Table 1

Characteristics of all of the included studies in order of reference

Author and year of publication	Setting	Study type	Study aims	Target group and no of participants (n)	Outcomes measured
Quantitative
Health outcome priorities
Fried 201139	USA—three senior centres and one assisted living facility	Quantitative: Cross-sectional Study.	To explore the use of a simple tool to elicit older persons’ health outcome priorities.	All volunteers included (n=357).	The prioritisation by participants of 4 universal health outcomes, namely: keeping alive maintaining independence reducing or eliminating pain reducing or eliminating other symptoms.
Fried et al, 201140	USA—recruited from participants in a larger study, where they had been recruited from age-aggregated community housing 74	Quantitative: Cross-sectional Survey.	To determine the feasibility of using a simple tool to elicit the preferences of older persons based on their prioritisation of universal outcomes.	Patients aged 65 and over with a known diagnosis of hypertension or use of antihypertensive medications, and having a known risk of falls (n=81).	>Rankings given by participants to four universal health outcomes in the outcome prioritisation tool: -keeping alive maintaining independence reducing or eliminating pain reducing or eliminating other symptoms >Feasibility of the use of outcome prioritisation tool.
Mantelli et al, 201857	Switzerland—GPs working in Switzerland who had previously taken part in case-vignette studies	Quantitative: Cross-sectional Survey.	To determine whether, how and why GPs deprescribe in frail oldest-old patients with multimorbidity and polypharmacy, and to identify factors that influenced their decision to deprescribe.	GPs (n=157).	Percentage of GPs willing to de-prescribe at least one medication in the case of frail older patients with cardiovascular disease and compared with frail older patients without cardiovascular disease. Reasons for deprescribing Importance ratings given to factors influencing decision to deprescribe.
van Summeren et al, 2017 46	Netherlands—general practice centres	Quantitative: Cross-sectional and implementation study.	To determine proposed and observed medication changes when using an outcome prioritisation tool during a medication review in older patients with multimorbidity and polypharmacy. A secondary aim was to explore the relationship between the prioritised health outcome of patients and the type of medication change, such as a stop, a dose adjustment, or a switch.	Patients aged 69 or over with two or more chronic diseases (one of which had to be cardiovascular disease) and daily use of five or more medications. (n=59) GPs (n=17).	>Patients’ priority rankings of the four health outcomes in the outcome prioritisation tool: Maintaining independence Remaining alive Reducing other symptoms Reducing pain >Medication changes proposed by the GP, and observed in the patient records following incorporation of the priority rankings given by patients, into a medication review consultation.
van Summeren et al, 201645	Netherlands—general practice centres	Mixed-methods: Cross-sectional survey pilot and qualitative interviews to assess acceptability (semistructured and in-depth).	To explore whether an outcome prioritisation tool is appropriate in the context of medication review in family practice, focusing on its acceptability and practicality.	Patients aged 69 or over with two or more chronic diseases (one of which had to be cardiovascular disease) and daily use of five or more medications (n=60) GPs (n=13).	>Patients’ prioritisation of the four domains of the outcome prioritisation tool: Maintaining independence Remaining alive Reducing other symptoms Reducing pain >Family practitioners views on the acceptability and practicality of using the outcome prioritisation tool for medication review.
Problem-based priorities
Junius-Walker et al, 201252	Germany—general practice centres	Quantitative: RCT	To investigate whether a structured priority-setting consultation reconciles the often-differing doctor–patient views on the importance of problems.	Patients aged 70 or over (n=317) GPs (n=40).	Baseline importance rankings given by patients and clinicians to a list of problems generated from a geriatric assessment for each patient. Importance rankings given again after a structured consultation incorporating the baseline problem list and importance rankings and degree of reconciliation in doctor–patient agreement after the structured consultation.
Junius-Walker et al, 201153	Germany—general practice centres	Quantitative: Cross-sectional Survey.	To gain insight into setting individual priorities with older patients using a priority definition that was coherent to the patients’ life and doctors’ work context.	Patients aged 70 or over and living at home (n=123) GPs (n=11).	Importance rankings given by patients and clinicians to a list of problems generated from a geriatric assessment for each patient.
Voigt et al, 201054	Germany—general practice centres	Quantitative: Cross-sectional Survey.	To ascertain health priorities of older patients and treatment priorities of their GPs on the basis of a geriatric assessment and to determine the agreement between these priorities.	Patients aged 70 or over and at least one contact with the GP in the preceding 3 months (n=35) GPs (n=9).	Importance rankings given to problems generated from a geriatric assessment by patients and clinicians Degree of agreement between patients and clinicians on the above.
