Skip to main content

Table 4 Studies applying consent processes across the capacity trajectory in serious illness (n = 30 studies reported in 32 publications)

From: Processes of consent in research for adults with impaired mental capacity nearing the end of life: systematic review and transparent expert consultation (MORECare_Capacity statement)

Authors, country, EPOC Year Study design and aim Setting Sample description Consent process for adults with declining, impaired, or lacking capacity No. patients/eligible (%recruited) Key findings, challenges, and solutions
Palliative care
    Abernethy et al. [51], Australia, A3 2006 Cluster RCT 2 × 2 × 2 to test GP educational outreach visiting and case conferencing to improve patient outcomes, e.g. pain management. Community settings Palliative care Assent personal consultee, or nominated consultee, e.g. GP capacity assessment MMSE ≤ 24 indicate require assent. n = 461/607 (76%)
Sample not differentiated by capacity/lack capacity
Key findings: strategies used enabled recruitment of largest community study in palliative care.
Solutions: (1) study design pragmatic 2 × 2 × 2 cluster RCT and methods, broad inclusion criteria, defined recruitment plan, detailed intervention; (2) minimise patient burden, and clinician burden, e.g. research nurses collect data.
Challenges: (1) time and resource complex trial design. (2) 7 patients ineligible no caregiver available/no pain.
    Gardiner et al. [49], UK, B3 2013 Cross-sectional survey to explore palliative care need in hospital and agreement between informants. All hospital inpatient adult ward Palliative care Assent personal consultee ALC (n= 38). Capacity assessment MCA criteria: HCP, family if available. n = 654/1359 (48%) (mean age 78 years) Key findings: 36% patients’ palliative care needs. Low identification medical (16%) and nursing (17%) staff. In 23 cases, consultees completed questionnaires on behalf of patients who lacked capacity to consent, and responses given via consultee may not be accurate.
Challenges: non-consenting patients (n = 582)—(1) patient/consultee declined (e.g. too ill) (n = 407); (2) consultee not contactable (n = 109).
    Rees et al. [54], UK, A3 2003 Feasibility RCT on anti-muscarinic medication (hyoscine versus glycopyrronium) for ‘death rattle’ to develop a process of advance consent enabling research to be undertaken in the terminal phase. Palliative care ward Palliative care Advance and process consent: patients identified on admission, asked if prepared to enter study if develop secretions, informed consent documented medical notes, consent reconfirmed at readmissions. n = 58/107 (54%) (n = 15 developed death rattle randomised) Key finding: Advance consent is a viable and acceptable method to consent for trials in dying phase. Patient accrual rates to date are lower than needed to recruit adequate numbers in the time allotted to answer the research question.
Challenges: (1) complexities of recruiting patients, e.g. too unwell (n = 15), died elsewhere (n = 16), died pre-randomisation (n = 15); (2) resource intense—estimate takes 3 years to meet sample size; (3) consent process time-consuming and emotionally draining.
Solutions: (1) study across care settings, e.g. hospices; (2) involve all HCP disciplines; (3) patient acceptability trials dying phase (16 patients declined).
    Whelan et al. [52], UK, A3 2013 RCT exploring the impact of the requirement for ‘proxy assent’ on recruitment in a trial of antibody response to influenza vaccination and use of a booster dose when indicated versus usual care in care homes (FEVER Trial). All care homes in three London boroughs Older people in care homes Personal consultee advice (n = 82, 14%) if unavailable professional consultee care home staff (n = 40, 13%). Capacity assessment informal (study pre-dates MCA 2005). Lack capacity n = 122/557 (22%)
Capacity n = 155/411 (38%)
Overall = 277/968 (29%)
Key findings: difficulties attaining consultee advise cause of recruitment bias with lower recruitment, e.g. older participants. Care home staff as consultee rarely used, e.g. reluctance of responsibility. Further research required independent risk/benefit expert panels, e.g. Independent Mental Capacity Advocates (MCA 2005).
Challenges: (1) high lack of capacity to consent (62%, n = 602); (2) no contactable consultee (n = 304, 55%); (3) relative declined permission enrolment (n = 27, 5%); (4) researcher considered patient likely to resist procedures (n = 146, 26%).
