Increased Opioid Exposure in Older Adults During COVID‐19: An Interrupted Times Series Analysis

Summary

Why does this matter?

Increases in opioid exposures during the COVID‐19 pandemic highlight the importance of providing medication administration support for older adults aged 65 and older as well as the importance of caregiving and appropriate opioid prescribing.

• This interrupted time series analysis examined opioid exposures among adults aged 55 and older whose exposures were reported to any U.S. Poison Centers from 2016 to 2021.

• Results showed significant increases in opioid exposures for licit formulations of opioids among Americans aged 65 and older following the onset of the COVID‐19 pandemic.

• Increases in opioid exposures were primarily related to medication errors among Americans aged 65 and older.

1. Introduction

Older adults in the United States have been dramatically impacted by the opioid overdose crisis. Over the past two decades, rates of opioid‐related overdose deaths and opioid use disorder have risen dramatically among Americans aged 50 and older, particularly among Americans enrolled in Medicare— the federal insurance program for Americans aged 65 or older in the United States [1, 2, 3]. In 2021, over 50,000 Medicare beneficiaries experienced a fatal or nonfatal opioid‐related overdose [3]. Further, over 12 million Medicare Part D beneficiaries received an opioid prescription in 2021 and roughly 200,000 beneficiaries received high amounts of opioids, defined as an average morphine equivalent dose of more than 120 milligrams per day for at least 3 months [3]. Notably, the COVID‐19 pandemic exacerbated the opioid overdose crisis, with deaths related to opioids exceeding 80,000 in 2021 [4].

Additionally, older adults experienced increased loneliness and social isolation during the COVID‐19 pandemic [5]. Both loneliness and social isolation are associated with negative physical and mental health outcomes among older adults including increased risk of mortality, depression, and opioid use disorder as well as reductions in cognitive function and overall quality of life [5, 6, 7, 8, 9, 10, 11].

While the COVID‐19 pandemic exacerbated the overdose crisis, few studies have examined how the pandemic impacted opioid‐related harms among older adults [11, 12]. To address this gap in the literature, this study examines changes in opioid exposures among older adults during the COVID‐19 pandemic.

2. Methods

2.1. Study Sample and Data

This study includes 47,599 opioid exposures among adults aged 55 and older whose exposure was reported to any U.S. Poison Centers from 2016 to 2021 (National Poisoning Data System, NPDS) [13]. Opioid exposures are cases reported to U.S. Poison Help Centers by individuals or health care facilities (i.e., hospital‐based patient care unit or emergency department, free‐standing emergency medical clinic, first aid station, physician's office, or clinic) which involve suspected opioid‐related poisonings, whether accidental, intentional, or of unknown intent, including negative reactions to medications taken as prescribed. Data include all reported exposures to oxycodone, hydrocodone, fentanyl, and heroin, grouped into licit and illicit formulations. Licit versus illicit classifications are based on formulations, thus all pharmaceuticals are considered licit formulations. Further, misuse of a licit formulation is captured by a measure indicating the reason for exposure (e.g., a patient may have the exposure substance reported as “hydrocodone,” but the reason reported may be “intentional – misuse”). All formulations of oxycodone, hydrocodone, and fentanyl, including those co‐formulated with other drugs, are represented in the data. Data include broad reasons for exposure (intentional, unintentional, or other) and specific reasons for exposure, within each broad category (ex. “suspected suicide” is a subcategory of “intentional”). Reasons for exposure are reported by callers and coded by Specialists in Poison Information (SPIs) according to definitions outlined by America's Poison Centers. Definitions for type of formulation and reasons for exposure are published by NPDS [14]. See Supplementary Figure 1 for a diagram illustrating how NDPS data are collected. Estimates of the state population are taken from the Area Health Resource File (2016–2021).

We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines for observational studies. The University of Georgia Institutional Review Board approved this study and waived the need for informed consent due to the deidentified nature of the data.

2.2. Measures

Our outcome variables measured the rate of opioid exposure per 100,000 population, stratified by age (55–64; 65 and older), opioid type (licit and illicit), and exposure reason (medication error, intentional use/misuse, and suspected suicide attempt). All confirmed non‐exposures (i.e., cases initially thought to be an exposure which were later determined not to be an exposure) were excluded from the analyses.

