Monitoring Data Collection
As programs were being designed and rolled out, IHME worked in close collaboration with global partners and individual implementation teams to design, coordinate, and execute a flexible monitoring system that aligned with the continuum of care. Timely reports and feedback were generated by IHME regarding the progress of program execution in order to accumulate longitudinal information for implementation teams and to facilitate the scale-up of intervention programs to meet targets over the implementation period.
Survey Data Collection
All survey questionnaires were designed by IHME. Patient exit interviews with biomarkers and health facility surveys were implemented in India and in South Africa in intervention and comparison areas. In the US and in Brazil, due to financial constraints, IHME did not collect quantitative data; rather, the analysis relies on data that was collected and collated by the grantees and shared with IHME.
From all sources of quantitative data, unless indicated otherwise, the average value across groups or the percentage of individuals or facilities included in a given category (e.g., percentage of facilities that stocked at least one key blood pressure medication; percentage of patients who were diagnosed with diabetes and had an A1c measure less than 8%) was estimated. For all indicators of interest, 95% confidence intervals (CIs) were computed. Confidence intervals aim to capture the range of likely values for a given measure while accounting for how much the measure might vary among individuals or facilities. When CIs are relatively narrow, this can mean there is less variation among individuals or facilities for a given measure; it can also mean a large enough number of individuals or facilities were included in the analysis and thus provided a more precise estimate of the indicator. When CIs are wide, it can mean individuals or facilities might vary a lot on a given measure; it can also mean that a relatively small number of individuals or facilities could be included and thus it was more difficult to be “confident” about the estimate. Due to the nature of the HealthRise program – with a community-based focus – and smaller sample sizes among some sites, it was not uncommon for particular measures to have wide confidence intervals.
Qualitative Data Collection
Qualitative data were comprised of a combination of focus group discussions and key informant interviews with various project stakeholders and participants. These were analyzed using thematic coding to distill major themes arising by country, site, and perspective (provider, patient, etc.), and to draw comparisons with baseline findings and conditions in comparison areas, depending on the data available for each country. The qualitative findings help to contextualize the quantitative results and elucidate impacts of HealthRise programs that cannot be captured with quantitative data. By comparing the qualitative findings from intervention facilities at endline with those from intervention facilities at baseline, as well as with comparison facilities, it is possible to draw inferences about what some of the effects of the HealthRise programs may have been. Key themes arising from the intervention site qualitative data at endline are presented for each country, reflecting the perspectives of patients, providers, other facility staff, and policymakers, and comparisons are drawn to baseline and non-HealthRise sites, depending on the data available for each country.