    Henwood et al. [56], Australia, B3 2014 Cross-sectional study with random sampling. Aim to establish the prevalence and risk factors to sarcopenia among older adults with compromised well-being residing in residential aged care (RAC). Residential aged care Very old (mean age 84.5 years) residents of the care facility All participants were required to give informed consent directly or by the substitute decision-maker, or by the service manager or director of nursing following discussion with the substitute decision-maker. 102/273 (37%)—91 provided consent and 11 were consented by proxy Key findings/challenges: highest reason of non-recruitment was not wanting to participate (79%), followed by ‘My GP does not want me to’ (7%), changes in health status, and death.
Solutions: To facilitate recruitment, requires involvement of key staff members in planning and execution of the research study to support the study and understanding of the study protocol.
    Myers et al. [50], USA, B3 2018 Prospective cohort study to determine whether unplanned hospital transfer can be avoided. One large county with a single system of emergency medical services Assisted living residents who fall Each patient’s primary care physician informed the patient or his or her healthcare Power of Attorney of the study during usual care. Patients choosing to participate or their powers of attorney signed a written informed consent document. 953/1473 (65%) Key findings: of the 953 residents in the study, 359 had 840 falls during 43 months. The protocol recommended non-transfer after 553 falls. Eleven of these patients had a time-sensitive condition. 549 of the 553 patients (99% [CI, 98 to 100%]) with a protocol recommendation for non-transfer received appropriate care.
Challenge: participant accrual was slower than anticipated and time-sensitive conditions were less prevalent.
    Irwin et al. [55], USA, C1 2008 Pilot study using a convenience sample to assess cognition and evaluate the presence of cognitive impairment in alert and awake hospice patients who did not have a current or past diagnosis of a cognitive disorder or cognitive impairments. Hospice Hospice inpatients/palliative care patients approaching the end of life (prognosis of < 6 months) Written (n = 14) or oral informed consent (n = 16) 30 Key findings: cognitive impairments are common among inpatient hospice patients (12/30 patients were diagnosed with dementia—DSM V). Findings highlight the under-recognition of cognitive impairment in this population.
Solutions: (1) psychoeducation for family members and caregivers, (2) psychopharmacological treatment for the patient, and (3) more timely final preparations to maximise function, adequate understanding and coping mechanisms, complete personal affairs, e.g. wills, advanced directives, legacy work and saying ‘good-byes’.
    Davies et al. [53], UK, A3 2018 A feasibility cluster RCT. Aim ‘can a definitive (adequately powered) study be done?’ Hypothesis was that adequate clinically assisted hydration during the last few days of life would maintain renal perfusion and prevent hyperactive delirium (‘terminal agitation’). Cancer centres and hospices Cancer patients at the end of life with delirium Informed consent from the patient where possible, if not personal consultee assent from a relative or a friend or nominated consultee assent from the site Study Guardian (an independent senior clinician). Process consent—if patient lost capacity during the study, personal/nominated consultee was required to confirm continued involvement in the study. 200/219 (91%) Key findings: the study recruited 91.3% of eligible patients by using multiple consent processes. Only 13/219 (6%) declined to take part and no withdrawals. Data collection burden from the patient and family members was minimised. Informed consent was received from patients (16, 8%), advice from personal consultees (161, 81%), and nominated consultees (23, 12%).
Challenge: Unbalanced trial arms due to differences in eligible participants and ‘competitive’ recruitment strategy.
Solutions: Multiple consent processes and minimise data collection burden.
Dementia and geriatric care
    Baskin et al. [76], USA, A3 1998 RCT to identify challenges to informed consent in research involving subjects with advanced dementia to check purpose of bid. Inpatient hospital Advanced dementia Personal consultee advice n = 75/146 (51%) Key findings: this study is the first to examine barriers to research in patients with advanced dementia. 49% (n = 71) eligible participants could not be enrolled in a study on palliative approaches to care. 68 because of an inability to engage the proxy in the consent process (22 because of the absence of a suitable proxy), and 4 because proxy declined consent. The findings indicate barriers in clinical research in end-stage dementia, and implications for medical decision-making in this vulnerable population.
