Pre-stroke Functional Status in Patients Undergoing Mechanical Thrombectomy: How Relevant Are False Estimations in the Emergency Setting?

TL;DR

The study finds that emergency evaluations of pre-stroke function often overestimate patients' abilities, potentially affecting outcome predictions.

Contribution

This study quantifies the frequency and impact of false estimations of pre-stroke functional status in emergency settings.

Findings

Disagreement between emergency and post-admission pre-stroke mRS occurred in 18.6% of patients.

ED-overestimation was linked to older age, diabetes, prior stroke, and worse 3-month outcomes.

Overestimation may falsely reduce expected favorable outcomes in stroke patients.

Abstract

The modified Rankin scale (mRS) is frequently used in the emergency setting to estimate pre-stroke functional status in stroke patients who are candidates to acute revascularization therapies (ps-mRS). We aimed to describe the agreement between pre-stroke mRS evaluated in the emergency department (ED-ps-mRS) and pre-stroke mRS evaluated comprehensively post-admission (PA-ps-mRS).

Retrospective study of consecutive ischemic stroke patients undergoing mechanical thrombectomy, with available ED-ps-mRS and PA-ps-mRS. ED-ps-mRS was evaluated by the treating neurologist and documented in the emergency stroke treatment protocol. PA-ps-mRS was retrospectively evaluated with information registered in the clinical record. Collection of baseline characteristics and 3‑month outcomes. Patients with ED-overestimated pre-stroke functional status (ED ps-mRS ≤ 2 and PA-ps-mRS ≥ 3) were compared to correct low and high ED-ps-mRS groups.

We included 409 patients (median age 77 years, 50% female, median NIHSS 14). Concordance of dichotomized ED-ps-mRS and PA-ps-mRS (0–2 vs. 3–5) was found in 81.4% (Cohen’s kappa = 0.476, p < 0.001). ED-overestimated pre-stroke functional status was found in 69 patients (17%). Patients with ED-overestimated pre-stroke functional status were older (p < 0.001), more frequently presented diabetes (p < 0.001), previous stroke (p = 0.014) and less frequently presented 3‑month functional independence (p < 0.001) compared to patients with correct low ED-ps-mRS. No differences in pre-stroke baseline characteristics between overestimated and correct high ED-ps-mRS was found.

Disagreement between dichotomized ED-ps-mRS and PA-ps-mRS (0–2 vs. 3–5) occurred in 1/5 of patients. Overestimation of pre-stroke functional status may falsely reduce the expected proportion of patients achieving favourable 3‑month functional outcomes.

The online version of this article (10.1007/s00062-024-01449-5) contains supplementary material, which is available to authorized users.

Figures