Emergency department utilization for substance use disorders and mental health conditions during COVID-19.

BACKGROUND: As the emergency department (ED) has evolved into the de-facto site of care for a variety of substance use disorder (SUD) presentations, trends in ED utilization are an essential public health surveillance tool. Changes in ED visit patterns during the COVID-19 pandemic may reflect changes in access to outpatient treatment, changes in SUD incidence, or the unintended effects of public policy to mitigate COVID-19. We use a national emergency medicine registry to describe and characterize trends in ED visitation for SUDs since 2019.
METHODS: We included all ED visits identified in a national emergency medicine clinical quality registry, which included 174 sites across 33 states with data from January 2019 through June 2021. We defined SUD using ED visit diagnosis codes including: opioid overdose and opioid use disorder (OUD), alcohol use disorders (AUD), and other SUD. To characterize changes in ED utilization, we plotted the 3-week moving average ratio of visit counts in 2020 and 2021 as compared to visit counts in 2019.
FINDINGS: While overall ED visits declined in the early pandemic period and had not returned to 2019 baseline by June 2021, ED visit counts for SUD demonstrated smaller declines in March and April of 2020, so that the proportion of overall ED visits that were for SUD increased. Furthermore, in the second half of 2020, ED visits for SUD returned to baseline, and increased above baseline for OUD ever since May 2020.
CONCLUSIONS: We observe distinct patterns in ED visitation for SUDs over the course of the COVID-19 pandemic, particularly for OUD for which ED visitation barely declined and now exceeds previous baselines. These trends likely demonstrate the essential role of hospital-based EDs in providing 24/7/365 care for people with SUDs and mental health conditions. Allocation of resources must be directed towards the ED as a de-facto safety net for populations in crisis.

Introduction

The Centers for Disease Control and Prevention (CDC) first documented a 42% decline in emergency department (ED) visits across the United States (US) during the early COVID-19 pandemic [1]. Decreased ED presentations for medical emergencies such as acute myocardial infarction (MI) and stroke, for which care-seeking may be discretionary in those with occult or atypical symptoms, have also been documented [2]. However, emergency medical services (EMS) and syndromic surveillance data has suggested an increase in opioid overdoses during the early months of the pandemic. In some regions, this may reflect worsening access to substance use disorder (SUD) treatment options amidst state and local shutdowns or the psychosocial effects of COVID-19 that place people at higher risk of overdose or returning to use [3-7]. Prior to the identification of COVID-19, the US was facing the highest reported number of drug overdose deaths ever reported, with 72,224 drug overdose deaths provisionally reported by the CDC in the twelve months preceding January 2020 [8]. More recently, the CDC has provisionally released that 96,779 drug overdose deaths have occurred in the twelve months preceding the end of March 2021, a new record for the highest number of drug overdose deaths in a twelve-month period ever recorded. Despite this, work characterizing changes in ED visits for SUD have documented decreased visit rates through July in a large contract management group sample of 108 EDs across 18 states, with a more muted decline in arrivals by EMS for opioid related reasons [9]. Other work using administrative claims data found similar declines in ED visits for substance use disorders through April 20, 2020 [10], and work using the National Syndromic Surveillance Program data found that ED visits were above the 2019 baseline for mental health, suicide attempts, overdose, and violence through October, 2020 [11]. While these concurrent crises have been described as an epidemic within a pandemic, little is known about how ED presentations for a broader array of SUDs or for mental health conditions, known to cooccur in nearly 40% of Americans with SUD [12,13], have been impacted during the evolving pandemic into 2021.

Given the gradual emergence of a ‘new normal’ and ongoing concerns about excess mortality both directly attributable to COVID-19 and related to SUDs, we aimed to provide an updated characterization of ED visitation for SUD in a national sample of hospital-based EDs and differentiate from prior work with a more granular exploration of visits for OUD as well as mental health conditions in comparison to commonly studied emergent medical conditions. As the ED has evolved into the de-facto site of care for a variety of SUD presentations including overdose, intoxication, acute withdrawal, and related mental health conditions, and given the finding that SUD patients were much more likely to present to the ED with COVID-19 [14], trends in ED visits are an essential public health surveillance tool for monitoring the prevalence and severity of SUD and facilitating ED-based efforts at treatment initiation and linkage [15]. Changes in ED visit rates for SUD may reflect changes in access to outpatient treatment for SUD or mental health conditions, broader trends in SUD incidence, the intended or unintended impact of COVID-19 public policy, or the effect of initiatives to address SUD through better treatment and social service coordination.

Accordingly, we utilized a national emergency medicine clinical quality registry of aggregated electronic health record data to describe trends in ED visitation for SUDs since 2019 through the current COVID-19 pandemic across a wide range of clinical and geographic settings.

Methods

Study design

We conducted an observational study of ED visits included in national emergency medicine clinical quality registry with data from January 1, 2019 through June 30, 2021.

Study setting and dataset

We included all ED visits identified in the Clinical Emergency Data Registry (CEDR), a voluntary national emergency medicine clinical quality registry maintained by the American College of Emergency Physicians (ACEP) with complete data from January 1, 2019 through June 30, 2021. In general, the EDs participating in the ACEP CEDR reflect community, hospital-based EDs of similar regional, case and insurance mix to national datasets with notably fewer academic hospitals. In order to characterize hospital-based EDs in the sample according to their urban-rural status, as well as their status as an academic hospital (an Accreditation Council for Graduate Medical Education accredited program), control status (public, private for-profit, not-for-profit), number of inpatient beds, sites were linked to data from the 2018 American Hospital Association Annual Survey, the most recently available data [16]. We furthermore linked this registry data to publicly available data on daily active COVID-19 cases per 1,000 population from the University of Maryland COVID-19 Impact Analysis Platform [17] to estimate the burden of COVID-19 in the county where each hospital-based ED was located.

Definitions

We defined SUD conditions based on ED diagnosis of International Classification of Disease, Tenth Revision, Clinical Modification (ICD-10-CM) diagnosis codes. Consistent with prior work, we defined SUD using diagnosis code groupings of the Clinical Classification Software Revised (CCSR) system, maintained by the Healthcare Cost and Utilization Project [9,18] to provide useful groupings of diagnosis codes, as well as select subsets of public health interest within that group including opioid overdose and opioid use disorder (OUD) (MBD018) and alcohol use disorders (AUD) (MBD017) [19], as well as other SUD inclusive of cannabis (MBD019), sedative/hypnotic/anxiolytic (MBD020), cocaine/other stimulant (MBD021), hallucinogenic (MBD022), inhalant (MBD023), and other psychoactive substance-related disorders (MBD025). As reference and to characterize changes in ED visitation for SUDs, we also characterized ED visit counts for two common acute medical emergencies, acute myocardial infarction and stroke (CCSR codes CIR009 and CIR020). Given the common co-occurrence of mental health conditions alongside SUDs, and evidence that conditions such as depression and anxiety that often cooccur with SUD and have escalated in incidence during COVID-19 due to social isolation, unemployment and other factors [20], we also identified ED visits for mental health conditions, broadly defined. We identified mental health conditions according to International Classification of Disease–10 (ICD-10) root codes F2 through F9, inclusive of psychotic and mood disorders, behavioral syndromes, and disorders of personality. .

Outcomes

The primary outcome was the weekly counts of ED visits for SUDs including subsets of OUD and AUD. As secondary outcomes to provide additional context we also calculated weekly counts of ED visits for acute myocardial infarction, stroke, and mental health conditions.

