Methods	RCT, single centre Study recruitment: June 1996 to January 1998. Duration of follow up: 12 weeks.
Participants	Country: Canada Participants: 100 patients (56 males, 56%) in comparison group, 92 (49 males, 53%) in intervention group. Actual age of study subjects: mean age 76 years, median age 77 years (range 33-93 years). Male sex: 55% Ethnicity: not given. Severity of heart failure in study subjects: NYHA class, n (%) at baseline: controls: I 2 (2%), II 20 (20%), III 69 (69%), IV 8 (8%); intervention group: I 0, II 21 (23%), III 60 (65%), IV 11 (12%) LVEF: not given Study inclusion criteria: Patients admitted with heart failure to the two study units in one centre during recruitment period. Residing in the residential home care radius (60km). Admitted for more than 24 hours. Patients expected to be discharged with home nursing care French or English speaking. Not cognitively impaired (score <8 on Short Portable Mental Status Exam) Study exclusion criteria: Coming from or being discharged to a long term care facility. On life support, having palliative care or in intensive care. Died shortly after admission
Interventions	Duration of intervention: 2 weeks following discharge. Intervention Group: Transitional care (TC) Before discharge: Standard discharge planning and care (see below). Comprehensive, evidence based education programme for heart failure self-management (PCCHF). A nursing transfer letter to the home care nurse detailing clinical status and self-management needs After discharge: Phone call from hospital nurse to patient within 24 hours of discharge. Minimum of two community nurse visits within two weeks of discharge Content of PCCHF: Patient workbook covering: the disease, self-monitoring, management of medication, diet, exercise, stress, support systems and community resources. Allowed tailoring for individual needs. Also contained an education plan and served as a patient held documentation tool Comparison Group: usual care discharge planning and ‘optimal’ usual post discharge care. Before discharge: Ideally a multidisciplinary discharge plan within 24 hours of admission and weekly discharge planning meetings. Regional home care co-coordinator consults with hospital team as required and may meet patients and their families. Immediately before discharge physician completes referral form for home care and necessary services and supplies are communicated with the home nursing agency After discharge: Number of home visits scheduled to match those received by TC group
Outcomes	Primary outcome: HRQL measured by the MLHFQ 6 and 12 weeks post discharge. Secondary outcomes: QOL measured with SF-36. Number of all-cause emergency room visits. Number of all-cause hospital re-admissions. Analysis done on intention to treat basis?: no “Eleven individuals were readmitted to hospital during the intervention period. Timing for outcome measures then began on second discharge and followed for 3 months. ” Also, eight patients who dropped out after randomisation were not included in the study
Notes	Data source: published data and information from author*. Generalisability: 483 patients thought to have heart failure admitted; 212 (44%) were eligible for the study; 12 (6%) refused consent. After randomisation 8 more patients (5 TC and 3 usual care) did not enter the study: four died or became too ill; two refused home care; one changed diagnosis and one was discharged to long term care. 192 patients (40% of admissions) were entered into the study, of these only 157 (82% of those considered entered into study, 78% of those randomised) were followed up Consort flow chart: not supplied Rationale for sample size: given Other points: Reasons for exclusions (271 patients): coming from/ being discharge to long term care (38%); living outside the catchment area (23%); too ill or deceased shortly after admission (15%); first language not French or English (12%); discharged with in 24 hours (6%); diagnosis changed (3%); other (3%) Not clear whether, or how many of, the eleven readmitted and re-entered patients were in intervention or control groups. Generation of randomisation sequence: computer generated schedule. Allocation concealment: “pre-packaged, consecutively numbered, sealed opaque envelopes containing the group allocation were prepared for each nursing unit and administered from the research office. Neither the patients nor the members of the study team were aware of treatment assignment until after randomisation.” Care giver performance bias comment: bias possible since hospital nurses provided both experimental and control interventions and all care givers in the community were informed that patients were in the intervention or control groups after randomisation*. Risk of attrition bias: likely. Only 157 patients (82% of those considered entered into study, 78% of those randomised) were followed up. 23 UC patients and 12 TC patients not followed up, reasons: died or too ill (11 UC, 9TC); withdrew (7 UC, 1 TC); lost to follow up (5 UC, 2 TC). Risk of detection bias: low, outcome assessors masked*.
Risk of bias
Item	Authors’ judgement	Description
Allocation concealment?	Yes	A - Adequate