Int J Health Policy Manag. 2023;12:6896. doi: 10.34172/ijhpm.2023.6896. Epub 2023 Feb 13.
ABSTRACT
BACKGROUND: Digital information management systems for health financing are implemented on the assumption thatdigitalization, among other things, enables strategic purchasing. However, little is known about the extent to which thesesystems are adopted as planned to achieve desired results. This study assesses the levels of, and the factors associated withthe adoption of the Insurance Management Information System (IMIS) by healthcare providers in Tanzania.
METHODS: Combining multiple data sources, we estimated IMIS adoption levels for 365 first-line health facilities in2017 by comparing IMIS claim data (verified claims) with the number of expected claims. We defined adoption as abinary outcome capturing underreporting (verified<expected) vs. not-underreporting, using four different approaches.We used descriptive statistics and analysis of variance (ANOVA) to examine adoption levels across facilities, districts,regions, and months. We used logistic regression to identify facility-specific factors (ie, explanatory variables) associatedwith different adoption levels.
RESULTS: We found a median (interquartile range [IQR]) difference of 77.8% (32.7-100) between expected and verifiedclaims, showing a consistent pattern of underreporting across districts, regions, and months. Levels of underreportingvaried across regions (ANOVA: F=7.24, P<.001) and districts (ANOVA: F=4.65, P<.001). Logistic regression resultsshowed that higher service volume, share of people insured, and greater distance to district headquarter were associatedwith a higher probability of underreporting.
CONCLUSION: Our study shows that the adoption of IMIS in Tanzania may be sub-optimal and far from policy-makers’expectations, limiting its capacity to provide the necessary information to enhance strategic purchasing in the healthsector. Countries and agencies adopting digital interventions such as openIMIS to foster health financing reform areadvised to closely track their implementation efforts to make sure the data they rely on is accurate. Further, our studysuggests organizational and infrastructural barriers beyond the software itself hamper effective adoption.
PMID:37579470 | DOI:10.34172/ijhpm.2023.6896