Condition-focused priorities
Moore et al, 201436	Canada—databases of all practising nurse practitioners, family practitioners and geriatricians in Ontario	Quantitative: Cross-sectional Survey.	To quantify how family physicians, nurse practitioners and geriatricians prioritise syndromes, diseases and conditions when caring for seniors.	Nurse practitioners (n=68) Family practitioners (n=84) Geriatricians (n=27).	Frequency and importance rankings given by family practitioners, nurse practitioners and geriatricians to 41 health issues known to arise in elderly patients
Zulman et al, 201044	USA—scheduled primary care visit for patients at nine veteran affairs facilities	Quantitative: Prospective Cohort Study.	To understand patterns of patient–provider concordance in the prioritisation of health conditions in patients with multimorbidity.	Patients with diabetes and hypertension who had their primary diabetes care provider enrolled in the study (n=1169) Primary care providers that is, physicians, physician assistants or nurse practitioners (n=92).	Patient rankings given in terms of their most important health concerns and providers rankings in terms of conditions most likely to affect each patient’s outcomes Concordance between the importance ratings of patient-provider ‘pairs’.
Herzig et al, 201956	Switzerland—primary data were from ‘Multimorbidity in Family medicine’ study.75 Patients enrolled by GPs during scheduled consultations.	Quantitative: Cross-sectional Survey.	To describe FPs’ medical priority ranking of conditions relative to their prevalence in patients with multimorbidity.	Patients suffering from at least 3 of 75 chronic conditions on a predefined list (based on the International classification of primary care 2 (n=888) GPs (n=100).	Importance rankings given by family practitioners to the list of chronic conditions that each patient had on the day of their inclusion in the study.
Déruaz-Luyetet al, 2018 58	Switzerland—primary data were from ‘Multimorbidity in Family medicine’ study.75 Patients enrolled by GPs during scheduled consultations.	Quantitative: Cross-sectional Survey.	To evaluate whether GPs could identify the condition that their patients with multimorbidity considered most important.	Patients suffering from at least 3 of 75 chronic conditions on a predefined list (based on the International classification of primary care 2, and receiving follow-up from their GP for at least the preceding 6 months (n=572 for main analysis, 585 for sensitivity analysis) GPs (n=100).	Whether there is agreement between what patients considered to be their most important health condition and what GPs thought patients considered to be their most important health condition.
Treatment priorities
Caughey et al, 201747	Australia—multidisciplinary ambulatory consulting service clinics at tertiary teaching hospitals	Mixed-methods: Structured quantitative interviews with patients then semistructured qualitative interviews with patients and clinicians.	To investigate how older patients with multimorbidity balance the benefits and harms associated with medication for prevention of CVD, and in the presence of competing health outcomes. To investigate the factors that clinicians consider when making treatment decisions for older patients with multimorbidity.	Patients aged 65 or older with 2 or more chronic conditions (n=15) Clinicians (n=5).	Patient willingness to take a medication when presented with different scenarios with variable degree of benefit, impact on daily living, adverse outcomes and impact on other comorbid conditions Patient-reported data during semistructured interviews where they were asked about their treatment preferences, medication effects and shared decision making Clinician reported data during semistructured interviews on treatment decisions, patient preferences and polypharmacy.
Qualitative
Kuluski et al, 2013 37	Canada—A Family Health Team in Ontario	Qualitative: semistructured interviews	To examine patient goals of care from the perspectives of older persons with multimorbidities, their family physicians and informal caregivers (ie, family member or friend who provides ongoing support) and then examine the extent of alignment between these three perspectives.	Patients aged 65 or older with a diagnosis of at least two chronic health conditions (n=28) Informal Caregivers of included patients (n=28) Family physicians (n=4).	>Patient, caregiver and physician reported data on goals of care for the patients >Degree of alignment of goals of care across patient, caregiver and physician ‘triads’
Schoenberg et al, 2009 38	USA—senior centres, low-income senior housing complexes, churches and a civic meeting hall	Qualitative: in-depth interviews	To understand how vulnerable older adults with multimorbidity prioritise and manage their chronic conditions.	Patients aged 55 or older with a diagnosis of at least two chronic illnesses, from low-income backgrounds (n=41).	Patient-reported data from in-depth interviews, regarding their medical history, self-care procedures, patient prioritisation by means of health-related areas of worry and health-related ‘expenditures’ in terms of money, time and need for reliance on others.
