Loading...
After the U.S. Preventive Services Task Force (USPSTF) published its final recommendation opposing prostate-specific antigen (PSA) screening, we conducted an online poll of readers' reactions. A total of 177 readers responded to the question “Please choose the statement that best fits your reaction to the USPSTF recommendation against PSA screening.”
As shown in the table, 78% of respondents agreed with the USPSTF recommendation, but a substantial proportion still will offer screening selectively. I'm guessing — based on informal discussions with other physicians — that some of these responses reflect concerns about litigation for failure to diagnose cancer.
Many readers shared their perspectives by posting “reader remarks” in response to our recent summary of the USPSTF recommendation (JW Gen Med Jun 7 2012). One reader will continue to screen because she “has probably saved the life” of several men with screening. She might be right: Even skeptics must concede that individual lives occasionally are saved by screening. After all, if enough men undergo prostatectomy, somewhere in the mix are men who eventually would have died of prostate cancer. But, the important question is this: How many people must undergo screening, biopsies, prostatectomies, and radiation therapy to benefit one person? The Task Force concluded that the number needed to screen (NNS) and number needed to treat (NNT) are too high and result in adverse outcomes for too many people, whereas advocates of screening believe otherwise. Nothing is wrong with arguing that a particular NNS or NNT is too high, as long as we remember that selection of “appropriate” cutoffs are value judgments and not scientific truths.
Another reader entitled his remark “Can't tell who is saved.” He correctly implies that when a man survives (cancer-free) after treatment for PSA-detected cancer, we can't determine whether his particular life was saved by screening. From the European randomized screening trial (JW Gen Med Mar 14 2012), we can infer that roughly 1 of every 30 patients who received treatment for screening-detected cancer had his life extended, but we don't know which particular man was “saved” and which 29 underwent treatment unnecessarily.
Several respondents claimed that mismanagement of PSA results and overtreatment of patients with low-risk prostate cancer — and not PSA screening — are the real problems. In my view, both the PSA test and overtreatment are problematic. No screening test has perfect sensitivity and specificity, but PSA test accuracy is especially poor: Fully 25% of men with PSA levels between 2 and 4 ng/mL have prostate cancer (JW Gen Med Jun 8 2004), and many men with PSA levels between 4 and 10 ng/mL don't have prostate cancer. Sensitivity and specificity can be refined somewhat by using age-specific cutoffs, change in PSA level over time, or other variations; but so far, these other approaches have not been tested rigorously in controlled studies. Management approaches are all over the map because clinicians don't quite know what to do when PSA levels go up a little: Biopsy now? Repeat in 1 year? Repeat in 6 months? Give antibiotics for “prostatitis,” and repeat in 1 month? And, regarding overtreatment of men with low-risk cancer, thoughtful urologists have told me, “I agree that we overtreat. But if a patient who doesn't really need surgery says, ‘I want my cancer treated,' what are we supposed to do? If we don't do the surgery, he'll go elsewhere.”
Another reader suggested that PSA screening is most beneficial in men older than 75. However, in the European screening trial, among men 70 or older at the time of randomization, researchers noted a trend toward higher mortality in screened versus nonscreened men. And, in the largest treatment trial (prostatectomy vs. watchful waiting in men with localized cancer; JW Gen Med Sep 16 2008), prostatectomy was associated with lower mortality only in men younger than 65.
One final interesting comment: A physician reader notes that when patients ask him whether he gets PSA tests himself, he replies that he does not, even though “my father and father-in-law had prostate cancer.” In some clinical encounters, it might be appropriate to share one's personal medical decisions. But, I believe that when patients ask about PSA testing, physicians should explain why they agree or disagree with screening and leave their own healthcare decisions out of the discussion.
