Monitoramento de tempos de sobrevivência com fração de cura: aplicação a portadores de câncer de boca e orofaringe no RN

Abstract

Control charts, ideally based on production processes, can be used in areas related to health to verify a change in the quality of Medical procedures. However, it is important to admit heterogeneity in the data, taking into that particular characteristics of patients before undergoing a procedure can lead to different risks of death.In this context, CUSUM charts (Cumulative sum) are introduced, which incorporate the heterogeneity of the data through regression structures. Some of these charts are intended to monitor data from the medical area in which the response is the time to the occurrence of an event and is subject to censorship on the right. In this context, the RAST CUSUM chart was proposed to monitor the life time of patients, incorporating accelerated failure time models into their risk adjustment process. The RACUF CUSUM chart has been proposed as an extension of RAST to monitor data that have a fraction of curing through a promotion time model. This study aimed to study the RAST CUSUM and RACUF CUSUM control charts for monitoring survival times and the application of these graphs to oral and oropharyngeal cancer data from patients treated at a referral hospital for cancer in the RN, that presented evidence of the existence of cured individuals. As a result, the RACUF chart was more efficient than RAST detecting an increase in the quality of hospital services.