Acute stroke is a medical and subsequent surgical emergency, which needs time-dependent evaluation and therapy by highly trained, skilled neurologists or neurosurgeons, nurses and other paramedics. Off-hour time may first lead to delay of hospital presentation after acute stroke. Off-hours may also increase time interval from computed tomography (CT) scanning to interventional procedure [17]. Because specialists are usually at home during off hours, fast decision-making and subsequent intervention or operation may be delayed [25]. Off hours may decrease performance skills of interventionalist or neurosurgeon owing to change of normal biological rhythm and physical or mental fatigue [11]. As a result, off-hour hospital presentation may increase the probability of unexpected complications [1] or frequent medical errors [3]. This phenomenon is well known as “off-hour effect” which been believed to result in worse outcome in time-dependent critical conditions such as acute stroke [4], acute myocardial infarction [16].

There have been incoherent results among countries on impact of off-hour admission on clinical outcome after acute str oke [4,5,10,13,15,18,19,22,24]. Its relation has not yet been thoroughly studied in Korea. We analyzed nationwide administrative data to verify off-hour effect in different subtypes of acute stroke in Korea.

Of the 11347 stroke subjects included, off-hour hospital presentation was more frequent than working hours (59.5% vs. 40.5%); 4319 (58.0%) in IS, 1504 (62.0%) in ICH, and 933 (63.0%) in SAH, respectively. Table 1 shows baseline data of 11347 stroke subjects according to subtypes from 26 hospitals of wide regional or regional emergency medical centers in Korea.

In patients with IS, off-hour hospital presentation had significantly more elderly patients ≥65 years (35.4% off hours vs. 32.1% working hours, p=0.029) and significantly more frequent intensive care unit (ICU) admission (32.5% off hours vs. 29.9% working hours, p=0.017) than in working hours. However, we could not observe delay by time to emergency room (ER), inter-hospital transfer and ER stay in patients with off-hour admission. There were also no significant differences between off hours and working hours in other medications or treatments including anti-platelets (74.0% off hours vs. 73.2% working hours, p=0.443), anti-coagulants (28.7% off hours vs. 27.1% working hours, p=0.132), vessel recanalization (14.5% off hours vs. 14.1% working hours, p=0.675) and decompressive surgery. In patients with off-hour admission, we did not detect meaningful abandonment of treatment (2.5% off hours vs. 2.4% working hours, p=0.807) or significant contraindication of vessel recanalization including risk of hemorrhage (3.4% off hours vs. 3.3% working hours, p=0.822), bleeding tendency (2.2% off hours vs. 1.8% working hours, p=0.261) and uncontrolled hypertension (0.7% off hours vs. 0.5% working hours, p=0.240). In patients admitted with ICH, offhour hospital presentation had significantly lower GCS scores (13 off hours vs. 14 working hours, p=0.024) than those in working hours. However, we did not discover significant differences between off hours and working hours in surgical procedures including hematoma aspiration (9.3% off hours vs. 8.9% working-hours, p=0.743), extra-ventricular drainage (17.5% off hours vs. 16.1% working hours, p=0.373) and decompressive surgery. In subjects with SAH, off-hour admission had significantly longer ER stay (5.0 hours off hours vs. 4.4 hours working hours, p=0.038) than that in working hours. However, there were no significant differences between off hours and working hours in endovascular coiling (42.8% off hours vs. 46.1% working hours, p=0.214), surgical clipping (34.3% off hours vs. 33.2% working hours, p=0.652), extraventricular drainage (18.8% off hours vs. 15.5% working hours, p=0.156) and decompressive surgery. Table 2 shows clinical characteristics of off-hour hospital presentation in different subtypes of acute stroke.

