Abstract TP220: Premorbid Anxiety and Depression as a Predictor of Length of Hospital Stay in Stroke Patients

Introduction: There is increasing interest in the impact of premorbid social factors on stroke outcomes. In this prospective observational survey of stroke survivors hospitalized at Stony Brook University Hospital, we aimed to evaluate the impact of premorbid anxiety and depression on length of hospital stay. Methods: At bedside, trainees recruited eligible hospitalized patients who experienced an Acute Ischemic Stroke (AIS, n=62) or Intracerebral Hemorrhage (ICH, n=12). Patients completed RedCap-based self-reported questionnaires, including premorbid Generalized Anxiety Disorder-7 (GAD-7) and the 2-item Patient Health Questionnaire-2 (PHQ-2) for depression. Additional variables were extracted from patient electronic health records (See Table 1). For the statistical analysis, predictor variables were dichotomized or trichotomized, and entered in a backward stepwise logistic regression to predict length of hospital stay (1= >7 days, 0= <7 days). Predictor variables included age, sex, premorbid Lawton-Brody Index, premorbid GAD-7, premorbid PHQ-2, NIH Stroke Scale (NIHSS) on admission, treatment received, and prior history of diabetes, hypertension, atrial fibrillation, and coronary artery disease. Logistic regressions were performed in SPSS v.29, with P=.05 required for entry and P=.10 for exit from the model. This study underwent IRB approval (5/2023). Results: 69 patients (median age=70 years [38-98]; 58% male; median NIHSS=3 [0-26]) were available for analysis (see Table 1). 5 patients were excluded for missing data. Higher pre-morbid anxiety significantly predicted a longer hospital stay (P=0.026, odds ratio [OR] 5.369), as did presence of one or more depressive symptoms (P=0.012, OR 13.196) (see Table 2). Discussion: In conclusion, our study highlights a significant impact of a patient’s psychological status on their length of hospital stay, a fact that may allow clinicians to better personalize patient discharge planning and incorporate early psychosocial interventions that could improve stroke recovery trajectories. Limitations include the recall bias for patients reflecting on their premorbid GAD-7 and PHQ-2, small number of patients describing one or more depressive symptoms on the PHQ-2, and the large representation of mild strokes. Further research can be conducted for a larger, more heterogeneous sample.