COVID-19 immunisation among individuals with opioid use disorder in Ontario: a population-based cohort study

Abstract

Background

The COVID-19 pandemic exacerbated health inequities, particularly among individuals with opioid use disorder (OUD). Disparities in vaccine uptake among people with OUD remain poorly understood. This study assessed COVID-19 immunisation rates among individuals with OUD compared with the general population in Ontario, Canada.

Methods

This population-based retrospective cohort study used linked administrative health data to compare COVID-19 vaccination rates between individuals diagnosed with OUD and a 10% random sample of individuals without OUD. Ontario residents aged >15 years with continuous healthcare coverage as of the censor date, 16 March 2020, were included. Inverse Probability of Treatment Weighting (IPTW) was applied to balance confounders, and Cox proportional hazards models estimated adjusted HRs (aHRs) for receiving two and three or more vaccine doses.

Results

The cohort included 105 733 individuals with OUD and 1 185 993 without OUD. Individuals with OUD had a lower hazard of receiving two vaccine doses (aHR: 0.75, 95% CI 0.73 to 0.76) and three or more doses (aHR: 0.69, 95% CI 0.67 to 0.70). The rate of two-dose and three-dose vaccination was also lower among those with OUD (115.3 vs 149.0 per 100 000 person-years and 44.7 vs 77.5 per 100 000 person-years).

Conclusion

Individuals with OUD had lower COVID-19 vaccination rates, suggesting barriers to access and uptake. Addressing these disparities through targeted interventions is crucial for equitable public health responses.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Individuals with opioid use disorder (OUD) face substantial health inequities, including barriers to accessing healthcare. Previous studies have suggested that marginalised populations may have lower COVID-19 vaccination rates, but data on individuals with OUD are limited.

WHAT DOES THIS STUDY ADD

• This study provides population-based evidence that individuals with OUD in Ontario had lower COVID-19 immunisation rates compared with the general population, despite controlling for confounders.

HOW DOES THIS STUDY AFFECT RESEARCH, PRACTICE OR POLICY

• These findings underscore the need for tailored vaccine outreach strategies, including integrating vaccination with harm reduction and substance use treatment services. Public health policies should prioritise equitable vaccine distribution to reduce disparities in emergency and routine immunisation.

Introduction and background

The COVID-19 pandemic exposed and worsened health inequities among people with opioid use disorder (OUD).¹ While widespread immunisation efforts were instrumental in reducing the burden of COVID-19, uncertainty remains regarding vaccine access and uptake among equity-seeking populations, such as people with substance use disorders.²

Individuals with OUD face significant social and structural barriers to healthcare, including stigma, unstable housing and incarceration.3,5 In a variety of jurisdictions, observational studies and knowledge syntheses demonstrate that people with OUD are at heightened risk of morbidity and mortality due to other preventable diseases amenable to vaccinations, such as diphtheria, tetanus and pertussis, tuberculosis, hepatitis A, hepatitis B and influenza.6,10 Understanding these disparities can help inform equitable public health policies, including vaccine distribution, and help to plan for future emergencies.

To alleviate COVID-19-related harms, Ontario implemented an extensive COVID-19 vaccination campaign, prioritising at-risk populations. However, the extent to which these efforts reached individuals with OUD remains unclear. Prior research indicates that marginalised groups, including people experiencing homelessness, often encounter logistical barriers to healthcare, such as difficulty scheduling appointments, transportation issues and simply not knowing where to go.¹¹ Vaccine hesitancy driven by mistrust in the healthcare system, past negative experiences and misinformation may have contributed to lower vaccination rates in this population.^{11 12}

This study aims to evaluate whether OUD in Ontario was associated with differential rates of COVID-19 immunisation compared with the general population.

