Validation of data in the USAID ASSIST Project HIV care improvement activities in Tanzania
This report summarizes findings of a validation exercise for data collected and reported for the USAID ASSIST Project in Tanzania as part of PEPFAR-funded technical assistance to improve HIV care in several regions. The USAID ASSIST Project mandates participating countries perform external validation exercises for improvement indicators and collection of data from control sites that do not receive support from USAID. ASSIST engaged the Harvard T.H. Chan School of Public Health as an independent organization to perform these activities in Tanzania.
This report investigates the validity of health indicator data used to inform HIV care improvement activities in three health facilities in the Morogoro Region of Tanzania: one regional hospital and two health centers. The indicators validated addressed both pediatric and adult HIV care and PMTCT in these public health facilities. We sought to determine if the data collected on the indicators during the course of implementing improvement activities were accurate, complete, and indicative of the health care being delivered and the health status of patients receiving care at the participating facilities.
The review of the CTC2 cards found a high proportion of data missing in the sampled cards in all three facilities. There were significant challenges to storage, indexing, and organization of the CTC2 cards and facility registers. Given the severe shortage of clinical staff in these facilities and the demands of a very high patient load, it is important that the system for access to medical records be as efficient as possible. The report offers recommendations to improve clinical effectiveness, data validity, and quality improvement processes at health facilities like these in Tanzania.