Disparities in Emergency Department Visits for Opioid and Stimulant Overdoses in Florida During COVID-19

Abstract

Emergency department (ED) visits for drug overdoses increased nationally during COVID-19 despite declines in all-cause ED visits. The study purpose was to compare characteristics of ED visits for opioid and stimulant overdoses before and during COVID-19 in Florida. This study tested for disparities in ED visits for opioid and stimulant overdoses by race/ethnicity, age, and insurance status. The study identified ED visits for opioid and stimulant overdose in Florida during quarters two and three of 2019 and compared them with quarters two and three of 2020. Overall, there was an increase in the number of opioid and stimulant overdoses during COVID-19. Combined with the decline in the number of all-cause ED visits, drug overdoses represented a larger share of ED visits during COVID-19 compared with before COVID-19. The study did not find evidence of disparities by race/ethnicity, as each group experienced similar increases in the likelihood of ED visits involving drug overdoses during COVID-19. Differences emerged according to age and insurance status. ED visits involving those under age 18 were more likely to involve opioid or stimulant overdose, and ED visits among those over age 65 were less likely to involve opioid overdose during COVID-19. ED visits among those with vulnerable insurance status were more likely to involve opioid overdose during COVID-19. Patterns of behavior change during periods of restricted activity due to a pandemic. These changes in behavior change the mix of risks that people face, suggesting the need for a reallocation of population health management resources during pandemics.

Introduction

National data have demonstrated that emergency department (ED) visits for drug overdoses increased after the implementation of COVID-19 restrictions, despite large declines in all-cause ED visits.^1
–3 Recent data from the Centers for Disease Control and Prevention (CDC) reported a record number of overdose deaths during the pandemic, with >100,000 deaths from April 2020 to April 2021.⁴ Vulnerable populations are potentially at risk for increased drug use during the pandemic shutdown, and this problem is potentially compounded by having drug treatment or prevention efforts interrupted by the pandemic.^5,6

Opioids constitute the majority of drug overdoses nationally.⁷ Although the opioid epidemic had historically been a problem mostly for non-Hispanic White populations, African American and Hispanic populations are increasingly at risk for opioid overdoses.⁸ The purpose of this study was to test for disparities by race/ethnicity, age, and insurance status in ED visits for opioid and stimulant overdoses during COVID-19.

Published studies have documented the increase in opioid and drug overdoses after the COVID-19 shutdown, but evidence is mixed on the differential impact of this increase on racial/ethnic minority populations. Friedman et al found a disproportionate increase in opioid-related deaths in California among African Americans, whereas Kelley et al found that opioid-related deaths in Los Angeles County occurred mostly among the non-Hispanic White population.^9,10

Results from single-center studies of ED visits for opioid overdoses during COVID-19 are also mixed, with some finding differential increases for non-Hispanic African American populations^11
–13 and others not observing differential impacts by race/ethnicity.¹⁴ No study, to the authors' knowledge, has considered other population dimensions along which disparities in drug overdoses during COVID-19 may have occurred.

This study adds to the literature by comparing characteristics of ED visits for opioid and stimulant overdoses before and during COVID in the state of Florida using a statewide hospital database. Florida is the third most populous state in the United States, has a racially diverse population,¹⁵ and is one of the states that declared a public health emergency due to the opioid epidemic.¹⁶ This study further adds to the literature on disparities in ED visits for drug overdoses during COVID-19 by considering ED visits for both opioid and stimulant drug overdoses. In addition, whereas most studies focus on disparities by race/ethnicity, this study also tests for disparities by age and insurance status.

The results of the study have implications for population health management strategies. Given the concerns over elevated drug use, coupled with interruptions to drug treatment options during the pandemic, it is important to determine which groups are at the highest risk for overdosing so that resources can be directed to these populations. Similarly, with activities being severely restricted in the first year of the COVID-19 pandemic, the mix of conditions for which patients present in the ED is likely to change. This suggests a need to reallocate population health management strategies and resources away from those events that are less likely to occur and toward those that are at elevated risk for occurring during a pandemic.