Analysis

To characterize changes in ED utilization for each condition, we plotted the 3-week moving average ratio, and associated 95% confidence intervals, of ED visit counts in 2020 and 2021 as compared to the same weeks in 2019. Due to the dramatic changes in overall ED visit counts throughout the pandemic, we identified different periods for descriptive analysis by applying a segmented linear regression approach on overall ED visit counts, with time as the only input into the model, using a previously described iterative linearization process for finding unknown breakpoints () [21]. These simply-derived breakpoints were used to identify time periods for subsequent descriptive analysis. Given distinct epidemiologic implications, we also report ED utilization in each of these identified periods for each condition in three ways including: 1) average weekly counts, 2) proportion of overall ED visits, 3) hospital-based ED visit incident rate ratios (IRR) for weekly visit counts comparing different periods in 2020/2021 to identical periods in 2019. IRRs for site-level ED visit counts across conditions, with 95% confidence intervals, were obtained via unadjusted Poisson regression models for hospital-based daily ED visit counts in each period compared to the same weeks in 2019.

As a secondary analysis to examine the impact of known geographic and temporal variation in COVID-19 outbreaks across the US, which may impact care-seeking for SUD, we conducted additional analyses stratified by geographic regions, defined according to census regions [22]: Northeast, Midwest, South, West. We utilized the daily active COVID-19 cases per 1,000 population for the county of each ED, averaged across counties represented in our sample in each census region, and plotted these along with SUD to provide context for trends in ED visits. Analyses were performed in R (4.0.2). Access to the CEDR data is restricted by agreement with ACEP, given potentially identifying features, and the study was classified as exempt by the Institutional Review Board at Yale University.

Results

Among a total of 174 EDs across 33 states included (), median ED visit volumes in 2019 was 29,296 [interquartile range 15,989 to 47,082], with the most sites in California (41) and Texas (15). Of all EDs, 55 (31.6%) are located in rural areas while 119 (68.4%) are located in urban areas [23,24]. 79 (45.4%) were sites participating in an Accreditation Council for Graduate Medical Education accredited program and 95 (54.6%) did not. 83 (47.7%) were nongovernment, not-for-profit, 33 (19.0%) were for-profit entities, 21 (12.1%) were church-operated, and the remaining 37 (21.2%) were various nonfederal government entities. Median inpatient beds was 138, with a minimum of 11 and maximum of 1,031 (interquartile range 49 to 239). Across sites, the proportion of visits covered by Medicare, Medicaid, and private insurance were 25.7%, 26.9%, and 24.6%, respectively.

As in , overall ED visit counts declined in the early pandemic period of March and April 2020, to a nadir of 52% of levels in the same period in 2019, before returning to 75–85% of visit counts in the period from July 2020 through June 2021. ED visits for SUD and mental health exhibited declines in counts as well, to 72% and 71%, respectively, much closer to 2019 levels than overall visits throughout the second half of 2020 and into 2021. Shown in , while visit counts for AUD remained low relative to 2019, ED visits for OUD returned to 2019 levels by May, and were subsequently above 2019 counts for much of 2020 and into 2021. depicts weekly average visit counts, the proportion of overall ED visits for each diagnosis group, and the incident rate ratio for those visits as compared to the same weeks in 2019. Weekly ED visit counts for OUD and other SUD (non-alcohol and non-opioid) were statistically significantly above the 2019 baseline counts in the first 12 weeks of 2021 (IRR 1.06 [1.05–1.08] and IRR 1.03 [1.01–1.04], respectively), while visits for alcohol remained lower (IRR 0.94 [0.92–0.95]). As with OUD, ED visit counts for mental health, as well as acute myocardial infarction and stroke, were at or above the 2019 levels in the final periods of 2020. Visit counts were lower following the arrival of COVID-19 for each of OUD, AUD, and other SUD (), but visits rose as a proportion of overall ED visits ().

Trends in weekly ED visits for substance use disorders and mental health conditions, 2020.

Notes: Figure depicts 3-week moving average ratio of emergency department visit counts in 2020 and 2021 compared to identical weeks in 2019. Sample includes 174 hospital-based EDs across 33 states from a national emergency department clinical quality registry. Visits for substance use disorders and mental health conditions were identified by International Classification of Disease-10 codes.

Trends in weekly ED visits for OUD, AUD, and other SUD-related, 2020.

Notes: Figure depicts 3-week moving average ratio of emergency department visit counts in 2020 and 2021 compared to identical weeks in 2019. Sample includes 174 hospital-based EDs across 33 states from a national emergency department clinical quality registry. Visits for opioid use disorder (OUD), alcohol use disorder (AUD) and other substance use disorder (SUD)-related were identified by International Classification of Disease-10 codes.

The secondary analysis of ED visitation based on geography demonstrated similar patterns for both overall ED visits as well as for overall SUD, AUD, and OUD across each regions despite substantially different covid outbreak patterns. In general, visit counts for SUD were high relative to 2019 in January and February, 2020, especially in the Northeast and Midwest, and fell to below 2019 levels briefly before again rising, in the second half of 2020.

Trends in weekly ED visits for OUD, AUD, and other SUD-related, 2020, By US census regions.

Note: Figure depicts 3-week moving average ratio of emergency department visit counts in 2020 and 2021 compared to identical weeks in 2019 in each U.S. Census Region. Sample includes 174 hospital-based Eds across 33 states from a national emergency department clinical quality registry. Visits for opioid use disorder (OUD), alcohol use disorder (AUD) and other substance use disorder (SUD)-related were identified by International Classification of Disease-10 codes.

Discussion

Unlike overall ED visitation, which has been shown to have substantially declined and then returned to modestly lower counts during the COVID-19 pandemic relative to 2019, declines in ED utilization for SUD were more muted and rebounded to 2019 levels earlier than even emergency conditions, such as acute MI and stroke [2,25]. Furthermore, despite the common cooccurrence of SUDs, ED visits for OUD have been increasing at a substantially higher rate than AUD or other SUDs since the beginning of the COVID-19 pandemic. This suggests that the pandemic may have disproportionate condition specific impacts as a result of fundamentally different patterns of acute care access and outcomes for different clinical scenarios. For example, while the pandemic may have resulted in impeded access to outpatient treatment programs (OTPs) or office-based buprenorphine prescriptions that in turn resulted in poor OUD management and increased ED visitation, the pandemic may not have had as much relative effect on services for AUD that are less medication or healthcare facility dependent. It is likely that a shift of focus away from the opioid epidemic as systems sought to manage and mitigate the COVID-19 pandemic may have also expanded supply chain gaps, worsened contacts with healthcare providers necessary to ensure access to naloxone, potentially driving increases in opioid-associated fatalities and exacerbating inequities in access to care.

Interestingly, we found little variation in ED utilization trends for each condition based on geography despite substantial variation in covid outbreak timing, public health policies and known geographic variability in SUD incidence. This likely supports the notion that initial declines in ED visitation reflect a broad change in care seeking behavior as opposed to a clinically nuanced decision by people. This finding also demonstrates that COVID burdens of EDs and hospitals were not related to local care seeking patterns or the local epidemiology of SUD. CDC provisional death data, though, show geographic variability in the burden of SUD mortality, especially with disproportionate impact in California, Arizona, and Washington and much policy attention to addressing the crisis [26]. Interestingly, amongst the four US census regions reported in , the West exhibited the smallest increases in ED presentations for OUD in our analysis. While this may represent a limitation of our work and ED visitation in our sample may not be representative generally, California was the best represented state in our sample (with 41 hospital-based EDs). This disconnect between ED visitation and population-level mortality may be related to delayed activation of EMS services. Given the value of EDs as sites for the initiation of medication-assisted therapy for OUD [27], failure to initiate transport in cases of opioid or other overdose not only risks mortality in the acute phase but also risks failure to initiate treatment and avoid a subsequent overdose.

