I read with interest the article by Hull et al1 concerning the cost-effectiveness of different strategies to diagnose pulmonary embolism. In the article, Hull and colleagues conclude that a strategy employing serial noninvasive testing for deep vein thrombosis in the lower extremity of a subgroup of patients with possible pulmonary embolism will be the most cost-effective approach to the diagnosis. Their statements concerning this are solely based on the cost of the evaluation, and their cost data certainly support this thesis. It is 23% less expensive to pursue this non-invasive strategy.
The effectiveness of noninvasive serial testing, however, is not so clearly defined. In their Table,1 the authors list the number of patients requiring treatment; however, they do not comment on the differences in the numbers of patients in whom pulmonary embolism is detected by each strategy. Strategies associated with noninvasive assessment are also associated with a smaller