OBJECTIVES

: To demonstrate the difference of relative frequencies in sustained remission at 2 years, between an institution with non-centralized management or standard of care (Hospital Militar Central. HMC) compared to another institution with centralized management or clinic of excellence (BIOMAB-IPS)

METHODS

: In Colombia, every institution that attends patients with RA, must follow up of them with the mandatory clinical registry of the High Cost Account (Cuenta de Alto Costo CAC). The 2-year clinical records for CAC were compared between an institution with non-centralized care in comparison with another institution with centralized care, performing a sociodemographic description, measuring control of the disease with DAS28 clinimetry, Fisher's test , simple lineal regression, average treatment effect (ATT) by Propensity score Matching (PSM)

RESULTS

: 7711 patients were obtained for the centralized care program and 371 patients in the non-centralized care program. Most of them corresponded to women, with time of illness of 9.5 years and 10.2 years, respectively, without statistically significant differences.

Difference was observed in the 2 programs to maintain remission at 2 years, in favor of the centralized program (p <0.00). Binomial generalized linear regression model, confirms this difference, but great bias between the groups was observed. Performed a PSM, finding a match with a propensity score of 1: 1, with a caliper of 0.065, obtaining a pseudo population with well balanced covariates. In the common support area, statistically significant differences were documented in favor of the centralized care group (ATT).

CONCLUSIONS

: With information of national clinical registries, novel statistical strategies can be used to evaluate important differences in centralized care programs, observing favorable results in these care modalities.