Several phenomena may explain these findings. First, given the importance of the ED in providing access to care for SUDs in both the acute phase, such as overdose or skin abscess, as well as initiating treatment for SUDs and serving as a conduit to limited outpatient SUD treatment resources, people with SUDs likely continued to utilize the ED to access services. This was likely compounded during the COVID-19 pandemic by the commensurate closure or limitation of many outpatient care venues because of government shutdown and social distancing policies. Second, the relatively higher utilization of the ED for SUDs, particularly OUD, may reflect compensatory increases in substance use or exacerbations of mental health conditions because of social policies. Furthermore, patients with injection drug use may have had increasingly limited access to harm reduction services and sterile syringes, leading to increased skin and soft tissue infections for which ED care was needed. In fact, our analysis may underestimate the degree of increased SUD associated harms as EMS data during the pandemic suggest both increased transport refusals for a myriad of conditions as well as increased out-of-hospital overdose-related cardiac arrests, which may not be classified as opioid-related and may have resulted in transport to the ED previous to the pandemic [7,28]. Third, our findings may represent less patient discretion in seeking emergency care for acute symptoms and concerns related to SUDs than even other medical emergencies. For example, a person experiencing an overdose may be more likely to be involuntarily transported to an ED in comparison to a person with chest pain who may underestimate the clinical significance of symptoms indicative of an acute myocardial infarction and overestimate the infectious risk of seeking care in hospital EDs.

Importantly, increasing social isolation and unemployment associated with COVID-19 may be exacerbating US rates of “deaths of despair,” which has been identified as the etiology of decreased US life expectancy from 2014–2017 when midlife mortality increased across all racial groups, caused by drug overdoses, alcohol use, suicides, and a diverse list of organ system diseases [29-32], with opioid overdoses exhibiting racial and ethnic disparities as well [33]. Health systems and overburdened EDs continue to adapt operations to the COVID-19 pandemic, particularly given increased trends of extremely prolonged ED length of stay, particularly the “boarding” of ED patients in the ED awaiting inpatient bed assignment [34]. Patients may increasingly seek care for conditions previously delayed or deferred, and the essential role of the hospital-based ED in providing 24/7/365 access to care for people with SUDs and mental health conditions becomes more evident.

Limitations

These findings should be interpreted within the confines of the study dataset and design. First, our sample includes primarily community based EDs similar to most EDs and hospitals in the US, however may not be generalizable to teaching or specialty centers that are disproportionately less likely to participate in the ACEP CEDR registry. Second, while we acknowledge the likely important interaction between race and ethnicity and these observed trends, the ACEP CEDR registry does not currently capture this data reliably and future work should explore whether observed changes in visitation trends reflect widening disparities in access to SUD care for disadvantaged populations. Third, our dataset is limited to examining ED visitation without an ability to capture other concurrent care setting such as telemedicine that have increased access to care for select SUD populations during the pandemic [35]. However, given that among SUDs OUD is likely the most amenable to telemedicine management (e.g ability to initiate and prescribe buprenorphine), but we observed the greatest increase in ED visits for OUD, these innovative care models are not likely to have had widespread impact.

Conclusions

As health systems continue to adapt to the COVID-19 pandemic, including substantial changes in ED and hospital volumes for SUD care which may have been delayed, deferred or exacerbated by the pandemic, the essential role of the hospital-based ED in providing 24/7/365 care for people with SUDs and mental health conditions has heightened importance. Allocation of resources must be directed towards the ED as a de-facto safety net that minimizes the morbidity and mortality for populations in crisis.

ACEP clinical emergency data registry sample sites.

(TIF)

Click here for additional data file.

Trends in overall weekly ED visits with segmented linear regression.

(TIF)

Click here for additional data file.

Visit counts for SUD: OUD, AUD, and other SUD.

(TIF)

Click here for additional data file.

Proportion of overall visits for SUD: OUD, AUD, and other SUD.

(TIF)

Click here for additional data file.

(DOCX)

Click here for additional data file.

Transfer Alert

This paper was transferred from another journal. As a result, its full editorial history (including decision letters, peer reviews and author responses) may not be present.

18 Jun 2021

Submitted filename: CEDR-OUD-PLOS-ONE-RESPONSE-REVIEWERS(1).docx

Click here for additional data file.

20 Sep 2021

PONE-D-21-20023Emergency Department Utilization for Substance Use Disorders and Mental Health Conditions During COVID-19PLOS ONE

Dear Dr. Venkatesh,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please address the following journal requirements and additional editor and reviewer comments.

Please submit your revised manuscript by Nov 04 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Fernando A. Wilson, PhD

Academic Editor

PLOS ONE

Journal requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf.

2. In ethics statement in the manuscript and in the online submission form, please provide additional information about the patient records/samples used in your retrospective study. Specifically, please ensure that you have discussed whether all data/samples were fully anonymized before you accessed them and/or whether the IRB or ethics committee waived the requirement for informed consent. If patients provided informed written consent to have data/samples from their medical records used in research, please include this information.

3. Thank you for stating the following in the Acknowledgments Section of your manuscript:

“This work was supported by the HHS Office of the Secretary Patient Centered Outcomes Research Trust Fund (PCORTF) under IDDA# ASPE-2018-001 and NIDA UG1DA015831-18S2. In addition, Dr. Venkatesh was supported by KL2 TR000140 from the National Center for Advancing Translational Sciences of the NIH and the American Board of Emergency Medicine – National Academy of Medicine Anniversary Fellowship. The contents of this work are solely the responsibility of the authors and do not necessarily represent the official view of NIH.”

We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

“This work was supported by the HHS Office of the Secretary Patient Centered Outcomes Research Trust Fund (PCORTF) under IDDA# ASPE-2018-001 and NIDA UG1DA015831-18S2. In addition, AKV was supported by KL2 TR000140 from the National Center for Advancing Translational Sciences of the NIH and the American Board of Emergency Medicine – National Academy of Medicine Anniversary Fellowship. The contents of this work are solely the responsibility of the authors and do not necessarily represent the official view of NIH.

The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.”

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

4. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

5. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For more information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially sensitive information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

Additional Editor Comments (if provided):

- The manuscript is not formatted according to PLOS ONE requirements, eg, include page and continuous line numbers. Refer to the following link:

https://journals.plos.org/plosone/s/submission-guidelines

- References are also not formatted following PLOS ONE requirements. Please review each reference and verify their formatting with the following instructions. Use square brackets for citations. Use NCBI journal abbreviations. For #2, please update this reference since it’s now published.

https://journals.plos.org/plosone/s/submission-guidelines#loc-references

https://www.ncbi.nlm.nih.gov/nlmcatalog/journals

- Add a dash within “co-occurrence”.

- Use either “U.S.” or “US” consistently throughout the manuscript.

- Use “1,000” instead of “1000” to be consistent with formatting of other numbers.

- Please provide more details on the regression analyses and regression tables in the Supporting Information section.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: I Don't Know

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: In this revised paper, the authors were very responsive to two previous reviewers’ comments and suggestions and addressed their concerns. The revised paper is solid in its analyses and presents interesting findings. Figures of changes overtime (compared to the 2019 baselines) are helpful for readers. However, writing in some parts, especially in the Discussion section, is difficult to understand (as previous reviewers also pointed out) and confusing. I also have a question about the findings related to geographic differences.

1. The major source of confusion is that ED visit counts and rates are described in the same paragraph and it is not clear which one is which. I gathered that the % in the Y axis in the Figures is the proportion of ED visit counts in 2020 as compared to the same weeks in 2019. In the Results section (first paragraph—please specify page numbers), it is stated that “ED visits for SUD and mental health exhibited similar, however, more muted declines in ED visitation and were much closer to baseline than overall visits by in the second half of 2020.” And then, “While visit rates for AUD remained below 2019, ED utilization for OUD returned to 2019 levels by June, and were subsequently above 2019 baseline for the remainder of the year.” The “visit rate” is confusing, as it is not clear what you mean by that. Also clarify that the “baselines” are the ED visit counts in the corresponding weeks in 2019. Also, in “ED visits for SUD and mental health exhibited similar, however, more muted declines in ED visitation and were much closer to baseline than overall visits by in the second half of 2020,” you do not have to repeat “in ED visitation.”