Challenges: (1) relative declined permission enrolment (n = 4); (2) proxy uncontactable (n = 41/63); no suitable proxy identified (n = 22).
    Sampson et al. [40]*, UK, B3 2018 A 9-month prospective cohort study aiming to describe (1) physical and psychological symptoms, (2) health and social care service utilisation, and (3) care at end of life in people with advanced dementia. 14 nursing homes in the UK or participant’s own homes People with advanced dementia (Functional Assessment Staging Scale 6e and above), aged 65 or over Personal consultee advice, consultee uncontactable approach professional consultee 85/159 (54%) Key findings: recruitment target not reached with 30% of patients/carers declining. Of 157 eligible care home residents, 80 people recruited by carer consultee (62, 1 patient died), or professional consultee (18). 32 declined, 28 uncontactable consultee, and 17 died. 42 people residing at home, 6 recruited by carer consultee. 16 consultees declined, 19 uncontactable, and one died.
    Jones et al. [39]*, UK, B3 2012 Protocol for the above cohort study. Care homes and own home Dementia (early stage) Personal consultee advice, consultee not contactable approach professional consultee Study protocol (recruitment target n = 100) Personal consultee, consultee not contactable approach professional consultee, e.g. staff not involved in study.
    Black et al. [73], USA, B3 2006 Cohort study to describe health problems and comorbidities in nursing home residents, and correlations with staff identified pain. Nursing homes Advanced dementia Personal or legally appointed representative
Capacity assessment informal HCP
n = 126/289 (44%) Key findings: physician high recognition symptoms at EoL, but detection decreases with increasing cognitive impairment indicate sample bias.
Challenges: (1) consultee denied enrolment permission (18%); (2) no response consultee (5%); (3) patient died (23%); (4) physician denied recruitment permission (5%).
    Goodman et al. [74], UK, B3 2011 Cohort mixed method design to understand EoLC needs and support for people with dementia in care homes. Care homes (residential, 6 homes) Advanced dementia Continuous consent adults able consent in moment (n = 65); personal consultee advice adults lacking capacity (n = 68). Capacity assessment MCA criteria researcher and care staff. n = 133/215 (62%) Key findings: 65 (74%) people with dementia who had capacity to understand the study agreed to participate, and those lacking capacity personal consultees advised participation (68/126, 54%).
Solutions: greatest impact on recruitment care home culture, e.g. working practices.
Challenges: (1) consultees declined permission participant enrolment (n = 9); (2) consultee uncontactable (n = 49); (3) resource—5 months full-time researcher recruit 133.
Solutions: (1) engagement families and care staff; (2) accessibility study information, e.g. A4 summary sheet.
    Hanson et al. [75], USA, A3 2010 Cluster RCT of a decision aid to inform and support the choice between tube feeding and assisted oral feeding in dementia. To describe recruitment strategies in nursing homes and ethical approaches to minimise harm and maximise benefits. Nursing homes (n = 24) Advanced dementia Personal consultee, e.g. son/daughter (68%). Capacity assessment: Cognitive Performance Scale and Global Deterioration Scale. n = 256/425 (60%) (paired resident and personal consultee) Key findings: clinical trials involving adults lacking capacity require multiple strategies to engage consultees and recruit in nursing homes. Recruitment approaches ensured adequate time to address concerns of family surrogates, and compensation for their time commitment to the study interviews.
Challenges: (1) care home culture recruitment rate varied by nursing homes (3 to 36, 30–94%).
Solutions: (1) study design clear potential benefits for residents; (2) engage consultees, e.g. monetary reimbursement for time; researcher frequent calling to book interview and flexibility—interviews convenient time and location/by telephone, repeated provision accessible study information; (3) engage care home staff, e.g. prior contacts, time and resource, minimal burden for care staff.