Methods

We conducted a population-based retrospective cohort study of Ontario residents aged over 15 years with at least 5 years of continuous Ontario Health Insurance Plan (OHIP) coverage as of 16 March 2020, when Ontario declared a state of emergency related to COVID-19 (index date). OHIP is Ontario’s publicly funded health coverage provided to all Canadian citizens, permanent residents and certain refugee claimants resident in Ontario. The start of the pandemic was chosen as the index date, as this date provided us with the most comprehensive set of data on vaccination, including data for those eligible for early vaccine rollout. We used administrative data housed at ICES, including the Canadian Institute for Health Information’s Discharge Abstract Database, the National Ambulatory Care Reporting System and the Ontario Mental Health Reporting System, for acute inpatient hospital admission details, emergency department visits and mental health-related hospitalisations, respectively. ICES is an independent, non-profit research institute whose legal status under Ontario’s health information privacy law allows it to collect and analyse healthcare and demographic data, without consent, for health system evaluation and improvement. The OHIP claims that database was used to identify outpatient claims for physician visits, while the Registered Persons Database (RPDB) was used to identify demographic and vital statistics information. These datasets were linked using unique encoded identifiers and analysed at ICES. The use of the data in this project is authorised under section 45 of Ontario’s Personal Health Information Protection Act (PHIPA) and does not require review by a Research Ethics Board.

The exposure was OUD diagnosis within 3 years prior and including the index date, with outpatient visits identified through OHIP billing codes for OUD assessments and management, consistent with similar studies (Appendix 1).¹³ Inverse probability of treatment weighting (IPTW) was used to balance the characteristics of individuals with OUD versus a 10% sample of individuals without an OUD diagnosis based on the following covariates at the index date: age, sex, neighbourhood income quintile, urban versus rural location, primary care attachment, inpatient admission in the past year, emergency department visit in the past year, recent history of homelessness, Ontario marginalisation indices (ON-Marg), history of asthma, chronic obstructive pulmonary disease, HIV and quartile of the total number of the Johns Hopkins ACG System (V.10) Aggregated Diagnosis Groups (Appendix 1) . The ON-Marg index, a tool developed to measure and understand the levels of marginalisation in different communities across Ontario, was also used to balance exposed and unexposed groups using four domains: material income quintiles were used to reflect access to basic material needs, educations attainment and unemployment; household dwelling quintiles were used to reflect housing stability and living environment; racialised and newcomer population quintiles were used to reflect social and cultural marginalisation based on immigrants, minority groups and language barriers; and age and labour force quintiles were used to reflect demographic composition and workforce engagement.¹⁴

The primary outcome was time to full COVID-19 immunisation, defined as receiving two vaccine doses on or after the index date. The secondary outcome was time to receiving three or more doses. We calculated the rate of vaccine uptake per 100 000 person-years and fit Cox proportional hazards models to generate HRs and IPTW-adjusted HRs (aHRs) using a type-1 error threshold of 5%. Individuals were censored on vaccination, death or at the end of the study period (16 March 2022).

Results

People with OUD were younger, more frequently male, had more comorbidities, greater healthcare utilisation and more indicators of marginalisation (based on the domains of ON-Marg) than the sample of people without OUD (table 1). IPTW procedures appropriately controlled for these differences, including healthcare utilisation patterns, creating comparable groups with respect to known confounders (online supplemental appendix A and table A). Among 105 733 with OUD and 1 185 993 individuals without OUD, 69 160 (115.3 per 100 000 person-years) and 930 513 (149.0 per 100 000 person-years) received two COVID-19 vaccinations, respectively (table 2).

Table 1. Characteristics of OUD and non-OUD groups prior to IPTW.

	OUD N=105 733	Non-OUD N=1 185 993	Standardised difference
Mean age (SD)	41.98 (13)	48.34 (19.4)	0.386
Median age (IQR)	40 (32–51)	48 (32–63)

