Numeracy and the shortcomings of utility assessment in head and neck cancer patients

Background. Because survival differences between surgical and nonsurgical treatment for head and neck cancer (HNC) are hard to detect, increasing focus has been placed on quality of life (QOL) differences after treatment. Utility assessment provides insight into COL. Evidence suggests that a patient's comfort with numerical concepts (numeracy) may influence utility measures. We hypothesize that patients who are nonnumerate provide inconsistent utility data in COIL studies. Methods. New HNC (n = 18) patients were recruited to participate. Patients completed a numeracy questionnaire, a utility assessment, and a global COIL questionnaire. Higher scores reflect better function. Interviewers rated the functional level of each patient. For both numerate and nonnumerate patients, utility scores were compared with global QOL (good vs poor) and observer-rated function. Results. Half of the patients were numerate. Numerate patients who rated their COIL as good had significantly higher utility scores than did patients with poor global COIL (0.95 vs 0.43, p =.03). In contrast, nonnumerate patients with good COIL had lower utility scores than did patients with poor COIL (0.45 vs 0.77 NS). Utility scores for numerate patients correlated well with observer-rated function (r= 0.41 to r= 0.57), whereas those of nonnumerate patients did not (r = -0.16 to r = 0.06). Conclusions. QOL evaluation through utility assessment may provide inaccurate and contradictory data about patient functioning for nonnumerate patients. This may confound COIL assessment when interpreting utility data. (C) 2004 Wiley Periodicals, Inc.*