    Sampson et al. [42]*, UK, B3 2015 Cohort study to define the prevalence of pain using self-rated and observational pain scales in people with dementia in the general hospital, and to examine demographic and clinical factors associated with pain and to explore a hypothesised association between pain and behavioural and psychiatric symptoms of dementia (BPSD). Two large acute general hospitals in London Patients with dementia who were admitted to general medical wards of two large acute hospitals Consent process: patient to give written informed consent or with an informal carer or ‘professional consultee’. The consent procedure followed the England and Wales MCA 2005. No formal capacity assessment. A personal consultee was identified to give agreement either verbally over the telephone, and posted an agreement form to sign and return. If forms were not returned, participants’ data were destroyed. If a personal consultee could not be reached in 48 h of screening, a professional consultee advice (e.g. geriatrician not involved in the study or patient care). 230/292 (79%) Challenge: 62 excluded because they did not fulfil the inclusion criteria or because carers who gave telephone agreement did not return signed forms.
Solutions: consent process practical and feasible in the acute setting.
    Scott et al. [41]*, UK, B3 2011 Protocol for the above cohort study. Hospital inpatient Dementia and unplanned hospital admission Assent personal consultee, if no consultee identified in 48 h, nominated ‘professional’ consultee—senior member clinical team not involved study. Capacity assessed MCA criteria identify consent/assent. Study protocol (recruitment target 250) Key finding: protocol developed MCA guidance, i.e. assess the person using structured assessment for capacity, seek assent from a carer for person lost capacity, and use professional consultees when carer not contactable.
    Warner et al. [72], UK, A3 2007 RCT of dementia treatment, Ginkgo biloba, to assess the proportion of people with capacity to consent, describe use of the MCA 2005 to assess capacity in a research setting, and identify variables associated with the presence of capacity. Community settings Dementia Informed consent or personal consultee advice for adults lacking capacity. Capacity assessed MCA criteria and MMSE. n = 42/176 (24%) Key findings: cognitive testing alone is insufficient to assess the presence of capacity. Recommend researchers record outcome capacity assessment and basis for the finding; cognitive measures considered poor proxy for judging capacity. Research may be undertaken with incapacitous participants who nevertheless appear to agree to participation, but is only permissible if there is some potential direct or indirect benefit to the patient and the research cannot be carried out on patients who have capacity.
Challenges: (1) high cognitive impairment, 76% participants assessed lacked capacity for informed consent; (2) MMSE did not predict the presence of capacity.
    Ellis-Smith et al. [78], UK, B2 2018 Prospective, mixed methods feasibility study to explore the mechanisms of action, feasibility, acceptability, and implementation requirements of a measure, the Integrated Palliative care Outcome Scale (IPOS-Dem), used in routine care to support comprehensive assessment of symptoms and concerns of care home residents with dementia and their family members. Three residential care homes registered to provide care for people aged 65 and over in a London borough, UK Palliative/residential care home residents with dementia The research team met with residents to ascertain willingness to participate and assess mental capacity to consent for themselves. Those residents that had capacity gave written informed consent. Advice of consultees was sought if ALC. The care home therefore sent a letter on behalf of the research team to a close friend or family member to invite them to advise on whether the resident should participate in research (personal consultee). Two letters were sent. If no response was received after 1 week of the second letter being sent, a nominated consultee was asked to advise on resident participation. The nominated consultee was independent from the research study and used all available information (including meeting with the resident, reviewing case notes, and speaking to care home staff) in order to give advice on resident participation. 10 family members approached. 22 professionals approached. 47 residents approached. 32 baseline and 30 final time point patient data. 20 professionals involved in focus groups and serial interviews. Six family members involved in a focus group and interviews Key findings: of 47 participants approached, one patient provided informed consent, 45 were assessed as not having capacity to consent. Consultee approached for n = 39. Personal consultees provided advised for n = 15, and professional consultees for n = 24. One resident declined, four NoKs uncontactable. Not recruited on consultee advice (n = 3), advised against participation (n = 2), professional consultee (external to the care home and study) advised NoK involvement, but NoK did not respond (n = 1). Findings showed that possible to introduce a measure into routine care of residents and change care processes to improve resident and family outcomes.
