Predictors of good functional outcome at discharge in stroke patients with depression in the Florida Puerto Rico Collaboration to Reduce


Background: Stroke is the fifth leading cause of death in the US and can result in major disability. It is important to identify factors that increase risk of poor outcome so that they may be target-ed for modification. Depression is related to worse post-stroke outcomes, but there is a dearth of re-search in the Hispanic population, who have a higher burden of stroke and depression than non-Hispanics. The goal was to determine how ethnicity and prior depression affect patient outcomes after stroke in a large, multicenter stroke registry in Florida and Puerto Rico (FL-PR).

Methods: We analyzed 80,687 cases from 84 institutions that participate in the FL-PR Stroke Registry (1/2014-12/2017). Previous medical history of depression and anti-depressant prescription prior to admission identified patients with depression. Univariate analyses of patient characteristics (e.g., age, race/ethnicity, insurance), previous medical history (PMH) [e.g., stroke or TIA, hyper-tension (HTN), diabetes mellitus (DM), peripheral vascular disease (PVD), heart failure (HF), and chronic renal insufficiency (CRI)], mode of arrival, initial NIHSS, and independent ambulation prior to stroke were completed. Multilevel multivariate logistic regressions using discharge modified Rankin score (mRS) and ambulatory status were created adjusting these factors.

Results:Out of 80,687 cases analyzed, 10,085 (12.5%) had depression. By race, 7,502/50,037 (15%) were white, 1,031/14,723 (7%) were black, 1,202/10,822 (11%) were Florida Hispanic, and 350/5105 (7%) were Puerto Rico Hispanic. In the multivariate models, depression decreased the likelihood of good clinical outcomes (mRS 0-2, OR: 0.85, 95% CI: 0.78 - 0.92; independent ambulation at discharge, OR: 0.86, 95% CI: 0.80 – 0.94). In addition, age ≥ 65, black, female, Medicare, no insurance or Medicaid, initial NIHSS ≥ 5, PMH of stroke or TIA, HTN, DM, atrial fibrillation, PVD, HF, CRI, and arrival via EMS were negatively associated with mRS 0-2; whereas PMH of dyslipidemia and the ability to ambulate independently prior to stroke showed positive association (p

Conclusions: Depression decreased the likelihood of a good functional outcome at discharge, based on mRS and independent ambulation. Many of the common stroke comorbidities also decreased likelihood of good functional outcome, whereas independent ambulation prior to stroke in-creased likelihood of good outcome. Surprisingly, dyslipidemia increased likelihood of good outcome in both multivariable models and smoking increased likelihood of good outcome in one model. It will be interesting and important to investigate how these traditionally negative comorbidities may con-tribute to better outcomes. For the other comorbidities, it remains essential for patients to be properly treated for these conditions.

Publication Date


Presented At:

14th Annual BHSF Research Conference

Content Type


This document is currently not available here.