Fried et al, 200841	USA—senior centres, doctors’ practices and a congregate housing site	Qualitative: focus groups	To examine the ways in which older persons with multiple conditions think about potentially competing outcomes, in order to gain insight into how processes to elicit values regarding these outcomes can be grounded in the patient's perspective.	Patients aged 65 or older and were taking 5 or more medications (participants also had a minimum of 3 chronic conditions).	Patient-reported data regarding their perceptions of the interactions between their different illnesses and treatment regimens, goals of treatment and decisions regarding treatment.
Naik et al, 201642	USA—qualitative data from the VETCARES study,76 in which participants recruited from the VA tumour registry	Qualitative: open-ended questions as part of mixed-methods interviews which also included structured questions.	To identify a taxonomy of health‐related values that frame goals of care of older adults with multimorbidity who recently faced cancer diagnosis and treatment.	Veterans with a diagnosis of head and neck, gastric, oesophageal, or colorectal cancer, and diagnosis fell 1 month prior to the study’s opening eligibility window (6 months) (n=146).	Patient-reported data regarding their priorities or concerns regarding their future healthcare decisions
Elliott et al, 200743	USA—Harvard Pilgrim Health Centre, a health maintenance organisation in New England	Qualitative: semistructured interviews.	To explore how older adults with multiple illnesses make choices about medicines.	Patients taking more than three medicines with purposive sampling to reflect symptomatic comorbidities and asymptomatic comorbidities and mental health issues (participants had a minimum of 3 comorbidities) (n=20).	Patient-reported data regarding beliefs about medicines, medicine-taking behaviour, historical versus potential choices between different medicines, and factors influencing these choices.
Turner et al, 201648	Australia—long-term care facilities in South Australia	Qualitative: nominal group technique.	To use nominal group technique to generate then rank factors that general medical practitioners, nurses, pharmacists and residents or their representatives perceive are most important when deciding whether or not to de-prescribe medication.	Residents/representatives of residents (n=11) GPs (n=19) Nurses (n=12) Pharmacists (n=14).	Generated factors important for deprescribing according to residents/resident representatives, GPs, nurses and pharmacists Priority rankings given by groups containing representatives from all of the above, to the list of priorities generated previously.
Lindsay, 200949	UK—participants recruited from CHD registries in Greater Manchester as part of a larger RCT77	Qualitative: focus groups and two interviews.	To use the concepts of ‘chronic illness trajectory’ and ‘biographical disruption’ to examine how patients self-manage multiple chronic conditions and especially how they prioritise their conditions.	Participants from the parent study who had more than one chronic condition (ie, at least two) (n=53).	Patient-reported data regarding how they prioritised their multiple conditions, what strategies they used to cope with their conditions and barriers in being able to manage their illnesses.
Cheraghi-Sohiet al, 201350	UK—secondary analysis of qualitative data from four other studies78–81	Qualitative: in-depth interviews.	To explore how and why people with multimorbidity prioritise some long-term conditions over others and what the potential implications may be for self-management activity, and in turn, suggest how such information may help clinicians negotiate the management of multimorbidity patients.	Participants from original studies who had two or more long-term conditions, and had given data regarding prioritisation (n=41).	Patient-reported data pertaining to prioritisation of their long-term conditions.
Morris et al 51	UK—general Practices in North-West England	Qualitative: semistructured interviews.	To examine what influences self-management priorities for individuals with multiple long-term conditions and how this changes over time.	Patients with more than one chronic condition and at least one of Chronic Obstructive Pulmonary Disease, Irritable Bowel Syndrome or Diabetes (n=21).	Patient-reported data on management strategies and experiences with primary healthcare, and data from follow-up interviews on any changes in their illness management.
Hansen et al, 201555	Germany—participants recruited from the ‘Multicare cohort study’82	Qualitative: Focus groups	To identify reasons for disagreement regarding illnesses between patients and their GPs.	Patients who had 3 or more chronic conditions from a list of 29 conditions (n=21). GPs of the recruited patients (n=15).	Data from separate focus groups for patients and clinicians in which any communication problems and reasons for disagreement between patients and clinicians were explored.

CVD, cardiovascular disease; GP, general practitioner; RCT, randomised controlled trial.