Health Services Research & Policy ■ Clinical Practice Management ■ Training & Education ■ Leadership
Health Services Research & Policy ■ Clinical Practice Management ■ Training & Education ■ Leadership
Health Services Research & Policy
■ Clinical Practice Management
■ Training & Education ■ Leadership

The Illusion of Diagnostic Imaging Choice

Due to confirmation bias, providers may overestimate the benefits of diagnostic imaging, especially of inappropriate and unnecessary studies.

A few days ago I saw a tweet with a link to a recently-published article entitled "The Science of Choosing Wisely – Overcoming the Therapeutic Illusion." The author explains that providers and patients alike tend to "overestimate the effects" of their decisions, especially therapeutic ones, leading to an "illusion of control," not unlike a superstitious gambler. If red-36 is a winner three times in one visit to the roulette table, was that just statistical chance? If you reject randomness and erroneously ascribe magical powers to your choice, you may decide that 36 is your lucky number, especially if you win with that number again shortly thereafter.

The author provides examples of the phenomenon of confirmation bias entrenched in medical treatment decision-making. He also explains its psychological underpinnings: the illusion of control by taking some action (any action) may confer beneficial positive feelings (provider confidence boosting, enhanced patient faith in the medical system overall and in the individual provider, reduced patient anxiety, insurance against litigation, etc) and is therefore (often) preferable to inaction.

So I applied the author's thinking to diagnostic imaging: What if (at least some of) diagnostic imaging has little actual impact on outcomes? When the test results confirm what was clinically suspected, the test may be venerated as a powerful diagnostic tool in the eyes of provider and patient alike. Confirmation bias leads you to order the test next time, perhaps when clinical suspicion is lower, because it was helpful before. Or was it just a lucky roll of the dice? Or was it even completely unnecessary because you didn't need actually it, since your clinical diagnosis correct before any test was performed?

Unnecessary diagnostic imaging trades off a batch of potential harms, such as ionizing radiation, intravenous contrast, incorrect diagnosis, incidental findings, etc., for other concerns when no study is performed, namely patient anxiety, lack of faith in the provider, provider fear of litigation, etc. The perceived value of any test may be more "valuable" to the individual (i.e. receiving the "peace of mind" of a negative test)than value defined as outcome divided by cost. When diagnostic imaging tests are fetishized in this way, it's not easy to change the status quo. Anybody who has ever unsuccessfully tried to talk a provider out of ordering rib radiographs for indications like "chest wall point tenderness after coughing" (or just the single word "pain") knows what I'm talking about.

During internship, my roommate and I had a mantra: "Get a test you don't need, get a result you don't want." You learned very quickly to have an unassailable, robust reason for requesting advanced imaging. If not, one "result" was that the attending would upbraid you for ordering an unnecessary test. Another "result" might be that you get an incidental finding that the patient might fret over and that you would have to manage (in the days before a handy collection of JACR articles was available for guidance), diverting a portion of your attention and energies from solving the initial clinical problem the patient presented with.

But that was over two decades ago and much has changed, such as the faster tempo of turnaround times and tests, greater temporal and spatial resolution of cross-sectional imaging studies, and increased defensive medicine. Tests are commonly performed before a visit to a specialist, rather than after initial in-office evaluation, such as the "pulmonary function test-chest radiograph combo" before you ever step foot into a pulmonologist's office.

Designed to assist providers in making the most appropriate imaging or treatment decision for a specific clinical condition, the ACR Appropriateness Criteria® (ACR-AC) forms the basis of ACR Select®, which is clinical decision support software that can be integrated into the electronic medical record and provide a low-level combination consultative-gatekeeping function. But as the name of the ACR-AC implies, these products only provide relative appropriateness of imaging choices. The decision-making power still ultimately lies in the hands of referring providers and institutions.

Therefore, the ACR is also involved in proactive efforts to support better medical decision making. For example, the Image Wisely campaign created by the ACR-RSNA Joint Task Force on Adult Radiation Protection aims to curb the trend of increased societal ionizing radiation exposure by minimizing radiation dose in medically-necessary imaging, but also strives to decrease utilization (eliminate unnecessary imaging).

Better choices and lower radiation. What more could patients and providers ask for? How about an answer to the question "Does my patient need this test (or any test for that matter)?" That judgment remains in the domain of the ordering provider. The ACR can't (and shouldn't) police the entire medical community to provide the best and most efficient care possible.

However, tangible steps are being taken to change provider behavior. For example, the American Board of Internal Medicine (ABIM) Foundation's Choosing Wisely program was created to help physicians and patients choose care that is characterized as: "supported by evidence," "not duplicative of other tests or procedures already received," "free from harm", and "truly necessary." With respect to the last of the four elements, numerous medical societies have submitted additional recommendations to the ABIM regarding what NOT to do. Here are the five proffered by the ACR:

This is an important, albeit small, step on a long and necessary march toward the goal of pulling back the curtain on the illusion of control that diagnostic imaging studies conjure. By providing unambiguous evidence-based expert consensus statements that advise when it's acceptable and appropriate NOT to image, provider and patient anxieties may be reduced when it turns out that the best imaging study is no imaging study at all.

Such published authoritative statements regarding when imaging can be avoided cannot be left to chance, a mere gamble with fingers crossed for luck. It takes self-reflection and deliberate, thoughtful action at the level of individuals, institutions and medical societies. The ACR is leading the charge by building brick-by-brick on the foundations of the ACR-AC, such as instituting the productive Incidental Findings committee, the Imaging 3.0™ initiative, Patient and Family Center Care Commission, and many other college actions.Providing motivated and committed people with the resources of the college is making a difference. The ACR is helping people not simply pick among a more restricted set of lucky numbers ("appropriate studies") on the diagnostic imaging roulette wheel of chance, but to avoid the game completely by making the one best choice.

Additional reading from JACR

Johnson PT, Mahesh M, Fishman EK. Image Wisely and Choosing Wisely: Importance of Adult Body CT Protocol Design for Patient Safety, Exam Quality, and Diagnostic Efficacy. JACR 2015; 12: 1185–1190. 

Mahesh M, Durand DJ. The Choosing Wisely Campaign and Its Potential Impact on Diagnostic Radiation Burden. JACR 2012; 10; 65–66.

April Madness
Imaging Informatics Resource Center: A One-Stop Sh...

Comments 1

Guest - VIVAN Life Sciences on Friday, 07 April 2017 11:56

Hey it was good to read this post which is a great guide. Appreciate all these tips.

Hey it was good to read this post which is a great guide. Appreciate all these tips.
Already Registered? Login Here
Wednesday, 22 May 2019

Captcha Image

Comments are moderated. Click here for comment posting guidelines.