Overall in-hospital mortality of acute stroke was 1074 (9.5%). Those of subtypes of stroke were 351 (4.7%) for IS, 448 (17.6%) for ICH, and 370 (23.3%) for SAH. Overall in-hospital mortality of acute stroke in patients with off-hour admission was significantly higher than that in working hours (10.3% off hours vs. 8.2% working hours, p<0.001). In patients admitted with IS, off-hour hospital presentation resulted in significantly more frequent unfavorable outcome, not discharge to home, 30-day mortality and in-hospital mortality than those in working hours, but not in subjects with ICH and SAH. Table 3 shows clinical outcomes of off-hour hospital presentation in different subtypes of acute stroke.

In patients with IS, off-hour admission was associated with unfavorable outcome (OR, 1.244; 95% CI, 1.106-1.400), 30- day mortality (HR, 1.359; 95% CI, 1.079-1.712) and in-hospital mortality (HR, 1.402; 95% CI, 1.124-1.747) even after adjustment for compounding risk factors. Off-hour hospital presentation increased death at discharge by about 40% in patients with IS. However, off-hour admission was not related to worse outcomes in subjects with ICH and SAH. Table 4 shows unadjusted and adjusted OR or HR of off-hour hospital presentation in different subtypes of acute stroke.

This study demonstrated that off-hour hospital presentation was associated with increased short-term morbidity and mortality in patients with IS in Korea and death at discharge was increased by about 40% in off hours than in working hours, but that was not related to poor outcome in subjects with ICH and SAH. Also, off-hour admission had more elderly and more frequent ICU admission than working hours.

Several observational studies have assessed the ‘off-hour effect’ in patients with IS. Prior studies have reported incoherent results on whether off-hour admission is associated with worse outcome after IS [5,22,24,28] or not [4,10,13,15,19,27,29]. In a retrospective study of 26676 IS subjects from the Hospital Morbidity Database of 606 hospitals in Canada, Saposnik et al. [25] reported that weekend admission was associated with early mortality, which was similarly observed among subjects presented to urban versus rural medical centers or specialist versus a general practitioner. Similarly, in the analysis of the Get With The Guidelines (GWTG)-Stroke program data of 187669 IS patients from 857 hospitals in America, Reeves et al. [24] showed that off-hour hospital presentation was associated with increased mortality, but the effect was very small. Since then, a retrospective cohort study of 82219 IS patients from 115 Dutch hospitals reported that weekend admission was related to an increase of 7-day mortality for IS patients, which is due to an underlying off-hour risk present during the week days [22]. A prospective database of Stroke Improvement National Audit Programme of 45726 stroke subjects of 130 hospitals in England, Campbell et al. [4] reported that weekend hospital presentation was associated with higher 30-day mortality. They suggested that unevenness in the use of stroke management during working hours as well as during off hours was common in contemporary England, but the impact on mortality is small and largely due to higher disease severity in patients presented during off hours. Recently, Korean nationwide claims data of 8957 IS subjects revealed that weekend admission was significantly associated with higher 30-day mortality [5].

On contrary, a retrospective cohort study of 51643 prospective data from 23 ICUs in the Paris metropolitan area disclosed that off-hour presentation was not related to higher death rate [18]. A study of the Nationwide Inpatient Sample (NIS) Database of 599087 IS patients showed that there is a slight off-hour effect on thrombolysis and length of stay, but no association with higher mortality [10]. A Korean study of 7075 IS patients of tertiary hospitals from the Clinical Research Center for Stroke registry-5 reported that off-hour admission was not associated with an unfavorable 3-month functional outcome and off-hour effect might be controlled with well-organized stroke management strategies [15].

