Computer adaptive assessment tool for monitorng the use of high cost drugs in civil servant medical beneftt scheme

Abstract

Objectives of this study were to develop computer adaptive assessment tool (CAAT) for monitoring the use of high cost drugs (HCDs) in Civil Servant Medical Benefit Scheme (CSMBS), to validate CAAT, and to evaluate the use of HCDs in CSMBS concerning pharmaceutical spending and rational utilization. Retrospective review of CSMBS outpatients initiating Atorvastatin and Rosiglitazone during November 2006 and April 2007 was conducted. The documentation review for CAAT algorithm and the analysis of data from electronic dispensing, laboratory, and ICD-10 database were employed for CAAT development. To validate CAAT, the results from CAAT analysis were compared with the results from conventional drug use evaluation. Rational use of HCDs and their expense were evaluated by using the developed CAAT. The results revealed that the development of CAAT algorithm and the data management were essential elements. CAAT guidelines development, selecting key variables and their cut points were needed for CAAT algorithm development. Data extraction, data linkage (data relation) and data completeness were required for data management. When validating CAAT for Atorvastatin, its sensitivity, specificity and accuracy were 81.3%, 100.0%, and 98.0% respectively. For Rosiglitazone, its sensitivity, specificity, and accuracy of CAAT were 82.3%, 100.0%, and 95.9% respectively. Using CAAT to evaluate rational use based on pattern analysis, Atorvastatin use of 76 (13.0%) patients from four regional hospitals and Rosiglitazone use of 15 (17.2%) patients from two hospitals were rational. It was found that 15.0% (448,500 baht) of Atorvastatin’s expenditure and 15.3% (94,500 baht) of Rosiglitazone’s expenditure were rational. In conclusion, CAAT can be applied for monitoring the use of HCDs efficiently. Finding of only 15% of expenditure of Atorvastatin and Rosiglitazone that were rational indicated the need for the improvement of HCD’s prescribing pattern. It was recommended that assessing acceptability and applicability of using CAAT for monitoring HCD use among health care providers and assessing policy acceptance of the Comptroller’s General Department are needed for further study. In addition, the national coding system of drugs, laboratory, and patient characteristics should be established.