Brazilian Heart Insufficiency With Telemedicine

1. AIM The aim of BRAHIT is to test a new model of care for patients with heart failure in Rio de Janeiro since the investigators can apply well-defined international guidelines for high-quality management of heart failure patients as success criteria, which have been endorsed and adapted in Brazil. Ultimately to provide a policy advice for innovative health management with telemedicine.
2. BACKGROUND Previous clinical trials that have randomized patients to conventional care or home monitoring with telemedicine solutions have failed to demonstrate a benefit of telemedicine. This is probably due to a selection bias. Accordingly, if only patients with high e-literacy, education and social level are selected for participating in telemedicine studies, it is probable that they have maximum advantage from any treatment, thus making it difficult to assess a benefit with telemedicine. In contrast, weaker patients who truly need a close care remain out of reach for cardiology expertise, remaining under primary care management, with scarce resources. Another problem is that, once the participants have access to the new healthcare management to be studied, conventional management gets unavoidably changed and comparison is no longer possible. Therefore, in BRAHIT, randomization will not occur at the individual patient level, but through cluster randomization at the organizational level of the primary care clinics. Successful multifaceted programs for heart failure are based on multiple visits to cardiology outpatient clinics, and patients can be safely transferred to Primary Care, but can be too strenuous for frail patients or impossible in remote areas BRAHIT includes quality indicators that allow monitoring quality of care; telemedicine-based tools for patient education, to enhance e-health literacy and innovative communication and data-exchange to support integrated care between health sectors. Telemedicine in BRAHIT can bring cardiology expertise to socially frail patients and, thereby, overcome social inequality. The Brazilian primary care sector is challenged with long waiting lists for referrals to specialist visits, elective surgeries, and to access higher technology services . This is aggravated in the last years, as Brazil is facing austerity measures. The Brazilian primary Health Care Model has evolved to almost 60% coverage, based on Family Health Teams, composed by specialized general practitioners, a nurse, nursing technicians and community health agents, while the rest are Family Health Teams with "generalists" (under-graduated physicians). The municipality of Rio de Janeiro, as it is particularly suitable for this project, with roughly six million habitants and a large social and cultural diversity. About 80% of the population has access to internet, electronic medical records and communication and information technology and Homecare teams are able to use social media and blogs in communicating with the local population. Therefore, it is feasible to use telemedicine communication in association with homecare teams to reach the entire population living in Rio de Janeiro. Furthermore, in BRAHIT, Primary care physicians training programs on heart failure will be based on action mapping, with adaptive technology-enhanced distance learning.
3. METHODS Participants Patients discharged from Instituto Nacional de Cardiologia (INC) with the heart failure as the primary reason for admittance (ICD-10): (DI-11.0, 13.0, 42.0, 42.6, 42.9, 50.0, 50.1, 50.9) and requests from Primary care to INC on questions for management of patients with heart failure diagnosis. Background population Rio de Janeiro State has nowadays 2,888 family health/homecare teams , covering 9,695,782 (57.99%) people of the State Population, and Rio de Janeiro municipality has 1.164 Family Health Teams, covering 4,016,835 people (61.61%) of the city population. Hypothesis A new collaboration model across sectors, using telemedicine aid, can improve the quality of care management of chronic heart disease, and reach out to weak patients. Concretely, the investigators test the hypothesis that the proportion of patients receiving correct medications for heart failure will increase from 30% to 60% in patients managed according to the BRAHIT model. Furthermore, the investigators hypothesize, that use of the BRAHIT model will result in a reduction of number of readmissions for heart failure within 90 days after discharge from 50% (current proportion of readmissions) to 30% (secondary end-point). Project design Since BRAHIT intends a complex public health intervention, it is necessary to apply mixed methods in a pragmatic manner to analyze the development and impact of the intervention. During an initial feasibility phase, unmet needs will be described based on ethnographic field observations on patients belonging to one Family Clinic that agrees to take a role as pilot-test clinic. Equivalent observations will be registered from patients discharged from INC with the selected heart failure diagnoses, who belong to other Family clinics or Basic Clinics discharged from INC Hospital. Patients will be informed that new procedures are being tested at INC Hospital. Data will be gathered for all patients in current clinical settings. Cluster randomization will follow with telemedicine intervention or usual care stratified by Family clinics and Basic Clinics willing to participate. Patients will be followed according to which primary care the patients belong to. The patients will be informed on what cluster their Primary Care Clinic and invited to be followed, including interviews for at least one year, until the end of the study. Follow-up details To quantify adequate use of target medications in the subgroup of patients with heart failure and reduced ejection fraction, the investigators construct a Brazilian Heart Insufficiency with Telemedicine (BRAHIT) score based on changes of clinical parameters, where higher score indicates worsening:

