Reasons for over-reporting of routine immunization administrative data in the Democratic Republic of Congo: a mixed cross-sectional study to determine explanatory factors for poor data quality.

Background: Differences of more than 30% have been observed between the results of vaccine coverage surveys and routine vaccine coverage data. In the context of the organization and operation of the health system, the study focused on investigating explanatory factors for over-reporting.

Methods: This was a mixed-method, cross-sectional, and analytical study. Over-reporting of routine immunization data was defined as a discrepancy of ≥ 10% points between routine data and survey data or recount data (standards) for the Penta3 vaccine. Data were collected by questionnaire from 117 health centers, 30 health zone offices, and 13 provincial health offices. Bivariable and multivariable analyses (α = 5%) were used to find factors influencing over-reporting. Data from 30 in-depth interviews were collected to complement quantitative data.

Results: The phenomenon of over-reporting of routine immunization data was verified in the health zones (90% or 77%) and health centers (43%) surveyed in 2019 and 2020. At the health zone level, six explanatory factors emerged. The most significant tree variables being the pressure exerted on managers to achieve pre-established annual targets (p = 0.016), the availability of data collection tools (p = 0.010) and bearer message for manual transport of reports (p = 0.031). At the health center level, seven factors were found, and the four most significant were: availability of a cell phone (p = 0.002), existence of table or graph for coverage monitoring (p = 0.003), availability of a computer in the health center (p = 0.007) and designated health data collector (p = 0.015). Qualitative data revealed three over-reporting practices: deliberate inflation of vaccine delivery figures, readjustment of expected target population figures, and occasional errors in data transcription.

Conclusion: Over-reporting is essentially generated by providers. Solving this problem requires lifting the pressure exerted on managers at different levels of the health system, making data management more secure, and qualifying the staff responsible for managing immunization data.