Stroke Referrals from Nursing Homes: High Rate of Mimics and Late Presentation

Abstract

Background: Stroke has become a treatable condition with increasing evidence of treatment benefits in older people. However, stroke mimics in geriatric patients are especially prevalent, causing incorrect suspicion and consecutive burden to patients and emergency room resources. We therefore examined the dimension of this problem by investigating emergency room admissions from nursing homes for suspected stroke. Methods: We performed a retrospective cohort study of all nursing home residents who were admitted to the neurological emergency room of our primary and tertiary care university hospital between 2013 and 2015. Patients were further divided into those with confirmed stroke and stroke mimics after diagnostic stroke work-up. Results: Of 419 nursing home patients referred to the emergency room, nearly one third had suspected stroke (n = 126; mean age: 78 ± 14 years, polypharmacy rate: 77%). Of those, 43 (34%) had a confirmed stroke (ischaemic: n = 34; haemorrhagic: n = 9) and 83 (66%) had stroke mimics after diagnostic work-up. Only one patient underwent intravenous thrombolysis, followed by mechanical thrombectomy for middle cerebral artery occlusion. Prehospital delay (47%) and multimorbidity-associated contraindications (27%) were the main reasons for withholding recanalization therapy. Among the stroke-mimicking conditions, infectious diseases (24%) and epileptic seizures (20%) were the most frequent. Multivariate analysis identified focal deficits (OR 16.6, 95% CI 4.3–64.0), atrial fibrillation (OR 3.9, 95% CI 1.5–10.5) and previous stroke (OR 3.2, 95% CI 1.2–8.9) as indicators that were associated with stroke. Conclusions: In our region, nursing home referrals for suspected stroke have a high false positive rate and occur delayed, which most often precludes specific stroke treatment in addition to multimorbidity. Such problems may also exist in other centres and highlight the need for targeted educational and organizational efforts. Simple indicators as identified in this study may help to sort out patients with true stroke more efficiently.