Factors Influencing Emergency Department Length of Stay in College-Aged Patients

Varsha Radhakrishnan, MD

Beth Israel Deaconess Medical Center
Factors Influencing Emergency Department Length of Stay in College-Aged Patients

Scientific Abstract

Background: In our recent study, “Impact of Collateral on Emergency Department Length of Stay in College- Aged Patients”, results suggested a statistically significant difference in ED length of stay (LOS) for hospitalized patients enrolled vs. unenrolled in college. Unexpectedly, the study revealed that college students had decreased length of stay in the ED regardless of collateral when compared to college unenrolled peers, suggesting a possible underlying difference between the two populations. This objective of this study is to evaluate the influence of collateral on ED length of stay for patients within the last 5 years, while also exploring additional factors that may influence length of stay including college enrolment status, final disposition, psychiatric diagnosis, history of substance abuse, psychiatric comorbidity, history of prior psychiatric hospitalizations, frequency of emergency department visits, geographic location of college, status as international student, and support structure available to the primary clinician during the time at which consults were performed.

Methods: This IRB approved retrospective chart review of the Electronic Medical Record will be performed comprising of 3165 ED consults for all college-aged patients 18-25 years old over the last 5 years. Data will be analysed using Excel and Python.

Results: Results are to be determined. There were multiple limiting factors in our previous study including small sample size (834 consults) and multiple additional potentially confounders. Our current hypothesis is that history of multiple prior hospitalizations and ED visits, active enrolment in college, psychiatric comorbidity, substance abuse, geographical location, and clinician support structure at time of consult could all independently increase ED LOS. Correlations observed among these different factors could help clarify some of the underlying differences between college enrolled vs. unenrolled patients noted during our previous study.

Conclusions: Conclusions are to be determined. We hypothesize that collateral is associated with decreased ED LOS regardless of the absence vs. presence of the aforementioned factors.

SoundCloud Transcript

Dr. Radhakrishnan:

Hi everyone- my name is Dr. Varsha Radhakrishnan, I am a 3rd year psychiatry resident at the Harvard BIDMC psychiatry residency training program and I am accompanied by my two colleagues Dr. Sumita Sharma, a fellow 3rd year psychiatry residency at BIDMC and Dr. Kevin M. Hill, a second year psychiatry residency at BIDMC. We are here to talk to you about our study- Factors Influencing Emergency Department Length of Stay in College Aged Patients. We have no conflicts of interest or financial disclosures.

Dr. Sharma:

 College students comprise an at-risk population for a variety of mental health disorders [1-4]. While there have been many studies reflecting the prevalence of mental health disorders, there have been limited studies focusing on streamlining care while in the Emergency Department during an acute mental health crisis. Length of stay is one metric used to evaluate efficiency given association of length of stay with higher hospital costs, patient frustration with prolonged boarding times, and systemic inefficiencies in delivering patient care [5-7]. In our recent study, “Impact of Collateral on Emergency Department Length of Stay in College-Aged Patients”, results suggested a statistically significant difference in Emergency Department length of stay for hospitalized patients enrolled vs. unenrolled in college. Unexpectedly, the study revealed that college students had decreased length of stay in the Emergency Department regardless of collateral when compared to college unenrolled peers, suggesting a possible underlying difference between the two populations.  This objective of this study is to evaluate the influence of collateral on Emergency Department length of stay for patients within the last 5 years, while also exploring additional factors that may influence length of stay including college enrolment status, final disposition, psychiatric diagnosis, history of substance abuse, psychiatric comorbidity, history of prior psychiatric hospitalizations, frequency of emergency department visits,  geographic location of college, status as international student, and support structure available to the primary clinician during the time at which consults were performed.

Dr. Hill:

Per our previous study “Impact of Collateral on Emergency Department Length of Stay in College-Aged Patients”, psychiatric consultations for college-aged patients comprised 19% of all Emergency Department consultations. This retrospective chart review of the Electronic Medical Record will be performed on 3165 ED consults for all college-aged patients (ages 18-25 years old) over the last 5 years from October 2015 to April 2020. A chart review was performed to obtain the following information: time that psychiatry was consulted, time of discharge, final disposition, psychiatric diagnosis, history of substance abuse, psychiatric comorbidity, frequency of ED visits, geographic location of college, status as international student, and support structure available to the clinician at the time at which the consultation was performed. This study was IRB approved. We are currently on the data collection stage of our study.

Figure 1 here depicts a process map of emergency psychiatric evaluations which can help better understand the factors that influence length of stay.   This figure delineates the process involved in Emergency Department (ED) psychiatric evaluations which begins for all patients in the ED waiting room and ends in an appropriate and safe disposition. Each step in this process contributes to the patient’s length of stay which starts in the ED waiting room- leads to evaluation first by the primary ED physician, subsequent consultation of psychiatry after triage, followed by obtaining imaging/labs, OMR records, collateral, and results in disposition planning/final recommendations followed by discharge home vs. admission to a higher level of care.

Dr. Radhakrishnan: 

As depicted in table 1- Average length of stay results- Our previous study consisted of a total of 834 ED consults with 414  consults reviewed following a collaborative care intervention to improve dissemination of collateral from local universities. Results from our recent study  for n=414 suggested that average LOS in the ED was 15.7 hours for students with collateral, 14.6 hours for students without collateral, 15.1 hours for college enrolled hospitalized patients (in total, regardless of absence vs. presence of collateral), and 19.5 hours for college unenrolled peers. Of those discharged home, students with collateral, students without collateral, and unenrolled peers spent 9.7, 11.6, and 13.6 hours in the ED respectively. There was a statistically significant difference in LOS for patients hospitalized when comparing college enrolled versus unenrolled patients (t= 2.1690, p= 0.0314).  LOS of enrolled versus unenrolled patients discharged home trended towards significance (t=1.8002, p= 0.0731), but no statistical significance was found in relation to collateral (t=1.211, p=0.2278).

Results from our current study are to be determined. There were multiple limiting factors in our previous study including small sample size (a total of 834 consults with n=414 consults post collaborative care intervention) and multiple additional potential confounders. Our current hypothesis is that history of multiple prior hospitalizations and ED visits, active enrolment in college, psychiatric comorbidity, substance abuse, geographical location, and clinician support structure at time of consult could all independently increase ED LOS. Correlations observed among these different factors could help clarify some of the underlying differences between college enrolled vs. unenrolled patients noted during our previous study.

Dr. Sharma :

Conclusions are to be determined. We hypothesize that collateral is associated with decreased ED LOS regardless of the absence vs. presence of the aforementioned factors. We moreover anticipate that there are fundamental differences between the two populations- college enrolled vs. unenrolled as was suggested in our prior study- and that this study may help us better elucidate upon the differences between these two populations and ultimately help bring to awareness potential discrepancies or shortcomings in the efficiency and delivery of emergency psychiatric care. We hypothesize that higher degree of psychiatric comorbidity, comorbid substance abuse, frequency of emergency psychiatric visits, status as an international student, and the nature of the existing support structure for clinicians will all influence ED length of stay.

Dr. Hill:

 Thank you for your interest – please feel free to contact us with any questions or  comments at vradhakr@bidmc.harvard.edu.

research Areas

Authors

Varsha Radhakrishnan, MD, Sumita Sharma, MD, Kevin M. Hill, MD