Changes in emergency department utilization in vulnerable populations after COVID-19 shelter in place orders
1Cleveland Clinic Lerner College of Medicine at Case Western Reserve University, EC-10 Cleveland Clinic, Cleveland, OH, USA
2Department of Psychiatry and Psychology, Center for Behavioral Health, Neurological Institute, Cleveland Clinic, Cleveland, OH, USA
3Department of Quantitative Health Sciences, Cleveland Clinic, Cleveland, Ohio
4Department of Psychiatry, Yale University School of Medicine, New Haven, Connecticut, USA
*Corresponding author: Jeremy Weleff, DO, Yale University School of Medicine, Department of Psychiatry 300 George St, New Haven, CT, 06511, Phone/fax: 517-803-5675 jeremy.weleff@yale.eduAbstract
Purpose
To compare emergency department (ED) utilization and admission rates for patients with a history of mental health (MH), substance use disorder (SUD) and social determinants of health (SDOH) before and after implementing COVID-19’s shelter-in-place (SIP) orders.
Methods
This was a retrospective, multicenter study leveraging electronic medical record data from 20 EDs across a large Midwest integrated healthcare system from 5/2/2019 to 12/31/2019 (pre-SIP) and from 5/2/2020 to 12/31/2020 (post-SIP). Diagnoses were documented in the patient’s medical records. Poisson and logistic regression models were used to evaluate ED utilization and admission rate changes.
Results
871,020 total ED encounters from 487,028 unique patients were captured. 2,572 (0.53%) patients had a documented Z code for SDOH. Patients with previously diagnosed MH or SUDs were more likely to seek ED care after the SIP orders were implemented (RR: 1.20, 95% CI: 1.18 – 1.22, p<0.001), as were patients with SDOH (RR: 2.37, 95% CI: 2.19 – 2.55, p<0.001). Patients with both previously diagnosed MH or SUD and a documented SDOH had even higher ED utilization (RR: 3.31, 95% CI: 2.83 – 3.88, p<0.001) than those with either condition alone. Patients with MH and SUDs (OR: 0.89, 95% CI: 0.86 – 0.92, p<0.001) or SDOH (OR: 0.67, 95% CI: 0.54, 0.83, p<0.001) were less likely to be admitted post-SIP orders while patients with a history of diseases of physiologic systems were more likely to be admitted.
Conclusions
Vulnerable populations with a history of MH, SUD, and SDOH experienced increased ED utilization but a lower rate of hospital admissions after the implementation of SIP orders. The findings highlight the importance of addressing these needs to mitigate the impact of public health crises on these populations.
Article notes
Competing Interest Statement
The authors have declared no competing interest.
Funding Statement
This study did not receive any funding.
Background
Patients with mental health (MH) and substance use disorders (SUD) face significant challenges in accessing healthcare equitably.1 In addition, social determinants of health (SDOH) profoundly impact healthcare utilization, delivery, quality of life, and mortality. Identifying these patients for targeted interventions is crucial2; however, inconsistent and non-standardized documentation of SDOH across healthcare systems hinders public health surveillance efforts of this population. To address this issue, the Centers for Medicare and Medicaid Services and other groups have proposed using Z-Codes to standardize the documentation of social needs, but the utilization of these codes remains limited.3,4
The COVID-19 pandemic has exacerbated health disparities and disproportionately affected individuals with MH, SUD, and SDOH.5 While research has shown that shelter-in-place (SIP) orders have led to increased ED visits for mental health and substance use reasons6–8, few studies have examined the impact on ED utilization and admissions in individuals with a prior history of MH or SUD. Furthermore, the effects of the COVID-19 pandemic on ED utilization and admissions by patients with SDOH have had been poorly studied thus far. The primary aim of this study is to investigate the changes in ED utilization and hospital admission rates among vulnerable populations with a history of MH, SUD, and SDOH before and after SIP orders across a large integrated healthcare system. A secondary aim was to characterize the frequency of utilization of Z-codes to identify patients with SDOH in the ED.
Methods
Patient population
Patients who visited any of 20 EDs ranging from large hospital-associated EDs to freestanding EDs across a large Midwest integrated healthcare system Ohio from 5/2/2019-12/31/2019 (designated as the “pre-SIP order” time period) and 5/2/2020-12/31/2020 (the “post-SIP order” time period) were identified. Demographics and diagnoses were extracted from the electronic medical record (EMR). MH and SUDs were identified using the International Classification of Diseases 10th Revision (ICD-10) codes; similarly, SDOH were identified in patients’ medical histories using respective Z-Codes in the ICD-10. Historical use of theses codes means that these diagnoses and SDOH were documented in the patient’s chart at any time, and not solely reliant on ED encounter documentation. This study was approved by Cleveland Clinic IRB.
Statistical analysis
Poisson regression models were used to calculate absolute risk ratios (RR) to compare the total number of cases between the pre-and post-SIP periods. Logistic regression models were used to evaluate changes in admission rates between the two periods, and odds ratios (OR) were calculated.