	N (%)	N (%)
Sex
Male	65 857 (62.3)	577 410 (48.7)	0.276
Female	39 876 (37.7)	608 583 (51.3)
Neighbourhood income quintile
1	39 726 (37.6)	230 497 (19.4)	0.41
2	23 428 (22.2)	237 821 (20.1)	0.052
3	17 540 (16.6)	235 444 (19.9)	0.085
4	13 424 (12.7)	236 726 (20)	0.197
5	11 615 (11.1)	245 505 (20.7)	0.268
Urban/rural
Urban	91 028 (86.1)	1 064 579 (89.8)	0.113
Rural	14 705 (13.9)	121 414 (10.2)
Primary care attachment
Yes	97 070 (91.8)	1 041 947 (87.9)	0.131
Rostered	49 306 (46.6)	881 921 (74.4)
CHC contact	3877 (3.7)	9838 (0.8)
Virtually rostered	43 887 (41.5)	150 188 (12.7)
No	8663 (8.2)	144 046 (12.2)
Sum of ADGs quartile
1	8023 (7.6)	294 082 (24.8)	0.48
2	18 839 (17.8)	248 613 (21)	0.08
3	30 302 (28.7)	332 831 (28.1)	0.013
4	48 569 (45.9)	310 467 (26.2)	0.42
Inpatient admission in the past year
Yes	22 274 (21.1)	102 304 (8.6)	0.355
No	83 459 (78.9)	1 083 689 (91.4)
ED visit in the past year
Yes	72 713 (68.8)	396 965 (33.5)	0.755
No	33 020 (31.2)	789 028 (66.5)
Recent history of homelessness
Yes	9473 (9)	2087 (0.2)
No	96 260 (91)	1 183 906 (99.8)
Asthma
Yes	26 361 (24.9)	188 377 (15.9)	0.226
No	79 372 (75.1)	997 616 (84.1)
COPD
Yes	15 185 (14.4)	89 100 (7.5)	0.118
No	90 548 (85.6)	1 096 893 (92.5)
HIV
Yes	1127 (1.1)	1753 (0.2)	0.118
No	104 606 (98.9)	1 184 240 (99.9)

ON-Marg Index
Material resources quintile
1	11 450 (10.8)	243 251 (20.5)	0.269
2	14 498 (13.7)	256 187 (21.6)	0.208
3	16 897 (16)	242 100 (20.4)	0.115
4	21 469 (20.3)	216 641 (18.3)	0.052
5	37 218 (35.2)	221 940 (18.7)	0.378
Unknown	4201 (4)	5874 (0.5)	0.237
Households and dwellings income quintile
1	9258 (8.8)	258 401 (21.8)	0.368
2	12 895 (12.2)	221 188 (18.7)	0.179
3	16 980 (16.1)	214 123 (18.1)	0.053
4	24 984 (23.6)	219 885 (18.5)	0.125
5	37 415 (35.4)	266 488 (22.5)	0.288
Unknown	4201 (4)	5874 (0.5)	0.237
Racialised and newcomer populations quintile
1	21 071 (19.9)	188 086 (15.9)	0.106
2	21 197 (20.1)	203 372 (17.5)	0.075
3	22 045 (20.9)	212 612 (17.9)	0.074
4	20 458 (19.4)	253 297 (21.4)	0.05
5	16 761 (15.9)	322 752 (27.2)	0.279
Unknown	4201 (4)	5874 (0.5)	0.237
Age and labour force quintile
1	21 413 (20.3)	303 917 (25.6)	0.128
2	20 705 (19.6)	236 486 (19.9)	0.009
3	19 608 (18.5)	210 174 (17.7)	0.021
4	18 444 (17.4)	204 775 (17.3)	0.005
5	21 362 (20.2)	224 767 (19)	0.032
Unknown	4201 (4)	5874 (0.5)	0.237
COVID-19 vaccination
Received dose 1	76 005 (71.9)	947 230 (79.9)	0.187
Median (IQR) days to dose 1	435 (401–487)	412 (387–437)	0.485
Received dose 2	69 160 (65.4)	930 513 (78.5)	0.294
Median (IQR) days to dose 2	488 (468–543)	471 (458–490)	0.460
Received dose 3	32 666 (30.9)	626 729 (52.8)	0.456
Median (IQR) days to dose 3	661 (646–674)	653 (641–667)	0.287

ADG, aggregated diagnosis group; CHC, Community Health Centre; COPD, chronic obstructive pulmonary disease; ED, emergency department; IPTW, inverse probability of treatment weighting; IQR, Interquartile Range; ON-Marg, Ontario marginalisation; OUD, opioid use disorder; SD, Standard deviation.

Table 2. OUD and associated risk of COVID-19 immunisation with two doses (vs<2 doses) and 3+doses (vs<3 doses), adjusted for IPTW.

