Factors Associated With Preferring to Discuss Life Expectancy
| Characteristic | Unadjusted Odds Ratio (95% CI) | P Value | Adjusted Odds Ratio (95% CI)a | P Value |
|---|---|---|---|---|
| Age, per year | 1.01 (0.98-1.03) | .62 | – | – |
| Female sex | 0.80 (0.61-1.05) | .11 | – | – |
| Race | .05 | .55 | ||
| White | Ref | Ref | ||
| Black | 0.56 (0.33-0.93) | 0.80 (0.45-1.43) | ||
| Other | 0.79 (0.53-1.17) | 0.82 (0.53-1.28) | ||
| Education | <.001 | .004 | ||
| <High school | Ref | Ref | ||
| High school | 1.27 (0.80-2.01) | 1.09 (0.66-1.82) | ||
| Some college | 2.06 (1.28-3.33) | 1.66 (0.96-2.85) | ||
| >College | 3.26 (2.05-5.19) | 2.18 (1.25-3.80) | ||
| Health literacy,36 per point | 1.05 (0.98-1.12) | .15 | – | – |
| (scale 3-15) | ||||
| Numeracy,37 per point (scale 3-18) | 1.09 (1.05-1.14) | <.001 | 1.04 (0.99-1.09) | .11 |
| Predicted 10-year mortalityb,31 | 0.81 (0.46-1.43) | .47 | – | – |
| Self-perceived 10-year mortalityb | 1.08 (0.62-1.87) | .78 | – | – |
| Doctors can accurately predict | <.001 | <.001 | ||
| life expectancy21 | ||||
| Disagree | Ref | Ref | ||
| Neutral | 1.36 (1.01-1.83) | 1.59 (1.14-2.21) | ||
| Agree | 3.10 (2.05-4.69) | 3.06 (1.93-4.86) | ||
| Has had life-threatening illness32 | 1.39 (1.04-1.86) | .03 | 1.50 (1.07-2.09) | .02 |
| Has discussed life expectancy | 4.22 (3.05-5.85) | <.001 | 3.98 (2.82-5.62) | <.001 |
| of a loved one32 | ||||
| Complete trust in doctor35 | .36 | |||
| Disagree | Ref | – | – | |
| Neutral | 1.38 (0.80-2.37) | – | – | |
| Agree | 1.44 (0.87-2.39) | – | – | |
| Prefers shared decision making or leaving decision to doctor (vs making own decisions)33 | 1.04 (0.79-1.37) | .79 | – | – |
| Religion is important34 | 0.65 (0.48-0.87) | .004 | 0.69 (0.49-0.97) | .03 |
Note: Preference expressed in the hypothetical patient scenario. See Methods for description.
Ref = reference group.
↵a Only variables having P ≤.05 in univariate analysis were included in multivariate analysis.
↵b Analyzed as continuous variables with range of 0 to 1 in the regression model. For example, one person’s 10-year mortality risk may be 0.02 or 2%, and another person’s 10-year mortality risk may be 0.92 or 92%. Here, the odds ratio is per 1 unit of mortality risk, ie, comparing 100% mortality risk with 0%.