Delays in the diagnosis of tuberculosis in a town at the triple border of Brazil, Paraguay, and Argentina

Objective. To identify the factors linked to patients and health services in delays in the diagnosis of tuberculosis. Methods. Epidemiological study in Foz do Iguacu, Parana, Brazil, 2009. The Primary Care Assessment Tool, adapted for appraising tuberculosis treatment, was the instrument used. Descriptive statistics techniques were used, such as frequency distributions, central tendency and dispersion measurements (median and interquartile intervals), and odds ratios. Results. There were greater delays in seeking health services for those in the age group 60 years and older, for females, for patients with low levels of education, and for patients with poor knowledge of the disease. Clinical variables (being a new case and HIV infection) and behavioral variables (use of tobacco and alcohol consumption) were not linked with delays in diagnosis. The median time delays before diagnosis attributable to patients and to the health services were 30 days and 10 days, respectively. Emergency 24-hour medical services and primary health care services were not effective in identifying suspicious cases of tuberculosis and requesting tests to confirm the diagnosis, with a high percentage of referrals to the Tuberculosis Control Program clinic. Conclusions. Going to primary health care services for diagnosis increased the time before diagnosis of the disease was reached. The Tuberculosis Control Program clinic was more effective in diagnosis of tuberculosis, due to the training of the staff and to an organized process for receiving patients, including the availability of tests to support the diagnosis