We appreciate the responses to our article [1]. We agree that the concept of NNT raised by Dr. Wendel and by Madhok and colleagues is important. Using the formula NNT = 1/(risk_placebo – risk_misoprostol), we determined that the NNT was 263 for the 67 serious NSAID-related gastrointestinal events we observed. For those patients at higher risk, however, the NNT to avert a gastrointestinal complication becomes much smaller. For example, for patients with previous peptic ulcer disease, the NNT is 53.

We recently completed additional analyses of the 242 events reported as suspected NSAID-induced complications. These events include patients who present with gastrointestinal symptoms while receiving an NSAID that might indicate an NSAID-related complication and therefore might have both clinical and economic consequences. To understand the full effect of the cost of NSAID-related complications and the role of misoprostol in reducing them, these cases must also be considered because the symptoms resulted in a medical work-up, whether or not a complication actually occurred [2].

Dr. Grossman's calculations are limited, given that they only address drug costs and not actual health care utilization (for example, hospitalizations) by patients with NSAID-related complications. Dr. Hammer suggests using drugs such as antacids to reduce dyspeptic symptoms during NSAID treatment. However, symptoms and complications are not necessarily linked in NSAID-related injury, in which the onset of a complication such as bleeding or perforation is often silent.

Although study dropout rates caused by side effects were significantly higher for patients receiving misoprostol than for those receiving placebo during the first month, dropout rates were similar thereafter. Therefore, most side effects noted with misoprostol seem transient. Further, Raskin and colleagues [3] showed that misoprostol, 200 µg given twice or three times daily, provides mucosal protection with a significant reduction in side effects [3].

Health care providers are clearly aware of the risks associated with the use of NSAIDs and the need for prevention strategies to reduce them. Although the rate of serious NSAID-related complications is about 1% in 6 months, the widespread use of NSAIDs (> 15 million patients in the United States alone use them) translates into a substantial number of complications (about 150 000 per year). With a number this large, the 40% reduction in complications seen with misoprostol becomes clinically important. Finally, we agree that misoprostol—the only approved agent for the prevention of NSAID-induced ulcers—should be used in at-risk populations in whom the clinical and economic benefits are clearly substantial.