of life.	Semi-structured interviews.	16/20 were reluctant to think about or plan for the future. Many preferred to “cross that bridge” of decision making only when had to Felt planning couldn’t be successful because of uncertainty over the future Some felt matters to be in “good hands”	A) Medium B) Medium C) Medium Overall: Medium
Clarke34	J Pain Symptom Manag 2010;40:85 7–69	74 people older people, informal caregivers and community group representative s from across UK	To explore older people’s concerns about end-of-life issues.	Focus groups	Only three described having made a living will. One said her family were unwilling to talk Some preferred not to talk to their families about death.	A) Low B) Low C) Low Overall Low
Damato11	N J Med 1993;90:21 5–20.	86 community-dwelling senior citizens from Jersey City, US	To determine knowledge and interest in advance directives and attitudes towards end of life care.	Questionnaire	28% discussed terminal care with family but only 4% put wishes in writing and only 2/86 discussed with their physician 71% indicated desire to prepare instructions in advance	A) Medium B) Low C) Low Overall Low
Froggatt35	Palliative Med 2009;23:33 2–8	213 managers of care homes in North West and South West England	To describe current Advance Care Planning practices in English care homes for older people.	Questionnaire and telephone interview	89% recommended advanced care planning Managers identified barriers to discussions include physical problems (62%), dementia (81%), families (51%), staff confidence (66%) and communication (40%). Challenges implementing wishes include other health professionals (GPs, district nurses) & other hospitals.	A) Medium B) Low C) Low Overall Low
Gamble12	Arch Intern Med 1991;151:2 77–80.	75 elderly people in rural eastern North Carolina	To explore the knowledge, attitudes and behaviour of elderly persons regarding living wills.	Questionnaire	81% would like to discuss end of life care with their doctor but only 11% had. 79% thought these issues should be discussed when a person was well. 20% thoughts these discussions should happen when someone was 50 years.	A) Medium B) Low C) Low Overall: Low
Golden17	Am J Hosp Palliat Care 2009;26:13–7.	1569 home-bound but nursing home eligible older adults in Florida.	To study the prevalence of specific barriers that prevent home-bound older adults from obtaining advance directives.	Interviews	66.2% had an advance directive Of the 530 who didn’t almost 60% of the barriers consisted of the answer “never thought of it”, “chose not to execute” or “do not want to think about a negative topic”.	A) Medium B) Medium C) Low Overall: Medium
Gordon13	Arch Intern Med 1999;159:7 01–4.	5117 people aged over 65 years.	To find the proportion of seniors who had been asked about their end of life care preferences by a clinician & had completed an advance directive	Questionnaire	One third reported having an advance directive on file but only 15% talked with a clinician about end of life care preferences.	A) Medium B) Medium C) Medium Overall: Medium
Johnston28	Arch Intern Med 1995;155:1 025–30.	329 adults age 19 to 94 years 282 resident & 272 practicing physicians at 8 primary care clinics across Eastern and Mid-Western US	To assess the opinions of Primary Care Patients and Physicians on discussions on Advance Directives	Questionnaire	The majority of patients & physicians agreed it is the responsibility of the physician to initiate discussions. 91% agreed that AD should be discussed before patients are extremely ill The patients believed the discussions should occur earlier than the physicians	A) Medium B) Low C) Low Overall: Low
Malcomson14	J Am Acad Nurse Pract 2009;21:18–23.	20 healthy 60–94 year olds Massachusetts US	To explored the perspective of healthy elders to advance care planning.	Focus groups & questionnaire.	Only 20% reported having an advance care planning conversation Participants expressed comfort even enthusiasm for discussing end of life preferences. Healthcare providers should initiate discussions when a patient is relatively well. Barriers identified include time, an emphasis on treatment and cure and a lack of comfort among family, friends & providers.	A) Low B) Low C) Med Overall: Low
Markson23	J Am Geriatr Soc 1997;45:39 9–406.	653 physicians across US	Investigates how much experience physicians have had discussing and following advance preferences and how physicians perceive their role in the advance decision making process	Questionnaire	79% had discussed advance preferences with at least one patient 82% felt helping patients chose advance preferences was an important part of their responsibilities Most didn’t feel conversations were particularly stressful	A) Medium B) Medium C) Medium Overall: Medium
McCarth15	J Geront A-Biol 2008;63:95 1–9.	220 community dwelling elders all over 80 years old US	To describe advance care planning, health care preferences and health perceptions in a very elderly sample.	Interview	69% reported discussing their EOL medical care with someone but only 17% discussed their wishes with a physician or health care provider.	A) Medium B) Medium C) Low Overall:Medium
Moore16	J Gerontol Soc Work 1999;31:21–39.	20 low income community dwelling senior adults in relatively good health age 58–78 years in New York State US	To explore factors that influence an elder’s decision to complete an advance directive	Interview	Only 1/20 had discussed an advance directive with a doctor. Majority would be comfortable if doctor raised an advance directive during a routine visit 8/20 said family reluctant to discuss issues as didn’t want to recognise decline.	A) Low B) Medium C) Low Overall: Low
Morrison24	Arch Intern Med 1994;154:2 311–8.	277 residents and attending physicians at a large New York Hospital, US	To determine the impact of five proposed barriers to physicians using advance directives.	Questionnaire	40% of physicians hadn’t discussed advance directives with a patient in the last month Barriers included physicians questions about appropriateness, their lack of understanding, lack of comfort & time constraint.	A) Medium B) Medium C) Medium Overall: Medium
