Abstract
Drug expenditure is a lion share of health expenditure and grows rapidly in the Civil Servant
Medical Benefit Scheme (CS). This research analyzed electronic databases of drugs in 5 selected
therapeutic classes, including (1) antihypertensive drugs (angiotensin-converting enzyme inhibitors –
ACEIs and angiotensin-2 receptor blockers -ARBs); (2) antilipidemics (statins and other new
products); (3) antiplatelets; (4) antiulcerants (proton pump inhibitors –PPIs and histamine-2 receptor
antagonists -H2As); and (5) non-steroidal anti-inflammatory drugs (NSAIDs, both conventional and
coxibs) that were prescribed during 2003 – 2007 from 6 provincial hospitals. The analysis focused on
a 60-month, interrupted time-series of 2 expenditure components, including propensity of use and
intensity of use, before and after the direct disbursement policy of the CS and the 30-Baht
abolishment in the Universal Healthcare Coverage Scheme (UC). The drug expenditure of each
therapeutic class was forecasted for the years 2008 and 2009 and potential cost-saving for the CS as a
result of 3 cost-containment strategies, including (1) substitution policy; (2) reference pricing; and (3)
UC-like prospective payment system (PPS) was estimated.
The study found that the drugs deemed as the cost-containment target, including ARBs,
single-source statins, clopidogrel, single-source PPIs, and coxibs had a major impact on the drug
expenditure of CS and its growth. The PPS could reduce the CS drug expenditure by approximately
40% (for PPIs-H2As and coxibs-NSAIDs) and 100% (for antilipidemics), which was comparable to
the reference pricing strategy. Potential cost-saving from the substitution policy depended on
substitutability in the real world setting, for example, the 20% substitutability would reduce the CS
expenditure by 8% (for PPIs-H2As and coxibs-NSAIDs) and 20% (for antilipidemics); whereas the
saving of 30-80% was expected, given substitutability of 80%.