Solutions: mutliple procedures of approaching the NoK, and if uncontactable, a professional consultee is feasible in care homes. The staff members know the residents and could inform the researchers about preferences for research participation (or not).
    Mitchell et al. [77], USA, B3 2006 Cohort study to examine EoLC in advanced dementia and to describe how research challenges were met (CASCADE study). Nursing homes (n = 15) Residents with advanced dementia and their healthcare proxies—Cognitive Performance Scale score 5/6 Professional consultee (healthcare staff) involving consent for themselves and providing proxy consent for residents (data collection chart review, brief clinical examination, nursing interview). Capacity assessment: Cognitive Performance Score 5 or 6. n = 189/343 (55%) (patient/HCP dyad) Key findings: involving ALC in research on palliative care, including the dying phase, acceptable to staff surrogate decision-makers in nursing homes. Findings suggest that the emotional distress of families does not decline when patients with dementia are institutionalised.
Challenges: clinicians’ refusal (n = 154), e.g. too burdensome (n = 30), lack of interest (n = 53).
Solutions: (1) nursing home culture, e.g. track record conducting research; (2) engage care home staff, e.g. researcher trained geriatric nurse, participation low burden; (3) research team credibility.
    Davies et al. [84], UK, B3 2010 Cohort study to examine health trajectories and outcomes of people aged 85+ cohort and associations with biological, medical, and social factors. Usual place of residence, e.g. at home, care home Older people aged > 85 years Advance consent—consent document participant nominate a personal consultee should they lose capacity. Used continuous consent process, fluctuating capacity awaited re-gain capacity. Capacity assessment. MCA criteria researcher. n = 1042/1453 (72%) Key findings: 72% recruitment rate using consultees, separate consent protocols for participants in care homes and flexibility regarding those with fluctuating capacity. Cognitive impairment is common requiring consideration of the relevant ethical and legal issues.
Challenges: (1) consultee declined recruitment (n = 5); (2) consultee uncontactable (n = 9).
Solutions: (1) comprehensive protocols for consent and capacity assessment that anticipated high cognitive impairment guided by legislation; (2) training researchers on consent and capacity processes, sensitively handling complex situations; (3) time and resource to engage family and care home staff; (4) follow-up telephone call/visit 1 week after receipt study information; (5) design study information for older person, e.g. font size, language.
    Mason et al. [83], UK, A3 2006 RCT to determine effectiveness and cost-effectiveness of pressure relieving mattresses on pressure ulcers. 10 research centres involving inpatient units Patients with pressure ulcers aged > 55 years with or without cognitive impairment Personal consultee advice for ALC. Lack capacity n = 87/2445 (4%)
Capacity n = 1972/2395 (82%)
Key findings: a higher proportion of relatives declined enrolment, than patients themselves. Proxy consent allowed only a small increase in trial recruitment.
Recruitment challenges: (1) no relative available (n = 2286); (2) relatives declined permission to enrol (n = 72/159, 45%).
    Botker et al. [85], Denmark, A3 2018 RCT. Aim to examine if the addition of brain natriuretic peptide measurement to the routine diagnostic work-up by prehospital critical care team physicians improved triage in patients with severe dyspnoea. Prehospital critical care units at Central Denmark Region hospitals Patients with severe dyspnoea Deferred consent: included patients prior to informed consent. Subsequent oral and written informed consent was obtained from all participants or next of kin by one of 10 study investigators. 711/747 (95%) Key findings: 28 (4%) participants withdrawn consent based on patient or proxy request. 7 (0.9%) patients were excluded as consent could not be obtained. The routine addition of prehospital NT-proBNP measurement did not improve the triage of patients with dyspnoea of cardiac cause directly to the department of cardiology and did not significantly improve treatment or patient outcomes.
    Galeotti et al. [79], Italy, A3 2012 RCT. The ADCare study aimed to evaluate the long-term safety and efficacy profiles of three atypical antipsychotic drugs and one conventional antipsychotic drug in treating psychosis, aggression, and agitation in outpatients with AD.
The aim of this paper is to report the ADCare study experience and to analyse in depth the possible reasons for the low accrual.