	OUD		Non-OUD
Outcome	Number of events	Rate (95% CI) of events/100 000 person-years	Number of events	Rate (95% CI) of events/100 000 person-years	Unadjusted HR (95% CI)	IPTW HR (95% CI)*
2 vaccines	69 160	115.3 (114.4 to 116.1)	930 513	149.0 (148.7 to 149.3)	0.62 (0.62 to 0.63)	0.75 (0.73 to 0.76)
3+vaccines	32 666	44.7 (44.2 to 45.2)	626 729	77.5 (77.3 to 77.6)	0.49 (0.49 to 0.50)	0.69 (0.67 to 0.70)

^*Using inverse probability of treatment weighting with stabilised average treatment effect weights.

CI, Confidence Interval; IPTW, inverse probability of treatment weighting; OUD, opiod use disorder.

OUD was associated with a lower unadjusted hazard of receiving two COVID-19 vaccine doses (HR 0.62, 95% CI 0.62 to 0.63), or receiving three or more COVID-19 vaccine doses (0.49, 95% CI 0.49 to 0.50). This effect was retained but attenuated when adjusting for known confounders, with an IPTW-adjusted hazard of receiving two COVID-19 vaccine doses of 0.75 (95% CI 0.73 to 0.76) and an IPTW-adjusted HR of 0.69 for three or more vaccine doses (0.67 to 0.70) (table 2).

Discussion

By leveraging comprehensive administrative health data, this study provides a population-based assessment of vaccination disparities, controlling for key sociodemographic and health-related factors associated with lower vaccine uptake. Disparities in COVID-19 immunisation rates between individuals with OUD and Ontario’s general population suggest that vaccine distribution was less effective and timely among people with OUD. Lower vaccination rates may be due to factors such as barriers to accessing healthcare, lack of healthcare continuity, stigma, distrust or mistrust of vaccination, being unhoused and competing priorities, such as managing addiction and daily survival.^{5 9 13} Some evidence suggests that lower vaccination rates among people with OUD contributed to higher mortality throughout the pandemic.¹⁵ This study suggests similar disparities in access to this important public health intervention in Canada and underscores needs for strategic interventions that support vaccine uptake in this population in future public health emergencies. Our data show equivalent primary care and acute care interactions among people with OUD than without; therefore, limited vaccine uptake among people with OUD is not likely attributable to reduced healthcare interactions. There is, therefore, an urgent need for strategic interventions for people with OUD, including focusing on vaccine access and addressing barriers for marginalised communities in future vaccine rollouts. Integrating vaccination efforts with routine healthcare visits, substance use treatment programmes and harm reduction services may improve vaccine uptake.¹⁶

A strength of this study is its use of population-based data, capturing vaccination rates of most Ontarians diagnosed with OUD during the COVID-19 state of emergency. Unmeasured confounders may still influence findings. Additionally, future research might investigate whether interventions to treat OUD modify the effect of these findings or alter vaccine uptake patterns among people with OUD. This study did not examine immunisation of diseases other than COVID-19 in people with OUD. However, our findings align with the existing reviews showing limited vaccine uptake among people who use drugs.³ People without OHIP coverage may also be over-represented among those with OUD but not captured by ICES data.

Overall, equitable vaccine strategies are needed to reduce disparities, prevent disproportionate harms from vaccine-preventable diseases among individuals with OUD and serve as precedent for inclusive public health efforts.

Parts of this material are based on the data and information compiled and provided by MOH and Canadian Institute for Health Information. We thank IQVIA Solutions Canada Inc. for use of their Drug Information File and the Toronto Community Health Profiles Partnership for providing access to the Ontario Marginalisation Index. This document also used data adapted from the Statistics Canada Postal CodeOM Conversion File, which is based on the data licensed from Canada Post Corporation and/or data adapted from the Ontario Ministry of Health Postal Code Conversion File, which contains data copied under licence from Canada Post Corporation and Statistics Canada. The analyses, conclusions, opinions and statements expressed herein are solely those of the authors and do not reflect those of the funding or data sources; no endorsement is intended or should be inferred.