Lingler 2016.

Study characteristics
Methods	Aim of study: to develop and examine the efficacy of a tailored, problem‐solving intervention on informal caregivers' management of medications for community‐dwelling persons with memory loss Study design: RCT (unit of allocation: dyad) Number of arms/groups: 2
Participants	Description: both patient/consumer and carers (recruited as dyads) Geographic location: USA Setting: participant's home Inclusion criteria: PATIENT: self‐ or caregiver‐reported memory loss necessitating help with medication‐taking, ≥ 2 comorbid conditions requiring medication, living in community, provided informed consent. INFORMAL CAREGIVERS: family members or kin‐like friends, ≥ 18 years, participating in management of patient's medications, exhibiting ≥ 1 deficiency on any of 3 measures of ability to effectively manage the patient's medications, living within 75 miles of the University Exclusion criteria: paid caregivers, living in residential care setting Number of participants randomised: 83 pairs (42 vs 41) Number of participants included in analysis: 76 pairs (37 and 39) Age: mean ± SD: patients 79.67 ± 9.19 vs 80.15 ± 8.48, caregivers 66.00 ± 12.8 vs 67.80 ± 11.2 Gender: female = patients 28 (67%) vs 22 (54%), caregivers 29 (69%) vs 29 (71%) Ethnicity: white: patients 34 (81%) vs 37 (90%), caregivers 34 (81%) vs 37 (90%); black: patients 3 (7%) vs 3 (7%), caregivers 4 (10%) vs 3 (7%); other patients 5 (12%) vs 1 (2%), caregivers 4 (9%) vs 1 (2%) Number of medications: total medications (including OTC, supplements, etc.): 10.79 ± 5.52 vs 10.61 ± 5.89 Frailty/Functional impairment: not specified Cognitive impairment: baseline sample (n = 91) patients: MMSE = 17.62, carer blessed = 2.97 Comorbidities: number of comorbidities: patients 8.691 ± 3.57 vs 9.024 ± 4.21, carers 7.86 ± 3.69 vs 6.44 ± 3.59
Interventions	Group 1 ‐ Maximising Medication Management by Caregivers of Persons With Memory Loss: guided by intervention manual; sessions with nurse or social worker interventionist addressed 7 basic aspects of the caregiver's role in managing medications during home/telephone discussions. Caregivers provided with self‐study version of intervention manual. Initial 8‐week intervention, then 4 bi‐weekly calls over next 8 weeks (note: mean length of home visits ‐ 40.05 minutes (SD 13.22) ‐ and telephone sessions ‐ 13.42 minutes (SD 6.34)) Group 2‐ Usual care: at baseline, received pamphlet on medication safety. Received home visits for purpose of data collection only (medication errors were corrected). At completion of study, caregivers received intervention manual Co‐intervention: for safety, if any errors were noted during medication reconciliation, they were brought to the attention of both caregivers and prescribers regardless of group assignment. In addition, all participants received care as usual from their healthcare providers Provider: nurse or social worker Where: patient's home When and how often: 2 or 3 home visits 2 weeks apart, followed by 2 or 3 telephone sessions 7 to 10 days apart for 8 weeks; then 4 bi‐weekly phone calls for 8 weeks Intervention personalised: Yes ‐ guided by intervention manual but individualised based on caregivers needs/queries and patients medication
Outcomes	Timing of outcome assessment: baseline and 2 months post intervention Medication‐taking ability (objective): MedMaIDE: Medication Management Instrument for Deficiencies in the Elderly: uses interview and observation to assess ability to self‐administer medications using 3 areas: knowledge of medications, how to take medications, and how to procure medications. Each medication is reviewed during administration. Scores 0 to 13; max total deficiency score is 13 Satisfaction with intervention (subjective): set of Likert scaled questions (not specified), eliciting open‐ended comments during an exit interview at study completion Other (objective): Medication Deficiency Checklist (MDC): 15‐item, investigator‐developed instrument; uses caregiver interviews to assess for the presence of errors and problems (e.g. taking at the wrong time). Investigator‐developed tool ‐ not validated
Notes	Trial registration: N/A Consumer involvement: not specified Funding source: programme project grant: NIH/NINR P01 NR010949 Dropout: 3 pairs withdrew (2 vs 1), 4 pairs were lost to follow‐up (3 vs 1) Fidelity: good ‐ independent rater randomly selected 10% of cases for an audit of protocol fidelity. Percentage of agreement 91.6%, quality of interaction 4.5/5 Unpublished data included successful communication with trial authors; follow‐up MedMAIDE results provided
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Random group assignments were computer generated with permuted blocks within strata to ensure balance of nurse/social worker, relationship of caregiver, and race/ethnicity
Allocation concealment (selection bias)	Unclear risk	Order of consent/allocation not specified; no details on allocation concealment provided
Blinding of participants and personnel (performance bias) All outcomes	High risk	Unblinded ‐ necessary because it was not feasible to blind participating caregivers to group assignment
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	No blinding specified; unclear who did outcome assessment
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Minimal loss to follow‐up; clearly detailed. Erlen paper: includes instances where data are missing, so number of participants may be different. We did not impute data for these participants
Selective reporting (reporting bias)	Unclear risk	MedMaIDE results presented only in visual format; raw results not included in results text
Other bias	Low risk	None apparent