Abstract
BACKGROUND: Brazil has implemented systematic control methods for leishmaniasis for the past 30 years, despite an increase in cases and continued spread of the disease to new regions. A lack high quality evidence from epidemiological observational studies impedes the development of novel control methods to prevent disease transmission among the population. Here, we have evaluated the quality of observational studies on leishmaniasis conducted in Brazil to highlight this issue. METHODS/PRINCIPAL FINDINGS: For this systematic review, all publications on leishmaniasis conducted in Brazil from January 1st, 2002 to December 31st, 2012 were screened via Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist to select observational studies involving human subjects. The 283 included studies, representing only 14.1% of articles screened, were then further evaluated for quality of epidemiological methods and study design based on the STROBE (Strengthening the Reporting of Observational studies in Epidemiology) checklists. Over half of these studies were descriptive or case reports (53.4%, 151), followed by cross-sectional (20.8%, n = 59), case-control (8.5%, n = 24), and cohort (6.0%, n = 17). Study design was not stated in 46.6% (n = 181) and incorrectly stated in 17.5% (n = 24). Comparison groups were utilized in just 39.6% (n = 112) of the publications, and only 13.4% (n = 38) employed healthy controls. Majority of studies were performed at the city-level (62.9%, n = 178), in contrast with two (0.7%) studies performed at the national-level. Coauthorship networks showed the number of author collaborations rapidly decreased after three collaborations, with 70.9% (n = 659/929) of coauthors publishing only one article during the study period. CONCLUSIONS/SIGNIFICANCE: A review of epidemiological research in Brazil revealed a major lack of quality and evidence. While certain indicators suggested research methods may have improved in the last two years, an emphasis on observational research which employs comparison groups and representative samples is urgently needed.