[NO] Reassurance After Normal Diagnostic Testing

[Michael Scally MD]

Rolfe A, Burton C. Reassurance After Diagnostic Testing With a Low Pretest Probability of Serious Disease: Systematic Review and Meta-analysis. JAMA Intern Med. 2013;173(6):407-416. JAMA Network | JAMA Internal Medicine | Reassurance After Diagnostic Testing With a Low Pretest Probability of Serious DiseaseSystematic Review and Meta-analysisReassurance After Diagnostic Testing

Importance - Diagnostic tests are often ordered by physicians in patients with a low pretest probability of disease to rule out conditions and reassure the patient.

Objective - To study the effect of diagnostic tests on worry about illness, anxiety, symptom persistence, and subsequent use of health care resources in patients with a low pretest probability of serious illness.

Evidence Acquisition - Systematic review and meta-analysis. We searched MEDLINE, the Cochrane Central Register of Controlled Trials, EMBASE, PsychINFO, CINAHL, and ProQuest Dissertations electronic databases through December 31, 2011, for eligible randomized controlled trials. We independently identified studies for inclusion and extracted the data. Disagreements were resolved by discussion. We performed meta-analysis if heterogeneity was low or moderate (I2 < 50%).

Results - Fourteen randomized controlled trials that included 3828 patients met the inclusion criteria and were analyzed with outcomes categorized as short term (?3 months) or long term (>3 months). Three trials showed no overall effect of diagnostic tests on illness worry (odds ratio, 0.87 [95% CI, 0.55-1.39]), and 2 showed no effect on nonspecific anxiety (standardized mean difference, 0.06 [?0.16 to 0.28]). Ten trials showed no overall long-term effect on symptom persistence (odds ratio, 0.99 [95% CI, 0.85-1.15]). Eleven trials measured subsequent primary care visits. We observed a high level of heterogeneity among trials (I2 = 80%). Meta-analysis after exclusion of outliers suggested a small reduction in visits after investigation (odds ratio, 0.77 [95% CI, 0.62-0.96]).

Conclusions and Relevance - Diagnostic tests for symptoms with a low risk of serious illness do little to reassure patients, decrease their anxiety, or resolve their symptoms, although the tests may reduce further primary care visits. Further research is needed to maximize reassurance from medically necessary tests and to develop safe strategies for managing patients without testing when an abnormal result is unlikely.

 

[Michael Scally MD]

Diagnostic testing is enticing to patients and clinicians. It appears more objective and less pedestrian than a simple clinical interview and physical examination. Medical certainty is seldom solidified until all the tests results are in. Patients anxiously await the telephone call or letter announcing “your tests are all normal.” Indeed, the grander the technology, the more alluring. However, the testing imperative can become addictive. As noted in a 1991 cautionary essay:

Technology pounds upon the shore, but the danger is the undertow. The effacement of sand castles we abide; the relentless tug is another matter, sucking us deeper. Like systole and diastole, there is faint pause, endless indications. Imaging fits the metaphor, wave after wave: radionuclide scanning, computerized tomography, magnetic resonance imaging, duplex sonography. The very names captivate our diagnostic instincts, and yet excess appears inevitable. Endoscopy is equally irrepressible. To witness a cause transcends the more banal concerns of costs and therapeutic outcome. Follow-up is inconvenient. To wait and see whether growing suspicions will justify exploration or whether signs and symptoms prove merely transitory cannot compete with immediate visualization.

Like many of our treatments, however, diagnostic testing is not without its adverse effects. Increased health care costs are the most obvious: wide geographic variations in the use of expensive tests persist more than 30 years after such inexplicable variation was first exposed. Still more insidious consequences lurk. One is the problem of false-positive results. The prevalence of detecting a serious condition may be as low as 0.5% to 3.0% when diagnostic tests are ordered in patients with a low probability of disease, meaning that a diagnostic test with a 90% sensitivity and 90% specificity would yield 4 to 19 false-positive results for every true-positive result in patients for whom the test is ordered simply to rule out a disease for which the clinical suspicion is already low. This disproportionately high false-positive rate may then cascade into additional and sometimes invasive procedures, not to mention considerable patient anxiety that may persist months after a negative finding of a workup cancels out the initial test results. One might consider this PTSD (post test stress disorder) an iatrogenic variant of the traditional PTSD (posttraumatic stress disorder).

False-negative results can also be a concern. For example, the high diagnostic accuracy of abdominal computed tomography for appendicitis and renal colic does not generalize to patients presenting to the emergency department with undifferentiated upper abdominal pain, where the negative predictive value is only 64%. That means as many as 1 of 3 normal abdominal computed tomographic scans in this population may represent a false-negative test result, with the most commonly missed pathologic changes being inflammatory conditions of the biliary tract and upper gastrointestinal tract systems.

