Five-year survival curves: how much data are enough for patient-physician decision making in general surgery?

OBJECTIVE

To find out whether patients' preferences for a particular type of treatment are changed by the way in which the information is presented (point estimates-year 0 and year 5 data points-or two five-year survival curves).

DESIGN

Prospective study.

SETTING

University-based Department of Veterans Affairs Medical Center, USA.

SUBJECTS

236 Consecutive patients seen in a general medical outpatient clinic.

INTERVENTIONS

Patients were asked to choose between two treatments, one with a 10% immediate mortality and 37% five-year survival (treatment A) and one with no immediate mortality and 22% five-year survival (treatment B). They were randomised to be given the survival data either as point estimates or as two five-year survival curves.

MAIN OUTCOME MEASURES

Number of patients who chose each treatment.

RESULTS

The mean age was 67 years (range 40-84) and mean duration of education was 13 years (range 0-22). Significantly fewer patients chose the treatment with better immediate survival when they were given the data as survival curves (47/140, 34% compared with 93/140, 66% p < 0.0001) than when they were given data as point estimates (57/96, 59%, compared with 39/96, 41%).

CONCLUSION

Patients are significantly more likely to make different decisions when given more data (in the form of five-year survival curves which show short, medium, and long term results) than when they are presented with only point estimates.