Results
Patient demographics
871,020 total ED encounters from 487,028 unique patients were captured; 473,449 visits (54%) occurred in 2019, and 397,571 (46%) were in 2020. Our cohort was mostly White (65.4%), female (53.8%), privately insured (68.9%), and had a mean age of 46.0 ± 24.0. 2,572 (0.53%) patients had a documented Z code for SDOH. The most coded SDOH in ED encounters was problems related to housing and economic circumstances, followed by other problems related to the primary support group and other psychosocial circumstances (Supplemental Table 1). Compared to patients without SDOH, patients with a coded SDOH were more likely to be Black, younger, identify as male, and were more likely to be on government or self-pay insurance (Supplemental Table 2).
ED utilization rates before and after shelter-in-place orders
Compared to before the implementation of SIP orders, patients with previously diagnosed MH or SUDs were more likely to seek ED care after the SIP orders were put in place (RR: 1.20, 95% CI: 1.18 – 1.22, p<0.001) (Table 1). Among patients with SDOH, those with a history of problems related to the primary support group (RR: 2.89, 95% CI: 2.54 – 3.29) or housing and economic circumstances (RR: 2.50, 95% CI: 2.23 – 2.80) were more likely to present to the ED after SIP orders. Patients with both previously diagnosed MH or SUD and a documented SDOH (RR: 3.31, 95% CI: 2.83 – 3.88, p<0.001) had even higher ED utilization than those with either condition alone.
ED admission rates before and after shelter-in-place orders
Compared to pre-SIP, patients with a previously diagnosed MH or SUD (OR: 0.89, 95% CI: 0.86 – 0.92, p<0.001) or any documented SDOH (OR: 0.67, 95% CI: 0.54, 0.83, p<0.001) were less likely to be admitted to the hospital after presenting to the ED post-SIP orders (Table 2). Of the SDOH diagnosis categories, only patients with problems related to their primary support group and those with problems related to other psychosocial circumstances were significantly less likely to be admitted. The difference in admission rate pre- and post-SIP orders in patients with both MH/SUD and SDOH did not reach significance (OR: 0.65, 95% CI: 0.42, 1.02, p=0.061).
4.Discussion
The effect of shelter-in-place orders on vulnerable patient populations
The COVID-19 pandemic has underscored the importance of addressing social determinants of health, which are critical in shaping an individual’s health status and access to care. Our analysis of ED presentation and admission rates pre-and post-SIP orders found that patients with previously diagnosed MH or SUDs and those with a documented SDOH were more likely to present to the ED post-SIP. Prior studies that analyzed MH and substance use related ED visits demonstrate increased visit rates for these conditions as a primary diagnosis or chief complaint.6–8 This study extends previous work by demonstrating that SIP orders significantly impacted vulnerable patients with a history of MH, SUDs and SDOH.
Moreover, our analysis revealed that patients with co-occurring MH/SUD and SDOH had the highest increase in ED visits following SIP orders. These patients may have faced increased difficulties in accessing care and maintaining their overall health due to the closure or reduction of community-based resources9 and the exacerbation of existing socioeconomic inequities during the pandemic.5 Individuals with multiple disadvantages are at a significantly higher risk of premature mortality from avoidable causes than those with a single disadvantage10, highlighting the urgent need for tailored interventions that address the complex interplay of social, economic, and mental health-related factors contributing to poor health outcomes among these populations. Furthermore, patients with MH/SUDs or SDOH were less likely to be admitted to the hospital after presenting to the ED. While our study was not designed to examine the reasons behind these disparities, possible explanations include inadequate inpatient resources for patients with MH/SUD and SDOH or healthcare professional biases, as well as structural changes of healthcare resources being shunted to emergent medical care beds.
Limitations
Our results should be interpreted in the context of their limitations. First, the retrospective design of our study limits our ability to draw causal conclusions. Second, our data comes from one large integrated healthcare system, which may differ from other healthcare systems or geographic regions. Nonetheless, our study includes a large and diverse patient population from different levels of care within the state. Third, our findings may be influenced by institution-specific documentation practices that could affect SDOH coding rates and limit the generalizability of our findings. Finally, our analysis relied on patients’ past medical history documentation in the EMR, which may not reflect their current diagnoses at the time of their ED visit. However, given the time constraints of the ED, we considered capturing diagnoses from past medical history the best way to represent our sample’s health status.
Conclusion
Our study highlights the impact of the COVID-19 pandemic on vulnerable patient populations with a history of MH, SUDs, and SDOH. Our findings demonstrate that these patients were more likely to present to the ED but less likely to be admitted following SIP orders, indicating a need for tailored interventions that address the complex interplay of socioeconomic and mental health-related factors contributing to poor health outcomes among these populations. The findings highlight the need for greater standardization and consistency in documenting SDOH and addressing these factors in healthcare delivery.
Abbreviations
- MH
- Mental health
- SUD
- Substance use disorder
- SDOH
- Social determinant of health
- ED
- Emergency department
- SIP
- Shelter in place
Data Availability
All data produced in the present study are available upon reasonable request to the authors.