Training advances benefits of HIV data

The following is a guest post by Serge Bisore, MPH and Hypax Mbanye, MA, of Data.FI

Using data to improve performance is challenging when the quality of the data is uneven. As we move towards epidemic control, national AIDS control programs and ministries of health need complete, timely, and accurate information on HIV indicators so they can pinpoint low-performing sites and problem solve to address those.

In Burundi, on the Data.FI project, we are working to assist government staff involved in data management to identify and address HIV data quality issues, as well as health officials and supervisors who manage and coordinate HIV programmatic activities.

In early 2020, we met with the leadership of the National Health Information System and the National AIDS Program to address data quality issues that we felt were hindering efforts to track the performance of care and treatment services supported by the U.S. President’s Emergency Plan for AIDS Relief.

It was clear from aggregated reports at the national level that facility records weren’t regularly updated, suggesting that some clients might be lost to follow-up. We knew that supervisors of staff doing data collection and reporting in health facilities weren’t using an existing World Health Organization data quality review tool designed to help people see where there were gaps. Some could not pinpoint which districts were not reporting or were not reporting on time, or which districts were reporting data that were not consistent or accurate. How could they then supervise others to track and analyze testing and treatment indicators? 

The National AIDS Program leadership felt that training on data quality was key ― both at the central level and for supervisors of staff doing data collection and reporting in facilities. Together, we devised a stepwise plan to strengthen the ability of supervisors to understand the story that the data were telling, so they could better help facility-based staff improve data quality. Having observed that the data quality tool was not used, we felt that it was essential for supervisors at the central level to properly organize and prepare field supervisors to better identify and address the problems to be solved during supervision. We proposed instituting “desk reviews” of data quality before supervisors travelled to the field for site data reviews, and that these be made a regular part of the National AIDS Program monitoring and evaluation system. 

In July 2020, with the National AIDS Program, we convened a training with 14 central-level supervisors to conduct desk reviews on data quality. In the training, we used the data quality review tool and a related PLNS manual for supervisors involved in overseeing HIV data in Burundi. We asked the supervisors to look at new data dashboards and to identify the facilities that had problems. We asked how they arrived at that list and to show us where the facilities fell short—whether, for example, a structure had missing or aberrant data to correct. The supervisors then had to link these data issues with HIV indicators to be able to monitor clinical cascade outcomes. These desk review meetings are now held regularly to prepare supervisors to conduct quarterly data review meetings at the district level.

We see hope. We already see a change in consciousness among our colleagues at the National Health Information System directorate on the importance of data quality. The National AIDS Program leadership, likewise, is starting to ask supervisors at implementing partners to probe for the underlying causes of the data quality issues. Our colleagues are also finding advantages in a stronger focus on preparation prior to field visits. Supervision is simplified because it targets data quality issues which are identified in advance, and reporting is simplified because it is based on the desk review report.

We will work on formalizing the use of desk reviews as a standard operating practice. Data.FI will continue to coach supervisors who go to the facility level to make sure those visits are well prepared and the teams well oriented. Supervisors need to know what their questions are, and where the data problems appear to lie. In time, these new habits will translate to curbing transmission, better treatment outcomes, and lives saved.

In our view, as soon as people at all levels use the data quality tool and the dashboards they generate ― and understand the benefits of good data ― decision makers will be better positioned to improve the health system, and thereby improve the care and treatment of people living with HIV.

Data.FI, funded by PEPFAR through USAID, is a global project that helps countries improve their data systems to strengthen prevention, testing, treatment, and lab services to end the HIV epidemic and to combat COVID-19.