BALTIMORE—For a diagnostic or screening test to be accepted for widespread use today, it must be not only accurate but also cost-effective. Helical computed tomography (CT) has gained a measure of acceptance for pulmonary embolism (PE) diagnosis, and it is currently being considered for lung cancer screening in smokers. But is either use cost-effective?

Two recent studies examined this question. The PE study suggests that the helical CT is not an appropriate choice when used as a single test for PE diagnosis, but it may be cost-effective when used in combination with D-dimer tests and lower-limb ultrasonography.[1] In contrast, the use of helical CT cannot be justified as a screening tool for lung cancer—even in high-risk patients.[2] In fact, the harm from such screening may far outweigh the benefits.

COST-EFFECTIVENESS IN PE

Arnaud Perrier, MD, and colleagues from Geneva University Hospital used decision analysis software to estimate the direct costs—those involving the health care system—of various PE diagnostic strategies. To increase the applicability of their findings, they used economic data from the United States, Canada, and Switzerland.

Unlike previous studies of the cost-effectiveness of helical CT, this investigation weighed patients’ clinical probability of PE (low, intermediate, or high). To further ensure the accuracy of their findings, Dr. Perrier and colleagues conducted two separate sets of analyses for each clinical probability. In the first set, helical CT was assigned a sensitivity of 70%—the degree of accuracy associated with older, single-detector scanners. In the second set, sensitivity was raised to 85% to correspond to the performance of newer, multidetector CT scanners.

The investigators found that when the clinical probability of PE was low, the most cost-effective approach was a sequential strategy that began with D-dimer testing and added ultrasonography and, if necessary, ventilation-perfusion (V/Q) scanning. (In this strategy, a patient with a nondiagnostic V/Q scan was not given treatment for thromboembolic disease.) V/Q scanning could be replaced by helical CT with only a small increase in cost and no loss of benefit.

When the clinical probability of PE was intermediate or high, the most cost-effective approach depended on the sensitivity of the CT scanner. For institutions with older equipment, the best strategy was to begin with D-dimer testing, ultrasonography, and V/Q scanning; a nondiagnostic V/Q scan would be followed by helical CT. (A pulmonary angiogram could be substituted for the CT scan at a small increase in cost.) For institutions with newer equipment, helical CT could replace the V/Q scan; in fact, doing so might eliminate the need for angiography, the investigators concluded.

Because of its low sensitivity, helical CT as a single test was the least cost-effective strategy in all clinical settings.

HELICAL CT IN LUNG CANCER SCREENING

The lung cancer screening study, like the PE diagnosis study, used computer simulations to estimate the cost-effectiveness of helical CT. However, it also assessed the potential benefits and harms of such a screening approach. The hypothetical cohort included 100,000 heavy smokers (patients with more than 20 pack-years of smoking). Their age was set at 60, and 55% were men. To account for the effect of smoking cessation, the authors performed three analyses—one with the hypothetical cohort as current smokers, another in which they had just quit, and a third in which they had not smoked in five years.

“We made a big assumption about helical CT’s effectiveness,” acknowledged lead author Parthiv J. Mahadevia, MD, MPH, a Research Scientist at the Johns Hopkins School of Medicine in Baltimore. “We assumed a 50% stage shift, meaning that half the lung cancers that would have presented in an advanced stage without screening were instead found in localized stages.” In addition, only non–small cell lung cancers underwent this stage shift. Because small cell lung cancer metastasizes early and has a low probability of cure, it did not undergo a stage shift.

Under these conditions, helical CT screening reduced lung cancer mortality in current smokers by 13%, from 4,168 to 3,615 deaths per 100,000 patients during a 20-year period. However, it also resulted in 1,186 needless invasive procedures per 100,000 patients due to false-positive results.

In 2001 US dollars, helical CT screening for lung cancer in current smokers cost $116,300 per quality-adjusted life-year (QALY) gained. In subjects who had just quit smoking, screening cost $558,600 per QALY gained, and in those who had not smoked in five years, the cost was $2,322,700 per QALY gained. Thus, the longer it had been since a patient quit smoking, the less cost-effective CT screening became.

“The estimates reported by Mahadevia et al represent the reality of current data and should lead to some hesitation and thoughtful consideration of whether to use this screening modality at the present time,” remarked Victor R. Grann, MD, and Alfred I. Neugut, MD, in an editorial.[3] They advised physicians, patients, and policy makers to be conservative about accepting helical CT as a lung cancer screening tool until more data are available.n

—Timothy Begany

References
1. Perrier A, Nendaz MR, Sarasin FP, et al. Cost-effectiveness analysis of diagnostic strategies for suspected pulmonary embolism including helical computed tomography. Am J Respir Crit Care Med. 2003;167:39-44.
2. Mahadevia PJ, Fleisher LA, Frick KD, et al. Lung cancer screening with helical computed tomography in older adult smokers: a decision and cost-effectiveness analysis. JAMA. 2003;289:313-322.
3. Grann VR, Neugut AI. Lung cancer screening at any price? JAMA. 2003;289:357-358.

Return to table of contents