Methods of the CLASSIC cohort

Practice recruitment

Ethics approval was obtained from the National Research Ethics Service (NRES) North West Lancaster (Research Ethics Committee reference 14/NW/0206).

Not all practices were invited to participate as they either had low numbers of patients aged ≥ 65 years or were affiliated with care homes with high numbers of dementia patients. Forty-seven practices were invited and 33 (70%) agreed to participate.

We used FARSITE [a tool for recruitment to research; see http://nweh.co.uk/products/farsite (accessed 18 May 2018)] to generate a list of eligible patients. Each practice was then asked to identify patients meeting exclusion criteria (i.e. in palliative care, those with conditions that reduce capacity to consent).

Practices did not receive incentives but did receive support costs to reimburse their time.

Patient recruitment and retention

Eligible participants were those aged ≥ 65 years and registered as having at least one long-term condition at a general practice.

A total of 12,989 patients were eligible and surveyed through general practice between November 2014 and February 2015. If they did not respond, they were sent a reminder and a second copy of the questionnaire 3 weeks later. Participants were offered an incentive of a £10 voucher.

Response was taken to indicate consent to further surveys at 6, 12 and 18 months.

To increase retention, patients were called by a researcher to offer completion of the questionnaire over the telephone. Patients were offered a £5 gift voucher for the completion of the 12- and 18-month follow-ups.

Cohort measures

The following list of measures was used in one or more of the surveys (Table 3).

• Baseline assessment included sociodemographic questions from the General Practice Patient Survey,56 including sex, age, work situation and qualifications; ethnicity using 17 2011 Census categories;57 a single-item health literacy measure;58 a measure of the number and impact of long-term conditions;59 and use of local community assets.60
• The Patient Assessment of Chronic Illness Care (PACIC),61 which includes items in five subscales (patient activation, delivery system design and decision support, goal-setting, problem-solving, and co-ordination); we used the short 11-item version.62
• The Patient Activation Measure (PAM) of patient knowledge, skills and confidence in self-management for long-term conditions;54,63 we used the short 13-item version.64
• The Summary of Diabetes Self-Care Activities (SDSCA) measure assesses the number of days per week on which respondents engage in healthy and unhealthy behaviours.65
• The Multimorbidity Illness Perceptions Scale (MULTIPleS) assesses patient experience of managing multimorbidity;66 we used 16 items from the MULTIPleS.
• The Quality, Innovation, Productivity and Prevention (QIPP) measure of personalised long-term condition care.67
• The ENRICHD Social Support Instrument (ESSI).68
• The EuroQol-5 Dimensions (EQ-5D) measure of health-related quality of life;69 we used the new EuroQol-5 Dimensions, five-level version (EQ-5D-5L).70
• The Mental Health Inventory – 5 (MHI-5) is a five-item scale that measures general mental health.71
• The ICEpop CAPability measure for Older people (ICECAP-O) index of capability measures quality of life for people aged ≥ 65 years in terms of attachment, security, role, enjoyment and control.72
• The World Health Organization Quality of Life-BREF (WHOQOL-BREF) 26-item measure of global quality of life across four domains (physical health, psychological health, social relationships and environmental), as well as a single-item scale.73
• Health-care utilisation, based on our previous studies;74 continuity of care and care planning from the General Practice Patient Survey;56 and patient experience of safety from the ICPs evaluation.
• We also used single-item measures assessing other issues, including items assessing issues of interest to stakeholders (e.g. internet use and accommodation).

We used a short assessment for carers, including EQ-5D, Patient Health Questionnaire-9 items, ICECAP-O and the Modified Caregiver Strain Index.75 Carer data are presented in Appendix 1.