Efetividade da terapia de ressincronização cardíaca no Sistema Único de Saúde do Brasil: análise pelo relacionamento probabilístico de dados

Abstract

The impact of cardiac resynchronization therapy (CRT) in the public health system of middle-income countries is unknown. This scenario is a great challenge for health care administrators who wish the rational use of available resources and quality health services. The aim of the study was to evaluate the effectiveness of CRT in the Unified Health System (SUS, acronym in Portuguese) of the country, from 2002 to 2007, analyzing the impact of the procedure on hospitalizations for heart failure (HF) and overall survival of patients undergoing implantation. Methods: the probabilistic record linkage strategy was used to match 3526 records from National Database of Multisite Pacemaker (SIH-MPMS) with Mortality Information System (SIM, acronym in Portuguese) and later, with the Hospital Information System of the SUS (SIH-SUS, acronym in Portuguese), of which were selected only data relative to hospitalizations for HF (SIH-HF). The SIH-MPMS consisted of all patients undergoing CRT alone (CRT-P) or combined with cardioverter defibrillator (CRT-D), carried out at SUS hospitals or hospitals that have agreement with the SUS, in the period from January 1st, 2002 to December 31st, 2007. Kaplan-Meier survival curves were plotted and Cox regression model was applied to evaluate predictors of survival. Hazard-ratios and their confidence intervals (CI) of 95% were calculated. After linkage between SIH-MPMS and SIH-HF, only patients who had hospitalizations for HF, twelve months before or twelve months after implantation, were selected for analysis. Paired t test was applied for dependent groups. Results: After linkage between SIH-MPMS and SIM, the mean age of patients at implant was 59.8±13.3 years. The CRT-P (79.7%) predominated over CRT-D (20.3%). At 1 and 5 years, overall survivals were 80.1% (95% CI 79.4-80.8) and 55.6% (95% CI 54.6-56.6) respectively, and were better in women (p <0.001 by log-rank). Overall survivals of CRT-P and CRT-D were not statistically different in this group of patients (p=0.237 by log-rank). Age in decades (HR, 1.04; 95% CI 1.00- 1.09; p=0.038), sex (HR, 1.42; 95% CI 1.26-1.60; p<0.001) and years of implant (HR, 1.21; 95% CI 1.07-1.38; p=0.003) formed final model of Cox regression. The comparison between the number of hospitalizations for HF, twelve months before and twelve months later implantation, analyzed after the linkage between SIH-MPMS and SIH-HF, revealed a reduction of 1.66 ± 1.79 hospitalizations for HF per year to 1.08 ± 1.56 (p<0.001), i.e., an average reduction of 0.58 ± 2.12 hospitalizations for HF/year. Conclusions: The effectiveness of CRT verified in the real world is similar to that of the large clinical trials. There was reduction in hospitalizations for HF and, throughout the years assessed, there was improvement in survival of the patients submitted to implant in the Brazilian public health system. Economic analyses are still needed to evaluate cost-effectiveness of the procedure.