Dr. Elijovich suggests that the negative predictive value of 100% for plasma free metanephrines used in our study may be misleading because of the low incidence of pheochromocytoma in the general hypertensive population. He further states that the important issue in the diagnosis of pheochromocytoma is how well a test confirms, rather than excludes, the presence of a tumor.

As with other biochemical tests, it is neither appropriate nor cost-effective to use plasma metanephrines as a screening test for the hypertensive population at large, and we did not advocate this practice in our report. However, in patients who do not respond to antihypertensive medication, who have symptoms suggestive of a pheochromocytoma, or who have an increased risk for the tumor because of hereditary factors (and thus have a higher pretest probability of pheochromocytoma than that in the general hypertensive population), plasma metanephrines offer a useful initial screening test. As shown in the bottom panel of Figure 4 of our report, the post-test probability of a pheochromocytoma remains 0%, even when the prevalence or the pretest probability of a pheochromocytoma increases. This finding contrasts sharply with the considerable increase in post-test probability for plasma catecholamines. Thus, at low prevalence rates, there is little difference between plasma metanephrines and catecholamines for exclusion of pheochromocytoma; at higher pretest probabilities, plasma metanephrines offer a clear advantage over plasma catecholamines.

We disagree with Dr. Elijovich's statement, and with the implication of the editorial accompanying our paper [1], that the positive predictive value of a test is far more important than the negative predictive value. This depends on the disease. In pheochromocytoma, a diagnosis missed because of a false-negative result can have serious consequences. In contrast, a single false-positive test result can be refuted by the findings of further tests. The essential advantage of plasma metanephrines over catecholamines is that they appear to reliably exclude a pheochromocytoma in suspected cases and thus remove the need for further tests. Apart from avoiding a missed diagnosis, the use of metanephrines entails obvious potential cost benefits for any health care system. We agree with Dr. Elijovich that neither catecholamines nor their metabolites are specific markers for pheochromocytoma. Thus, when plasma metanephrines provide a positive test result, further tests are needed to confirm the presence of a tumor.