Association of Antidepressant Medication Therapy With Inpatient Rehabilitation Outcomes for Stroke, Traumatic Brain Injury, or Traumatic Spinal Cord Injury

Objective: To study whether outcomes in patients who have undergone inpatient rehabilitation for stroke, traumatic brain injury (TB!), or traumatic spinal cord injury (TSCI) differ based on antidepressant medication (ADM) use. Design: Retrospective cohort study of 867 electronic medical records of patients receiving inpatient rehabilitation for stroke, TBI, or TSCI. Four cohorts were formed within each rehabilitation condition: patients with no history of ADM use and no indication of history of depression; patients with no history of ADM use but with a secondary diagnostic code for a depressive illness; patients with a history of ADM use prior to and during inpatient rehabilitation; and patients who began ADM therapy in inpatient rehabilitation. Setting: Freestanding inpatient rehabilitation facility (IRF). Participants: Patients diagnosed with stroke (n=625), TBI (n=175), and TSCI (n=67). Interventions: Not applicable. Main Outcome Measures: FIM, rehabilitation length of stay (LOS), deviation between actual LOS and expected LOS, and functional gain per day. Results: In each impairment condition, patients initiating ADM therapy in inpatient rehabilitation had longer LOS than patients in the same impairment condition on ADM at IRF admission, and had significantly longer LOS than patients with no history of ADM use and no diagnosis of depression (P<.05). LOS for patients initiating ADM therapy as inpatients even exceeded LOS for patients without ADM history, but who had a diagnosis for a depressive disorder. Deviation in LOS was significantly larger in the stroke and TBI groups initiating ADM in IRF than their counterparts with no history of ADM use, illustrating that the group initiating ADM therapy in rehabilitation significantly exceeded expected LOS. Increased LOS did not translate into functional gains, and in fact, functional gain per day was lower in the group initiating ADM therapy in IRF. Conclusions: Explanations for unexpectedly long LOS in patients initiating ADM in inpatient rehabilitation focus on the potential for ADM to inhibit therapy-driven remodeling of the nervous system when initiated close in time to nervous system injury, or the possibility that untreated sequelae (eg, depressive symptoms or fatigue) were limiting progress in therapy, which triggered ADM treatment.