High Incidence of Tuberculosis, Low Sensitivity of Current Diagnostic Scheme and Prolonged Culture Positivity in Four Colombian Prisons. A Cohort Study

Luis Carlos Orozco,Zulma Vanessa Rueda,Diana Marín,Lázaro A Vélez,María Patricia Arbeláez,Henry Pulido,David W Dowdy,Lucelly López,Fernando Montes,Margarita Rosa Giraldo

doi:10.1371/journal.pone.0080592

Abstract

ObjectiveTo determine the incidence of pulmonary tuberculosis (TB) in inmates, factors associated with TB, and the time to sputum smear and culture conversion during TB treatment.MethodsProspective cohort study. All prisoners with respiratory symptoms (RS) of any duration were evaluated. After participants signed consent forms, we collected three spontaneous sputum samples on consecutive days. We performed auramine-rhodamine staining, culturing with the thin-layer agar method, Löwestein-Jensen medium and MGIT, susceptibility testing for first-line drugs; and HIV testing. TB cases were followed, and the times to smear and culture conversion to negative were evaluated.ResultsOf 9,507 prisoners held in four prisons between April/30/2010 and April/30/2012, among them 4,463 were screened, 1,305 were evaluated for TB because of the lower RS of any duration, and 72 were diagnosed with TB. The annual incidence was 505 cases/100,000 prisoners. Among TB cases, the median age was 30 years, 25% had <15 days of cough, 12.5% had a history of prior TB, and 40.3% had prior contact with a TB case. TB-HIV coinfection was diagnosed in three cases. History of prior TB, contact with a TB case, and being underweight were risk factors associated with TB. Overweight was a protective factor. Almost a quarter of TB cases were detected only by culture; three cases were isoniazid resistant, and two resistant to streptomycin. The median times to culture conversion was 59 days, and smear conversion was 33. ConclusionsThe TB incidence in prisons is 20 times higher than in the general Colombian population. TB should be considered in inmates with lower RS of any duration. Our data demonstrate that patients receiving adequate anti-TB treatment remain infectious for prolonged periods. These findings suggest that current recommendations regarding isolation of prisoners with TB should be reconsidered, and suggest the need for mycobacterial cultures during follow-up.

Highlights

The incidence of tuberculosis (TB) in prisons is estimated to range from 25.3 to 6799 cases per 100000 prisoners per year, and the risk for TB is greater in prisons than in the general population [median estimated annual incidence rate ratio for TB: 23.0 (IQR: 11.7 – 36.1)][1]
This study yielded the following main findings: 1) a high incidence of TB in prisons; 2) a low sensitivity of persistent cough, which underscores the importance of evaluating for TB in immunocompetent people with a cough of any duration and using culture methods for diagnosis; and 3) a long interval between treatment initiation and conversion to negative sputum by culture, which has implications for the duration of respiratory isolation in confined places such as prisons
The difference may be attributable to real difference in incidence of TB and/or the methods used

Summary

Introduction

The incidence of tuberculosis (TB) in prisons is estimated to range from 25.3 to 6799 cases per 100000 prisoners per year, and the risk for TB is greater in prisons than in the general population [median estimated annual incidence rate ratio for TB: 23.0 (IQR: 11.7 – 36.1)][1]. A study conducted in Cambodia, Thailand, and Vietnam, on patients with HIV infection showed that the presence of a cough for two weeks or more had a sensitivity of 33% for TB. When those authors used the presence of cough of any duration in the preceding 4 weeks, the sensitivity increased to 71%[3]. Given the low sensitivity of the current recommendations, these results suggest that broadening the criteria for testing will result in an increased number of TB cases detected in high-risk prisoner populations

Methods

Results

Conclusion