Methods

Data

Data for the study were obtained from Florida's Agency for Health Care Administration. The data set includes all ED visits in the state reported by 67 unique facilities. The period before COVID includes quarters two and three of 2019, and the COVID period includes quarters two and three of 2020. Owing to seasonality in ED data, it was important to compare the same quarters before COVID and during COVID. Quarters two and three of 2019 occurred before the first reports of an outbreak in Wuhan, China,¹⁷ whereas quarters two and three of 2020 occurred after Florida instituted its stay-at-home order¹⁸ and once lifted while activity remained relatively restricted, thereby allowing a valid comparison. The data set includes diagnosis codes and patient demographics. The unit of analysis is the ED visit. Only visits by Florida residents were included. The study was determined to be exempt by Nova Southeastern University's Institutional Review Board.

Variables

The primary outcome measures were ED visits for opioid overdose and stimulant overdose. Each outcome measure was coded as 0 or 1 if the visit included a diagnosis for opioid or stimulant overdose, respectively. Based on the CDC surveillance case definition guidance, opioid overdose in this study includes poisoning by opium (T40.0X), heroin (T40.1X), other opioids (T40.2X), methadone (T40.3X), synthetic narcotics (T40.4X), unspecified narcotics (T40.60), and other narcotics (T40.69); stimulant overdose consists of poisoning by cocaine (T40.5X), unspecified psychostimulants (T43.60), caffeine (T43.61), amphetamines (T43.62), methylphenidate (T43.63), ecstasy (T43.64), and other psychostimulants (T43.69).¹⁹ These were identified using ICD-10 codes.

Control variables included patient race/ethnicity (non-Hispanic White as the reference group, non-Hispanic Black, Hispanic, and other races), gender (males as the reference group), age (0–17, 18–34, 35–64 as the reference group, and 65 years old and above), insurance status (private as the reference group, Medicare, Medicaid, uninsured, and other), whether the visit occurred on the weekend (Saturday or Sunday) or after hours (5 pm–8 am), and the Charlson Comorbidity Index, generated based on listed diagnosis codes. The Charlson Comorbidity Index is a validated and commonly used summary measure of comorbidities.^20,21 Charges for opioid and stimulant overdose ED visits were computed. Charges from 2019 were converted to 2020 dollars using the medical care consumer price index.²²

Analyses

Multivariable logistic regressions were estimated separately for each outcome. To determine whether African American, Hispanic, or other race ED visits were more likely to involve an opioid or stimulant overdose during COVID-19 compared with the reference population, models included dummy variables for race/ethnicity, interactions between the race/ethnicity dummy variables and a COVID-19 dummy variable, and the set of control variables. Statistical significance on any of the interaction terms indicates an elevated likelihood of an ED visit being due to an overdose during COVID-19 relative to the non-Hispanic White reference group.

Similar analyses were performed to test for disparities by age and insurance status. Owing to the high correlation of the age 65 and over variable and Medicare coverage dummies, models included one but not both in the regressions. For regressions by age group, age 65 and over and its interaction with the COVID-19 indicator were used and Medicare was excluded from insurance type controls. For insurance type regressions, models included Medicare and Medicare interacting with the COVID-19 indicator and excluded age 65 and over from the age controls. All regressions included county and year-quarter fixed effects. Error terms were clustered at the facility level.

The study was approved by Nova Southeastern University's Institutional Review Board.

Results

Summary statistics

Summary statistics are presented in Table 1. During the time periods studied, a total of 7,126,308 ED visits by Floridians were included in the analyses. Less than half of the ED visits involved non-Hispanic White patients (48%), and more than half (57%) involved females. Ages 35–64 accounted for the largest share of ED visits (37%), and those over age 65 accounted for the smallest (17%). Forty eight percent of visits involved public insurance (28% Medicaid and 20% Medicare).

Table 1.