2. First paragraph of the Discussion section: “…ED visits for OUD have been increasing [at is missing here] a substantial higher rate than AUD or other SUDs since the beginning of the COVID-19 pandemic. These distinct condition specific trends demonstrate the fundamentally different patterns of acute care access and outcomes for different clinical scenarios. Most importantly, it is likely that a shift of focus away from the opioid epidemic as systems sought to manage and mitigate the COVID-19 pandemic expanded gaps in OUD treatment and naloxone access, driving increases in opioid-associated fatalities and exacerbating inequities in access to care.” I found last two sentences very confusing to read, as the second sentence is not specific about which the fundamentally different patterns of acute care access and outcomes are referred to. I expected that the third sentence would provide that explanation; however, the third sentence, starting with “More importantly” sounds like that it has moved to a different subject matter.

3. Again in the Discussion section, the second part of the following sentence is not clear: “In fact, our analysis may underestimate the degree of increased SUD associated harms as EMS data suggests increased out-of-hospital overdose-related cardiac arrests, which may in turn have not resulted in ED visits among people fearful of seeking emergency or hospital care (7).” Do you mean that the overdosed people with cardiac arrest refused transportation to ED or hospital care because they were afraid of contracting COVID?

4. “…increased trends of ED boarding (27)”: ED boarding may be an acceptable term among ED physicians, but it is not among non-ED readers.

5. Geographic differences: You cited the CDC’s provisional data on overdose death counts (Ahmad et al., 2021). The data show that many states in the West (CA, AZ, WA…) had higher increase in opioid and other drug overdose deaths in 2020. However, your Figure 3 data shows that the West region has the most muted increase of opioid and other drug related ED visits of all four regions. Does the difference suggest that fewer overdosed cases in the West were cared for at ED (i.e., EMT found the cases dead or DOA)? Or, is the difference attributable to that fact that your source of data are not representative of all EDs? Please refer to the following and provide some more discussion:

Baumgartner, J.C., and Radley, D.C., 2021. The spike in drug overdose deaths during the COVID-19 pandemic and policy options to move forward. https://www.commonwealthfund.org/blog/2021/spike-drug-overdose-deaths-during-covid-19-pandemic-and-policy-options-move-forward.

6. One last question: ED visit counts not number of unique individuals, right?

Reviewer #2: The authors analyzed a large sample of emergency department (ED) visits identified in the Clinical Emergency Data Registry (from 258 sites across 34 states) spanning 2019-2020. Key findings from the authors’ analyses were that overall ED visits declined in the early pandemic period and had not returned to baseline by the end of 2020; however, ED visits for substance use disorder (SUD) declined at slower rates in March and April 2020, such that the proportion of overall ED visits that were for SUD increased in the earlier months of 2020. The authors also found that ED visits for SUD returned to baseline in the second half of 2020. The authors concluded their paper with a meaningful discussion about hospital-based SUD care in 2020 within the context of resource allocation and availability concerns experienced by many hospital systems during 2020. Overall, the trends presented in this paper are meaningful within the context of public health surveillance and continuing to learn how the COVID-19 pandemic may have disrupted substance use disorder treatment.

However, there are several concerns and questions that should be addressed. I offer the following comments and questions for the authors’ consideration. I will first summarize key issues I have with the paper as currently written. Additional minor comments, suggestions, and questions follow.

Thank you for the opportunity to review your paper.

COMMENTS AND RECOMMENDATIONS

KEY ISSUES

A. General

1. My biggest concern is that other studies published prior to the submission of this paper have conducted similar analyses and highlighted some similar results. For example, Pines et al. (2021) examined ED visit trends for SUD from Jan-July 2020 and also found that ED visits for SUD decreased sharply early on, declined less than overall ED visits, and eventually began to return to 2019 levels. Pines et al. also explored differences by patient demographic factors. In turn, Holland et al. (2021) examined over 3,000 EDs spanning 48 states to explore trends in drug-related overdose/OUD and MH ED visit rates from Dec. 2018 to October 2020. Holland et al. described increases in OD-related ED visit rates, similarly noting that visits did not decrease in a manner similar to other ED visits, suggesting an increase in OUD ED burden in 2020.

I am not certain about the journal or editors’ stance on the novelty of findings or studies that coincide with or replicate other studies. However, it’s an important part of the scientific process, and surely the pandemic warrants much inquiry and better continued learning about how COVID-19 affects care delivery in various settings. Perhaps one thing the authors could do is cite studies like Holland et al., Pines et al., and Lucero et al. and differentiate in their paper how their analytic sample differs from others, if there are meaningful differences (e.g., significant differences in community-based EDs identified in ACEP CEDR?), and then reposition the Discussion to focus on the experience of those types of EDs, patient populations, etc., while also making sure to correctly summarize the gaps in the scholarly literature that are being filled.

References:

Pines, J. M., Zocchi, M. S., Black, B. S., Carlson, J. N., Celedon, P., Moghtaderi, A., & Venkat, A. (2021). How emergency department visits for substance use disorders have evolved during the early COVID-19 pandemic. Journal of Substance Abuse Treatment, 108391.

Holland, K. M., Jones, C., Vivolo-Kantor, A. M., Idaikkadar, N., Zwald, M., Hoots, B., ... & Houry, D. (2021). Trends in US emergency department visits for mental health, overdose, and violence outcomes before and during the COVID-19 pandemic. JAMA Psychiatry, 78(4), 372-379.

Lucero, A. D., Lee, A., Hyun, J., Lee, C., Kahwaji, C., Miller, G., ... & Pan, L. (2020). Underutilization of the emergency department during the COVID-19 pandemic. Western Journal of Emergency Medicine, 21(6), 15.

2. The authors seem to suggest their data will not be made publicly available. PLOS journals require authors to make all data necessary to replicate their study’s findings publicly available without restriction at the time of publication. Authors do not need to submit their entire data set if only a portion of the data was used in the reported study. When specific legal or ethical restrictions prohibit public sharing of a data set, authors must indicate how others may obtain access to the data. For studies involving third-party data, PLOS encourage authors to share any data specific to their analyses that they can legally distribute. PLOS recognizes, however, that authors may be using third-party data they do not have the rights to share. When third-party data cannot be publicly shared, authors must provide all information necessary for interested researchers to apply to gain access to the data. I.e., https://journals.plos.org/plosone/s/data-availability.

B. Introduction Section

1. As described in a later comment, I think the Introduction section nicely motivates the need for this type of analysis. However, it would be helpful if the authors directly stated the study objective in the second paragraph of the Introduction section. I understand you are just examining (important) trends and that the research question/study objective (and possibly any hypotheses) may be implicit. Although it is not necessary to test any explicit hypotheses in this type of study, given the journal’s broad scope and readership it would be helpful for the reader if you directly stated the study objective in relation to the gap you are ostensibly filling. E.g., perhaps open the second paragraph with an explicit statement of the objective, which would lead nicely into the rest of the paragraph supporting the rationale for that objective.

C. Methods Section

1. Did you use all ICD-10 codes within F10-F19 as inclusion criteria for SUD visits? Did you also include other relevant codes outside of that range, e.g., T40.-, G62.1, etc.? As supplemental content, could you include a table of all ICD-10 codes used to flag SUD and comparison conditions in this study, similar to Holland et al. (2021) and Pines et al. (2021)? This will be helpful for assessing/knowing what visits were included in the trend analyses, comparing to other studies, etc.

2. If I am interpreting this correctly, you summarized and described your analytic sample in sentences 2-4 of your Study Setting and Dataset (Methods) subsection. This is important information, but its placement in the Methods section is uncommon in health services research. These descriptive results would be better situated at the beginning of the Results section, helping the reader understand and assess your analytic sample characteristics leading into the main analyses (e.g., assessing generalizability and applicability of your analytic sample vis-a-vis achieving your study objective).

