Background: Despite system-level focus on avoidance of potentially preventable hospitalizations, the decision to admit a patient with low medical acuity has been understudied. Our study sought to identify factors associated with admission versus discharge from the Emergency Department (ED) for patients considered unlikely to be medically appropriate for admission.
Methods: We conducted a retrospective cross-sectional study of admission calls to the acute care Medicine service of a public county teaching hospital between March 1, 2018 and March 1, 2019. A “triage hospitalist” logs an assessment of appropriateness for admission based on medical acuity (“Definitely”/”Probably”/Probably NOT”/”Definitely NOT) and contributing admission factors. Calls were included in this study if they came from the ED and if the hospitalist assessed appropriateness as any category other than “Definitely”. Data abstracted from the medical chart related to patient demographics, socioeconomic information, measures of illness, and system-level factors such as previous healthcare utilization and day/time of presentation. The outcome of interest was admission to the hospital versus discharge from the ED. Categorical variables were compared using a χ2 test and continuous variables were compared by Student’s t-test. Variables meeting significance in univariate analysis were fit to a binary logistic regression model. This study was approved by the UW IRB.
Results: From 3,875 logged Triage Database admission calls, 452 (11%) met inclusion criteria. Of these, 253 (56%) were assessed as “Probably” appropriate; 158 (34.9%) as “Probably NOT” appropriate; and 41 (9.1%) as “Definitely NOT” appropriate. There were no statistically significant differences in patient demographics, socioeconomic factors, or measures of illness between appropriateness categories. A total of 349 (77.2%) of 452 calls resulted in admission to the hospital and 103 (22.8%) resulted in discharge from the ED. Predictors of admission included age over 65 (OR 3.5 [95%CI 1.1-11.6], p = 0.039), homelessness (OR 3.3 [95% CI 1.7-6.4], p < 0.0001), night/weekend presentation (OR 2.0 [95%CI 1.1- 3.5], p = 0.020), and admission appropriateness category (“Probably NOT”: OR 0.02 [95%CI 0.01-0.06], p < 0.0001; “Definitely NOT”: OR 0.03 [95%CI 0.01-0.09], p < 0.0001). The most common contributing factors to the decision to admit reported by the triage hospitalist included: lack of outpatient social support (35.8% of admissions), homelessness (33.0% of admissions), substance abuse (23.5% of admissions), and lack of outpatient medical support (23.2% of admissions).
Conclusions: A large proportion of patients assessed as lacking definite medical appropriateness for admission were ultimately admitted to the acute care Medicine service at our institution. This appears to be related to physician medical decision-making that incorporates consideration of individual patient characteristics, social setting, and system-level barriers.