Data quality auditing for the Tanzania vitamin A supplementation program

Background: Distribution of Vitamin A Supplementation (VAS) in Tanzania is conducted twice-yearly and has contributed to the reduction in under-five mortality from 147/1000 to 112/1000. After each campaign, reports generated by districts using tally sheets show consistently higher coverage rates than other assessments including Tanzania Demographic Health Surveys (TDHS) and Post Event Coverage surveys (PECs). These discrepancies called for a systematic review to identify reasons for over-reporting through tally sheet data.

Methodology: A random sampling of health facilities was conducted within the eight health zones in Tanzania. Tally sheet data and submitted coverage reports were collected from each health facility, and coverage was recalculated to determine if it was reported correctly on reports from that facility and from higher levels.

Results: Tally sheets and reporting forms were available in 84% and 71% of facilities, respectively. Only 50% of the visited facilities had an operational definition of how to calculate coverage. Reports could not be found in 47% of the visited facilities. For the reports that were found, 24% of them had no missing information and 69% had data transferred correctly from the original reports to reports from higher levels. When coverage was recalculated, discrepancies were noted in 25% of reports.

Conclusion: Coverage reports based on tally sheets face multiple sources of error including incorrect calculations by district officials, incomplete reports and incorrect data capture when transferring information from lower to higher levels. Given these limitations, additional monitoring is necessary to confirm the accuracy of tally sheet data.