ASSUMPTIONS PATIENTS MAKE WHEN INTERPRETING GRAPHICAL DISPLAYS OF SURGICAL DATA

Previous research on patient preferences in surgical decision making h as suggested that most patients are willing to exchange a probability of immediate mortality of 10% for a 15% better 5-year survival rate. T o study whether patients base their choices on extrapolations about fu ture trends beyond year 5, we studied 101 consecutive patients seen in a university-based Department of Veterans Affairs Medical Center. Mea n age was 66.7 years (SD = 8.87, range = 40-84); mean length of educat ion was 12.8 years (SD = 11.40, range = 4-22 years). All patients were provided with a graph comparing two 5-year survival curves: one curve (worse shortterm, better long-term survival) had a 10% immediate risk of death and 37% 5-year survival; the other (better short-term, worse long-term survival) had 0% chance of immediate death and a 22% 5-year survival. Patients were then presented with two additional comparison s (a 6-year and a 10-year survival comparison); each included the same baseline 5-year survival data. In the 6-year comparison, both treatme nt curves had a O% chance of survival beyond year 6. In the 10-year co mparison, the first curve had a 10% 10-year survival, and the second c urve had a 5% 10-year survival. Patients were asked to give their trea tment preference on each comparison and to report any assumptions they made when interpreting the original 5-year survival curve. Fewer than 15% of patients showed preference patterns consistent with a signific ant influence of data beyond 5 years. For most patients, 5-year result s appear to be sufficient for choosing among alternative treatments. F or a minority of patients, explicit information about data extrapolati on leads to preference change. Ignoring this aspect of data interpreta tion could lead to miscommunication about patient preferences.