eTabelle 3. Studies with improved adherence to medications and improved clinical outcomes in the intervention group.

Article	Study type recruit ment	Population number, age, male, NYHA (I/II/III/IV), comorbidities	Comparison Intervention (IG) vs. control (CG)	Risk of bias (I/II/III/IV /V/VI)	Patient adherence (measurement, followup) IG vs. CG	Conclusions on primary outcome, clinical outcomes and adherence
Antonicelli 2010 (e7)	RCT Italy 16 months	57 hospitalized HF patients 78±7 years, 61% male NYHA: 0/58/37/5% Exclusion of patients with severe dementia, debilitating psychiatric disorders or chronic renal failure requiring dialysis	IG (n=28): Reassessment of therapeutic regime on the basis of telemonitoring and telephone contacts Training course for patients and caregivers in the hospital to use equipment CG (n=29): similar course in the hospital on the importance of adherence	unclear/ unclear/ high/ low/ low/ low	Adherence to prescribed treatment: 12-months: 89.7 vs. 35.7%	Intervention can improve the composite endpoint of mortality and hospitalization and medication adherence (RD 0.47; 95%CI 0.25 to 0.70), but not mortality and quality of life.
Brotons 2009 (e9)	RCT Spain 01/2004 to 09/2005	283 hospitalized HF patients 76±8 years, 45% male NYHA 49/45/5/1% Diabetes: 42% Hypertension: 76% Exclusion of patients with a cognitive deficit	IG (n=144): patient education in the hospital, booklet monthly visits for one year (education, assessment of adherence to prescribed medications and lifestyle habits) Nurses contacted the family physician or cardiologist when necessary Contact per telephone every 15 days to evaluate clinical status by nurses CG (n=139): referral to family physician and/or cardiologist	low/ low/ high/ low/ low/ low	Adherence to pharmacological treatment (high scores are better) 12-months: 86.1 vs. 75.5%.	Intervention can reduce mortality and hospital readmissions, improve QoL and medication adherence (RD 0.11; 95%CI 0.01 to 0.21).
Galbreath 2004 (e11)	RCT USA 1999 to 2003	1069 patients with HF symptoms identified through lists from partner institutions 71±10 years, 71% male NYHA: 19/57/21/3% Diabetes: 28% Hypertension: 72% Hyperlipidemia: 50%	IG (n=710): Assignment of a disease manager and a specialized cardiac nurses Telephone administration of a DM program (first weekly, later monthly over 18 months) Mailed educational material smoking cessation instructions For instable patients contact to a nurse Fax with a call summary to the primary physician toll-free telephone number CG (n=359): usual care by their physicians	unclear/ unclear/ unclear/ unclear/ low/ low	Adherence to guideline-based medications in systolic HF patients 18-months: 54.4 vs. 43.3%	Intervention can decrease mortality, but not event-free survival and improve longtime medication adherence (RD 0.11; 95%CI 0.03 to 0.19). -
GESICA 2005 (e12)	RCT Argentin a 06/2000 to 11/2001	1518 ambulatory stable HF patients 65±13 years, 71% male NYHA III-IV: 49% Diabetes: 21% Hypertension:59% Exclusion of patients with primary pulmonary hypertension	IG (n=760): education booklet telephone follow-up by trained HF nurses (14-day frequency, later according to the needs over 12 months) with monitoring and education nurses could adjust doses of diuretic or recommend non-scheduled medical or emergency visits CG (n=758): followed by their attending cardiologist	low/ low/ unclear/ low/ high/ low	Adherence to medication and diet (mean follow-up of 16 months): beta-blocker: 59 vs. 52% spironolactone: 27 vs. 23% digoxin: 33 vs. 29% furosemide: 77 vs. 70% ACE-inhibitors: 78 vs.6% Drug stop: 8 vs. 18%. dietary transgressions: 20 vs. 65%	Intervention can decrease mortality, readmissions and the probability of worsening HF and improve QoL and medication adherence (no drug stops of any drugs: RD 0.10; 95%CI 0.07 to 0.14 and diet: RD 0.45; 95%CI 0.40 to 0.49).
Sadik 2005 (e6)	RCT United Arab Emirate s	221 HF patients from general medical wards and from cardiology and medical outpatient clinics 59 years, 50% male NYHA: 30/50/16/4% Diabetes: 18% Hypertension: 23% Exclusion of patients with low cognitive status	IG (n=109): rationalization of therapy or simplification of dosage regimes by a research pharmacist and the physician patient education, booklet instructions on a self-monitoring program with a monitoring diary card Discussion of the program by the research pharmacist and patient’s physician CG (n=112): usual care	low/ unclear/ high/ low/ low/ low	Compliance with the prescribed medicines: 12-months: 82 vs. 34% Lifestyle advice: baseline: 21 vs. 22% 12-months: 72 vs. 28%	Intervention can improve QoL and compliance to medications (RD 0.48; 95%CI 0.36 to 0.60) and lifestyle adjustments (RD 0.44; 95%CI 0.32 to 0.56) with no influence on mortality.
Wu 2012 (e5)	RCT USA	82 HF ambulatory and hospitalized patients 60±13 years, 57% male NYHA I-II/III-IV: 51/49% Charlson comorbidity index: 3.1±1.9 Exclusion of patients with impaired cognition	IG (n=54): education of major caregiver s and patients by a cardiovascular nurse expert (weekly, 4 dates) intensive group (n=27): additional feedback to medication-taking behavior CG (n=28): usual care	unclear/ unclear/ low/ unclear/ low/ high	Medication taking adherence: baseline: 70 vs. 59 vs. 64% 9-months: 74 vs. 65 vs. 36%	Intervention improved eventfree survival, hospitalization, but not mortality and QoL. Intervention can improve adherence in both intervention groups (RD 0.38; 95%CI 0.14 to 0.63 and RD 0.29; 95%CI 0.03 to 0.54).

CG, Control group; CI, confidence interval; DM, disease management; HF, heart failure; IG, intervention group; n, number of randomized participants; NYHA, New York

Heart Association; QoL, Quality of life; RD, risk difference; RCT, randomized control trial;

RD>0 describe better adherence in IG

Risk of bias: I, random sequence generation; II, allocation concealment, III, blinding of outcome assessment; IV, incomplete outcome data; V: selective reporting; VI: otherbias