2.3. Statistical Analysis

We calculated descriptive statistics for all variables (See Supplementary Tables 1–4 for descriptive statistics). We used a panel interrupted time series (ITS) analysis to assess the impact of COVID‐19 on opioid exposures. Analyses were stratified by age group (55 to 64; 65 and older), opioid type (licit and illicit), and exposure reason (medication error, intentional use/misuse, and suspected suicide attempt). Opioid exposures were aggregated to the state‐quarter level. Opioid exposures before March 1, 2020 (the beginning of the COVID‐19 pandemic) were compared to opioid exposures after March 1, 2020. To account for the impact of COVID‐19 across all states (i.e., there was no control panel because all individuals experienced COVID‐19), we used a fixed effects regression model where the outcome (opioid exposure) was regressed on time, the post‐COVID indicator, and the interaction term with year fixed effects and standard errors clustered at the state level. This model accounts for both the pre‐existing time trend and the immediate changes following the COVID‐19 pandemic, controlling for unobserved heterogeneity at the state level and year‐specific shocks.

To assess autocorrelation in our panel ITS models, we visually inspected residual plots and autocorrelation function (ACF) plots. All statistical tests were two‐sided. Full model diagnostics and sensitivity checks are provided in the Supporting information.

Data for this study were obtained in 2023; analyses were conducted between May 30, 2023, and April 25, 2025, using Stata 18.

3. Results

Of the 47,599 opioid exposures among older adults aged 55 and older, approximately 91.4% involved licit opioid exposures (n = 43,509) and 58.6% of total reported opioid exposures occurred among females [See Supporting information Table 1]. ITS results showed that reports of any opioid exposure in individuals aged 65 and older increased by 0.031 reports per 100,000, following the March 2020 COVID‐19 lockdown (slope difference: 0.031 [95% CI: 0.011, 0.052], p = 0.003) (Table 1; Figure 1). Reports of opioid exposures in individuals aged 65 and older increased significantly by 0.030 reports per 100,000 for licit formulations, following the onset of the COVID‐19 pandemic (slope difference: 0.030 [95% CI: 0.107, 0.049], p = 0.002). However, we did not find a statistically significant increase in illicit opioid exposures for individuals in this age group. We also found no change in opioid exposures among Americans aged 55 to 64 following the onset of the COVID‐19 pandemic.

Table 1.

Results of interrupted time series analysis.

Exposures per 100,000 population		βtime*post (SE)	95% CI	p value
Aged 65 and older	All opioids	0.031 (0.010)	0.011–0.052	0.003
	Licit opioids	0.030 (0.009)	0.107–0.049	0.002
	Illicit opioids	0.001 (0.002)	−0.003 to 0.006	0.559
	Medication errors	0.015 (0.005)	0.004–0.026	0.008
	Suspected suicide attempt	0.003 (0.005)	−0.006 to 0.012	0.503
	Intentional use or misuse	0.004 (0.003)	−0.002 to 0.011	0.186
Aged 55–64	All opioids	0.027 (0.017)	−0.006 to 0.061	0.113
	Licit opioids	0.018 (0.011)	−0.003 to 0.040	0.098
	Illicit opioids	0.009 (0.011)	−0.013 to 0.031	0.415

Note: Bold values indicate statistical significant.

Figure 1.

Comparison of Opioid Exposure Rate for Adults Aged 65 and Older: Pre‐ and Post‐COVID‐19 Onset in March 2020. shows the average rate (per 100,000 population) of individuals aged 65 and older reporting an opioid exposure in a given state‐quarter (top left), rate (per 100,000 population) of individuals aged 65 and older reporting an opioid exposure of a licit opioid formulation in a given quarter (top right), rate (per 100,000 population) of individuals aged 65 and older reporting an opioid exposure of an illicit opioid formulation in a given quarter (bottom left), and rate (per 100,000 population) of individuals aged 65 and older reporting an opioid exposure and listing the reason for exposure as “medication error” in a given quarter (bottom right). The solid diagonal lines and dashed lines represent the forecasted values from the time‐series models and the associated 95% CIs, respectively. The vertical black line corresponds with the onset of the COVID‐19 pandemic in March of 2020.

ITS models examining reason for opioid exposure among Americans aged 65 and older (medication errors, intentional use/misuse, and suspected suicide attempt) showed an increase of 0.015 reports due to medication errors per 100,000, following the March 2020 COVID‐19 lockdown (slope difference: 0.015 [95% CI: 0.004, 0.026], p = 0.008). Results showed no significant change in reported opioid exposures due to suspected suicide attempts or intentional use/misuse. See Supporting information Table 5 for complete ITS results.

4. Discussion

Our study found significant increases in reported opioid exposures among Americans aged 65 and older following the onset of the COVID‐19 pandemic. More specifically, we found increases in opioid exposures for licit opioid formulations (vs. illicit opioid formulations) and that increases in opioid exposures were primarily related to medication errors among older adults aged 65 and older. Our results are consistent with other studies finding that the pandemic was associated with increased opioid‐related harms [15, 16, 17]. While exposures related to intentional use/misuse and suspected suicide attempts increased during the pandemic, these increases were not statistically significant. This is an important area for future research as other studies have found increases in loneliness and social isolation among older adults during the COVID‐19 pandemic [5], both of which may increase the risk of mortality and depression [5, 6, 7, 8, 9, 10, 11].

