External validity of authors' conclusions is doubted

Editor-The logic of risk stratification in acute myocardial infarction is that it allows treatment (and finite resources) to be targeted on the group in whom the potential benefits are greatest. Yet Sumit Basu and colleagues seem to be recommending cardiac catheterisation in at least 68 of their 100 patients with myocardial infarction on the basis of abnormal scintigraphic findings, while reporting a positive predictive value of scintigraphy of only 49%.1 Given that this was a highly selected group of low risk patients, which excluded elderly patients, those with severe heart failure, and those with left bundle branch block, applying the authors' recommendations to all patients with acute myocardial infarction would inevitably lead to invasive investigation of the large majority. While this strategy is possible, it seems to turn the logic of risk stratification on its head.

A fundamental requirement of risk stratification is that it is done early enough after myocardial infarction to anticipate the period of greatest risk. Our own database of 1225 patients shows that, by delaying their studies for five to seven weeks, Basu and colleagues would have failed to stratify the 230-247 patients (18.8-20.2%) who died, had a reinfarction, or were admitted to hospital with unstable angina before that time. Even if analysis was restricted to 800 hospital survivors treated with thrombolysis, death or reinfarction during this early period of heightened risk occurred in 50-60 cases (6.3-7.5%). …