Mechanisms and Contextual Factors Affecting the Implementation of Animal Health Surveillance in Tanzania: A Process Evaluation.

A strong animal health surveillance system is an essential determinant of the health of animal and human population. To ensure its functionality and performance, it needs to be evaluated regularly. Therefore, a process evaluation was conducted in this study to assess animal health surveillance processes, mechanisms and the contextual factors which facilitate or hinder uptake, implementation and sustainability of the system in Tanzania. A mixed-method study design was used to evaluate the national animal health surveillance system guided by a framework for process evaluation of complex interventions developed by Moore and others. The system was assessed against standard guidelines and procedures using the following attributes: fidelity, adherence, exposure, satisfaction, participation rate, recruitment and context. Quantitative and qualitative data were collected using a cross-sectional survey, key informant interviews, document review, site visits and non-participant observation. Data from questionnaires were downloaded, cleaned and analyzed in Microsoft™ Excel. Qualitative data were analyzed following deductive thematic and content analysis methods. Fidelity attribute showed that case identification is mainly based on clinical signs due to limited laboratory services for confirmation. Data collection was not well-coordinated and there were multiple disparate reporting channels. Adherence in terms of the proportion of reports submitted per month was only 61% of the target. District-level animal health officials spent an average of 60% of their weekly time on surveillance-related activities, but only 12% of them were satisfied with the surveillance system. Their dissatisfaction was caused by large area coverage with little to no facilitation, poor communication, and lack of a supporting system. The cost of surveillance data was found to be 1.4 times higher than the annual surveillance budget. The timeliness of the system ranged between 0 and 153 days from the observation date (median = 2 days, mean = 6 days). The study pointed out some deviations in animal health surveillance processes from the standard guidelines and their implication on the system's performance. The system could be improved by developing a user-friendly unified reporting system, the active involvement of subnational level animal health officials, optimization of data sources and an increase in the horizon of the financing mechanism.