• NYHA: improved or I (0 points) / no improvement (1 point) / worsening (2 points)
• Weight: closer to target or on-target (0 points) / no change +/- 1.9 Kg since last measurement (1 point), increase 2 Kg or more (2 points)
• ACE-I/ARB: Target dose (0 points) / tried but not reached target or not tolerated (1 point) / Never used (2 points)
• Betablockers: Target dose (0 points) / tried but not reached target or not tolerated (1 point) / Never used (2 points)
• Mineralocorticoid antagonist: Target dose or not indicated (0 points) / tried but not reached target or not tolerated (1 point) / Never used (2 points) Each patient record will have the option "unknown", when no data are available at each follow-up timepoint, to provide an indicator for data quality in the trial. Qualitative measurements Ethnographic observations throughout the study will be recorded and summarized at the end of the feasibility phase, identifying patient groups and elements in the process needed to reach the patients who have difficulties to receive specialized assistance when needed and an equivalent summary will be repeated based on subsequent observations until the end of the study, aiming to elucidate how use of low-cost technologies can equally reach frail and resourceful patients, regardless of the level of health literacy. Open-questions interviews will be conducted during the cluster-randomization phase to evaluate how successful the approach is from the patient perspective. Statistics These are a priori calculations based on expected outcomes, and they will be re-calculated based on the observations acquired during the pilot phase. Categorical variables will be described as absolute and relative frequencies, and continuous variables as medians (interquartile range) or means (standard deviation) according to the distribution of the variable. Normality will be evaluated with the visual inspection of histograms and application of the Shapiro-Wilks test of normality. Comparison of categorical variables will be analyzed with the Fisher exact test or chi-square for the relationship between intervention and outcomes, whereas continuous variables will be analyzed with the Mann-Whitney U test. To assess the relationship between intervention and outcomes the investigators will construct models of analysis of variance (ANOVA). These tests will be calculated as two-tailed, and p values < 0,05 will be considered significant. For evolution over time the investigators will perform Kaplan-Meier tests. The investigators use SAS, R and STATA. Sample size calculations for primary outcome Mortality and risk of readmission can be decreased roughly by 10% when patients are managed by multidisciplinary teams, compared to primary care management. There are no currently available data for an accurate power calculation for the population of Rio De Janeiro. Yet, based on the BREATH registry, almost 50% of the patients admitted with heart failure diagnosis are re-admitted to the hospital within 90 days after discharge, in most cases due to lack of use of relevant heart failure medications. Accordingly, 30% of patients in the cohort were admitted due to inadequate medicines for heart failure. Assuming a similar success to that dedicated clinics, use of beta-blockers, the most triggered treatment compliance due to side effects, can rise from less than 40% to 89% (Danish experience) or from 35 to 42% (Germany). Based on these effects, and counting on two parallel clusters, managing 26 patients in the novel collaborative BRAHIT model, there should be an improvement of adherence to the treatment from 30% to 60%, three months after discharge, with 90% power and a significance level of 5%. This shall be re-calculated with most recent data at the end of BRAHIT feasibility phase. Secondary outcomes In addition to the 30% admissions due to inadequate medications, many re-admissions may be due to insecurity from the primary care side. The investigators expect that the BRAHIT intervention provides a 20% absolute reduction of readmissions.

1. PUBLICATIONS AND DISSEMINATION The protocol and the results of the pilot phase of BRAHIT will be published in international peer-reviewed journals, the latter according to the extended CONSORT checklist for quantitative measurements. The findings and the evolution of the project will be described following the Template for Intervention Description and Replication (TIDieR) checklist.

Additionally, the project aims a high outreach through presentation in popular scientific media for the broader public and to publish a short policy advise that can facilitate implementation of the findings in BRAHIT.