Adesão e sobrevida de pacientes em um programa de assistência medicamentosa para doença de Alzheimer no Rio de Janeiro: um estudo baseado em relacionamento de bases de dados

Abstract

Alzheimer's disease (AD) affects today more than one million people in Brazil and leads a large impact not only in health, but also social and financial. The proper use of medicines and integrated care programs are recommended as best practice in this disease. In 2002 the Ministry of Health created the Pharmaceutical Assistance Program for Persons with Alzheimer's disease that ensures to patients with this pathology access to expert evaluation and high cost drug treatment. However, for better planning of actions related to the program is important to understand the morbidity and mortality patterns in the target population, the quality of care offered and their potential impact. Directed to this scenario, this thesis aimed to study the adherence, survival and mortality in a population with Alzheimer's assisted in pharmaceutical assistance program. For presentation of this research, the results were divided into three articles. In the first article, it was evaluated the feasibility of using the refill frequency information collected in the construction of measures of adherence, a methodology already explored in the international literature but a limited experience in Brazil. Data was accessed in the form of authorization procedures High Cost (APAC) storaged in the database of the medicines access program of the Ambulatory Information System (SIA). As result, it demonstrated the possibility of creating multiple adherence measures with potential application in the field of pharmacoepidemiology and public health. In the second article, the focus was to understand how individual, treatment or service factors could be related to an increased risk of leaving the program. As result, factors such as female gender, number of comorbidities in APAC, first cholinesterase inhibitor prescribed and refilling irregularity were associated with a higher risk of dropping out of the program. In the third article, by using the techniques of probabilistic database, it was added information about the mortality patterns of this disease by relating it the death certificate data of the Mortality Information System (SIM). About 40% of the patients died during the follow-up period. The main cause of mortality was the same Alzheimer's disease (19%). However, overall, the disease was poorly reported in any of the fields of the certificate (31%). Older age and male gender were the only variables associated with a shorter survival. The persistence of the program in six months and the different levels of adherence, measured by the proportion of possession of medication, showed no association with survival. The conclusion was that the data available after applying the method allowed the investigation of the pattern of mortality and use of an Alzheimer's disease s assistance program. The ability to analyze large populations on an ongoing basis, with objective measures and a relatively low cost may outweigh the limited scope of the information on individuals and disease. Policies that support not only an increase in the availability, quality and scope of information but also advances in administrative databases research methodologies should be a health priority, as they contribute to the creation of information relevant to a rational use of resources and improvement in care practices.