3. More information is needed about the regression modeling you performed. Just to confirm, the piecewise regression estimation process was simply used to identify the breakpoints, then you report the ED utilization in those identified periods? Could you provide a bit more information about the Poisson regression models used to estimate the IRRs comparing the 2020 periods to the 2019 reference periods? What variables went into the Poisson regression models? These were estimated for each period identified by the piecewise regression approach? At a minimum, more details are needed to assess what was conducted, though you could certainly include statistical notes as supplemental content to help the reader understand and replicate your analytic approach.

4. Tables and Figures: It would be helpful to include notes sections for each table and figure to briefly describe things like what the outcomes are/derived from which analytic method, number of observations, etc.

MINOR ISSUES AND COMMENTS

1. Please paginate your manuscript file. It makes providing feedback easier, given the files compiled and presented to reviewers.

2. Introduction: The beginning of the Introduction section is effective. The authors effectively summarized relevant (preliminary) findings from the scholarly literature, CDC, etc. The authors also quickly pivoted to motivating their research objective in a convincing way, noting the convergence of several critical public health crises (i.e., the “epidemic within a pandemic”).

3. Methods, Study Design subsection: I would recommend saying you conducted an “observational study” instead of “observational analysis”, then describe your analytic methods in your Analysis section.

4. Methods, Study Setting & Data subsection: PLOS encourages authors to cite any publicly available research data in their reference list, such as the U of Maryland COVID-19 Impact Analysis Platform. Please see: https://journals.plos.org/plosone/s/data-availability.

5. Results section: This section is well written, and key findings are discussed. However, the section is difficult to follow in relation to the tables and figures. I think you have indicated where you would like the placement of each table/figure. But I would recommend directly saying, “Table 1 shows….” or parenthetically citing the table/figure at the end of each sentence talking about the table/figure. E.g., “While visit rate for AUD… for the remainder of the year (Figure 2).

6. Do the x-axes on Figures 1 and S2 align? In Figure 1 it appears the 3-week moving average of ED visits overall begins to increase at the end 2020; however, that is not shown in Figure S2 (just what appears to be a new trend of decreasing moving averages). If different, does the increase (weeks 51-52?) affect the piecewise regression breakpoints?

7. Results section: You say, “ED visits for SUD and mental health exhibited similar, however, more muted declines in ED visitation and were much closer to baseline than overall visits by in the second half of 2020.” However, my read of Figure 1 is that SUD and MH ED 3-week moving averages also experienced precipitous decreases vis-à-vis 2019. If I am interpreting that correctly, I am not sure “muted” is the right word or tone for the discussion of this finding. Perhaps consider including the exact percentages (per 2019 visits) for SUD and MH ED visits as you did for all ED visits in the previous sentence. That will help clarify the relative rate of decrease.

8. You say, “Compared to similar periods in 2019, ED visits for SUD and AUD were close to baseline in the final period (Nov. 11 to Dec. 31) of 2020 (IRR 0.88 [0.85-0.91] and IRR 0.87 [0.83-0.91], respectively), while visits for OUD were above baseline (IRR 1.07 [1.02-1.13]).” I see how that can be interpreted from the IRR estimates presented in Table 1. However, the error bars in Figure 2 suggest the 3-week moving averages for OUD ED visits vis-à-vis the 2019 averages were not greater (i.e., crossing over 100% baseline). Am I interpreting that correctly? Or were the estimates greater than baseline for the average of that entire final breakpoint period in 2020? This is where the table/figure notes sections and additional statistical modeling information could be useful.

9. Figure 3 is interesting, though it is noisy and a bit difficult to interpret with the error bars. It is certainly responsible to illustrate the ranges, but I wonder if it might make sense just to plot the means for better conveying the key takeaways.

10. Discussion section: I think you meant to say “substantially higher rate” in the second sentence, not “substantial”.

11. Table 1 is nice. Very intuitive presentation.

12. The Discussion of potential explanations for your trends findings was compelling.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

4 Nov 2021

Dear Dr. Venkatesh,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please address the following journal requirements and additional editor and reviewer comments.

Please submit your revised manuscript by Nov 04 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Fernando A. Wilson, PhD

Academic Editor

PLOS ONE

Journal requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf.

• We have updated the manuscript to meet the PLOS ONE style requirements

2. In ethics statement in the manuscript and in the online submission form, please provide additional information about the patient records/samples used in your retrospective study. Specifically, please ensure that you have discussed whether all data/samples were fully anonymized before you accessed them and/or whether the IRB or ethics committee waived the requirement for informed consent. If patients provided informed written consent to have data/samples from their medical records used in research, please include this information.

• We have added this to the end of the Methods section to clarify: “Analyses were performed in R (4.0.2). Access to the CEDR data is restricted given potentially identifying features, and the study was classified as exempt by the Institutional Review Board at Yale University.”

3. Thank you for stating the following in the Acknowledgments Section of your manuscript:

“This work was supported by the HHS Office of the Secretary Patient Centered Outcomes Research Trust Fund (PCORTF) under IDDA# ASPE-2018-001 and NIDA UG1DA015831-18S2. In addition, Dr. Venkatesh was supported by KL2 TR000140 from the National Center for Advancing Translational Sciences of the NIH and the American Board of Emergency Medicine – National Academy of Medicine Anniversary Fellowship. The contents of this work are solely the responsibility of the authors and do not necessarily represent the official view of NIH.”

We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

• We have removed this from the acknowledgements in the revision per editorial guidance.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

“This work was supported by the HHS Office of the Secretary Patient Centered Outcomes Research Trust Fund (PCORTF) under IDDA# ASPE-2018-001 and NIDA UG1DA015831-18S2. In addition, AKV was supported by KL2 TR000140 from the National Center for Advancing Translational Sciences of the NIH and the American Board of Emergency Medicine – National Academy of Medicine Anniversary Fellowship. The contents of this work are solely the responsibility of the authors and do not necessarily represent the official view of NIH. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.”

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

• We have removed funding-related text and updated the funding information in the submission portal.

4. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

• In the revised work, three separate data sources were included in this analysis. The University of Maryland COVID-19 Impact Analysis Platform, with county-level data on daily COVID-19 cases, is available on email request at this URL: https://data.covid.umd.edu/. The American Hospital Association Annual Survey data is available for purchase with data use agreement. However, the Clinical Emergency Data Registry (CEDR) dataset utilized is not a publicly available dataset, was obtained under a legal data use agreement to protect the identification of any site, physician group or physician and includes numerous easily identifiable elements were it available in a publicly accessible format. The dataset was not created or built upon any public funding or public disclosure requirements as well. Therefore, in accordance with the PLOS guidelines, this dataset cannot be available via public repository.

5. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For more information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a. If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially sensitive information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

• While the data fields for county-level COVID-19 cases are available via the University of Maryland COVID-19 Impact Analysis Platform, and data from the American Hospital Association Annual Survey are available for purchase, the Clinical Emergency Data Registry (CEDR) dataset is owned by a third party, the American College of Emergency Physicians (ACEP). Our data use agreement with ACEP permits the use of data for select research purposes under which no ED site, physician or physician group could be identified, which is not possible given the ease of reverse engineering identification if data were shared freely. Interested parties may seek to utilize the same dataset with permission by contacting the American College of Emergency Physicians at URL: https://www.acep.org/cedr.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

Additional Editor Comments (if provided):

- The manuscript is not formatted according to PLOS ONE requirements, eg, include page and continuous line numbers. Refer to the following link:

https://journals.plos.org/plosone/s/submission-guidelines

- References are also not formatted following PLOS ONE requirements. Please review each reference and verify their formatting with the following instructions. Use square brackets for citations. Use NCBI journal abbreviations. For #2, please update this reference since it’s now published.

https://journals.plos.org/plosone/s/submission-guidelines#loc-references

https://www.ncbi.nlm.nih.gov/nlmcatalog/journals

• We have corrected the formatting of the citations in this resubmission.

- Add a dash within “co-occurrence”.

- Use either “U.S.” or “US” consistently throughout the manuscript.