Summary Statistics (N = 7,126,308)

Variable	Proportion or mean	SD
Female	0.568	0.052
Non-Hispanic White	0.484	0.500
African American	0.254	0.435
Hispanic	0.233	0.422
Other races	0.029	0.169
Age 0–17	0.174	0.38
Age 18–34	0.283	0.45
Age 35–64	0.374	0.48
Age 65 and over	0.169	0.37
Charlson index^*	0.209	0.65
Privately insured	0.271	0.44
Medicaid	0.278	0.45
Medicare	0.204	0.40
Other insurance	0.078	0.27
Uninsured	0.169	0.37
Weekend visit	0.268	0.44
After hours visit	0.493	0.50
Quarter 2 2019	0.298	0.38
Quarter 3 2019	0.292	0.45
Quarter 2 2020	0.184	0.48
Quarter 3 2020	0.226	0.37
Charge per opioid overdose ED visit^**	$7702	$9100
Charge per stimulant overdose ED visit^**	$10,833	$10,545

Notes: Numbers in the “proportion” column are proportions in all cases except for the Charlson index^*, with mean of 0.209 and a range from 0 to 14. ^**Charges include data from all four quarters, converted to 2020 dollars.

ED, emergency department; SD, standard deviation.

As seen in national trends, the total number of visits declined sharply during COVID-19, with 60% of the total visits (in quarter two and three of 2019 and 2020) occurring in 2019 compared with only 40% occurring in 2020. The average charges for opioid and stimulant overdoses were $7702 and $10,833, respectively. In the pre-COVID-19 period, there were 8561 opioid overdoses, compared with 10,895 during, for an increase of 2334 overdoses. The number of stimulant overdoses was relatively unchanged (1451 before COVID and 1432 during), but given the large reductions in all-cause ED visits, stimulant overdoses made up a larger share of ED visits during COVID-19 compared with before.

Disparities by race/ethnicity

The non-Hispanic White population had the highest unadjusted rate of overdose per 1000 ED visits (Table 2). Opioid overdose was far more common than stimulant overdose for each group. All groups experienced an increase in the rate of overdose per 1000 visits during COVID-19, nearly doubling for each group.

Table 2.

Unadjusted Rates of Emergency Department (ED) Visits for Opioid and Stimulant Overdose per 1000 All-Cause ED Visits

Group	Quarter	Rate of opioid overdose visits	Rate of stimulant overdose visits
White	Before COVID-19	3.39	0.46
White	During COVID-19	5.96	0.63
African American	Before COVID-19	0.60	0.26
African American	During COVID-19	1.13	0.36
Hispanic	Before COVID-19	0.98	0.21
Hispanic	During COVID-19	1.80	0.33
Age 0–17	Before COVID-19	0.11	0.19
Age 0–17	During COVID-19	0.29	0.34
Age 18–34	Before COVID-19	3.10	0.60
Age 18–34	During COVID-19	5.08	0.73
Age 35–64	Before COVID-19	2.73	0.39
Age 35–64	During COVID-19	4.91	0.55
Over age 65	Before COVID-19	1.16	0.03
Over age 65	During COVID-19	1.32	0.05
Private insurance	Before COVID-19	1.14	0.26
Private insurance	During COVID-19	1.92	0.37
Medicare	Before COVID-19	1.64	0.12
Medicare	During COVID-19	2.16	0.17
Medicaid	Before COVID-19	0.97	0.30
Medicaid	During COVID-19	2.28	0.46
Uninsured	Before COVID-19	5.63	0.81
Uninsured	During COVID-19	10.89	1.13

Notes: The table contains unadjusted rates of opioid and stimulant overdose ED visits per 1000 all-cause ED visits. The data contain 7,126,308 all-cause ED visits.

In logistic regressions controlling for patient and visit characteristics, the coefficients of the group × COVID-19 interaction terms were not statistically significant, suggesting that none of the groups was differentially impacted by COVID-19 (Table 3). The increases in overdoses per 1000 ED visits during COVID-19 was common to all of the groups.

Table 3.