Medication errors among older adults may be related to disorientation or confusion, increases in self‐administration of medications due to decreased in‐person medical services during COVID‐19, and/or disruptions to in‐person caregiving associated with the COVID‐19 pandemic. For example, older adults face a higher risk of delirium from opioid use than younger adults, and are also more likely to be co‐prescribed medications which could exacerbate negative cognitive effects [18, 19]. Furthermore, the pandemic interrupted nonpharmacological pain interventions such as physical therapy services, which may have contributed to increased opioid prescribing as a substitute for nonpharmacological interventions [20].

This study has several limitations. First, our study includes a limited number of opioid formulations. However, oxycodone and hydrocodone are the most frequently used and diverted prescription opioids, and heroin and fentanyl are the most commonly misused street formulations [21, 22]. Relatedly, our study does not include opioid formulations used in the treatment of opioid use disorder. Second, NPDS captures only a subset of annual opioid exposures. NDPS is likely biased toward less severe or accidental cases and likely underrepresents cases involving illicit substances and exposures resulting in fatalities [23, 24]. However, NPDS is the only national data set that captures exposures regardless of healthcare facility utilization. Further, the national distribution and consistent use of NPDS over the past two decades make the data advantageous for conducting analyses of opioid exposures over time. Third, our analyses do not account for other time‐varying factors that may have impacted opioid exposure patterns during the study period, including changes in healthcare access, changes in individual behavior during lockdowns, or differential implementation of state‐level drug policies.

5. Conclusions

Overall, our findings suggest an increased need to provide medication administration support for older adults aged 65 and older, particularly in times of disruptions to healthcare. Our results also highlight the importance of caregiving and appropriate opioid prescribing for this age group. Taken together with increases in opioid‐related overdose deaths and opioid use disorder, additional efforts are likely needed to address the ongoing opioid overdose crisis among older adults in the United States.

Author Contributions

Amanda J. Abraham: conceptualization, funding acquisition, methodology, writing – original draft, writing – review and editing. Shelby R. Steuart: conceptualization, data curation, formal analysis, funding acquisition, methodology, writing – original draft, writing – review and editing. Samantha J. Harris: conceptualization, writing – original draft, writing – review and editing. Victoria Bethel: conceptualization, writing – original draft, writing – review and editing.

Disclosure

America's Poison Centers® (The Association) maintains the National Poison Data System® (NPDS), which houses deidentified records of self‐reported information from callers to the country's Poison Centers (PCs). NPDS data do not reflect the entire universe of U.S. exposures and incidences related to any substance(s). Exposures do not necessarily represent a poisoning or overdose and The Association is not able to completely verify the accuracy of every report. NPDS data do not necessarily reflect the opinions of The Association.

Conflicts of Interest

The authors declare no conflicts of interest.

Sponsor's Role

All authors have read and approved the final version of the manuscript. Dr. Steuart had full access to all of the data in this study and takes complete responsibility for the integrity of the data and the accuracy of the data analysis.

Impact Statement

• 1.We certify that this study is novel.
• 2.This study adds to the literature by examining how the COVID‐19 pandemic impacted opioid‐related harms among older adults.

Transparency Statement

The lead author Amanda J. Abraham affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.

Supporting information

Supplementary Figure 1: America's Poison Centers Call Triage Flowchart. Supplementary Table 1: Total Reported Opioid Exposures by Reason for Exposure and Formulation, 2016‐2021. Supplementary Table 2: Total Reported Opioid Exposures Pre‐ and During‐COVID‐19. Supplementary Table 3: Average Number of Opioid Exposures by State, Quarter, Stratified by Age Group, Formulation, and Reason for Exposure, 2016‐2021. Supplementary Table 4: Average Opioid Exposure Rates per 100,000, by Age Group, Formulation, and Reason for Exposure, 2016‐2021. Supplementary Table 5: Results of Interrupted Time Series Analysis. Supplementary Table 6: Durbin Watson Test Estimates. Supplementary Table 7: Autocorrelation Function Plots.

Abraham A. J., Steuart S. R., Harris S. J., and Bethel V., “Increased Opioid Exposure in Older Adults During COVID‐19: An Interrupted Times Series Analysis,” Health Science Reports 8 (2025): 1‐5, 10.1002/hsr2.71607.

Data Availability Statement

Data sharing is not allowed per the authors' DUA with the America's Poison Centers®.

The authors confirm that the data supporting the findings of this study are available within the article [and/or] its supporting materials.