Reference-dependent decision-making in colorectal cancer screening: Evidence from HINTS

Colorectal cancer is a pertinent health problem with 52,000 deaths annually in the US. While socioeconomic and access variables impact screening uptake, Whynes (2007) posits that human capital variables may only act as indicators of motive and fail to shadow the actual experience of screening utilization. This study extends the traditional expected-utility approach by exploring whether individuals with differing reference points, reflected by the extent to which they worry about acquiring colorectal cancer, exhibit differing behavior with respect to their decisions to opt for cancer screening with results compared across endoscopy and FOBT. Data was collected from the Health Information Trends Survey (HINTS) 2003. Following logistic regressions, marginal effects quantified the determinants of intention to screen in the next five years with previous screening behavioral, demographic and psychosocial variables included. Key explanatory variables were extent of worry, perceived susceptibility and an interaction term to allow differing effects. Intention to have endoscopy was higher than that of FOBT (43% v 32%). Evidence of differing screening behavior between the two groups supports reference dependence. ‘Worriers' who perceived themselves to have a high chance of acquiring cancer had a higher probability of endoscopy. Yet ‘worriers', with low perceived risk were statistically more likely to screen (ME=-0.234***), in contrast with expected utility theory. Our results suggest that, when seeking to understand the complexities of individuals' decisions to screen, heuristically based reference approaches such as prospect theory may be more suitable than the standard approach.