• We have made further stylistic corrections as well.

- Use “1,000” instead of “1000” to be consistent with formatting of other numbers.

- Please provide more details on the regression analyses and regression tables in the Supporting Information section.

• We have added language to the Analysis section to specify that the segmented linear regression approach for overall ED visit counts ‘with time as the only input into the model’ was generated ‘using a previously described iterative linearization process for finding unknown breakpoints.’ Incident rate ratios (IRRs), reported to describe difference in incidence of ED visits across conditions in different time periods of analysis, were ‘obtained via unadjusted Poisson regression models.’

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

________________________________________

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: I Don't Know

________________________________________

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No

________________________________________

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

________________________________________

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: In this revised paper, the authors were very responsive to two previous reviewers’ comments and suggestions and addressed their concerns. The revised paper is solid in its analyses and presents interesting findings. Figures of changes overtime (compared to the 2019 baselines) are helpful for readers. However, writing in some parts, especially in the Discussion section, is difficult to understand (as previous reviewers also pointed out) and confusing. I also have a question about the findings related to geographic differences.

• We appreciate these sentiments and have sought to improve the clarity of this manuscript in this revision.

1. The major source of confusion is that ED visit counts and rates are described in the same paragraph and it is not clear which one is which. I gathered that the % in the Y axis in the Figures is the proportion of ED visit counts in 2020 as compared to the same weeks in 2019. In the Results section (first paragraph—please specify page numbers), it is stated that “ED visits for SUD and mental health exhibited similar, however, more muted declines in ED visitation and were much closer to baseline than overall visits by in the second half of 2020.” And then, “While visit rates for AUD remained below 2019, ED utilization for OUD returned to 2019 levels by June, and were subsequently above 2019 baseline for the remainder of the year.” The “visit rate” is confusing, as it is not clear what you mean by that. Also clarify that the “baselines” are the ED visit counts in the corresponding weeks in 2019. Also, in “ED visits for SUD and mental health exhibited similar, however, more muted declines in ED visitation and were much closer to baseline than overall visits by in the second half of 2020,” you do not have to repeat “in ED visitation.”

• To improve clarity in this revision, we have removed the term “rate” and instead used language to describe the nominal/count of ED visits or described visits as “relatively” higher or lower when comparing to 2020 and 2019.

2. First paragraph of the Discussion section: “…ED visits for OUD have been increasing [at is missing here] a substantial higher rate than AUD or other SUDs since the beginning of the COVID-19 pandemic. These distinct condition specific trends demonstrate the fundamentally different patterns of acute care access and outcomes for different clinical scenarios. Most importantly, it is likely that a shift of focus away from the opioid epidemic as systems sought to manage and mitigate the COVID-19 pandemic expanded gaps in OUD treatment and naloxone access, driving increases in opioid-associated fatalities and exacerbating inequities in access to care.” I found last two sentences very confusing to read, as the second sentence is not specific about which the fundamentally different patterns of acute care access and outcomes are referred to. I expected that the third sentence would provide that explanation; however, the third sentence, starting with “More importantly” sounds like that it has moved to a different subject matter.

• We have corrected the typo mentioned above.

• We have attempted to clarify the intended point here, adding that:

“ED visits for OUD have been increasing at a substantially higher rate than AUD or other SUDs since the beginning of the COVID-19 pandemic. This suggests that the pandemic may have disproportionate condition specific impacts as a result of fundamentally different patterns of acute care access and outcomes for different clinical scenarios. For example, while the pandemic may have resulted in impeded access to outpatient treatment programs (OTPs) or office-based buprenorphine prescriptions that in turn resulted in poor OUD management and increased ED visitation, the pandemic may not have had as much relative effect on services for AUD that are less medication or healthcare facility dependent. It is likely that a shift of focus away from the opioid epidemic as systems sought to manage and mitigate the COVID-19 pandemic may have also expanded supply chain gaps, worsened contacts with healthcare providers necessary to ensure access to naloxone, potentially driving increases in opioid-associated fatalities and exacerbating inequities in access to care.”

3. Again in the Discussion section, the second part of the following sentence is not clear: “In fact, our analysis may underestimate the degree of increased SUD associated harms as EMS data suggests both increased transport re increased out-of-hospital overdose-related cardiac arrests, which may in turn have not resulted in ED visits among people fearful of seeking emergency or hospital care (7).” Do you mean that the overdosed people with cardiac arrest refused transportation to ED or hospital care because they were afraid of contracting COVID?

• We have added an additional reference regarding specifically deferred EMS transportation, and updated this to read and clarify:

“In fact, our analysis may underestimate the degree of increased SUD associated harms as EMS data during the pandemic suggest both increased transport refusals for a myriad of conditions as well as increased out-of-hospital overdose-related cardiac arrests, which may not be classified as opioid related and may have resulted in transport to the ED previous to the pandemic.”

4. “…increased trends of ED boarding (27)”: ED boarding may be an acceptable term among ED physicians, but it is not among non-ED readers.

• We have replaced this phrase to read “increased trends of extremely prolonged ED length of stay, particularly the “boarding” of ED patients in the ED awaiting inpatient bed assignment” and further edited the language to improve clarity.

5. Geographic differences: You cited the CDC’s provisional data on overdose death counts (Ahmad et al., 2021). The data show that many states in the West (CA, AZ, WA…) had higher increase in opioid and other drug overdose deaths in 2020. However, your Figure 3 data shows that the West region has the most muted increase of opioid and other drug related ED visits of all four regions. Does the difference suggest that fewer overdosed cases in the West were cared for at ED (i.e., EMT found the cases dead or DOA)? Or, is the difference attributable to that fact that your source of data are not representative of all EDs? Please refer to the following and provide some more discussion:

Baumgartner, J.C., and Radley, D.C., 2021. The spike in drug overdose deaths during the COVID-19 pandemic and policy options to move forward. https://www.commonwealthfund.org/blog/2021/spike-drug-overdose-deaths-during-covid-19-pandemic-and-policy-options-move-forward.

• This is an excellent point and an important part of contextualizing our results. First, we acknowledge that more muted increase in OUD and other substance use disorder in the West may represent a limitation of the sample. That said, California is the best-represented site in the analysis with 41 hospital-based EDs. Second, it may be that the disconnect between ED visitation for OUD and population-level mortality represents a failure to activate emergency medical services, with consequences in the immediate phase and for patients who might have benefited from the initiation of medication-assisted therapy during an ED encounter for OUD:

“Interestingly, amongst the four US census regions reported in Fig 3, the West exhibited the smallest increases in ED presentations for OUD in our analysis. While this may represent a limitation of our work and ED visitation in our sample may not be representative generally, California was the best represented state in our sample (with 41 hospital-based EDs). This disconnect between ED visitation and population-level mortality may be related failure to active or delayed activation of EMS services. Given the value of EDs as sites for the initiation of medication-assisted therapy for OUD, failure to initiate transport in cases of opioid or other overdose not only risks mortality in the acute phase but also risks failure to initiate treatment and avoid a subsequent overdose.”

6. One last question: ED visit counts not number of unique individuals, right?

• Correct, our analysis is of ED visits not patients as some patients may have repeated ED visits. However as our analysis seeks to describe trends and does not include any outcome at risk of bias from patient-level clustering, a visit level analysis was felt to be most appropriate.

Reviewer #2: The authors analyzed a large sample of emergency department (ED) visits identified in the Clinical Emergency Data Registry (from 258 sites across 34 states) spanning 2019-2020. Key findings from the authors’ analyses were that overall ED visits declined in the early pandemic period and had not returned to baseline by the end of 2020; however, ED visits for substance use disorder (SUD) declined at slower rates in March and April 2020, such that the proportion of overall ED visits that were for SUD increased in the earlier months of 2020. The authors also found that ED visits for SUD returned to baseline in the second half of 2020. The authors concluded their paper with a meaningful discussion about hospital-based SUD care in 2020 within the context of resource allocation and availability concerns experienced by many hospital systems during 2020. Overall, the trends presented in this paper are meaningful within the context of public health surveillance and continuing to learn how the COVID-19 pandemic may have disrupted substance use disorder treatment.

