Laboratory testing--routine or on demand?

Clinical judgment is highly vulnerable to exposes. It can be exposed as a fraud by an autopsy that refutes the antemortem diagnosis or discloses a missed diagnosis of a potentially curable disease. Less catastrophic, but embarrassing nevertheless, the expose can come during life when laboratory data are similarly in conflict with a clinical assessment. Elsewhere in this issue (p 571) Deglin and associates report on such an embarrassment. Having determined serum lidocaine concentrations in 84 samples obtained from 33 patients receiving intravenous lidocaine hydrochloride infusions in a coronary care unit, these investigators proceeded to correlate the determined values in each patient with those predicted by his physician. Specifically, the physician was asked whether the serum lidocaine levels would be within, below, or above the therapeutic range and, when the predicted level was above this range, whether it would be in the toxic range. In 49% of these samples, the predictions