Multivariable Logistic Regressions: Disparities in Opioid and Stimulant Emergency Department Visits Before and During COVID-19 by Race/Ethnicity

	Opioid overdose			Stimulant overdose
Variable	Coefficient	P	OR	Coefficient	P	OR
Female	−0.612	<0.001	0.542 (0.516, 0.570)	−0.659	<0.001	0.512 (0.476, 0.563)
African American	−1.905	<0.001	0.149 (0.126, 0.175)	−0.714	<0.001	0.490 (0.416, 0.575)
Hispanic	−1.159	<0.001	0.314 (0.272, 0.362)	−0.764	<0.001	0.466 (0.387, 0.561)
Other races	−0.830	<0.001	0.436 (0.362, 0.525)	−0.594	0.001	0.552 (0.388, 0.786)
Age 0–17	−2.754	<0.001	0.064 (0.053, 0.076)	−0.554	<0.001	0.575 (0.483, 0.683)
Age 18–34	0.095	0.001	1.100 (1.037, 1.166)	0.356	<0.001	1.428 (1.292, 1.579)
Age 65+	−0.026	<0.001	0.472 (0.426, 0.524)	0.067	<0.001	0.111 (0.085, 0.146)
Charlson	0.476	0.207	0.974 (0.935, 1.015)	0.375	0.132	1.069 (0.980, 1.167)
Medicaid	0.501	<0.001	1.610 (1.492, 1.736)	0.276	<0.001	1.456 (1.279, 1.656)
Other insurance	1.398	<0.001	1.650 (1.443, 1.886)	0.941	0.001	1.317 (1.122, 1.547)
Uninsured	0.166	<0.001	4.046 (3.782, 4.328)	0.201	<0.001	2.561 (2.282, 2.875)
Weekend	0.556	<0.001	1.181 (1.146, 1.217)	0.466	<0.001	1.223 (1.132, 1.320)
After hours	0.021	<0.001	1.744 (1.684, 1.807)	−0.025	<0.001	1.593 (1.469, 1.729)
African American during COVID-19	−0.058	0.757	1.021 (0.895, 1.165)	0.071	0.799	0.976 (0.806, 1.181)
Hispanic during COVID-19	−0.118	0.317	0.944 (0.843, 1.057)	0.041	0.527	1.074 (0.862, 1.338)
Other races during COVID-19	−0.612	0.235	0.889 (0.731, 1.808)	−0.659	0.870	1.042 (0.637, 1.704)

Notes: Regressions include county and year-quarter fixed effects. Error terms were clustered at facility level. Odds ratios (ORs) include the point estimate and 95% confidence interval. The total number of all-cause ED visits was 7,126,308. P-values <0.05 are considered statistically significant.

Overdoses were less likely to involve females, Hispanic patients, African American patients, patients of other races, and people under age 18 or over 65. Overdoses were more likely to involve non-Hispanic White patients, people ages 18–64, and those without private insurance. ED visits among the uninsured were four times as likely to involve opioid overdose relative to the privately insured reference group (odds ratio [OR] 4.052, P < 0.001) and more than twice as likely to involve stimulant overdose (OR 2.560, P < 0.001). Overdoses were more likely to occur on weekends and after hours. Similar results were observed for both opioid overdoses and stimulant overdoses.

Disparities by age

Unadjusted rates of overdose per 1000 ED visits were highest in the 18–34 and 35–64 age groups (Table 2). As in the preceding analysis, opioid overdose was more common than stimulant overdose, with the exception being the age 17 and under group where stimulant overdose was more common. All groups experienced an increase in the rate of overdose per 1000 visits during COVID-19, with the 17 and under age group and ages 35–64 age group experiencing the largest proportional increases during COVID-19. In logistic regressions controlling for patient and visit characteristics, the coefficients on the group × COVID-19 interaction terms confirm the differential effect for these groups (Table 4).

Table 4.

