Treatment delay affects clinical severity of tuberculosis: a longitudinal cohort study

Abstract

Objectives: To describe the risk factors for treatment delay and the effect of delay on the severity of tuberculosis (TB) in a prospectively followed TB cohort at the Bandim Health Project in Guinea-Bissau.

Background: Treatment delay in patients with TB is associated with increased mortality and transmission of disease. However, it is not well described whether delay influences clinical severity at diagnosis. Previously reported risk factors for treatment delay vary in different geographical and cultural settings. Such information has never been investigated in our setting. Change in delay over time is rarely reported and our prospectively followed TB cohort gives an opportunity to present such data.

Participants: Patients were included at the time of diagnosis at three local TB clinics and the national TB reference hospital. Inclusion criteria were age >15 years and diagnosis of TB by either sputum examination or by the WHO clinical criteria. Patients with extrapulmonary TB were excluded.

Primary and secondary outcome measures: The primary outcome was treatment delay. Delay was assessed by patient questionnaires. The secondary outcome was Bandim TBscore as a measure of TB morbidity and all-cause mortality.

Results: A total of 1424 persons were diagnosed with TB in the study area between 2003 and 2010. We included 973 patients with TB in the study. The median treatment delay was 12.1 weeks. Risk factors for delay were low educational level, HIV-1+HIV-2 dual infection and negative sputum smear. TB treatment delay decreased by 10.3% (7.9-12.6%) per year during the study period. Delay was significantly associated with clinical severity at presentation with 20.8% severe TB cases in the low delay quartile compared with 33.9% if delay was over the median of 12.1 weeks.

Conclusions: Long treatment delay was associated with more severe clinical presentation. Treatment delay in TB cases is decreasing in Guinea-Bissau.

Figure 1

The graph describes the relation between treatment delay and mortality. Strata were created from 25th centiles of median treatment delay.

Figure 2

Treatment delay was defined as the time from self-reported onset of TB symptoms to the initiation of specific antituberculosis treatment. All patients were in treatment by the end of the analysis period because only those patients who ultimately began treatment were included in this analysis.

Figure 3

Flowchart demonstrating included individuals.

Figure 4

Median treatment delay during the study period divided into 6 months intervals starting from November 2003. The last interval is 8 months to include all included cases in the figure.