The worst kind of guideline

Before critiquing the American College of Physicians' recent guidance statement on screening for prostate cancer, I will begin by saying that I generally hold the ACP's clinical guideline development process in high regard. They often base their guidelines on comprehensive and methodologically sound reviews of the evidence produced by Evidence-Based Practice Centers. In some cases, when several good-quality guidelines are already available, the ACP chooses not to reinvent the wheel and instead critically appraises the existing guidelines, as it did this time. The authors of the guidance statement are respected experts in evidence-based medicine, including Informed Medical Decisions Foundation president Michael Barry and current U.S. Preventive Services Task Force member Douglas Owens.

That being said, I won't mince words. On screening for prostate cancer, the ACP's guideline committee laid an egg.

The ACP's appraisal of the evidence on screening with the prostate-specific antigen (PSA) test was remarkably similar to the U.S. Preventive Services Task Force's assessment: harms of the test outweigh benefits. The ACP recommended that clinicians inform men between the age of 50 and 69 years of the rationale behind this conclusion, and provide a helpful, if overly extensive, list of "talking points with patients" which seem designed to discourage men from getting the test. All well and good. But what about men who absorb all of this information and still want to get screened? Well, the ACP says, go ahead and give them the test.

In what other area of medicine are physicians explicitly instructed to provide interventions that are judged do more harm than good because patients request them? To use a counter example, the vast majority of upper respiratory infections are due to viruses. Nonetheless, many of my patients request antibiotics. Antibiotics may benefit individual patients with upper respiratory infections in exceedingly rare cases, but on a population level clearly do more harm than good - leading, for example, to diarrhea, allergic reactions, and increasing bacterial resistance. But what if, after I finish patiently explaining all of these facts, the patient still "expresses a clear preference" for an ineffective and potentially harmful antibiotic prescription? Should I then go ahead and prescribe it? If I applied the ACP's approach to PSA screening, the answer would be yes.

In a thoughtful commentary published this week in JAMA Internal Medicine, L.A. County Department of Health Services Director Mitchell Katz asks the obvious question:

First, when as a profession did we decide that we had an ethical obligation to offer interventions that cause more harm than good? When we offer an intervention that is on the whole detrimental, are we not sending our patients a mixed message? Presumably there are any number of interventions that cause more harm than good. Should we be offering our patients a menu of ineffective interventions on the idea that they are better able than we are to determine effectiveness?

The ACP's guidance statement on PSA screening is even more perplexing when viewed in the context of the organization's participation in laudable efforts to prevent unnecessary or harmful medical care, including the High Value Care Initiative and the Choosing Wisely campaign. The success of these campaigns will depend on the ability of individual physicians and health systems to reduce waste and refocus resources on care that is most likely to benefit patients. Dr. Katz goes on to recount a conversation with a colleague about the hypothetical elimination of PSA screening tests in his health system:

If I cannot eliminate PSA testing, for which there is a USPSTF conclusion that the test cannot be recommended for men of any age, what is the likelihood that I could eliminate any low-value test, given that most things are not so well studied with such clear expert advice? ... Much of why the US health care system is so expensive for the benefits we gain is because we do not treat it like a system. If a new test or treatment is approved, we increase premiums to pay for it. Although we are slowly incorporating cost-effectiveness data into medical choices, especially when choosing among drugs used for the same indication or among diagnostic algorithms for evaluating a particular symptom, there is little dialogue about how to divert money from low-value care to higher-value care.

Good guideline recommendations must be concordant with the evidence upon which they are based. I may disagree vehemently with the American Urological Association about the value of PSA screening, but to their credit, their 2009 guideline at least interpreted the evidence as supporting a net benefit from the test. But to conclude, as the ACP did, that the PSA test causes more harm than good to patients, but allow physicians to provide it anyway, is nonsense. It is a huge step backward from high-value care. It is Choosing Unwisely. It is, in my opinion, the worst kind of guideline.