Post-stroke patients with moderate function have the greatest risk of falls: a National Cohort Study.

Published online

Journal Article

BACKGROUND: Stroke patients have increased risks of falls. We examined national registry data to evaluate the association between post-stroke functional level and the risk of low falls among post-stroke patients. METHODS: This retrospective cohort study analyzed data from national registries to examine the risk factors for post-stroke falls. Data for patients who suffered ischemic strokes and survived the index hospital admission was obtained from the Singapore National Stroke Registry and matched to the National Trauma Registry, from 2011 to 2015. The primary outcome measure was a low fall (fall height ≤ 0.5 m). Competing risk analysis was performed to examine the association between functional level (by modified Rankin score [mRS] at discharge) and the risk of subsequent low falls. RESULTS: In all, 2255 patients who suffered ischemic strokes had recorded mRS. The mean age was 66.6 years and 58.5% were men. By the end of 2015, 54 (2.39%) had a low fall while 93 (4.12%) died. After adjusting for potential confounders, mRS was associated with fall risk with an inverted U-shaped relationship. Compared to patients with a score of zero, the sub-distribution hazard ratio (SHR) increased to a maximum of 3.42 (95%CI:1.21-9.65, p = 0.020) for patients with a score of 2. The SHR then declined to 2.45 (95%CI:0.85-7.12, p = 0.098), 2.86 (95%CI:0.95-8.61, p = 0.062) and 1.93 (95%CI:0.44-8.52, p = 0.38) for patients with scores of 3, 4 and 5 respectively. CONCLUSIONS: An inverted U-shaped relationship between functional status and fall risk was observed. This is consistent with the complex interplay between decreasing mobility (hence decreased opportunity to fall) and increasing susceptibility to falls. Fall prevention intervention could be targeted accordingly. (263 words).

Full Text

Duke Authors

Cited Authors

  • Wei, WE; De Silva, DA; Chang, HM; Yao, J; Matchar, DB; Young, SHY; See, SJ; Lim, GH; Wong, TH; Venketasubramanian, N

Published Date

  • December 26, 2019

Published In

Volume / Issue

  • 19 / 1

Start / End Page

  • 373 -

PubMed ID

  • 31878876

Pubmed Central ID

  • 31878876

Electronic International Standard Serial Number (EISSN)

  • 1471-2318

Digital Object Identifier (DOI)

  • 10.1186/s12877-019-1377-7

Language

  • eng

Conference Location

  • England