Prevention of Mother to Child Transmission of HIV: Data Quality Assessment in Selected Health Facilities of the Southern Region of Zambia

Abstract

Background: Although substantial investments have been made in Zambia to expand prevention of mother-to-child transmission of HIV (PMTCT) services, the quality of patient data recording and reporting remains a challenge. The study aimed at assessing the quality of PMTCT programme data at selected PMTCT sites in the southern region of Zambia. Methods: This was a quantitative study which followed two protocols. The first was a retrospective record review which involved collecting PMTCT data on selected PMTCT indicators. The second one was a cross-sectional systems assessment of the relative strengths and weaknesses of the functional areas of the data management and reporting systems. It covered 66 PMTCT sites from four randomly selected districts. Data was collected using record review forms and structured questionnaires and was entered and cleaned in Epi-Data Version 3.1. Analysis involving descriptive statistics was done in SPSS version 16. Data quality was determined through assessment of dimensions of quality1. A Likert scale was used to score and categorize data. Results: The quality of PMTCT data was above average at just over two-thirds (67.11%). Data accuracy, completeness and timeliness levels were found to be below 50% for all indicators (48.23%, 49.23% and 44.65% respectively). However, confidentiality, reliability, integrity and precision levels of PMTCT data were 75.11%, 76%, 88% and 88.69% respectively. Data was least accurate for infant indicators (32.3%-34.33%) while data completeness was lowest for the antenatal indicators (26.50% - 33.33%) and was least timely for the antenatal indicators (23.85% - 31.43%). Discussion: This study underscores a low level of quality for the PMTCT programme data due to low levels of accuracy, completeness and timeliness on infant and antenatal indicators. It was found that the main problem revolved around CD4 count and PCR testing. Materials needed for testing were inadequately and erratically supplied to health facilities. Most of the test results were not being received and those received were rarely received in time and some were coming as invalid results. Other factors affecting data quality included: low staffing levels, lack of training for staff involved in data management at service delivery points and involvement of unqualified people in the recording and reporting of patient data. Lack of adequate storage facilities for documents also affected data quality. Conclusion: It can be concluded that the quality of PMTCT data was lower than expected. To improve data quality, healthcare data must be appropriate, accurate, timely, reliable, complete, precise, and must be handled confidentially with integrity. Data entry checks are also critical for accurate data reporting. Hence, having dedicated personnel in each facility would improve data recording and reporting significantly. The limitations of the study was that it only covered four districts due to limited funding, this can be addressed by further studies. Key words: Data quality, Likert scale, data verification, systems assessment, dimensions of quality