However, there are several concerns and questions that should be addressed. I offer the following comments and questions for the authors’ consideration. I will first summarize key issues I have with the paper as currently written. Additional minor comments, suggestions, and questions follow.

Thank you for the opportunity to review your paper.

COMMENTS AND RECOMMENDATIONS

KEY ISSUES

A. General

1. My biggest concern is that other studies published prior to the submission of this paper have conducted similar analyses and highlighted some similar results. For example, Pines et al. (2021) examined ED visit trends for SUD from Jan-July 2020 and also found that ED visits for SUD decreased sharply early on, declined less than overall ED visits, and eventually began to return to 2019 levels. Pines et al. also explored differences by patient demographic factors. In turn, Holland et al. (2021) examined over 3,000 EDs spanning 48 states to explore trends in drug-related overdose/OUD and MH ED visit rates from Dec. 2018 to October 2020. Holland et al. described increases in OD-related ED visit rates, similarly noting that visits did not decrease in a manner similar to other ED visits, suggesting an increase in OUD ED burden in 2020.

I am not certain about the journal or editors’ stance on the novelty of findings or studies that coincide with or replicate other studies. However, it’s an important part of the scientific process, and surely the pandemic warrants much inquiry and better continued learning about how COVID-19 affects care delivery in various settings. Perhaps one thing the authors could do is cite studies like Holland et al., Pines et al., and Lucero et al. and differentiate in their paper how their analytic sample differs from others, if there are meaningful differences (e.g., significant differences in community-based EDs identified in ACEP CEDR?), and then reposition the Discussion to focus on the experience of those types of EDs, patient populations, etc., while also making sure to correctly summarize the gaps in the scholarly literature that are being filled.

• We are sensitive to this concern and believe our work extends upon the prior cited work in several respects. First the Pines paper cited only captured data through July 2020, the Holland Paper through October 2020 and the Lucero paper through April 2020, which currently reflect only an incremental aspect of the initial pandemic. In this revision we include data through June 2021 demonstrating sustained changes throughout a dynamic pandemic, as well as heterogeneity in changes in visitation across OUD, alcohol-related visits, mental health visits, as compared to overall visits and visits for MI and stroke.

• Our paper also includes a more focused discussion and analysis of individual substance use disorders than some of the prior papers and includes a more heterogeneous mix of EDs than the paper of Pines and Lucero, both of which are based on the data of single contract management groups in EM. Therefore our findings may be more generalizable.

• We have updated the revision to not only include these citations but also discuss the importance of our study in extending these findings through June 30, 2021.

References:

Pines, J. M., Zocchi, M. S., Black, B. S., Carlson, J. N., Celedon, P., Moghtaderi, A., & Venkat, A. (2021). How emergency department visits for substance use disorders have evolved during the early COVID-19 pandemic. Journal of Substance Abuse Treatment, 108391.

Holland, K. M., Jones, C., Vivolo-Kantor, A. M., Idaikkadar, N., Zwald, M., Hoots, B., ... & Houry, D. (2021). Trends in US emergency department visits for mental health, overdose, and violence outcomes before and during the COVID-19 pandemic. JAMA Psychiatry, 78(4), 372-379.

Lucero, A. D., Lee, A., Hyun, J., Lee, C., Kahwaji, C., Miller, G., ... & Pan, L. (2020). Underutilization of the emergency department during the COVID-19 pandemic. Western Journal of Emergency Medicine, 21(6), 15.

2. The authors seem to suggest their data will not be made publicly available. PLOS journals require authors to make all data necessary to replicate their study’s findings publicly available without restriction at the time of publication. Authors do not need to submit their entire data set if only a portion of the data was used in the reported study. When specific legal or ethical restrictions prohibit public sharing of a data set, authors must indicate how others may obtain access to the data. For studies involving third-party data, PLOS encourage authors to share any data specific to their analyses that they can legally distribute. PLOS recognizes, however, that authors may be using third-party data they do not have the rights to share. When third-party data cannot be publicly shared, authors must provide all information necessary for interested researchers to apply to gain access to the data. I.e., https://journals.plos.org/plosone/s/data-availability.

• We addressed this above.

B. Introduction Section

1. As described in a later comment, I think the Introduction section nicely motivates the need for this type of analysis. However, it would be helpful if the authors directly stated the study objective in the second paragraph of the Introduction section. I understand you are just examining (important) trends and that the research question/study objective (and possibly any hypotheses) may be implicit. Although it is not necessary to test any explicit hypotheses in this type of study, given the journal’s broad scope and readership it would be helpful for the reader if you directly stated the study objective in relation to the gap you are ostensibly filling. E.g., perhaps open the second paragraph with an explicit statement of the objective, which would lead nicely into the rest of the paragraph supporting the rationale for that objective.

• We have made changes to the final paragraph of the Introduction to better frame the objective of the study:

“Given the gradual emergence of a ‘new normal’ and ongoing concerns about excess mortality both directly attributable to COVID-19 and related to SUDs, we aimed to provide an updated characterization of ED visitation for SUD in a national sample of hospital-based EDs and differentiate from prior work with a more granular exploration specifically identifying OUD, mental health visits, and comparing these with overall visits and visits for select emergent medical conditions.”

C. Methods Section

1. Did you use all ICD-10 codes within F10-F19 as inclusion criteria for SUD visits? Did you also include other relevant codes outside of that range, e.g., T40.-, G62.1, etc.? As supplemental content, could you include a table of all ICD-10 codes used to flag SUD and comparison conditions in this study, similar to Holland et al. (2021) and Pines et al. (2021)? This will be helpful for assessing/knowing what visits were included in the trend analyses, comparing to other studies, etc.

• In response to this thoughtful suggestion we have actually revised our methods to follow those of Pines et al. (2021) by making use of the Clinical Classification Software Revised codes for substance use as in their work now explained in the Methods section. This did add a small number of codes and visits to the sample, without a major effect on our overall findings

• We have included a supplemental table as requested by the reviewers for the diagnostic codes used to identify each group of visits in the analysis.

3. If I am interpreting this correctly, you summarized and described your analytic sample in sentences 2-4 of your Study Setting and Dataset (Methods) subsection. This is important information, but its placement in the Methods section is uncommon in health services research. These descriptive results would be better situated at the beginning of the Results section, helping the reader understand and assess your analytic sample characteristics leading into the main analyses (e.g., assessing generalizability and applicability of your analytic sample vis-a-vis achieving your study objective).

• We have moved these analytic sample descriptions to the Results Section as requested.

4. More information is needed about the regression modeling you performed. Just to confirm, the piecewise regression estimation process was simply used to identify the breakpoints, then you report the ED utilization in those identified periods? Could you provide a bit more information about the Poisson regression models used to estimate the IRRs comparing the 2020 periods to the 2019 reference periods? What variables went into the Poisson regression models? These were estimated for each period identified by the piecewise regression approach? At a minimum, more details are needed to assess what was conducted, though you could certainly include statistical notes as supplemental content to help the reader understand and replicate your analytic approach.

• We have added additional detail to the methods regarding the regression approach and clarifying that the piecewise regression estimation process was simply used to identify breakpoints to report ED utilization in those identified periods. The Poisson models were unadjusted, estimated at the level of the hospital-based EDs in the sample simply for the purpose of reporting incident rate ratios for visit counts comparing different time periods in 2020/2021 to 2019.

5. Tables and Figures: It would be helpful to include notes sections for each table and figure to briefly describe things like what the outcomes are/derived from which analytic method, number of observations, etc.

• We have updated this draft to includes footnotes for the Figures and Table to better clarify the results.