N/A Alzheimer’s disease patients Informed consent 83/~ 800 eligible patients from 19 clinical centres Key findings: ~ 800 eligible patients in 1 year were identified in the clinical centres. Although initially willing to participate, most of these patients/family members declined because a legal court nomination was required to participate in the trial. Court nomination was perceived by patients and family members as intrusive and potentially creating conflicts in families.
Challenges: only 9 clinical centres participating in the study enrolled patients with the involvement of a legal proxy, even with initiatives to disseminate information on legal agency for research participants. The main obstacles were caregivers’ reluctance to designate a legal proxy and court delays.
Intensive care
    Day et al. [109], UK, B3 2015 Pilot RCT single centre. Aim to report challenges to undertake the RCT and feasibility of delivering and evaluating a complex intervention in a critical care unit. Two mixed critical care units of a large inner London hospital Mixed medical, surgical, and trauma patient population requiring either level 3 (intensive) or level 2 (high dependency) care Researchers not involved in data collection were responsible for the consent process. In this study, patients were willing to consent verbally but deferred to their relative to provide written consent. If ALC, their relative provided advice, the researchers then ensured that when patients regained capacity and they requested an informed consent. 158/221 (72%) Key findings: rushing the decision-making process was a concern. Some patients approached felt unable to say no, but kept saying come back later. This may indicate concern that their care may be affected if decline. Viewing informed consent as a process and revisiting it throughout the trial period to ensure participants can recall providing consent.
Challenges: Most of those approached felt too tired or lacked the concentration to read the study information. Patients with no appropriate consultee were excluded.
Solutions: (1) avoid consent bias, risk of coercion and allow more time for decision-making. (2) Participant information sheet clear and concise. (3) To review capacity in patients with memory loss.
    Bench et al. [110], UK, A3 2015 Pilot RCT. Aim (1) providing an initial evaluation of a user-centred critical care discharge information pack (UCCDIP), (2) inform decisions about its further development and evaluation, and (3) estimate the sample size for a full trial. Two ICUs within a single teaching hospital in central London Mixed medical, surgical, and trauma patient population requiring either level 3 (intensive) or level 2 (high dependency) care Informed consent from the patient was then obtained prior to data collection on the ward. For ALC, informed written consent at the point of ICU discharge (deferred), personal consultee declarations, usually from the patient’s next of kin, were sought. The relatives of all recruited patients were given study information when they visited the ICU or telephoned and invited to participate. Written consent was obtained from relatives during their next hospital visit. 158/221 (36 declined to participate) (72%) Key findings/challenges: The a priori enrolment goal was not reached, and attrition was high leading to insufficient statistical power to determine outcome benefit. 101 (64%) patient participants provided primary outcome data at time point 1. A total of 48 (60%) patients’ relatives provided at least one set of outcome data. Twenty-seven (17%) patients and 32 (40%) relatives were lost to any follow-up.
    Higginson et al. [111], UK, B3 2016 Prospective ethnographic study. Aim to explore the nature and patterns of decision-making processes during ICU admission, including sources of conflict and resolution. Two ICUs in an inner city hospital serving an ethnically and socially diverse population ICU patients where clinicians had potential end of life concerns, discussions or a high risk of dying during the current admission Where possible, informed consent from patients. For patients who lacked capacity, following the Mental Capacity Act 2005, firstly capacity was assumed to be present, unless proven to be absent. Clinicians assessed capacity. If lacked capacity, approach for advice a personal consultee (e.g. family member), if not identified a nominated consultee (e.g. clinician) 16 patients and 19 relatives Solutions: the researchers looked for instances where capacity was present and discussed with the clinicians. The research team ensured that the nominated consultee was not part of the research team. Relatives also provided informed consent/assent for interviews and observations.
    Rouzé et al. [112], France, A3 2017 RCT to test the hypothesis that the use of an algorithm based on fungal biomarkers would increase the percentage of early discontinuation of empirical antifungal treatment among critically ill patients. Mixed 50-bed ICU department of the University Hospital of Lille, France ICU patients with Candida infection An informed written consent was obtained before randomisation from the patients or their proxies. 110/387 (28%) Key findings: 8/510 patients who were assessed for eligibility refused. One patient withdrew consent after receiving the biomarker strategy. The use of a biomarker-based strategy increased the percentage of early discontinuation of empirical antifungal treatment among critically ill patients with suspected invasive Candida infection.
