The recent editorial by Levine [1] implies that the use of misoprostol in combination with nonsteroidal anti-inflammatory drugs (NSAIDs) in patients with rheumatoid arthritis is justified only if the dollar value accrued by preventing hospitalizations due to serious gastrointestinal complications exceeds the costs of the medication for the entire population at risk. This "rudimentary" cost-effectiveness analysis does not reflect other potential benefits of misoprostol therapy, such as the prevention of death in patients with serious bleeding, the avoidance of a medical work-up for patients with gastrointestinal symptoms but no serious complications, and the avoidance of gastrointestinal symptoms in patients with or without serious complications.

Any evaluation of cost-effectiveness should reflect the full value of a proposed therapy. The costs of a medication are easily determined algebraically, and although it may be difficult to assess the costs of complications resulting from therapy, the benefits side of the equation typically suffers because of the difficulty in evaluating the value of more subjective benefits. The difficulty in assessing value does not justify the assignment of no value. Moreover, unless some value limits can be assigned to each cost or benefit, the analysis is useless.

For Levine, a medication must meet an absurdly rigid standard to gain approval. Many drugs, such as NSAIDs, are prescribed for relief of symptoms and have little or no beneficial effect on outcome. In this context, the 40% decrease in serious gastrointestinal complications reported by Silverstein and colleagues [2] is an impressive result for any pharmacologic intervention, and although good reasons exist to limit the population in which misoprostol is used, the reason given by Levine is unconvincing.