Adherence to treatment and supervision for tuberculosis in a DOTS programme among pastoralists in Uganda.

Abstract

To estimate the levels of adherence to anti-tuberculosis medications and associated risk factors in a pastoral community practising transhumance in North-Eastern Uganda. A cross-sectional retrospective study of a cohort of tuberculosis (TB) patients and their community volunteers registered from August 2008 to July 2009. Self-reported data were collected on doses missed, doses not supervised, sociodemographic situation, drug availability, perceived drug side effects and use of the directly observed treatment (DOT) card. Medication adherence was defined as having ingested at least 90% of doses given, and supervision adherence was defined as direct observation of at least 90% of doses ingested. Independent predictors of medication adherence were identified using binary and multivariate logistic regression. A total of 126 TB patients and an equal number of community volunteers were enrolled in the study. Medication and supervision adherence were estimated at respectively 72% and 63%. Independent predictors of medication adherence were perceived drug side effects (adjusted odds ratio [aOR] 5.0, 95%CI 1.86-13.6), running out of drugs (aOR 5.97, 95%CI 2.27-15.70) and DOT card not filled in (aOR 6.65, 95%CI 2.08-20.66). Adherence among pastoralists is less than optimal due to poor institutionalisation of the DOTS strategy in a pastoralist community.