The 15 June 1997 issue of Annals contained an interesting juxtaposition of articles. The first is a position paper produced by the Clinical Efficacy Assessment Project of the American College of Physicians [1] on diagnosing syncope. An expert panel reviewed the appropriate workup for patients with syncope and noted that "routine use of basic laboratory tests is not recommended ... ." Immediately following this position paper is an essay, "What's Wrong with This Picture?" [2], that describes a patient with isolated syncope. This patient has a thorough cardiac evaluation that fails to detect his colon carcinoma. One wonders whether a routine laboratory test, the complete blood count, if properly assessed, would have led to an earlier diagnosis of colon cancer. It is well recognized that colonic neoplasms are associated with gastrointestinal bleeding. Right-sided lesions, in particular, are often asymptomatic until significant anemia develops. It is likely that Dr. Brody's patient had iron deficiency. Even if the patient did not have a textbook picture of hypochromia and microcytosis, I would expect an increased red cell deviation. Any of these findings should have alerted his physicians to the possibility of gastrointestinal blood loss.

I was trained in an era when a complete patient evaluation included a set of routine laboratory tests. These tests were non-invasive and were designed to screen the patient for otherwise silent disease. Linzer and colleagues [1] noted that the most common cause of syncope is "unknown" (their Table 1. Routine laboratory testing may help diagnose clinically occult bleeding and other hidden conditions and should remain part of our approach to patients with serious medical symptoms.