How Choosing Wisely Decides a Test Is Low Value

Choosing Wisely never runs a test through a scoring formula to declare it low value. The judgment is delegated. Each participating specialty society writes its own list of tests and treatments that clinicians and patients should question, and every candidate item has to clear three rules the ABIM Foundation set at the outset: it must fall within that specialty's own control, it must be common or costly, and there must be evidence to support questioning it. The first two rules are easy to satisfy. The third one bends, ranging from randomized trial data on one item to committee judgment on the next, which is exactly why a Choosing Wisely line is best read as a well-sourced question, not a verdict.

Key points#

• The campaign delegates each list to the relevant specialty society; there is no central algorithm.
• Every item must pass three tests: specialty control, high volume or cost, and supporting evidence.
• The evidence test is the weak link, because "evidence" can mean trials or consensus, and the list rarely tells you which.
• Self-regulation gives the lists credibility but also a blind spot: societies find it easier to flag work at the edges than the high-volume procedures at the center.
• Publishing a recommendation barely moves care on its own; the tools built around it do the work.

The three tests an item must pass#

The ABIM Foundation asked societies to hold each candidate recommendation against three parameters.

First, the practice has to sit inside the specialty's own purview. The physicians who order the scan or perform the procedure are the ones who name it. This keeps the recommendations credible, since they come from the people who actually control the behavior.

Second, the practice has to be common or carry meaningful cost. That focuses attention on care delivered at scale rather than on rare situations, so a successful list item touches many patients.

Third, there has to be evidence to support questioning the practice. On paper this reads like the strongest of the three. In practice it is the loosest, and it is where you should slow down.

Why the third test is the one to watch#

"Evidence to support the recommendation" is a wide phrase. At one end it means a stack of randomized trials showing a test does not improve outcomes. At the other it means sound physiologic reasoning plus professional consensus that a test almost never changes what a clinician does next. Both are legitimate grounds for questioning a practice. They are not the same strength of claim, and the list itself usually does not label which one you are looking at.

Two familiar examples sit at opposite ends of that range. The recommendation against imaging for uncomplicated low back pain in the first weeks of symptoms rests on a deep evidence base, with trials and guidelines showing that early scans do not improve outcomes and often set off downstream tests and procedures that do harm. The long-standing advice against antibiotics for viral upper respiratory infections is similarly well supported. By contrast, a recommendation against a rarely useful lab panel may rest mainly on the reasoned judgment of the writing committee, with little trial data behind it. Neither kind of recommendation is wrong. A reader who treats every line as equally proven is simply overreading the method.

A method built on self-regulation#

Because each item must fall within a specialty's own control, the whole campaign is an exercise in self-regulation. That design has a genuine strength: the clinicians who understand when a test helps are the ones best placed to say when it does not, and asking them to name their own low-value practices is more honest than an outside body doing it for them.

It also has a limit worth stating plainly. Societies find it more comfortable to question practices at the periphery of their work than the high-volume, high-revenue procedures at its core. Independent analyses have noted that the lists tend to favor targets that are safe to name. The practical takeaway is that you should read a Choosing Wisely list as a floor for reducing overuse, a set of clear early wins, rather than a full map of everything that could be cut.

Does a published list actually change care?#

The most useful fact for anyone leaning on these lists is that writing one is not the same as changing behavior. A 2021 systematic review by Cliff and colleagues in the Milbank Quarterly examined 131 studies of Choosing Wisely interventions run between 2012 and mid-2019, and the pattern was consistent. When the intervention was simply telling clinicians the recommendation existed, only about 13 percent of efforts produced the intended reduction in low-value care. Active interventions worked far more often, around 65 percent. Programs that paired the recommendation with several tools, such as clinician education, decision support at the point of ordering, or performance feedback, succeeded about 77 percent of the time, compared with roughly 47 percent for single-tool efforts.

Two cautions keep that finding in perspective. Only about 17 percent of the studies used a control group, so much of the evidence comes from before-and-after designs that cannot fully separate the intervention from background trends. And published studies may tilt toward the ones that worked. Even so, the direction is clear and useful: the list names a target, but the machinery built around it, not the list on its own, is what shifts practice.

Reading a recommendation for what it claims#

Whether you are a clinician or an informed patient, the move is to treat a Choosing Wisely item as a starting question and then read it closely. Notice which society wrote it, because the recommendation reflects that specialty's vantage point. Follow the citations the society published alongside it, since the campaign asks societies to document their evidence, and see whether they point to trials, to guidelines, or to consensus. Check whether the wording is absolute or conditional, because most items say "avoid routinely" rather than "never," and the conditions carry the clinical meaning. A test that is low value for an average, low-risk person can be exactly right for someone at higher risk.

Read that way, Choosing Wisely is a durable and honest contribution to reducing overuse. Its method, transparent criteria applied by the specialists closest to the care, is a reasonable way to build such a list. It works best as the opening of an evidence-grounded conversation, and the quality of that conversation still depends on reading each recommendation for exactly what it claims.