สถาบันวิจัยระบบสาธารณสุข

The diagnosis related group (DRG) has been used for making retrospective payments to health-care providers under the Universal Health-care Coverage Scheme. The amount of allocated budget depends largely on the quality of the reported data; however, evidence has revealed unsatisfactory data quality.This article is aimed at describing the coding audit process of the Bureau of Claims and Medical Audit (BCMA), National Health Security Office (NHSO) as well as the findings from recent audit results. Coding audits comprise main steps: (1) provider selection, (2) medical record selection, (3) medical record audit, (4) summary and report of the findings, (5) provider’s assessment of the report, (6) provider’s appeal, and (7) final decision. In fiscal year 2008, BCMA audited 57,828 medical records from 931 providers in 75 provinces. The summary assessment was found to be correct in 57.63 percent of the records; of that proportion the main reason for errors was missing secondary diagnosis (28.4%). Coding assessment was correct in 49.11 percent of them; the main reason for errors was disagreement between coder and auditor (17.85%) and incorrect coding for principal diagnosis (14.49%).