No abstract

weekend or "Monday blues." This work investigates factors affecting weekly patterns in stroke hospitalizations. Methods: All hospitalizations of adults with stroke as main diagnosis were included from German health claims between 01/ 01/2013 and 25/06/2018 (286 days for each day of the week). We tested for the Monday peak non-parametrically (chi-squared test) and parametrically adjusting for patient characteristics (multinomial logistic regression). If Monday peaks were driven by unhealthy weekend behaviors, we would expect the Monday peak also on public holidays. We therefore estimated a multinomial logit for 17 weeks where Monday was a public holiday and compared the probability of strokes over weekdays. We further stratified our analyses by gender, age, and type of stroke. Results: Of 65,657 stroke hospitalizations (57,129 cerebral infarctions, 7,069 intracerebral hemorrhage, 1,459 subarachnoid hemorrhage), most occurred on Monday (16.92%) and Tuesday (15.41%), while the least occurred on Saturday (11.65%) and Sunday (11.60%). A chi-squared test rejected the null of uniform distribution over the week (p,0.001). A multinomial logit showed that, after adjusting for over 20 patient characteristics, a stroke was 50% more likely to occur on Monday compared to Sunday (p,0.001). When restricting to weeks where Monday was public holiday, the probability of stroke was highest on Tuesday (18.0%), lowest on Saturday (11.4%), Sunday (10.9%) and Monday (12.8%), and on average 15.3% higher on Tuesday compared to other working days (p,0.001). Similar patterns were observed when stratifying by gender, age, and stroke type. Conclusions: Stress factors may play a major role in weekly stroke patterns, calling for stroke care readiness measures in hospitals after weekends and public holidays, and strategies for a gradual transition into the working week.