Advance directives have been widely endorsed as a method to match medi
cal interventions to patients' preferences. However, applying advance
directives to unspecified future decisions may be problematic. We wond
ered how well scenario-based treatment choices in advance directives c
an be used to infer other choices specified by the same patients. We a
sked 495 out-patients from the Massachusetts General Hospital to state
11 treatment choices in each of four illness scenarios. We calculated
likelihood ratios (LRs) to see how well one choice could predict anot
her. Predictions within a given scenario were strong and followed a pa
ttern that reflected the invasiveness of the treatment. Thus, decline
of antibiotics predicted decline of major surgery, in the same scenari
o (LR 36.0-108.3). Requests for major surgery strongly predicted reque
sts for antibiotics (LR 90.4 - 244.1). Requests for major surgery and
decline of antibiotics were the two most predictive preferences; more
traditionally considered decisions about cardiopulmonary resuscitation
and mechanical ventilation ranged from the third to ninth most predic
tive among the 11 interventions. Predictions between scenarios were we
aker, but still potentially useful, and followed a pattern that reflec
ted the prognoses of the scenarios. Declining treatment in the best pr
ognosis scenario (coma-with-a-chance) predicted declining the same tre
atment in other scenarios with LRs of 2.5-6.1. Requests for treatment
in the worst prognosis scenario (dementia-with-terminal-illness) predi
cted the same request in the other scenarios with LRs of 5.2-30.5. The
se data suggest that patients' advance scenario-based treatment choice
s can provide potentially useful information to apply to unspecified d
ecisions if inferences follow considerations of prognosis and treatmen
t invasiveness. For maximal predictive power, treatment directives sho
uld use a range of illness scenarios and include choices, among severa
l others, on antibiotic use and major surgery.