Study on completeness and accuracy of the Diabetes reporting system in four provinces of Thailand, 2012-2013

Abstract

Background: In Thailand, the situation of chronic non-communicable diseases has been monitored by the Bureau of Epidemiology since 2003 and the reports have been increasing throughout these years. Diabetes was reported as the second highest rate of patients due to its complications. The reporting system for diabetes and its complications is being carried out in 4 provinces, including Angthong, Lampoon, Mukdahan, and Trang. This study aimed to evaluate the diabetes reporting system from the Standard Data Set in the “43 files” from 4 provinces and provide recommendations for further improvement of the reporting system.
Methods: A cross-sectional study was performed during June-August 2014. Data on demographic information and disease outcome from 16 district hospitals and 4 provincial hospitals during 2012 and 2013 were reviewed and compared with the reported information in the Bureau of Policy and Strategy.
Results: Completeness of the reported gender, age, occupation and address (subdistrict) were 85.93%, 85.93%, 83.34% and 75.76% respectively. Accuracy of the reported sex, age, occupation and address were 99.78%, 97.56%, 65.06% and 66.86% respectively. Completeness of the reported diabetic foot, diabetic nephropathy, stroke, diabetic retinopathy and ischemic heart disease were 69.08, 50.0%, 49.06%, 25.86%, and 0% respectively. Meanwhile, accuracy of the reported stroke, diabetic foot, diabetic retinopathy and diabetic nephropathy were 100%, 81.22%, 77.66%, and 67.19% respectively. Comparing two important files in “43 file” reporting system, “Diag file” and “Chronic file”, completeness in month and year of onset in “Diagnosis file” was 83.65% and the accuracy was only 60.69. The accuracy was increased to 92.62% when considering only the year of onset. Completeness and accuracy of the reported month and year of onset in “Chronic file” were 76.62% and 80.68%. From 1,621 records, only 28.75% of those were reported to both “Chronic file” and “Diagnosis file” with the same month and year of onset, and 57.96% with the same year of onset.
Conclusions: Completeness of diabetes-related reporting system through “43 file” system was quite low. Using combination of report in “Diagnosis file” and “Chronic file” would increase the completeness of diabetes report. However, the data quality and reliability could be improved by increasing knowledge, skill and technology for data management.