| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Massachusetts General Hospital Boston, MA 02114
Washington University Medical Center St. Louis, MO 63110
American Cancer Society Atlanta, GA 30329
The Mount Sinai Hospital New York, NY 10029
The recent article by Obuchowski et al. [1] attempted to summarize the basic principles of screening as they apply to imaging tests. The authors cited examples from the experience with breast cancer screening and used these to discuss lung and colorectal cancer screening. We disagree with a number of statements made regarding the basic principles of screening in general, and breast cancer screening in particular. Because of space limitations, we must confine our comments here to their comments concerning breast cancer screening.
The authors state that "mammography is clearly not an ideal screening model." In fact, it is the screening model. It has been tested in the most rigorous fashion and has undergone greater scrutiny than any other screening test. All the issues that pertain to new imaging screening tests have been addressed with regard to mammographic screening.
It is surprising that an article from a radiology department claimed that the effectiveness of mammography is controversial. The authors ignored a huge body of literature supporting mammographic screening, instead citing a single highly controversial and poorly reasoned review article [2] that has subsequently been discredited [3]. Some lingering disagreement still exists about the age at which screening should begin, the optimal interval between screens, and the magnitude of the benefit derived from screening, but the data are clear that mammographic screening reduces deaths from breast cancer.
Randomized controlled trials have validated the efficacy of screening mammography by showing a statistically significant mortality rate reduction from breast cancer among women invited to screening in the trials. Moreover, this reduction in mortality rate has now been shown to translate into general service screening, outside of scientific trials (an important point of efficacy overlooked by the authors). Over a 29-year period in the two counties that were host to the Swedish Two County Trial, overall breast cancer mortality fell by 50%. Among women who actually participated in screening, the reduction in the breast cancer mortality rate was 63% [4]. The decrease in deaths has been in direct proportion to the percentage of women being screened. Breast cancer mortality is also dropping in the United States, and the decline is accelerating. Much of this change is attributable to mammographic screening.
In their discussion of screening intervals, Obuchowski et al. [1] suggested that cost was the key factor. Cost is an important practical issue, and ultimately society will need to decide what is affordable. However, the authors have confused cost issues with medical—scientific decision making. The best screening interval is the one that provides the greatest benefit for the patient.
Finally, the authors [1] overlooked the most critical criterion for effective screening programs: Simply detecting cancers in the preclinical phase does not guarantee efficacy. Efficacy depends on whether the natural history of the disease can be altered by intervening at the time of early detection, resulting in treatment that either truly delays death or truly prevents death from the cancer. Unless surrogate end points can be validated, the only way to prove efficacy is through a randomized controlled trial in which death from the cancer is the measured endpoint.
Many lessons can be learned and applied to new screening tests from the decades of study and argument that went into proving the efficacy of mammographic screening. The issues are the same.
References
The Cleveland Clinic Foundation Cleveland, OH 44195
We thank Kopans, Monsees, Smith, and Feig for expressing their concerns about our criteria for effective screening vis-á-vis mammography. First, it is important to understand that our article [1] was never intended to critique mammography. Mammography is the only reasonable and efficacious imaging test that we have for this devastating disease of breast cancer, and therefore it should be used. However, as the only imaging study used for screening, mammography serves as an important point of reference for the possible role of multislice CT in screening for colorectal and lung carcinomas. If mammography is considered a reasonable, effective, and acceptable means in our society for screening for breast cancer, and it is not ideal as a screening test, then should higher standards be applied to CT in detecting lung or colorectal carcinomas? That was the question we posed in our article.
We state that mammography is not an ideal screening model [1]. Why is it not ideal? First, the prevalence of the disease in the screened population is less than 1%. As a result, both the pretest and posttest probabilities are low. Second, numerous instances of pseudodisease are detected with mammography, reducing the effectiveness of screening. Third, the specificity of the test is relatively low; false-positive rates range from 0.9-6.5%. As a result, many biopsies are performed that result in benign findings. Nevertheless, if mammography is accepted in our society as effective, as it is despite these issues, then the standards applied to multislice CT in detecting colorectal and lung cancers should be set at an equivalent level.
We have not confused financial costs with scientific decision making. Our article [1] clearly states that "costs" include "both monetary and patient outcome costs" (page 1359). Patient outcome costs include the consequences of false-positive findings and instances of pseudodisease, as well as the morbidity or mortality caused by the screening test.
We do not believe we have "overlooked the most critical criterion for effective screening programs." We are well aware that efficacy depends on whether the natural history of the disease can be favorably altered at the time of early detection; this was the message of our article [1]. Randomized controlled trials are certainly essential in demonstrating the efficacy of screening. Such trials should not, however, be the first step. Randomized controlled trials should be undertaken only after smaller scale, less expensive studies suggest that the 10 criteria of screening discussed in our article are reasonably met. In many cases, screening tests will never reach the randomized controlled trials phase; this outcome benefits everyone.
Our article [1] cited the work by Gotzsche and Olsen [2] and stated that conclusions from large randomized trials assessing the benefit of early treatment for women screened by mammography are "highly controversial." We could not cite the work of Duffy [3] and Tabar et al. [4], both published in 2001 after our paper was accepted. Regardless, our use of the term "highly controversial" was unwise, inflammatory, and overstates the case. We apologize and retract the statement. We agree with Kopans et al. that the reduction in breast cancer mortality is largely attributable to mammography (as well as to heightened awareness and more examinations).
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  C. D. Furtado, D. A. Aguirre, C. B. Sirlin, D. Dang, S. K. Stamato, P. Lee, F. Sani, M. A. Brown, D. L. Levin, and G. Casola Whole-Body CT Screening: Spectrum of Findings and Recommendations in 1192 Patients Radiology, November 1, 2005; 237(2): 385 - 394. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