These discrepancies across previous studies might be resulted from differences in the health care system including the availability of specialized stroke units [4,25], and the number of stroke physician [28] during off-hour time. On the other hand, Nationwide Survey of Acute Stroke Care Capacity for Proper Designation of Comprehensive Stroke Center in Japan study of 35685 subjects from 262 hospitals for IS, ICH, or SAH reported that the effects of off-hour presentation were disappeared after adjustment for consciousness level on admission [13]. They suggested that well-known off-hour effect might be attributed to the severely ill patient population during off hours. However, in this study, off-hour hospital presentation had significant impact on short-term mortality in patients with IS even after adjusting for outcome co-variates including consciousness level. Like most previous studies [2,9,14,31], we revealed no significant differences on clinical outcomes of IS going through delay of hospital presentation and treatment delay for intravenous or intra-arterial thrombolysis or surgery during ER stay between off hours and working hours. However, in this study, age ≥65 years and ICU admission after IS were even higher during off hours. This data implies that worse outcome might be attributed to old age or the severity of other medical illness apart from the severity of stroke during off hours in patients presented with IS.

Among the previous studies of ICH, early two studies of a retrospective cohort study of 13821 subjects of NIS [7] and analysis of the GWTG-Stroke program data of 34845 acute hemorrhagic stroke [24] showed the association between off-hour hospital presentation and higher short-term mortality, but most of studies have ever reported no off-hour effect in patients with ICH [6,20,21,26,32]. The Intensive Blood Pressure Reduction in Acute Cerebral Hemorrhage Trial of 2794 subjects disclosed off-hour hospital presentation was not related to higher mortality or morbidity in subjects with acute ICH [26]. Strict blood pressure control can give rise to similar therapeutic effect irrespective of admission hours [26,30]. Furthermore, a recent study from the Universitätsklinikum Erlangen Cohort of Patients With Spontaneous Intracerebral Hemorrhage registry of 1269 subjects suggested an off-hour effect in ICH is associated with functional outcome rather than death rate [21]. In this study, 7-day mortality was significantly higher in off hours than in working hours, but it disappeared after adjustment of compounding variables. Like most of prior studies, we could not detect an association between off-hour hospital presentation and clinical outcomes although the subjects admitted with ICH had even higher severe neurological deficit and more ICU admission during off hours.

The previous studies of SAH, the other constituents of hemorrhagic stroke, have yielded inconsistent findings on the impact of off-hour hospital presentation on clinical outcomes. Most of early previous studies have reported no association between off-hour hospital presentation and worse outcome in patients with SAH [6,23,32]. However, recent two studies of SAH suggested a relationship between off-hour hospital presentation and worse outcome in patients with SAH. A prospective study of 285 aneurysmal SAH from 12 large regional hospitals of the northwest of England reported higher short-term mortality of subjects with SAH, but there were no off-hour effects on long-term mortality [8]. Another retrospective cohort data of 54703 SAH from NIS [12] showed higher in-hospital mortality in SAH patients with weekend hospital presentation. In this study, there were no significant differences in clinical characteristics, severity of disease and treatments of SAH between off hours and working hours. We did not find an association between off-hour hospital presentation and clinical outcomes in subjects with SAH.

This study has several limitations. First, NEDIS data covers only non-capital area of Korea excluding a metropolitan area of Seoul-Incheon-Gyeonggi districts. Hence, the mortality of acute stroke could be overestimated due to greater quality level and accessibility of emergency medical centers of SeoulIncheon-Gyeonggi districts. Second, this NEDIS data is somewhat out of date and may be different from current medical conditions. Third, NEDIS database has some limited data on different diseases of acute stroke. It has no contents of National Institutes of Health Stroke Scale, individual subjects’ treatment, complications and causes of death. This limited information made it difficult for us to find an accurate cause about why this results occur.

This study suggests that off-hour hospital presentation may be associated with poor short-term morbidity and mortality in patients with IS, but that is not likely to be related to poor outcome in subjects with ICH and SAH. Excessive death seems to be attributed to old age or the higher severity of medical illnesses apart from that of stroke of patients admitted during off hours. Multidisciplinary management for elderly or frequent medical illnesses during off hours is important to improve clinical outcome of acute stroke. Further multicenter prospective study is required to investigate a relationship between off-hour hospital presentation and longterm outcome in patients with acute stroke.