Direct Observation (DO) for Drug-Resistant Tuberculosis: Do We Really DO?

Introduction Directly-observed therapy (DOT) is recommended for drug-resistant tuberculosis (DR-TB) patients during their entire treatment duration. However, there is limited published evidence on implementation of direct observation (DO) in the field. This study aims to detail whether DO was followed with DR-TB patients in a Medecins Sans Frontieres (MSF) tuberculosis program in Mumbai, India. Methods This was a cross-sectional, mixed-methods study. Existing qualitative data from a purposively-selected subset of 12 patients, 5 DOT-providers and 5 family members, were assessed in order to determine how DO was implemented. A questionnaire-based survey of DR-TB patients, their DOT-providers and MSF staff was completed between June and August 2014. Patients were defined as "following Strict DO" and "following DO" if a DOT-provider had seen the patient swallow his/her medications "every day" or "most of the days" respectively. If DO was not followed, reasons were also recorded. The qualitative data were analysed for theme and content and used to supplement the questionnaire-based data. Results A total of 70 DR-TB patients, 65 DOT-providers and 21 MSF health staff were included. Fifty-five per cent of the patients were HIV-co-infected and 41% had multidrug-resistant-TB plus additional resistance to a fluoroquinolone. Among all patients, only 14% (10/70) and 20% (14/70) self-reported "following Strict DO" and "following DO" respectively. Among DOT-providers, 46% (30/65) reported that their patients "followed DO". MSF health staff reported none of the patients "followed DO". Reasons for not implementing DO included the unavailability of DOT-provider, time spent, stigma and treatment adverse events. The qualitative data also revealed that "Strict DO" was rarely followed and noted the same reasons for lack of implementation. Conclusion This mixed-methods study has found that a majority of patients with DR-TB in Mumbai did not follow DO, and this was reported by patients and care-providers. These data likely reflect the reality of DO implementation in many high-burden settings, since this relatively small cohort was supported and closely monitored by a skilled team with access to multiple resources. The findings raise important concerns about the necessity of DO as a "pillar" of DR-TB treatment which need further validation in other settings. They also suggest that patient-centred adherence strategies might be better approaches for supporting patients on treatment.