Concordance of couples' prostate cancer screening recommendations from a decision analysis

Abstract

Objective: To determine whether different utilities for prostate cancer screening outcomes for couples, and husbands and wives separately, lead to incongruent screening recommendations.

Methods: We evaluated survey results of 168 married couples from three family practice centers in Texas, USA. Utilities for eight adverse outcomes of prostate cancer screening and treatment were assessed using the time trade-off method. We assessed utilities separately for each partner and jointly for each couple. Using a previously published decision-analytic model of prostate cancer screening, we input the husband's age (starting point) and utilities for outcomes from the husband's, wife's, and couple's perspectives (to adjust for quality of life). Both group-level and individualized models were run. We also asked husbands (and wives) if they intended to be screened (or have their husbands screened) for prostate cancer in the future.

Results: Husbands' lower tolerance for adverse outcomes (lower utilities) was associated with lower quality-adjusted life expectancy (than their wives) for the choice of screening versus not screening. Depending on the perspective, 48 husbands (28.6%), 89 wives (53.0%), and 58 couples (34.5%) preferred screening in the individual decision-analytic models. Comparing the three perspectives, agreement in model recommendations was greatest between the husbands and the couples (82.1%), intermediate between the wives and couples (63.7%), and lowest between the husbands and wives (55.4%). Using group-aggregated utilities in the decision-analytic model tended to mask the variation in recommended strategies amongst individuals. There was no relationship between screening preferences from the model and the husbands' and wives' reported desire for screening, as the majority of subjects wanted screening.

Conclusions: Discordant health preferences may yield conflicting recommendations for prostate cancer screening. The results have broad implications for informed healthcare decision making for couples.