Palker18	Nurse Pract 1995;20:7–8, 13, 17–8, passim.	104 nursing home residents, South Eastern US	To determine the prevalence of advance directives among residents of a Nursing Home, to identify barriers to documentation and to explore death anxiety.	Review of nursing home records plus interview with 17 residents	52% had as least one advance directive 13/17 didn’t know what an advance directive was & didn’t have one. 4/17 had an advance directive &gave largely practical reasons for preparing it	A) Low B) Low C) Low Overall: Low
Pfeifer29	J Gen Intern Med 1994;9:82–8.	43 primary care physicians and 47 ambulatory patients in 8 cities across US	To identify primary care patients and physicians attitudes to discussions of end of life medical care.	Interviews.	Patient preference for discussions in an honest & straight-forward manner. Holding discussions & making decisions early outweighed any discomfort over content. Physicians felt it was their responsibility to initiate discussions but were uncomfortable with early discussions citing risks to patient’s hope.	A) Medium B) Medium C) Medium Overall: Medium
Samsi31	Health Soc Care Comm 2011;19:52–9.	37 adults over 50 years in UK	To explore experiences, opinions and attitudes of older adults living in community in the context of the Mental Capacity Act.	Interviews	Although a large proportion recognised the benefits of planning they were keen to postpone it until they were older, in worse health or it was “more appropriate” Some were unsure they could rely on their GP, reporting not always seeing the same GP	A) Medium B) Low C) Low Overall: Low
Schonfeld25	Am J Hosp Palliat Care 2012;29:26 0–7.	Primary care physicians at University of Nebraska Medical Centre, Canada	To explore differences between end-of-life conversations with patients /families with multiple co-morbidities versus a single, terminal diagnosis.	Focus groups	46.9% had initiated end of life conversations where diagnosis was unspecified Overwhelming felt EOL conversations more difficult with multiple co-morbidities. Rely on physical and social cues to prompt discussions Lack of clear threshold or prompting event a barrier	A) Medium B) Medium C) Medium Overall: Medium
Seymour26	Soc Sci Med 2004;59:57–68.	32 individuals from older people’s community groups in Sheffield UK	To explore older people’s views on advance statements and the role these might play in end of life care decisions.	Focus groups	Recognition of opportunity advance statements offered to address care & treatment issues before cognitive impairment. Perceived risk of “leaving it too late” Most envisaged problems making decisions for future situations that were difficult to imagine	A) Low B) Low C) Med Overall: Low
Stelter19	Arch Intern Med 1992;152:9 54–9.	214 people over 65 years attending senior centres in Midwest US.	To learn the reasons why so few people had completed living wills	Questionnaire	15% had completed a living will. 66% without one had planned to complete one. 94% believe planning for the future is important. 61% wanted to discuss living wills with a physician 84% reported comfortable talking about the end of life	A) Medium B) Medium C) Low Overall: Low
Stewart20	Age Ageing 2011;40:33 0–4.	Staff & families from 34 homes for older people in London, UK.	To explore views on advance care planning in care homes for older people.	Interviews	Staff felt ACP promoted respect for residents’ wishes and aided their treatment decisions. Discussions should start early, in gradual stages before the onset of serious health problems. Barriers include lack of capacity, unforeseen medical scenarios, and the reluctance of some residents and staff to discuss end of life issues	A) Medium B) Medium C) Medium Overall:Medium
White32	J Am Acad Nurse Pract 2005;17:14–20.	13 new residents at a long-term care facility who had signed an advance directive in Midwest US.	Explore experiences of residents who had signed an advance directive on admission to a long-term care facility and apply author developed model.	Interviews	Sample had signed an advance directive on moving to care facility, 45% had consulted their physician regarding their advance directive. Need time and information to make decision.	A) Low B) Low C) Low Overall: Low
Winland-Brown33	Adv Pract Nurs Q 1998;4:36–40.	17 people over 65 with no formal advance directive Florida, US	To understand older people’s reasons for not having formalized their end-of-life decisions.	Interview	Living in & accepting the present valued over fear of death “If I ever have a terminal illness, I’ll think about advance directives.” “In God’s hands” Confidence family will fulfil wishes	A) Low B) Low C) Medium Overall: Low
Zronek30	JONAS Healthc Law Ethics Regul 1999;1:23–8.	51 people over 60 years with an advance directive prior to hospital admission in Mid-West US	To examine patients’ beliefs and level of understanding of the advance directives they had completed.	Survey	Most felt well informed about advance directives Benefits of advance directives include ensuring wishes will be heard & assisting family in making decisions	A) Medium B) Low C) Low Overall: Low

↵* Gough’s “Weight of evidence criteria”
Papers are assessed on four criteria:
1. Coherence & integrity of the evidence in its own terms
2. Appropriateness of form of evidence for answering review question
3. Relevance of the evidence for answering review question
4. Overall assessment of study contribution to answering review question (low, medium or high)
Criteria 1 involved an attempt to assess the risk of bias within individual studies. The weightings of each paper are shown in the final column of Table 1 with the weighting given for overall assessment of study contribution (criterion 4) in bold.