Factors Associated with Functional Recovery among Older Intensive Care Unit Survivors

Abstract

Rationale: Most of the 1.4 million older adults who survive the intensive care unit (ICU) annually in the United States face increased disability, but little is known about those who achieve functional recovery.

Objectives: Our objectives were twofold: to evaluate the incidence and time to recovery of premorbid function within 6 months of a critical illness and to identify independent predictors of functional recovery among older ICU survivors.

Methods: Potential participants included 754 persons aged 70 years or older who were evaluated monthly in 13 functional activities (1998-2012). The analytic sample included 218 ICU admissions from 186 ICU survivors. Functional recovery was defined as returning to a disability count less than or equal to the pre-ICU disability count within 6 months. Twenty-one potential predictors were evaluated for their associations with recovery.

Measurements and main results: Functional recovery was observed for 114 (52.3%) of the 218 admissions. In multivariable analysis, higher body mass index (hazard ratio [HR], 1.07; 95% confidence interval [CI], 1.03-1.12) and greater functional self-efficacy (HR, 1.05; 95% CI, 1.02-1.08), a measure of confidence in performing various activities, were associated with recovery, whereas pre-ICU impairment in hearing (HR, 0.38; 95% CI, 0.22-0.66) and vision (HR, 0.59; 95% CI, 0.37-0.95) were associated with a lack of recovery.

Conclusions: Among older adults who survived an ICU admission with increased disability, pre-ICU hearing and vision impairment were strongly associated with poor functional recovery within 6 months, whereas higher body mass index and functional self-efficacy were associated with recovery. Future research is needed to evaluate whether interventions targeting these factors improve functional outcomes among older ICU survivors.

Keywords: Medicare; functional status; intensive care; quality of life.

Figure 1.

Assembly and outcomes of the analytic sample from the parent cohort. The analytic sample included participants who had at least one intensive care unit (ICU) admission from their enrollment date through 2012. Eight participants were excluded owing to refusal to continue in the parent study (n = 6) or an ICU admission lasting less than 24 hours (n = 2). Participants could contribute more than one ICU admission over the 14 years of the study, but only the first ICU admission within an 18-month interval was eligible. To permit an evaluation of recovery, ICU admissions were included if the post-ICU disability count (from the first monthly interview after ICU admission) was greater than the pre-ICU disability count (from the monthly interview immediately before ICU admission). Over the more than 14 years of the study, 302 participants contributed 388 qualifying ICU admissions (i.e., first within an 18-month interval). Of these, 170 ICU admissions were not eligible for recovery because of death in the hospital (n = 72), death occurring postdischarge but before the first monthly interview (n = 28), or because the post-ICU disability counts (measured at a median of 14.5 d [interquartile range, 9–24 d] after hospital discharge) were equal to (n = 46) or less than (n = 24) the pre-ICU disability counts. The remaining 186 participants contributed a total of 218 ICU admissions. Within the 6-month follow-up period, 114 (52.3%) achieved functional recovery, 69 (31.7%) were living with increased disability, and 35 (16.1%) had died. The percentages sum to 100.1% because of rounding.