The Discrepancy in Real-World Evidence of New Drug Expenditure Between Actual and Estimated Figures From Manufacturers and Government Agencies in Taiwan

OBJECTIVES: Budget impact analysis is widely adopted in many countries where health technology assessment is implemented in new drug pricing and reimbursement. We aim to examine the real-world evidence in differences of budgetary impact stemming from existing applications and administrative decision-making.

METHODS: Reports of new drug pricing applications were extracted from the official website of the National Health Insurance Administration between 2014 and 2017. The estimated budgets from pharmaceutical manufacturers and the CDE/HTA agency were derived as the primary outcome. The real-world drug expenditure was also derived. The estimation of total expenditure was adjusted by the final approved new drug unit price. Adjusted predicted drug expenditures and the real-world expenditures provided by both were compared over a five-year time span. Common one-sample and two-sample t-tests, along with signed rank and Wilcoxon Rank Sum tests, were used.

RESULTS: From 2014 to 2017, 42 cases with complete expenditure information were derived and analyzed. A discrepancy was evident between both estimations and actual drug expenditures, particularly around years 1 and 2. Class I drugs showed notably higher estimated expenditures, nearly twice the actual expenditure in year one, while Classes 2A and 2B drugs exhibited significant overestimations of 1.2 and 1.4 times the actual amounts, respectively. From the third year onward, there was no difference between estimated and actual expenditures, however, the trend is statistically declining with p < .05. The comparison between estimated drug expenditures by manufacturers and the CDE/HTA showed no significant differences.

CONCLUSIONS: The estimated new drug expenditures in budget impact analysis from both the manufacturer and the government agency were inconsistent with the actual expenditures in the first and second years, being lower in actual new drug expenditure due to practical reasons. No significant differences were found in estimated new drug expenditures between the manufacturer and the agency.