ABSTRACT NUMBER: ESOC2026LB69 DAYLIGHT EXPOSURE FROM QUASI-RANDOMIZED BED ASSIGNMENT IN A STROKE UNIT IS ASSOCIATED WITH LENGTH OF HOSPITAL STAY AFTER STROKE

Abstract Background and aims Timed daylight exposure, the primary circadian zeitgeber, reduces delirium incidence and influences outcomes in critically ill, yet its role in stroke patients remains unexplored. We aimed to determine differences in daylight exposure between window- and door-side beds arising from quasi-randomized bed assignment and their association with hospital length of stay (LOS). Methods Light exposure was measured as melanopic equivalent daylight illuminance (mEDI), the main physiologically relevant metric for circadian-effective light, using a spectroradiometer under two weather scenarios (sunny and cloudy). Light intensity (photopic illuminance) was monitored over two years with luxmeters. Associations between bed assignment and LOS were assessed through log-linear regression in 550 large-vessel occlusion patients admitted to the stroke unit after endovascular treatment (Table 1). Results Door-side beds had significantly lower daylight levels than window-side beds under both weather conditions (p<0.001) (Fig. 1). Over two years, door-side beds remained under the recommended threshold of 250 lux on 89% of days [interquartile range, 77–94] versus 29% [25–33] for window-side beds (p<0.001). Median LOS was 1 day longer for door-side versus window-side patients (8 [6–11] vs 7 [5–10] days; p=0.035). In analyses adjusted for age, sex, and baseline NIHSS, window-side bed assignment was associated with a 10% shorter LOS (adjusted ratio 0.90, 95% CI 0.81–0.99; p=0.036) (Fig. 2). Conclusions This study provides the first characterization of circadian-effective light in a stroke environment and suggests differences in clinical outcomes driven by bed position. Optimizing (day)-light exposure may support recovery and serve as a modifiable factor in stroke care. Conflict of interest Table 1 - belongs to Conclusions Figure 1 - belongs to Conclusions Figure 2 - belongs to Conclusions