Variation in Stroke Care Quality by Day and Time of Admission

Study Questions:

What is the relationship between stroke quality measures and time and day of the week of hospital admission?


The authors used data from the Sentinel Stroke National Audit Program (SSNAP), the national stroke registry for England and Wales. The SSNAP collects demographic, clinical, and quality data for all acute-care stroke hospitalizations in England and Wales. This study included all acute ischemic stroke and intracerebral hemorrhage patients between April 2013 and March 2014. Quality measures were aligned with UK guidelines and included: prompt brain scan, prompt admission to a stroke unit or intensive care unit, administration of tissue plasminogen activator (tPA), tPA administration within 60 minutes of arrival (for patients treated with tPA), prompt dysphagia screening, prompt review by a stroke specialist physician and nurse, and prompt therapy assessments (physiotherapy, occupational therapy, and speech language pathology). Patients with contraindications were excluded from the measures.

The primary outcome was 30-day post-admission survival. Time-stratified analyses were done using two methods: 4-hour time blocks in each day of the week and comparing weekday daytime (8 a.m.–7:59 p.m.), weekend daytime (8 a.m.–7:59 p.m.), weekday overnight (8 p.m.–7:59 a.m.), and weekend overnight (8 p.m.–7:59 a.m.). Variation in quality measure adherence was quantified by the coefficient of variation (with lower values indicating less variation). Multilevel logistic regression was used to adjust for clinical and demographic factors.


The study included 74,307 patients admitted to 199 hospitals. The median age was 77 years, and 88% had an ischemic stroke. Monday was the most common day of admission (16% of patients) and there were fewer admissions on Saturday or Sunday (13% each day). Discharge rarely occurred on the weekend (Sunday 3%, Saturday 6%). There was tremendous variation in the adherence to quality measures with coefficients of variation ranging from 3.5 (prompt evaluation by a stroke nurse) to 18.2 (tPA within 60 minutes of arrival).

There were four patterns of temporal variation in the metrics. Overall, quality of care seemed to improve as the week went on and then decreased on the weekends. Prompt brain scanning, tPA, and dysphagia screening showed diurnal variation, with patients arriving in the morning more likely to receive these measures than patients arriving in the afternoon. There was lower quality of care on the weekends for prompt stroke physician or nurse assessment and therapy assessments. Administration of tPA <60 minutes after arrival was less common overnight and on the weekends.

There was no difference in 30-day survival between weekend and weekday admissions, though overnight admissions were associated with decreased odds of survival (adjusted odds ratio, 0.09; 95% confidence interval, 0.82-0.99).


There are variations in the quality of acute stroke care that occur not only between weekdays and weekends, but across the week.


Many studies of various disease processes have shown variations in care during off hours. By using national registry data, this work provides very granular data about how stroke care varies not just between days and nights, weekdays and weekends, but even across the week. This is valuable information because interventions aimed at improving stroke care during off hours will not impact these more complex trends. The reasons for the variation in care are not clear. Staffing models or bed capacity issues may underlie some of the findings, but given the different temporal variations, they are unlikely to fully explain the variability.

< Back to Listings