Mental health
    Ho et al. [97], Australia, B3 2018 Methodological paper aiming to describe an informed consent process used when recruiting persons with intellectual disability for a study which is currently investigating falls among people with intellectual disability, and to reflect on the methods of informed consent used. Community—participant’s home, small group homes with up to two to four co-inhabitants with paid support People with intellectual disability Informed consent and proxy advice from the participants themselves where possible. The process was designed to provide a collective perspective of the capacity of the person with intellectual disability to consent. The researcher uses repeated observations to establish if the person has capacity to provide consent, and consult with the caregiver. Use of a decisional questionnaire involving score of > 3 out of six suggest capacity to provide informed consent. The next of kin, family member, or caregiver is asked to be present during this process to provide a supportive, comfortable environment and to provide oversight to the discussion. [84] 40/68 (59%) Key findings: only 3 out of 40 participants were able to provide informed consent. 22 participants were able to have a discussion about their involvement in the study with the support of their caregiver. 15 participants were unable to engage in the process. The decisional questionnaire gave an accurate representation of participants’ decision-making capacity.
Challenges: more than 40% caregivers declined taking part in the study on behalf of the person. This consent process was time-consuming.
Solutions: Training and building relationships with patients and their caregivers.
    Ramerman et al. [96], Holland, B3 2018 Experimental study to examine challenging behaviour, physical symptoms, and quality of life associated with antipsychotic drug. Hospital intellectual disability mental healthcare clinics Adults with intellectual disabilities Written informed consent was obtained from participants and/or their legal representatives. 159 (no data on number of people approached) Key findings: mean age was 46.2 (SD 17.7). Health quality of life (HQoL) was negatively associated with both symptoms of challenging behaviour and physical symptoms associated with antipsychotic drugs. No data on differentiation proxies and patients in informed consent process.
    Marcantonio et al. [101], USA, A3 2010 RCT to determine whether a delirium abatement programme (DAP) can shorten duration of delirium in new admissions to postacute care (PAC). Hospital Patients with delirium older than 65 (mean age = 84); MMSE and CAM used to assess capacity In those who had delirium, family caregivers, acting as proxies, provided informed consent for trial participation. 457/667 (69%) Key findings: this study was the largest cohort of patients with delirium ever enrolled in a research study. Facilities received a small incentive based on their performance. Clustering effect widened the confidence intervals. Possible to ascertain adherence and outcomes only in trial participants whose proxies provided informed consent to allow medical records to be reviewed.
Challenges: Of 667 patients who were eligible, 138 relatives refused (21%), 56 relatives uncontactable (8%).
    Cole et al. [102], Canada, A3 2002 RCT to determine if systematic detection and multidisciplinary care of delirium in older patients could reduce time to improvement in cognitive status. Hospital Older patients 65 years old or more admitted to a general hospital medical service Informed consent was obtained from the patient or substitute decision-maker. Confusion assessment method was used to assess capacity. 227 /299 (76%) Key findings: 72 were excluded as they could not provide consent. Study was not powered to detect any statistically significant differences.
Emergency medicine
    Offerman et al. [108], USA, B3 2013 Observational prospective multicentre study to describe the rate of successful consent using an altered (deferred telephone) consent process in emergency department (ED) patients. Emergency department Adults who had attended ED with blunt head trauma Informed consent was obtained from the patient or substitute decision-maker. 506 Key findings: follow-up telephone contact was successfully accomplished in 501 of the 506 subjects (99.0%; 95% CI = 97.7 to 99.7%). Consent for study inclusion and conduct of the telephone survey was obtained in 500 of 501 subjects at time of the follow-up call (99.8%; 95% CI = 98.9 to 100.0%). Surrogate consent was obtained in 199 of the 501 subjects (39.7%; 95% CI = 35.4 to 44.2%)