Multivariable Logistic Regressions: Disparities in Opioid and Stimulant Emergency Department Visits Before and During COVID by Age Group

Opioid overdose				Stimulant overdose
Variable	Coefficient	P	OR	Coefficient	P	OR
Female	−0.613	<0.001	0.542 (0.516, 0.570)	−0.660	<0.001	0.517 (0.475, 0.562)
African American	−1.893	<0.001	0.151 (0.133, 0.171)	−0.725	<0.001	0.484 (0.422, 0.556)
Hispanic	−1.191	<0.001	0.304 (0.269, 0.343)	−0.726	<0.001	0.484 (0.421, 0.556)
Other races	−0.895	<0.001	0.408 (0.353, 0.473)	−0.573	<0.001	0.564 (0.441, 0.721)
Age 0–17	−2.898	<0.001	0.055 (0.043, 0.070)	−0.658	<0.001	0.518 (0.419, 0.639)
Age 18–34	0.137	<0.001	1.147 (1.067, 1.234)	0.430	<0.001	1.538 (1.370, 1.727)
Age 65+	−0.483	<0.001	0.617 (0.548, 0.694)	−2.219	<0.001	0.109 (0.072, 0.164)
Charlson	−0.026	0.217	0.975 (0.935, 1.015)	0.067	0.130	1.070 (0.980, 1.167)
Medicaid	0.477	<0.001	1.611 (1.494, 1.738)	0.377	<0.001	1.457 (1.281, 1.658)
Other insurance	0.501	<0.001	1.650 (1.443, 1.886)	0.277	0.001	1.319 (1.123, 1.549)
Uninsured	1.399	<0.001	4.052 (3.789, 4.335)	0.940	<0.001	2.560 (2.281, 2.874)
Weekend	0.166	<0.001	1.181 (1.145, 1.217)	0.201	<0.001	1.223 (1.132, 1.320)
After hours	0.557	<0.001	1.745 (1.685, 1.807)	0.466	<0.001	1.593 (1.469, 1.728)
Age 0–17 during COVID-19	0.310	0.035	1.364 (1.021, 1.821)	0.255	0.050	1.290 (1.000, 1.664)
Age 18–34 during COVID-19	−0.074	0.040	0.929 (0.865, 0.997)	−0.149	0.029	0.862 (0.754, 0.985)
Age 65+ during COVID-19	−0.514	<0.001	0.598 (0.524, 0.683)	0.041	0.889	1.041 (0.590, 1.839)

Notes: Regressions include county and year-quarter fixed effects. Error terms were clustered at facility level. Odds ratios (ORs) include the point estimate and 95 percent confidence intervals. The total number of all-cause ED visits was 7,126,308. P-values <0.05 are considered statistically significant.

ED visits were statistically significantly less likely to involve overdoses for the age 18–34 group during COVID-19, relative to the reference group (ORs for opioid and stimulant overdoses 0.929, P = 0.040, and 0.862, P < 0.029, respectively). ED visits were significantly less likely to involve opioid overdoses during COVID-19 for those over age 65 relative to the reference group (OR 0.598, P < 0.001). For those age 17 and under, ED visits were significantly more likely to involve overdose during COVID, compared with the reference group (ORs for opioid and stimulant overdose were 1.364, P = 0.035, and 1.290, P = 0.05, respectively).

The signs and magnitude of the coefficients on control variables were similar to the preceding analysis of racial/ethnic disparities. Similar results were observed for both opioid overdoses and stimulant overdoses, with the one exception being that there was not a statistically significant difference in stimulant overdoses during COVID-19 for those over age 65, as stimulant overdoses were not common for this group before or during COVID-19.

Disparities by insurance status

Unadjusted rates of overdose per 1000 ED visits were most common among those with uninsured (Table 2). Opioid overdose was again far more common than stimulant overdose for each group. All groups experienced an increase in the rate of overdose per 1000 visits during COVID-19, with Medicaid patients experiencing the largest proportional increases during COVID-19. In logistic regressions controlling for patient and visit characteristics, the coefficients on the group × COVID-19 interaction terms suggest differential impacts of COVID-19 by insurance status for opioid overdoses (Table 5).

Table 5.

