Extent of routine diagnostic cardiac work-up at certified German stroke units participating in the prospective MonDAFIS study

Abstract

Abstract Background About 25% of all ischaemic strokes are related to cardio-embolism, most often due to atrial fibrillation (AF). Little is known about the extent and standardization of routine cardiac diagnostic work-up at certified stroke-units in Germany. Methods The MonDAFIS study included non-AF patients with acute ischaemic stroke or transient ischaemic attack (TIA) at 38 certified stroke-units in Germany. Here, we analysed routine diagnostic work-up and disregarded study-related Holter-ECG monitoring. We compared duration of stroke-unit stay, number of 24-h Holter-ECGs, and echocardiography performed between university-based comprehensive stroke centres (UCSC, 12 hospitals, 1606 patients), non university-based comprehensive stroke centres (nUCSC, 14 hospitals, 892 patients), and primary stroke centres at non-university hospitals (PCS, 12 hospitals, 933 patients) using multivariable mixed regression analyses. Detection of a first AF episode in-hospital was also compared between hospitals of different stroke-unit levels. Results In 3431 study patients (mean age 66.2 years, 39.5% female, median NIHSS = 2 on admission), median duration of the stroke-unit stay was 72 h (IQR 42–86). Stroke-unit stay was longer (categorised ≤ 24/ > 24- ≤ 72/ > 72 h) for patients with severe stroke (NIHSS score ≥ 5/ < 5: OR = 1.6, 95%CI 1.3–2.0) and for patients with ischaemic stroke vs. TIA (OR = 1.7, 95%CI 1.4–2.1). Overall, 2149/3396 (63.3%) patients underwent at least one additional 24-h Holter-ECG (median 1 [IQR 0–1], range 0–7). Holter-ECG rate was 47% in UCSC, 71% in nUCSC, and 84% in PCS. Compared to PCS, AF was less often detected in-hospital in UCSC (OR = 0.65, 95%CI 0.45–0.93) and nUCSC (OR = 0.69, 95%CI 0.46–1.04). Transoesophageal echocardiography (TEE) only was performed in 513/3391 (15.1%) study patients, transthoracic echocardiography (TTE) only in 1228/3391 (36.2%), and TEE combined with TTE in 1020/3391 (30.1%) patients. Patients younger than 60 years (vs. ≥ 60 years) underwent TEE more often than those older than 60 years (OR = 3.44, 95%CI 2.67–4.42). TEE (IQR 34–65%) and TTE rate (IQR 40–85%) varied substantially among study centres. Echocardiography rate (TTE and/or TEE) was 74.0% in UCSC, 85.4% in nUCSC, and 90.3% in PSC, respectively. Conclusions In the MonDAFIS study, the routine use of echocardiography and Holter-ECG monitoring varied in participating stroke centres and at stroke-unit level, if grouped according to stroke-unit certification grade and hospitals´ university status. Trial registration Clinical Trials, NCT02204267. Registered 30 July 2014, https://clinicaltrials.gov/ct2/show/NCT02204267.