The biggest limitation of this study was that the researcher was not involved in the design of the BES study. Selected BES data was used and this was provided by a third party. Thus the researcher did not involve and have control in data collection, data entry and data validation procedures. However, proper training for data collector prior to data collection, hierarchy of supervision and guidance were given to those interviewers to minimize the likelihood of interviewer bias. Data that could be used to provide extended descriptions of the relationship between TB prevalence and domestic environment conditions were not generated by the BES survey.