MINOR ISSUES AND COMMENTS �

1. Please paginate your manuscript file. It makes providing feedback easier, given the files compiled and presented to reviewers.

2. Introduction: The beginning of the Introduction section is effective. The authors effectively summarized relevant (preliminary) findings from the scholarly literature, CDC, etc. The authors also quickly pivoted to motivating their research objective in a convincing way, noting the convergence of several critical public health crises (i.e., the “epidemic within a pandemic”).

• We appreciate the positive feedback.

3. Methods, Study Design subsection: I would recommend saying you conducted an “observational study” instead of “observational analysis”, then describe your analytic methods in your Analysis section.

• Agreed, language adjusted as suggested.

4. Methods, Study Setting & Data subsection: PLOS encourages authors to cite any publicly available research data in their reference list, such as the U of Maryland COVID-19 Impact Analysis Platform. Please see: https://journals.plos.org/plosone/s/data-availability.

• Agreed, the citation has been added.

5. Results section: This section is well written, and key findings are discussed. However, the section is difficult to follow in relation to the tables and figures. I think you have indicated where you would like the placement of each table/figure. But I would recommend directly saying, “Table 1 shows….” or parenthetically citing the table/figure at the end of each sentence talking about the table/figure. E.g., “While visit rate for AUD… for the remainder of the year (Figure 2).

• Agreed, the results section has been edited to reflect this suggestion.

6. Do the x-axes on Figures 1 and S2 align? In Figure 1 it appears the 3-week moving average of ED visits overall begins to increase at the end 2020; however, that is not shown in Figure S2 (just what appears to be a new trend of decreasing moving averages). If different, does the increase (weeks 51-52?) affect the piecewise regression breakpoints?

• The Figures have been adjusted so that x-axis align. There is a decline in the last few weeks of 2020 exhibited in both Figure 1 and Supplemental Figure 2.

7. Results section: You say, “ED visits for SUD and mental health exhibited similar, however, more muted declines in ED visitation and were much closer to baseline than overall visits by in the second half of 2020.” However, my read of Figure 1 is that SUD and MH ED 3-week moving averages also experienced precipitous decreases vis-à-vis 2019. If I am interpreting that correctly, I am not sure “muted” is the right word or tone for the discussion of this finding. Perhaps consider including the exact percentages (per 2019 visits) for SUD and MH ED visits as you did for all ED visits in the previous sentence. That will help clarify the relative rate of decrease.

• Agreed, the results section has been edited to reflect this suggestion.

8. You say, “Compared to similar periods in 2019, ED visits for SUD and AUD were close to baseline in the final period (Nov. 11 to Dec. 31) of 2020 (IRR 0.88 [0.85-0.91] and IRR 0.87 [0.83-0.91], respectively), while visits for OUD were above baseline (IRR 1.07 [1.02-1.13]).” I see how that can be interpreted from the IRR estimates presented in Table 1. However, the error bars in Figure 2 suggest the 3-week moving averages for OUD ED visits vis-à-vis the 2019 averages were not greater (i.e., crossing over 100% baseline). Am I interpreting that correctly? Or were the estimates greater than baseline for the average of that entire final breakpoint period in 2020? This is where the table/figure notes sections and additional statistical modeling information could be useful.

• This is an excellent point for clarification. In our updated manuscript, viewed in aggregate and estimated with a simple unadjusted Poisson regression model for the periods for weeks 28-53 of 2020 and weeks 1-12 of 2021, OUD visits were above baseline. These differences, when comparing individual 3-week moving averages, were generally not statistically significantly different from 2019 levels as depicted in Figure 2.

9. Figure 3 is interesting, though it is noisy and a bit difficult to interpret with the error bars. It is certainly responsible to illustrate the ranges, but I wonder if it might make sense just to plot the means for better conveying the key takeaways.

• Agreed, for the purpose of readability Figure 3 no longer includes the error bars.

10. Discussion section: I think you meant to say “substantially higher rate” in the second sentence, not “substantial”.

• Fixed.

11. Table 1 is nice. Very intuitive presentation.

• We appreciate the positive feedback.

12. The Discussion of potential explanations for your trends findings was compelling.

• We appreciate the positive feedback.

Submitted filename: CEDR-OUD-PLOS-ONE-RESPONSE_TO_REVIEWERS_revision.docx

Click here for additional data file.

17 Dec 2021

Emergency Department Utilization for Substance Use Disorders and Mental Health Conditions During COVID-19

PONE-D-21-20023R1

Dear Dr. Venkatesh,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Fernando A. Wilson, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

- At the next available opportunity during production, please ensure that Supplemental Table 1 is uploaded, and insert "of" between "myriad" and "conditions" in Discussion.

Reviewers' comments:

5 Jan 2022

PONE-D-21-20023R1

Emergency Department Utilization for Substance Use Disorders and Mental Health Conditions During COVID-19

Dear Dr. Venkatesh:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Fernando A. Wilson

Academic Editor

PLOS ONE

Table 1

ED utilization for substance use disorder and mental health.

		2020				2021
		Weeks 1–10	Week 11–13	Weeks 14–27	Weeks 28–53	Weeks 1–12	Weeks 13–25
SUD	Weekly Average Counts	7977	6651	6162	6581	6865	6423
	Proportion	6.81%	7.64%	8.06%	7.20%	7.35%	6.50%
	IRR	1.16	0.95	0.85	0.92	0.99	0.94
	IRR 95% CI	1.12–1.2	0.79–1.14	0.81–0.89	0.89–0.95	0.98–1.01	0.92–0.96
Alcohol	Weekly Average Counts	2360	1964	2097	2194	2183	2142
	Proportion	2.02%	2.26%	2.74%	2.40%	2.34%	2.17%
	IRR	0.97	0.76	0.77	0.84	0.94	0.89
	IRR 95% CI	0.92–1.01	0.63–0.93	0.72–0.82	0.81–0.87	0.92–0.95	0.88–0.91
Opioid	Weekly Average Counts	619	517	579	607	610	595
	Proportion	0.53%	0.59%	0.76%	0.66%	0.65%	0.60%
	IRR	1.14	0.97	1.00	1.06	1.06	1.01
	IRR 95% CI	1.10–1.18	0.86–1.10	0.93–1.08	1.01–1.12	1.05–1.08	0.99–1.04
Other SUD-related	Weekly Average Counts	4998	4169	3486	3780	4072	3687
	Proportion	4.27%	4.79%	4.56%	4.14%	4.36%	3.73%
	IRR	1.29	1.06	0.88	0.95	1.03	0.96
	IRR 95% CI	1.25–1.33	0.88–1.29	0.84–0.92	0.92–0.98	1.01–1.04	0.94–0.98
Mental Health-related	Weekly Average Counts	5050	3981	4037	4350	4700	4427
	Proportion	4.31%	4.58%	5.28%	4.76%	5.03%	4.48%
	IRR	1.07	0.82	0.84	0.94	1.02	0.96
	IRR 95% CI	1.05–1.09	0.71–0.94	0.80–0.87	0.90–0.98	0.98–1.01	0.94–0.99
MI and Stroke	Weekly Average Counts	1166	896	966	1081	1160	1063
	Proportion	1.00%	1.03%	1.26%	1.18%	1.24%	1.08%
	IRR	0.97	0.75	0.86	0.97	0.98	0.97
	IRR 95% CI	0.94–1.00	0.70–0.80	0.83–0.90	0.95–0.99	0.97–1.00	0.95–0.99
Overall	Weekly Average Visits	117,064	87,011	76,457	91,414	93,446	98,776

Note: Sample includes 174 hospital-based EDs across 33 states from a national emergency department clinical quality registry. Visits for different diagnosis categories were identified by Clinical Classification Software Revised and International Classification of Disease-10 codes. Incident rate ratios (IRRs) are estimated with unadjusted Poisson regression models for hospital-based ED daily visit counts in each period as compared to the same weeks in 2019. SUD = substance use disorder, MI = myocardial infarction.