Abstract
BACKGROUND: The ability to determine which episodes of delirium are likely to lead to poor clinical outcomes has remained a major area of challenge.
OBJECTIVE: To quantify delirium severity and course over an entire hospitalization using several measures, and to evaluate their predictive validity for 30- and 90-day outcomes post-discharge.
DESIGN: Two prospective cohort studies.
PARTICIPANTS: Analysis was conducted in two independent cohorts of adult patients aged ≥70.
MAIN MEASURES: Nine delirium episode severity measures were examined: (1) measures reflecting delirium intensity (peak Confusion Assessment Method-Severity [CAM-S] and mean CAM-S score), (2) a measure reflecting delirium intensity and duration (sum of all CAM-S scores, sum of all CAM-S scores on delirium days only, peak CAM-S score x days with delirium), (3) measures requiring information on delirium duration and delirium at discharge (total number of delirium days, percentage of delirium days, delirium at discharge), and (4) a measure of cognitive change. Associations of the delirium episode severity measures with 30- and 90-day post-hospital outcomes (death, nursing home placement, and readmission) relevant to delirium were examined.
KEY RESULTS: The delirium episode severity measure that required information on both delirium intensity and duration (sum of all CAM-S scores) was the most strongly associated with 30- and 90-day post-hospital outcomes. Using this measure, the relative risk [95 % confidence interval] for death at 30-days increased across levels of sum of all CAM-S scores from 1.0 (referent) to 2.1 [0.8, 5.4] for 'low,' to 2.9 [1.2, 7.1] for 'moderate,' to 6.4 [2.9, 14.0] for 'high' (p for trend <.01).
CONCLUSIONS: The delirium episode severity measure that included both intensity and duration had the strongest association with important post-hospital outcomes.