Patients who are admitted to the emergency unit for decompensated heart failure at the three largest clinics in Stockholm, Sweden, are screened for recruitment. For patients to be eligible for study participation, they shall be discharged from the hospital to the general practitioner without being referred to a nurse led heart failure clinic, which is considered the gold standard in terms of improving patients' self-care behaviour. At the point of discharge, patients are approached about study participation and those who accept will be provided study information and forced to provide a written consent to participate. Continuous sampling is employed. Baseline characteristics are noted, as well as baseline data collected for the following end-points: Self-care, using the European Heart Failure Self-care Behaviour Scale (EHFScB), general quality of life using the SF-36 questionnaire, disease-specific quality of life using the Kansas City Cardiomyopathy Questionnaire (KCCQ) and knowledge about heart failure using the Dutch Heart Failure Knowledge Scale (DHFKS). After this data is collected patients will be randomized, using sealed envelopes, to either receive standard care or receive the home-based tool that constitutes the intervention investigated. Within one week after discharge, the patients randomized to the intervention group will have he tool installed in their home.
The tool consists of a tablet computer locked to a specialized software, and a Bluetooth connected weight scale. Every day the patient is encouraged by the tablet computer to step up on the weight scale. The weight is wirelessly transferred to the tablet computer and today's dose of diuretics is shown on screen together with a short tip about how to improve living with heart failure. Every five days the patient assess their self-perceived symptoms, as prompted on the tablet computer. For every patient equipped with the tool, a patient-specific normal dose of diuretics, as well as a high dose is input in the tool at the point of randomization. If the weight increases by 2kg or more in 3 days or less, the high dose is recommended for two days, after which the effect on the weight is assessed. If the weight was decreased back to where it was prior to the rapid weight gain, the dose is returned to normal and if it is still elevated, the patient will be encouraged to call his/her health care provider at the number presented on the table computer. If the self-assessed symptoms deteriorate, the same type of warning will appear on screen, urging the patient to contact the health care provider. Both the intervention patients and the control patients will be given the same discharge information in writing, describing the most important parts of self-management for heart failure patients. The data collected at baseline will be collected again after 3 months and 6 months after randomization. After 6 months data about hospitalizations will be retrieved from the medical records and adjudicated as either heart failure-hospitalization or not. The adjudication will be done by the cardiologists participating in the study and as the focus of the study is in estimating the net/marginal incidence rate ratio of in-patient hospitalization, absorbing competing events (e.g. death) will be handled by censoring.