In addition, reflexive test ordering may marginalize the clinical examination. Preliminary data suggest that that the history typically accounts for 75% or more of the diagnostic yield when evaluating common symptoms; the physical examination, 10% to 15%; and testing, generally less than 10%. Ironically, the US reimbursement system financially incentivizes these components in the reverse order. Diagnostic testing and procedures receive the highest remuneration and, even within the clinical encounter, evaluation and management coding favors from a billing standpoint the physical examination of more bodily parts (even if irrelevant to the presenting complaint) over a detailed and more diagnostically informative interview.

Despite these limitations of diagnostic testing in patients with a low probability of disease, a conventional justification is reassurance of the patient. However, the meta-analysis by Rolfe and Burton suggests that even this benefit may be overestimated. The authors included only trials in which patients with a low probability of disease were randomized to receive initial diagnostic testing vs a nontesting approach. The patient sample is appropriate because most would agree that diagnostic testing is warranted to rule in or confirm a suspected disease, determine its extent or severity, establish its prognosis, and/or facilitate treatment decisions. The inclusion of only randomized clinical trials reduces confounding by patient and physician factors that influence test ordering in naturalistic settings and that could bias perceived benefits.

Most of the 14 trials that met the authors' inclusion criteria involved recent-onset rather than persistent symptoms. If anything, this might have enhanced the yield of diagnostic testing because symptom chronicity makes serious causes less likely. Also, none of the trials blinded the patient or clinician. This could have inflated group differences in reassurance because some patients randomized to nontesting probably still desired a test, as might their treating clinician. This disappointment in both parties could have heightened anxiety in the control group.

The meta-analysis found no benefits of diagnostic testing on reducing illness worry or anxiety, although only a few studies examined these outcomes. Moreover, no effect on symptom status was found. A small decrease occurred in the likelihood of having a return clinic visit in the group randomized to initial testing. However, the number needed to investigate to prevent a return clinic visit ranged from 16 to 26 patients. If the cost of these expensive diagnostic tests is estimated conservatively to be $250 to $500 per test, $4000 to $16 000 would be expended to prevent a $100 primary care visit.

Although the meta-analysis focused on higher-cost tests (imaging, ultrasonography, and cardiac testing), the findings likely apply to other diagnostic tests as well. However, most studies were performed in Europe, where a more socialized health care system rather than a fee-for-service system might already constrain testing. If expectations for testing are higher in the United States, initiatives to reduce testing may have more deleterious effects on reassurance. Because the meta-analysis is based on a small number of trials across a variety of tests, many might claim “more research is needed” before advocating for more restrained diagnostic testing. Alternatively, one could argue more research is needed in defense of diagnostic testing. Although rigorous evidence is required before therapeutic drugs or devices can be marketed and introduced into practice, diagnostic tests tend to be adopted more indiscriminately and with potential adverse consequences of their own.

How might this meta-analysis as well as related research on reassurance inform clinical practice?
First, rather than turning off the faucet completely, one might order diagnostic tests selectively based on patient factors such as greater illness anxiety, symptom persistence, or complexity.
Second, we should not automatically assume most patients want a test. Indeed, physicians tend to order more investigations than patients desire. Just as we are still trying to reduce patient expectations for antibiotics induced by a generation of overprescribing, reversing the tide of testing will be a slow process.
Third, when we do test, we might enhance reassurance through written or verbal information on the meaning of normal results before or after the test.
Fourth, evidence-based guidelines for diagnostic testing in common conditions should be developed, such as have been developed for low back pain. Less desirable and more onerous options would be increasing patient copayment or requiring preauthorization for certain types of testing.
Fifth, we must address reasonable physician concerns about malpractice (ie, defensive test ordering) and patient demand.
Finally, because most test results are normal when ordered for common symptoms, simply stating “everything is normal” is insufficient. In the aftermath of a normal test result, effective communication and symptom management remain paramount. -

Kroenke K. Diagnostic Testing and the Illusory Reassurance of Normal Results: Comment on “Reassurance After Diagnostic Testing With a Low Pretest Probability of Serious Disease”. JAMA Intern Med. 2013;173(6):416-417. JAMA Network | JAMA Internal Medicine | Diagnostic Testing and the Illusory Reassurance of Normal ResultsComment on “Reassurance After Diagnostic Testing With a Low Pretest Probability of Serious Disease”Diagnostic Testing