Explicitly Modeled Variables of Decision-Tree Analysis
Algorithm | Variable | Value of variable in base-case analysis | Supporting literature or other justification |
---|---|---|---|
C, P | Prob. of initial comprehensive H&P (including neurologic examination) | 1.00 | Required part of evaluation |
C, P | Prob. of needing CT or MRI | 0.625, 1.00 initial evaluation | Set to be equal in both algorithms |
(0.10 in comprehensive follow-up evaluation) | |||
C, P | Prob. of needing special tests or outside consultation (blue-shaded boxes in Figs. 1 and 2) | 0.035 (initial evaluation only) | Refs. (48,49); set to be equal in both algorithms |
C | Prob. of obtaining neuropsychologic testing | 0.59 | Ref. (50) |
C | Prob. of having multiple cognitive domains affected (including patients undergoing neuropsychologic testing) | 0.67 initial evaluation | Based on number of clinical dementias per documented AD case in Ref. (10) and adjusting AD prevalence to rate in Ref. (50) |
Additional 0.01 of remaining patients per month on follow-up evaluation | Based on conversion rates as documented in Refs. (51,52) | ||
C | Prob. of functional decline (among those with multiple cognitive domains affected) | 0.90 | Estimate based on clinical experience (no published values found) |
C, P | Prob. of potentially reversible cause of dementia evident on H&P or common laboratory tests | 0.22 | Refs. (48,49,53,54) |
C, P | Prob. of identifying potentially reversible dementia by H&P or common laboratory tests and completely reversing it with therapy | 0.045 | Refs. (48,49,53,55) |
C, P | Prob. of identifying potentially reversible dementia by CT or MRI, not found by H&P or common laboratory tests, and completely reversing it with therapy | 0.00 | Ref. (53) |
C = conventional; P = proposed; Prob. = probability; H&P = history and physical examination; Ref. = reference.