post by Bill Gardner
Six years ago, after celebrating my 50th birthday, I chose to begin an annual ritual of having my prostate-specific antigen (PSA) level checked... One year later, I had my PSA level checked again; it was 2.5 ng/mL... I chose to be an informed patient and got an opinion from leading medical, surgical, and radiation oncologists throughout the country. In the end,... I chose surgery, viewing the risks as small and the benefits as great. Fast forward 5 years: cancer free. However, as a result of the surgery, my right arm and right leg are permanently weak, with this deficit appearing immediately after surgery. The reasons for this outcome are unclear. My PSA level remains 0, but my daily 5-mile jog is no longer possible.
He is an example of the problem that has led many to question the wisdom of PSA screening: death is not necessarily quick or certain if you have prostate cancern, and the cost of being "cancer-free" is often very high. The oncologist continues,
Where am I now on the PSA dilemma in light of the recent US Preventive Services Task Force recommendations? It is clear that prostatectomy results in a very high chance of 20-year prostate cancer–specific survival, but even when the procedure is performed by an expert urologist, it can also result in significant rates of sexual, bladder, and bowel dysfunction and other less common adverse effects, such as my weakness. Active surveillance with longitudinal PSA tests and physical examination is associated with very low rates of bowel, bladder, and sexual dysfunction and has a high probability of correctly identifying when to move from surveillance to treatment. If I could do it all over again, I would not undergo the surgery; instead, I would opt for active surveillance.
Emphasis mine. He is not regretting the decision to be screened, he is regretting the decision to have immediate surgery, instead of closely monitoring the cancer and having surgery when he clearly needs it. Without further study, we can't say with confidence that patients like the oncologist would do better with active surveillance. Moreover, he might not have been able to endure the concern about cancer. The point, though, is that the screen result was just information. What mattered is what the patient and his doctors decided to do with the screen.
A screen in and of itself provides little benefit: it only benefits the patient if it leads to a beneficial treatment. In a manuscript under review, I discuss how the potential benefit of depression screening programs is largely wasted because the clinical processes that convey patients from positive screen results to successful treatments break down at myriad points. Lots of mundane clinical process issues -- for example, does the doctor have time to check the EHR for a screen result before talking to the patient -- affect the results of screening as surely as does the biochemistry of the screening assay. I do not think that it has been sufficiently stressed that, as Aaron said on Twitter, "Screening is a process, not a test."