Evaluation of data completeness of Client Intake Form (CIF) before and after Data Quality Audit (DQA) of a Voluntary Medical Male-Circumcision (VMMC) program in Zimbabwe

Abstract

Abstract Purpose To evaluate and compare the data completeness of Client Intake Forms (CIFs) collected both before and after a Data Quality Audit (DQA) in February 2015 and to identify the factors contributing to incompleteness of CIFs used for the Voluntary Medical Male Circumcision (VMMC) procedure in Zimbabwe. Methods Four out of ten sites that received a DQA in February 2015 were selected based on convenience sampling. Secondary data analyses were conducted using data from CIFs, which were filled out by the staff during the VMMC procedures at the sites. Two-levels of completeness were evaluated in two months (November 2014 and May 2015). The record availability of CIFs was assessed by measuring the percentage of VMMC clients whose CIF was on file at each site. Then, a data completeness evaluation tool was used to measure the completeness of 34 key variables on the CIF, and a comparison of the performance on pre- and post- data-completeness in each sampled site was conducted. Results Record availability at the Norton and Mutoko sites were generally maintained in high standard (around 100%); the record availability in site Muvonde was 88.0% before the DQA and then improved 8.0% after the DQA; the record availability in site Gokwe South was 79.8% before the DQA and than improved 9.7% after the DQA. The average monthly completeness rate on CIFs prior to the DQA studied for site Muvonde, Gokwe South, Mutoko and Norton is 88.0%, 95.6%, 94.6% and 95.3%, respectively. The average monthly completeness rate after the DQA for site Muvonde, Gokwe South, Mutoko and Norton is 96.6%, 98.3%, 98.3% and 93.6%, respectively. Three of the four sites had statistically significant improvements in data completeness across 34 key indicators, with mean improvement of 8.6%, 2.7%, 3.8%, respectively (p<0.05). For one site, Norton, the data completeness of CIFs in May 2015 decreased 1.7% (p<0.05) compared to November 2014. Conclusion Our findings suggest that the data quality of CIFs in terms of record availability and data completeness were good and generally improved after the DQA. However, there were still some remaining gaps pertaining to the completeness of certain key indicators. Further study should confirm the gaps and offer possible solutions to improve data completeness and to assist quality program implementation. Furthermore, the results suggest the evaluation of data accuracy and authenticity should be integrated into routine data information system in VMMC program in the future. Keywords: data quality; data completeness, quality improvement, circumcision