| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Yale University School of Medicine New Haven, CT 06520
I read with interest the perspective entitled "CT Screening: Why I Do It," by M. Brandt-Zawadzki [1]. I have often discussed with residents and fellows the possibility that cancer induction associated with radiation might negate the potential increase in the cancer cure rate attributable to CT screening. I was surprised to see that the discussion in this article did not result in the numbers I had previously calculated on the basis of similar data. I then realized that an arithmetic error of 100-fold was made in the calculation of the potential benefit to patients based on a 0.005% increase in the cure rate. The number of patients who would benefit per 23,000 would be 1.15 (0.005% x 23,000) not 115, as is stated in the article. Thus, on the basis of the numbers presented, if a screening program using CT were instituted and early detection resulted in a cure rate of 0.005%, 40 people would have cancer induced because of the increased radiation dose, whereas 1.5 would be cured because of early detection. It is possible that the error was typographic and that 0.005% should have been written as 0.05%, but this early detection and cure rate for the subgroup of patients with cancer would be optimistic.
Cancer was found in only 1% of the patients presented in the article (18/1777, table 2) [1]. Assuming 23% of people in the original group would likely die of cancer, 408 would be expected to die of cancer. A cure rate of 0.5% would indicate that two patients in the screened group would be cured because of early detection. If the case of metastatic disease is excluded, using this projected cure rate means that 12% (2/17) of the patients in whom cancer was detected would need to be cured who would not have been cured if they had not undergone CT screening. If patients undergo multiple CT screening examinations, the number of cancers detected per screening examination will likely be lower because the incidence rate of cancer is typically lower than the prevalence rate in screening programs. Therefore, an even higher percentage of patients in whom cancer was detected would need to be cured for the cure rate to be 0.5%. Multiple screening examinations would also result in an increased cumulative radiation dose—also likely increasing the chance of cancer induction. Assuming a linear increase in the cancer rate with cumulative radiation dose, we could expect the number of patients with induced cancers per 100,000 to be 40 multiplied by the average number of screening CT examinations per patient. Although all of these numbers are approximations and may be off by orders of magnitude, the important point is that it is quite plausible that a CT screening program may result in the deaths of more patients than are saved.
The importance of accepting the fact that we may be doing more harm than good when we make decisions on diagnosis and treatments without adequate data to support our recommendations has been underscored by the recent findings of increased health risk due to hormone replacement therapy in women [2]. It is essential that large well-designed studies support the belief that the benefit of CT screening outweighs the risk before we recommend this procedure.
References
Hoag Memorial Hospital Newport Beach, CA 92663
I thank Dr. McCauley for pointing out the typographic error in my article [1]. In the original drafts, 0.5% was the number used; this somehow became 0.005% in the published manuscript. The derivation of the number 115 (of hypothetically cured patients) is based on the 0.5% of individuals who might be saved with screening, of the 23,000 who are likely to die from cancer per 100,000 population [1]. This calculation did not specifically derive from the 1777 patients examined in our study, in whom a 1% cancer prevalence was found; the true prevalence of cancer in that population is likely larger, given the abnormalities within that group for which we do not have follow-up. Because lung cancer is the single most common cause of cancer death and colon cancer is the second leading cause of cancer death and because a proven screening program (for lung cancer) or adherence to recommended screening programs (for colon cancer) is clearly of benefit, the expectation that 0.5% of all victims of cancer death might be expected to derive benefit is, in my opinion, not an unreasonable speculation. My article was not meant to provide a statistical basis for the value of CT screening; its intent was that of a perspective on how one individual places such a diagnostic option for patients within the current medical and social context.
The 40,000-foot view of cancer statistics suggests that use of diagnostic X rays does not contribute to carcinogenesis. The American Cancer Society's statistics document a decreasing incidence of cancer since the early 1990s, despite the dramatic increase in the use over the past two decades of nuclear medicine, diagnostic and interventional radiography, and CT (33 million scans obtained in 2001 alone). The incidence of and death rates from cancer are falling except in the African American population; the epidemiologists compiling these data suggest that such racial disparity reflects diagnosis at a later stage—perhaps due to lack of access to earlier diagnosis [2]. In fact, a recent review of the risks of low-dose radiation suggests that Dr. McCauley's fears of carcinogenesis caused by radiation in the range of a CT screening examination may be groundless [6].
Despite the speculation of Dr. McCauley and published estimates discussed in the article, we do not truly know whether diagnostic radiation causes cancer or not. We also do not know that the use of diagnostic radiation prolongs life. What we do know is that use of diagnostic radiation allows detection of cancer and other diseases in their early stages, that use of CT is replacing the physical examination in daily practice, and that cancer incidence rates and survival statistics have improved in the recent decades despite dramatic growth of diagnostic radiation use. That we may be doing more harm than good is speculation; the reality is that most medical practice lacks rigorous scientific proof of value. For instance, it has been estimated that less than 20% of what is done in routine medical practice has a basis in best practice and published scientific research. [3, 4] Others have shown that expert consensus cannot be achieved for medical management decisions when a scientific basis is absent [5]. Radiologists in clinical practice do studies and interventions every day that are not based on double-blind controlled studies of outcome (imaging the spine and intervention therein for back pain, as an obvious example) and that may potentially harm. No double-blind controlled study documents improved mortality rates as a result of coronary artery stenting or even bypass graft procedures (except for left main coronary artery disease); nevertheless, those treatments with their complications are ubiquitous.
I understand McCauley's hesitance to accept widespread use of CT for early disease detection until he knows for sure that he is not harming more people than he is helping. I wonder if he applies that criterion throughout his practice. Given how the rest of medicine is practiced, and a thorough personal review of the data at hand, I am obviously less hesitant than he. I hope that readers can learn from this interchange, and realize that, in this arena, rational people can differ.
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 |
