|
Improving Data for PMTCT programme management
W. Mphatswe1, K. Mate2, B. Bennett2, Z. Luvuno3, J. Reddy1, N. Moodley4, N. Rollins1, P. Barker2,5
1Nelson R Mandela School of Medicine, Paediatrics and Child Health, Durban, South Africa, 2Institute for Healthcare Improvement, Cambridge, United States, 3University of KwaZulu Natal, Durban, South Africa, 4Department of Health, KwaZulu Natal, Pietermaritzburg, South Africa, 5University of North Carolina, Department of Paediatrics, Chapell Hill, United States
Background: Recent studies on the District Health Information System (DHIS) in South Africa indicate sub-optimal data to track process performance and outcomes for PMTCT care. We report efforts to improve the completeness and accuracy of data in the DHIS to assist a large-scale intervention to improve PMTCT care in KwaZulu Natal.
Methods: We assessed the completeness of data reported to the DHIS from 316 clinics and hospitals in three districts for six key PMTCT data elements at two points in time: pre-intervention (Sept-Oct 2007) and post-intervention (Jun-Aug 2008). We independently surveyed clinic registers and monthly site reports for the same six elements and compared the accuracy of these data with what had been reported to DHIS at the same two time-points. The data improvement intervention consisted of a series of data training days, data monitoring site visits, and improved data collection tools.
Results: Aggregate analysis of all six data elements from 316 sites reported to the DHIS showed that data were complete no more than 64.9% of the time pre-intervention. After the intervention, data completeness had improved to 72.2% across all facilities in the three districts. Similarly, the percent of facilities reporting accurate data (i.e. within 10% of true value) improved from 9.0% pre-intervention to 30.5% afterwards. All data elements showed improvement, however the element “Nevirapine dose to baby born to woman with HIV” improved most (13.6% before versus 58.3% after). In contrast, “Nevirapine dose actually taken by woman in Labour” showed least improvement in accuracy (11.6% before vs 17.5% after).
Conclusions: Systematic data improvement interventions can improve the completeness and accuracy of DHIS data. Improvements in data quality can assist programme managers to focus on key obstacles to improving clinical care for pregnant women interacting with the PMTCT programme.
Download the e-Poster
|