Table 2 Outcomes and results

Boult 2011 [35]

RCT

USA

Primary outcome: Health service use

Adjusted ratio of service use: hospital 30 day readmissions 1.01 (95% CI 0.83 to 1.23); hospital days 0.79 (0.53 to 1.16); skilled nursing facility admissions 1.00 (0.77 to 1.30); skilled nursing facilities days 0.92 (0.6 to 1.4); emergency department visits 0.84 (0.48 to 1.47); primary care visits 1.04 (0.81 to 1.34); speciality care visits 1.02 (0.91 to 1.14); home healthcare episodes 1.07 (0.93 to 1.23)

PACIC (Patient Assessment of Chronic Illness Care) score at 18 months adjusted mean difference (aMD) 0.2 95% CI 0.07 to 0.33, p = 0.002

Satisfaction: no difference between groups

Provider satisfaction with care mixed effects

Contant 2019 [27] (Fortin 2016)

RCT

Canada

Primary outcome: Self-management using the Health Education Impact Questionnaire (heiQ) 8 domains

The intervention group showed improvement in 4 of the 8 heiQ domains in multivariate analysis: These four domains were: health-directed behaviour: OR 1.98, 95% CI 1.07 to 3.66, p −0.03; constructive attitudes and approaches: 3.92, 95% CI 1.73 to 8.89, p = 0.001; skill and technique acquisition OR 2.48, 95% CI 1.32 to 4.65, p = 0.005; health service navigation OR 2.73, 95% CI 1.2 to 6.22, p = 0.02.

There were no significant improvements in positive and active engagement in life, emotional well-being, self-monitoring and insight and social integration and support.

Secondary outcomes were not reported in this secondary data-analysis study of Fortin 2016

Gonzalez Ortega 2017 [30]

RCT

Spain

Primary outcome: Emergency admissions

After 6 months, urgent visits per patient decreased in intervention 1.27 baseline versus 0.89 follow-up, p = 0.091 and control 1.06 baseline versus 0.86 follow-up, p = 0.422, mean difference 0.18 [95% CI −0.48 to 0.84].

HRQoL SF12 significant effect on physical component score (aMD −4.71, 95% CI −9.03 to −0.41, p = 0.02) but no effect on the mental component score (aMD 2.6, 95% CI −3.9 to 9.11, p = 0.42)

No significant effect on clinic visits; Charlson score; Function (Barthel); HRQoL; Cognitive status (Pfeiffer test); Pressure Ulcer risk (Norton scale); Social risk (Gijon Test); Caregiver Burden (Zarit test); chronic treatment (number of repeat medicines) or resource use (direct costs)

Hochhalter 2010 [31]

RCT

USA

Primary outcome: Patient activation measure (PAM).

PAM Intervention 66.8 (18.5) vs Control 66.2 (13), no significant difference, all groups had significant improvement from baseline

Significant improvement in self-efficacy compared to usual care (but attention control group also had a significant improvement).

No difference in total unhealthy days and self-rated health

Mercer 2016 [38]

Cluster RCT

(exploratory)

Scotland

Primary outcomes: Health-related quality of life (EQ-5D-5L) and well-being (W-BQ12)

EQ5D Index scores: 0.06 (95% CI −0.02 to 0.14, p = 0.15)

EQ-5D-5L area under the curve over the 12 months was higher in the CARE Plus group (p = 0.002).

CARE Plus significantly improved one domain of well-being (negative well-being), with an effect size of 0.33 (95% confidence interval [CI] 0.11–0.55) at 12 months (p = 0.0036). Positive well-being, energy, and general well-being (the combined score of the three components) were not significantly influenced by the intervention at 12 months.

No significant difference in anxiety and depression (HADS); self-efficacy self-esteem and medications

Cost Effectiveness Analysis: Within-trial cost-utility analysis based on the EQ-5D-5L utility scores, and on health service utilisation: Adjusted mean difference in cost of GBP929 (95% CI 86 to 1788) per patient

Gain in QALY 0.076 (95% CI 0.028‑0.124)

Cost effectiveness ratio (CER) GBP12,224 per QALY

Salisbury 2018 [40]

Cluster RCT

UK

Primary outcome: HRQoL (EQ-5D-5L)

No difference between groups with EQ-5D-5L aMD 0·00, 95% CI –0.02 to 0.02; p = 0·93.

PACIC score: aMD 0.29 (95% CI 0.16 to 0.41)

Continuity of care score: adj MD 0.081; 95% CI 0.02 to 0.13

Mean Consultation and Relational Empathy (CARE) score (for doctor consultations): aMD 1.2; 95% CI 0.28 to 2.13

Mean CARE score (for nurse consultations): aMD 1.11; 95% CI 0.03 to 2.19

Higher proportion of intervention patients were very satisfied with their care (42%) compared to those receiving usual care (39%) (MD 1.58, 95% CI 1.19 to 2.08, p = 0.0014).

No significant differences in Self-rated health; Bayliss measure of illness burden; depression and anxiety (HAD scale); Treatment burden (MTBQ); Medication adherence (Moriskey measure) and number of medications; Number high risk prescriptions; Healthcare utilisation (GP and nurse visits, OPD visits and admissions) and Quality of care (QOF indicators)

Cost-effectiveness: 50.8% chance of being cost-effective at a willingness-to-pay threshold of GBP20 000 per QALY (55.8% at £30 000 per QALY). Reported as ‘equivocal cost-effectiveness’

Schafer 2018 [41]

Cluster RCT

Germany

(HRQoL data from author)

Primary outcomes: Number medications and HRQoL (EQ-5D)

No difference group in the change of the number of medications taken: 0.43, 95% CI −0.07 to 0.93; p = 0.094

No difference in EQ-5D index score: aMD 0.03; 95% CI −0.03 to 0.09; p = 0.302.