Multivariable Logistic Regressions: Disparities in Opioid and Stimulant Emergency Department Visits Before and During COVID by Insurance Status

	Opioid overdose			Stimulant overdose
Variable	Coefficient	P	OR	Coefficient	P	OR
Female	−0.610	<0.001	0.543 (0.517, 0.571)	−0.656	<0.001	0.519 (0.477, 0.564)
African American	−1.875	<0.001	0.153 (0.135, 0.174)	−0.697	<0.001	0.433 (0.433, 0.573)
Hispanic	−1.172	<0.001	0.310 (0.274, 0.350)	−0.712	<0.001	0.491 (0.427, 0.564)
Other races	−0.879	<0.001	0.415 (0.358, 0.481)	−0.569	<0.001	0.566 (0.442, 0.724)
Age 0–17	−2.665	<0.001	0.070 (0.058, 0.0.083)	−0.485	<0.001	0.616 (0.516, 0.735)
Age 18–34	0.177	<0.001	1.194 (1.124, 1.270)	0.432	<0.001	1.540 (1.388,1.708)
Charlson	−0.079	<0.001	0.924 (0.885, 0.970)	0.034	0.474	1.035 (0.942, 1.137)
Medicaid	0.604	<0.001	1.830 (1.635, 2.048)	0.420	<0.001	1.522 (1.271, 1.822)
Medicare	0.173	0.001	1.189 (1.543, 2.141)	−0.782	<0.001	0.457 (0.360, 0.581)
Other insurance	0.598	<0.001	1.818 (1.544, 2.141)	0.256	0.023	1.292 (1.036, 1.610)
Uninsured	1.540	<0.001	4.663 (4.227, 5.144)	1.016	<0.001	2.763 (2.335, 3.268)
Weekend	0.165	<0.001	1.1790 (1.144, 1.215)	0.200	<0.001	1.222 (1.132, 1.320)
After hours	0.571	<0.001	1.770 (1.709, 1.833)	0.481	<0.001	1.618 (1.491, 1.755)
Medicaid during COVID	0.152	0.007	1.164 (1.043, 1.300)	0.008	0.945	1.008 (0.809, 1.254)
Medicare during COVID	−0.233	<0.001	0.792 (0.710,0.884)	−0.028	0.860	0.972 (0.708, 1.334)
Other insurance during COVID	0.139	0.072	1.149 (0.987, 1.338)	0.029	0.838	1.029 (0.782, 1.353)
Uninsured during COVID	0.132	0.006	1.141 (1.037, 1.255)	−0.042	0.680	1.618 (1.491, 1.755)

ED visits were statistically significantly less likely to involve opioid overdose for those on Medicare during COVID-19, relative to the reference group (OR 0.792). ED visits for those on Medicaid or uninsured were significantly more likely to include opioid overdose, relative to the privately insured reference group during COVID-19 (ORs 1.164, P = 0.007, and 1.141, P = 0.006, respectively). The signs and magnitude of the coefficients on control variables were similar to the preceding analyses. No differential impacts were detected for stimulant overdoses during COVID-19.

Discussion

This study tested for disparities in the likelihood of ED visits involving opioid or stimulant overdoses during COVID-19. Consistent with national data, there was an across-the-board increase in drug overdoses during COVID-19. Also consistent with national data, the non-Hispanic White population had the highest share of ED visits that involved overdoses. In terms of insurance status, in models controlling for patient characteristics, ED visits among those who were uninsured were four times as likely to involve an overdose, suggesting that low socioeconomic status and reduced access to drug prevention and treatment are a problematic combination.

The study did not find evidence of differential impacts during COVID-19 in overdoses by race/ethnicity. Overdoses made up a larger share of all-cause ED visits for all groups, suggesting an overall need to increase prevention and treatment resources during pandemics. However, differences emerged by age group. ORs suggest that ED visits were 36.4% more likely to involve opioid overdose and 29.0% more likely to involve stimulant overdose during COVID for those age 17 and under, relative to the reference group. Post hoc analysis of the age 17 and under group revealed the majority of overdoses occurred in older youths, ages 11–17, with ∼85% of opioid overdoses and 63% of stimulant doses occurring in this age group.

