Current issues of ACP Journal Club are published in Annals of Internal Medicine


Health Economics

It is unclear whether misoprostol is cost-effective in preventing NSAID gastropathy

ACP J Club. 1995 Mar-April;122:53. doi:10.7326/ACPJC-1995-122-2-053


Source Citation

Stucki G, Johannesson M, Liang MH. Is misoprostol cost-effective in the prevention of nonsteroidal anti-inflammatory drug-induced gastropathy in patients with chronic arthritis? A review of conflicting economic evaluations. Arch Intern Med. 1994 Sep 26;154:2020-5.


Abstract

Objective

To measure the cost-effectiveness of misoprostol in the prevention of gastropathy induced by nonsteroidal anti-inflammatory drugs (NSAIDs) in patients with chronic arthritis.

Data Sources

Economic studies published in English up to December 1993 were identified using MEDLINE.

Study Selection

Studies published in peer-reviewed journals that analyzed the economic benefit of misoprostol in patients who received NSAIDs.

Data Extraction

Type of analysis, daily dose and cost of misoprostol, endoscopically detected symptoms, ulcer rates, compliance-adjusted ulcer rates, and the derivation of hospitalization and ambulatory costs.

Main Results

5 studies were identified; 4 studied osteoarthritis and 1 studied rheumatoid arthritis. All studies used the same analytic model and compared the costs and probability of developing a symptomatic NSAID-induced gastric ulcer with and without misoprostol use. Each study used a different method for estimating resource utilization. Direct ambulatory and hospitalization costs or charges were used; units of service were costed using accounting costs or charges. 2 studies used misoprostol, 800 mg/d, and 3 used 400 mg/d. 4 studies found that misoprostol was cost saving. The results of the first study were sensitive to rates of silent ulcers and compliance and were less sensitive to rates of hospitalization and surgery. In the second study, misoprostol was cost saving but results were sensitive to rates of silent ulcers, patient compliance, and ambulatory costs. The third study found misoprostol cost saving for patients with osteoarthritis and epigastric pain but results were sensitive to cost of misoprostol, compliance rate, cost of ambulatory care, risk for ulcer, and reduction of that risk. The fourth study judged that the costs per life-year gained ($95 600) and per bleed avoided ($5300) were high. The fifth study found that misoprostol was cost saving for high-risk older patients ($625/symptomatic ulcer prevented); results were sensitive to ulcer complication rate and cost of ambulatory treatment and misoprostol.

Conclusion

The diverse methods and conflicting results of the 5 economic evaluations reviewed do not provide a definitive answer to the questions of whether misoprostol prevents clinically important gastropathy induced by non-steroidal anti-inflammatory drugs or whether it is cost-effective in patients with chronic arthritis.

Source of funding: National Institutes of Health.

For article reprint: Dr. G. Stucki, Robert B. Brigham Multipurpose Arthritis and Musculoskeletal Disease Center, Department of Rheumatology/Immunology, Brigham and Women's Hospital, 75 Francis Street, Boston MA 02115. FAX 617-731-9032.


Commentary

The article by Stucki and colleagues and the 5 studies it reviews highlight an important feature of models used for economic evaluation. This feature is the use of sensitivity analysis in drawing a conclusion when the qualitative result (e.g.,misoprostol is cost saving or is associated with an attractive cost-effectiveness ratio) is sensitive to quantitative estimates of variables used in the model. Cost-effectiveness analysis should not be used to make conclusions unless extensive sensitivity analyses have been done first. The work summarized by this review article clearly shows that the uncertainties in the variables of the model affect the results to such an extent that no conclusion can be drawn.

Since this article was published, the results of 2 other studies have become known. The MUCOSA trial has shown that the reduction of endoscopically proven ulcers at 3 months attributed to misoprostol, the end point used in the 5 papers reviewed, is translated into a small but significant reduction in gastrointestinal ulcer complications (1). A second economic evaluation incorporated utility as an outcome variable in its model, an end point not used in the 5 studies reviewed, and concluded that misoprostol may actually worsen the quality of life for some patients (2).

Cost-effectiveness assessments of a health care intervention assist in allocation decisions made by those responsible for setting clinical and reimbursement policies. Is misoprostol cost saving or associated with an attractive cost-effectiveness ratio in the prevention of NSAID-induced gastropathy? Stay tuned, because the answer is still unknown. Research in progress may soon enlighten us.

Matthew W. Morgan, MD
Allan S. Detsky, MD, PhD University of Toronto The Toronto Hospital Toronto, Ontario