DrPH Dissertation: Angela Porter

Dissertation Title: Operationalizing “DATA FOR ACTION”: Evaluation of data triangulation use and impact on immunization programs in three African countries

Dissertation Chair: Dr. Kimberly Freire

Dissertation Abstract:

Introduction: Immunization programs routinely collect data from multiple sources, yet these data are often managed and analyzed separately, limiting their use for timely programmatic decision-making. Data triangulation provides an approach for integrating and interpreting multiple data sources to identify program gaps and inform action. This study evaluated how data triangulation was operationalized and integrated into routine immunization and vaccine-preventable disease (VPD) surveillance programs in Nigeria, Rwanda, and Uganda.

Methods: A mixed-methods cross-country analysis was conducted using desk review, key informant interviews (KIIs), journey mapping, and quantitative assessment of dashboard usability. KIIs included 55 national and sub-national program staff across the three countries. In Nigeria and Rwanda, usability of automated triangulation dashboards was assessed using the 19-item Computer System Usability Questionnaire (CSUQ). Uganda was excluded from the usability assessment because an automated dashboard had not been implemented.

Findings: Countries operationalized triangulation through different technical approaches. Rwanda implemented an automated, decentralized dashboard integrated into DHIS2; Nigeria implemented an automated dashboard with more targeted sub-national access; and Uganda primarily used manual and semi-automated analyses. Across countries, triangulation was most consistently integrated through existing program processes, particularly routine data reviews and coordination meetings. Automated dashboards were perceived positively, with overall CSUQ mean scores ranging from 6.02 to 6.59 on a seven-point scale. However, usability did not ensure consistent routine use, which was influenced by staff capacity, infrastructure, access, leadership, indicator relevance, and program ownership. Triangulation was used to identify data quality issues, immunization performance gaps, un- and under-immunized populations, surveillance gaps, and VPD outbreaks. Findings informed actions including supportive supervision, data-quality correction, outreach, vaccine management, and outbreak response.

Conclusion: Data triangulation can strengthen the routine use of immunization and VPD surveillance data to identify program gaps and inform action. Automation can reduce analytic burden but is not required to operationalize triangulation. Institutionalization depends on embedding triangulation within routine workflows, building staff capacity, ensuring access and infrastructure, and establishing a sustained pathway from identification and investigation of program signals to action and follow-up. Data triangulation should therefore be approached as a routine data-to-action function rather than a stand-alone technology.