This finding suggests the need to devote attention, resources, and programming to middle and high school age youths at risk for substance use.²³ ED visits were ∼40% less likely to involve opioid overdose among those over age 65 relative to the reference group. ED visits among those with vulnerable insurance status were more likely to involve opioid overdose during COVID-19. For Medicaid and uninsured patients, ED visits were 16.4% and 14.1% more likely to involve opioid overdose, respectively, relative to the privately insured reference group.

The study had limitations. As an observational study of hospital ED visits, causality cannot be inferred from the findings. It is possible that the changes observed during COVID-19 were a continuation of pre-existing trends. However, national data have shown a spike in drug overdoses during COVID-19, so it is likely that time trends do not account for all of the changes observed in the data. The study included data from Florida and, therefore, may not generalize to other parts of the country.

Despite this potential threat to external validity, Florida is the third largest state in the United States in terms of population, and the sample yielded more than seven million ED visits during the time period studied. Features of the study design mitigate the impact of those who may reside in Florida only in the winter months. First, the sample is limited to Florida residents. This eliminates concerns of a differential mix of seasonal residents before and during COVID-19 and the reduced tourism that occurred during COVID-19. Second, the same two quarters of the year were compared before and during COVID-19 to mitigate the effects of seasonal factors.

The results of the study have implications for those managing high-risk populations during pandemics. The proportion of ED visits involving drug overdoses nearly doubled during COVID, and this increase was not specific to any particular racial or ethnic group. The largest proportional changes in ED visits involving overdoses were seen among Medicaid and uninsured patients, suggesting that socioeconomic status and access to treatment services contribute to health disparities and may be especially problematic for these populations during a pandemic.

Disproportionately large increases in the likelihood of ED visit rates involving overdoses occurred in those age 17 and under. As patterns of behavior change during periods of restricted activity due to a pandemic, so do changes occur in the mix of risks that people face. Overdoses were found to comprise a larger share of all-cause ED visits for these populations, suggesting municipal and state lawmakers should appropriate additional resources toward prevention and treatment for those at high risk for drug use or abuse during pandemics. Although disparities are often studied along racial/ethnic dimensions, the results of this study underscore the importance of examining disparities along other population dimensions.

The authors recommend leveraging interventions that have been developed to educate and support children who are at high risk for substance and opioid misuse, while instituting enhanced and more frequent monitoring of those children. In addition, the provision of virtual mental and behavioral health services through telehealth should be prioritized at the local and state level, as well as in-home treatment options for those populations most at risk for developing substance use problems or relapsing. Future study should extend the modeling to include community-level factors as predictors of elevate risk of overdose so that accurate resource allocation models can be developed for decision support.

There was a large increase in the raw number of opioid overdoses during COVID-19. Although not all of the increase can be attributed to COVID-19, the approximate additional 2500 overdoses observed in the data result in approximate additional health care charges of nearly $19 million during a 6-month period. These additional charges were no doubt borne by a variety of public and private insurers, and some may have resulted from care for which facilities were never compensated. Thus, there is also a business case to be made for targeted management of those at risk for overdosing during pandemics.

Footnotes

Authors' Contributions

T.F.P. made substantial contributions to conception and design, literature review, drafting the article for critically important intellectual content, and final approval of the version to be published. W.C. made substantial contributions to conception and design, data analysis, and final approval of the version to be published. J.A.J. and F.S. made substantial contributions to conception and design, drafting the article for critically important intellectual content, and final approval of the version to be published.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was received for this article.

References

Slavova

, Rock

, Bush

, Quesinberry

, Walsh

. Signal of increased opioid overdose during COVID-19 from emergency medical services data. Drug Alcohol Depen, 2020; 214:108176.

Soares III

, Melnick

, Nath

, et al. Emergency department visits for nonfatal opioid overdose during the COVID-19 pandemic across 6 US healthcare systems. Ann Emerg Med, 2022; 79:158–167.

Friedman

, Akre

COVID-19 and the drug overdose crisis: uncovering the deadliest months in the United States, January–July 2020. Am J Public Health, 2021:e1–e8.

