Data saturation was reached in each respondent category after 30 interviews: 12 health-care receiving, nine health-care providing and nine key informant. Saturation was defined as occurring when themes that emerged in each respondent group were reiterated by different participants; and when additional participant transcript analysis revealed no new themes. Respondents included MDR-TB patients currently in the MSF home-based treatment and care programme (n = 7), their family members (5), health-care workers (6), village health team members (3) and identified key informants from the MoH (2) and non-governmental organisations (7). The two MDR-TB patients who did not participate in the study did so due to logistical reasons regarding time availability and working hours. All other approached participants voluntarily agreed to being interviewed.