Increase in prescribing of analgesics in the intervention group (Adjusted RR 2.043, P = 0.019)

No significant differences in

patient satisfaction; patient empowerment; depression; healthcare utilisation or in direct costs reported using Leipzig supply and Cost Instrument

Sommers 2000 [42]

RCT

USA

Primary outcome: Health service use

Odds ratio admissions/patient/year 0.63 (95% CI 0.41 to 0.96); ≥ 1 60 day readmissions 0.26 (0.08 to 0.84).

Not fully reported for seven other outcomes, non-significant for six. Difference in adjusted mean scores, social activities count 0.50 (95% CI 0.02 to 1.00). Symptom scale 0.50 (−3.20 to 0.16), SF-36 self-rated health 0.10 (−0.27 to 0.02), not reported for four other outcomes, non-significant

Social activities count: Int = 0.2 vs Con −0.3, p = 0.04

No significant differences in patient reported health status; social activities count; HRQoL (SF36); depression scores; nutrition checklists and drug adherence

Eakin 2007 [28]

RCT

USA

Primary outcome: Dietary behaviour, and physical activity

Adjusted means (SE):

dietary behaviour (lower score better) 2.20 (0.05) v 2.41 (0.05), p < 0.5; change minutes walking/week 8 (22) v −10 (27), p > 0.5

Support for healthy lifestyle (higher score better) 2.98 (0.06) v 2.68 (0.06), p < 0.05

Garvey 2015 [29]

RCT

Ireland

Primary outcome: Activity participation

Frenchay Activities Index aMD at immediate follow up 4.22, 95% CI 1.59 to 6.85.

Significant improvements in perceptions of activity performance and satisfaction, self-efficacy, independence in daily activities and HRQoL (EQ-5D VAS scores only).

The intervention group demonstrated significantly higher levels of goal achievement, following the intervention.

No significant differences in anxiety, depression, HeiQ scores or healthcare utilisation.

O’Toole 2019 [33]

RCT

Ireland (data from authors)

Primary outcomes: HRQoL (EQ5D) and Activity Participation (FAI)

At 6-month follow-up there were no differences in primary outcomes:

EQ5D index score aMD = 0.1; 95% CI −0.02 to 0.22

FAI aMD = 1.20; 95% CI −0.89 to 3.29

No significant difference in Activities of daily living (NEADL); Anxiety and depression (HADs); Self-efficacy and healthcare utilisation. One of the two occupational performance domains (COPM) showed a significant difference.

There were two pre-planned sub-group analyses for the primary outcomes. There was no difference in effects by number of conditions but there was a significant improvement in the EQ5D VAS in those aged < 65 compared to those ≥ 65 years, a 23.13, 95% CI = 3.19 to 43.06, p = 0.0284.

Reed 2018 [34]

RCT

Australia

Primary outcome: Self-rated health

intervention were more likely than control participants to report improved self-rated health at 6 months: Odds Ratio (2.50, 95% CI, 1.13 to 5.50, p = 0.023).

No significant differences in Fatigue; Pain; Health distress; Energy; Depression; Illness intrusiveness; Exercise; Medication adherence; Self-Efficacy; Health Education Impact (HEiQ); Healthcare utilisation (GP visits, Emergency Department (ED) visits and admissions)

Jager 2017 [36]

Cluster RCT

Germany

Primary outcome: Summary score of 10 prescribing indicators

The increase in the degree of implementation was 4.2 percentage points (95% CI −0.3 to 8.6) higher in the intervention group compared to the control group (p = 0.1).

Harms were not expected or reported

No significant difference in Patient Activation Measure (PAM-13D); Medication Adherence Report Scale (MARS); Beliefs About Medicines Questionnaire (BMQ-D) and % Potentially Inappropriate Medicines (PIMs)

Koberlein Neu [37]

cRCT 2016 (stepped wedge design)

Germany

Primary outcome: Quality of medication therapy (mean MAI score)

Mean MAI score: Intervention phase 1 vs Control Phase, aMD −4.51, 95% CI −6.66 to −2.36

Mean reduction in drug-related problems of −0.45, 95% CI −0.81 to −0.09

No significant difference in Number of drug-related problems (DRPs); Potentially inadequate medication (PIM); Number of prescribed medicines per patient; HRQoL (SF12); Function (Barthel Index); Instrumental Activities of Daily Living (IADL); Gait stability/risk of falling (Tinetti score)

Level of social support results not reported.

Krska 2001

RCT [32]

UK

Primary outcome: Pharmaceutical care issues. [outcome trial specific]

Pharmaceutical care issues (%) resolved after intervention: 82.7% v 41.2%, p < 0.001

No significant differences in medicine costs, HRQoL (SF36 scores) and health service use

Muth 2018 [16]

Cluster RCT

Germany

Primary outcome: Medication Appropriateness Index (MAI) at 6 months

No significant effect on mean MAI sum scores with aMD of 0.7 (95% CI −0.2 to 1.6)

Functional status (Vulnerable Elderly Survey-13) MD 0.4, 95% CI 0.0 to 0.8, p = 0.047

No significant difference in all other secondary outcomes including MAI at 9 months; HRQoL EQ-5D (aMD 2.3; 95% CI −1.6 to 6.2, p = 0.247); All-cause hospitalisation; Severity of chronic pain (von Korff Index); Satisfaction with shared decision-making (Man-Son-Hing Scale); Patient’s future expectation, expected/desired lifetime duration; Years of Desired Life (YDL); Medication adherence: Observed adherence: drug score, dose score, regimen score; Self-reported adherence (Morisky); Patient Beliefs about Medicines Questionnaire (BMQ);

Medicines prescribed; Medication Regimen Complexity Index and number of prescriptions/single doses