Background: It has been recently shown that patients admitted at night and during the weekend have worse mortality outcomes. The impact of those variables on other important outcomes such as length of stay (LOS) and readmission is less clear. We sought to analyze this relationship in a large cohort of inpatients.
Methods: We performed a retrospective analysis of records from patient admitted to our institution between 1/2016 and 12/2018. We selected only inpatient status and excluded observation status. We recorded demographic variables and the outcomes LOS and 30 day readmission. We performed both univariate and multivariate analyses.
Results: 78,826 patients were initially selected, of which 12,694 did not have time of admission appropriately recorded and were therefore excluded resulting in 66,132 patients included in the final analysis. 55% of patients were female. 24.6 % were African American, 15.3% Hispanics and 54.3% Whites.The average age was 55.4 (SD 18.8). The average LOS was 5.5 (SD 7.5) days. 1, 443 patients expired in the hospital (2.2%) and 8,622 (11%) were readmitted within 30 days. 12,692 patients were admitted during the week-end (19.2%). The median admission time was 14.8 (2: 48 pm). We divided the admission time into equal quartiles based on frequency (25% in each quartile): Quartile 1 (Q1): 00:00-8:42, Q2: 8.43-14:48, Q3: 14:49-19:12, and Q4 >19:13. (SD 6.2 hours). The proportion of patients readmitted varied based on their admission time (Q1 to Q4): 11.2 %, 14.8%, 14.6% and 14.2%. (P<0.001). LOS increased from Q1 to Q4: 4.6, 5, 6, and 6.2 days (P<0.001). Patients admitted on the weekend had a higher LOS (5.4 vs 5.9 days, P<0.001) but slightly lower readmission risk (13% vs 13.9%, P=0.017). After performing a multivariate regression analysis, we found that later admission times were significantly associated with longer LOS (Q1 comparator): Q2 Odds ratio 1.4 [1.3-1.5], Q3 1.9 [1.8-2] and Q4 2.3 [2.1-4], all P <0.001. We also found a higher readmission risk (Q1 comparator): Q2 1.3[1.2-1.4] P<0.001, Q3 1.1[1.01-1.2] P=0.018, but not Q4 1 [0.9-1.1]. Week-end admissions had a significantly higher LOS (1.2[1.1-1.3], P<0.001) but lower readmission risk 0.8[0.8-0.9], P<0.001
Conclusions: Admission that occur later during the day and on weekends were found to be associated with worse LOS. Patients admitted during mid-day had a higher readmission risk but not later and week-end admission had a lower readmission risk. It’s unclear whether hospital staffing and resources or factors specific to the admitted patients would explain these results. These findings suggest that health systems may need to address operations and deployment of resources to later in the day or on week-ends to match the demand of admitted patients.