Center for Disease Control and Prevention National Center for Health Statistics. Drug Overdose Deaths in the U.S. Top 100,000 Annually, 2021. https://www.cdc.gov/nchs/pressroom/nchs_press_releases/2021/20211117.htm Accessed November 28, 2021 .

Johnson

, Tiako

, Flash

, Lamas

, Alba

. Disparities in the recovery from critical illness due to COVID-19. Lancet Psychiat, 2020; 7:e54–e55.

Rosales

, Janssen

, Yermash

, et al. Persons from racial and ethnic minority groups receiving medication for opioid use disorder experienced increased difficulty accessing harm reduction services during COVID-19. J Subst Abuse Treat, 2021; 108648.

Centers for Disease Control and Prevention. Drug Overdose Deaths, 2020. https://www.cdc.gov/drugoverdose/deaths/index.html Accessed November 12, 2021 .

Shiels

, Freedman

, Thomas

, Berrington de Gonzalez

. Trends in US drug overdose deaths in non-Hispanic black, Hispanic, and non-Hispanic white persons, 2000–2015. Ann Intern Med, 2018; 168:453–455.

Friedman

, Hansen

, Bluthenthal

, Harawa

, Jordan

, Beletsky

. Growing racial/ethnic disparities in overdose mortality before and during the COVID-19 pandemic in California. Prev Med, 2021; 153:106845.

10.

Kelley

, Lucas

, Stewart

, Goldman

, Doctor

. Opioid-related deaths before and after COVID-19 stay-at-home orders in Los Angeles County. Drug Alcohol Depen, 2021; 228:109028.

11.

Ochalek

, Cumpston

, Wills

, Gal

, Moeller

. Nonfatal opioid overdoses at an urban emergency department during the COVID-19 pandemic. JAMA, 2020; 324:1673–1674.

12.

Khatri

, Pizzicato

, Viner

, et al. Racial/ethnic disparities in unintentional fatal and nonfatal emergency medical services–attended opioid overdoses during the COVID-19 pandemic in Philadelphia. JAMA Netw, 2021; 4:e2034878.

13.

Patel

, Walter

, Li

. Opioid overdose crises during the COVID-19 pandemic: implication of health disparities. Harm Reduct J, 2021; 18:89.

14.

Glober

, Mohler

, Huynh

, et al. Impact of COVID-19 pandemic on drug overdoses in Indianapolis. J Urban Health, 2020; 97:802–807.

15.

United States Census Bureau. U.S and World Population Clock. https://www.census.gov/popclock/ Accessed November 14, 2021 .

16.

National Governors Association. Using emergency declarations to address the opioid epidemic, 2018. https://www.nga.org/center/publications/using-emergency-declarations-to-address-the-opioidepidemic/ Accessed November 14, 2021 .

17.

World Health Organization. WHO's COVID-19 Response. https://www.who.int/emergencies/diseases/novel-coronavirus-2019/interactive-timeline Accessed November 14, 2021 .

18.

Desantis

https://www.flgov.com/wp-content/uploads/orders/2020/EO_20-91-compressed.pdf Accessed November 14, 2021 .

19.

Vivolo-Kantor

, Pasalic

, Liu

, Martinez

, Gladden

. Defining indicators for drug overdose emergency department visits and hospitalisations in ICD-10-CM coded discharge data. Inj Prev, 2021; 27(Suppl. 1):i56–i61.

20.

Austin

, Wong

Y-N

, Uzzo

, Beck

, Egleston

. Why summary comorbidity measures such as the Charlson Comorbidity Index and Elixhauser score work. Med Care, 2015; 53:e65–e72.

21.

Quan

, Li

, Couris

, et al. Updating and validating the Charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol, 2011; 173:676–682.

22.

United States Bureau of Labor

Statistics

. Measuring Price Change in the CPI: Medical care. https://www.bls.gov/cpi/factsheets/medical-care.htm Accessed November 14, 2021 .

23.

Jacko

, Cohen

, Appadoo

, et al. A call to action: preventing opioid and substance abuse in South Florida Youth. HCA Healthc J Med, 2021; 2:1.