Investigation Momentum: The Relentless Pursuit to Resolve Uncertainty

Debate regarding the prostate-specific antigen (PSA) screening test centers around test reliability and whether screening reduces mortality.^1-3 We consider yet another potential downside to the widespread use of unreliable screening tests: the downstream effect of receiving inconclusive or ambiguous results. When receiving information from screening tests, we usually want to know whether the result is a “yes” or a “no.” Receiving an inconclusive result amounts to a “don't know”; this situation should have a level of uncertainty regarding the diagnosis similar to that of not conducting the test at all. Yet, we propose that the psychological uncertainty experienced after an inconclusive test result can lead to investigation momentum: additional, and potentially excessive, diagnostic testing. In contrast, not conducting the unreliable test would result in no further action. To investigate this, we evaluated whether receiving an inconclusive result from an unreliable test (the PSA screening), compared with undergoing no test, motivated more individuals to undertake an additional, more invasive and costly, test (a prostate biopsy).

We recruited 727 men aged between 40 and 75 years to an online survey through e-mail solicitation (data were subsequently collected and analyzed anonymously). Participants received information regarding prostate cancer and answered some background questions (see Table). They were randomized to 1 of 4 conditions. In the first condition, “no PSA,” participants were given information about the risks and benefits of prostate biopsies and asked whether they would have a biopsy (yes or no) and their certainty, ranging from −100 (most certain they would not undergo biopsy) to +100 (most certain they would). In the other 3 conditions, participants were given information about PSA tests, as well as prostate biopsies, and were then presented with a scenario that asked them to imagine that they had just received their PSA test result at 1 of 3 PSA levels: normal, elevated, or inconclusive—the latter result stated “this result provides no information about whether or not you have cancer.” After getting the information about the PSA test and a particular outcome, participants were asked to indicate whether, under these conditions, they would undergo a biopsy and their level of certainty in that decision.

We conducted 2-sided χ² tests and analyses of variance with planned contrasts to examine differences between the 4 conditions. P < .05 was considered statistically significant.

There were no significant differences among the 4 conditions in participant age, believed likelihood of developing prostate cancer, previous screening examinations, race, and family history of prostate or breast cancer (Table). As hypothesized, significantly more men indicated that they would undergo a prostate biopsy if they received an inconclusive PSA test result (40%) than if they had no PSA test (25%; χ² = 8.80; P = .003). Men in the elevated and normal PSA level conditions also responded significantly differently: Those assigned an elevated PSA test result were more likely to state that they would undergo a biopsy (62%) compared with those who had no PSA test (χ² = 47.76; P < .001) and compared with those assigned an inconclusive PSA test result (χ² = 17.89; P < .001) (although 38% of men with an elevated PSA test result still would not opt for a biopsy). Those assigned a normal PSA test result were less likely to state that they would undergo a biopsy (13%) compared with those who had no PSA test (χ² = 8.47; P = .004) (demonstrating some, but not total, reassurance from receiving a normal PSA test result)⁴ and compared with those assigned an inconclusive PSA test result (χ² = 35.85; P < .001) and those assigned an elevated PSA test result (χ² = 97.80; P < .001). Similar results, and significance, were obtained with participants' certainty ratings (Table).

These results are likely not confined to the PSA test and provide evidence that an inconclusive test result sparks investigation momentum. When tests give no diagnostic information, rationally, from an information perspective, it should be equivalent to never having had the test for the purpose of future decision making. Yet, we find that more men would opt to undergo the more invasive biopsy after receiving a meaningless test result than when they have no result at all. This has financial and clinical implications owing to the cost and invasive nature of further investigations. Furthermore, such a tendency is likely to have an impact not only on those physicians who incorporate patients' preferences into medical decision making⁵ but on all physicians, because they are vulnerable to the same psychological processes that encourage patients to resolve ambiguity.⁶ These results suggest that the ubiquitous use of simple but unreliable screening tests may lead to consequences beyond the initial cost and patient anxiety of inconclusive results; they could also lead to investigation momentum. Furthermore, it is possible that just ordering a test may lead to a commitment to pursue and find abnormalities.⁷ As we have also seen previously, sometimes “less is more” when it comes to health care procedures with incremental benefit but definite risks such as imaging for low back pain (a symptom likely to generate investigation momentum).⁸ These findings need to be replicated in broader populations and clinical settings to allow us to better understand how ambiguous information affects medical decision making.

Correspondence: Dr Sah, Georgetown University, McDonough School of Business, 37th and O Street, Rafik B. Hariri Bldg, Washington, DC 20057 (ss3250@georgetown.edu).

Published Online: April 15, 2013. doi:10.1001/jamainternmed.2013.401

Author Contributions: Dr Sah had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: All authors. Acquisition of data: Elias. Analysis and interpretation of data: Sah and Elias. Drafting of the manuscript: Sah. Critical revision of the manuscript for important intellectual content: All authors. Statistical analysis: Sah and Elias. Obtained funding: Ariely. Study supervision: Sah and Ariely.