TO TEST OR NOT TO TEST: WHAT IS THE VALUE OF KNOWING?

PURPOSE:  To develop a conceptual model for understanding the impact of the “value of knowing”, defined as the value of information from medical tests exclusive of treatment or life-planning decisions, on a patient’s decision to undergo testing.

 

METHODS: We drew upon the behavioral economic, loss-aversion, cost-benefit and willingness-to-pay literatures to develop a mathematical model how a medical diagnostic test affects patients’ sense of well-being, and how this phenomenon affects their decision to undergo testing.  The model allows simultaneous evaluation of the impact of baseline (pre-test) disease risk, test inaccuracy, prior information, worrying over disease onset, time preference, and the degree of loss aversion on patients’ net assessment of the value of knowing.  We then simulate the net value of knowing under alternative hypothetical scenarios about test accuracy and patient characteristics, and also provide examples of how the value of knowing helps explain observed over- and under-utilization of selected diagnostic tests. 

 

RESULTS: Patients agree to testing when the expected benefits from good news (measured by willingness to pay) exceed the psychic costs of bad news (measured by willingness to accept).  The value of knowing from testing is shown to depend on test accuracy, pre-test disease risk, the patient’s discount rate, time to disease onset, and the patient’s aversion to receiving bad news (loss).  Simulation results indicate that the value of knowing increases (and testing becomes more likely) when: tests are more accurate; the baseline expectation of a positive test is low and the adverse consequences of a positive test are either small or occur far in the future; or patients do not worry about onset of future disease.   The model’s results are consistent with low observed testing rates for Huntington’s Disease, and low colorectal and breast cancer screening rates.  They also help explain the persistently high rate of imaging studies for diagnosing acute low-back pain despite evidence that such studies are rarely medically necessary.

                                                                                                                                                                  

CONCLUSION: The value of knowing has largely been ignored in the literature, but has important implications for understanding the testing decision, explaining observed testing patterns, evaluating the cost-effectiveness of new diagnostic tests, and predicting how patient and diagnostic test characteristics affect the propensity to test.  Further theoretical and empirical research on the value of knowing is needed.