COVID-19 vaccine uptake among Ontario physicians: a descriptive population-based retrospective cohort study

Introduction

Vaccines against SARS-CoV-2 have been crucial in countering the COVID-19 pandemic. Current vaccines are highly effective in preventing severe infections that lead to hospitalisation and death.1 Due to the reduced likelihood of becoming infected, vaccinated individuals have a lower probability of contributing to ongoing transmission of SARS-CoV-2.1 Over 13 billion COVID-19 vaccines have now been administered globally, with more than 97 million doses given in Canada.2 Despite this, there are limited data on the vaccination status of Canadian physicians. Vaccine hesitancy, which is the delay in acceptance or refusal of safe vaccines despite their availability,3 has been noted in healthcare workers during the COVID-19 pandemic.4–6

Health Canada authorised the use of Pfizer-BioNTech’s BNT162b2 vaccine on 9 December 20207 and the first COVID-19 vaccines were administered in Ontario on 14 December 2020.8 Later authorised vaccines include Moderna’s mRNA-1273 on 23 December 2020,9 AstraZeneca’s ChAdOx1 on 26 February 202110 and Janssen’s (Johnson & Johnson) JNJ-78436735 on 5 March 202111 (although delivery of Janssen did not begin until 8 November 202112). Front-line healthcare workers were among the first groups eligible to be vaccinated.13 During the first 6 months, vaccination was largely voluntary for healthcare workers, and uptake depended on available supply. However, on 31 May 2021, the Ontario government mandated COVID-19 vaccination for healthcare workers in long-term care homes.14 All long-term care staff were to be fully immunised (ie, primary series/two doses), and each staff member must do one of the following: provide proof of vaccination of each dose, give a documented medical reason for not being vaccinated, or participate in an educational programme about benefits of vaccination and the risks of being unvaccinated. On 16 July 2021, a news release from the Ontario Medical Association (OMA) stated that all healthcare workers should receive the vaccine.15 On 17 August 2021, the Ontario government issued Directive #6 mandating hospitals and home and community care providers to have a COVID-19 vaccination policy for all staff by 7 September 2021.16 The requirements are the same as the previous mandate for long-term care workers, and as a result, hospitals in Ontario announced their own vaccination policies.17 Hospitals initially required all physicians and staff to be vaccinated or regularly tested for the virus; stricter policies were later implemented, where unvaccinated employees were placed on leave or faced termination.18 Eventually, by 14 March 2022, all mandates were terminated.19

Vaccine uptake refers to the number of people vaccinated with a dose of a vaccine over a certain period of time.20 Healthcare providers’ recommendation of vaccinations is associated with increased vaccine uptake by the public.21 Therefore, physicians’ willingness to be vaccinated could influence and strengthen trust in the COVID-19 vaccines. Factors leading to increased vaccine acceptance in healthcare workers include male gender, older age, history of receiving seasonal influenza vaccinations in preceding years and occupation as a physician.22 Physicians were also more likely to accept COVID-19 vaccination than other health professions.22 In a news release by the OMA in July 2021, it was announced that the OMA had conducted a survey of Ontario physicians, in which 98% of respondents indicated they had both doses of a COVID-19 vaccine.15 However, details of the survey were not provided in the news release and are also not available elsewhere online. Therefore, it is unclear exactly how many Ontario physicians were vaccinated, and whether vaccination trends varied by physician characteristics that might correlate with determinants of vaccine uptake. Therefore, our study examined trends in COVID-19 vaccine uptake among Ontario physicians from 14 December 2020 to 13 February 2022. Our primary outcomes were the proportions of physicians who were recorded to have received at least one, at least two and three doses of a Health Canada-approved COVID-19 vaccine by our study end date. Secondary outcomes were how uptake varied by physician characteristics (including age, sex, specialty and residential location) and time elapsed between doses. Measuring time elapsed between doses would allow us to determine whether governmental policies or supply-chain issues had any influence on physicians receiving their vaccines.

Methods

We conducted a descriptive, population-based retrospective cohort study of physicians in Ontario, Canada using linked administrative healthcare databases held at ICES (formerly known as the Institute for Clinical Evaluative Sciences, ices.on.ca). The study period was from 14 December 2020 (the date of the first vaccine administered in Ontario8) to 13 February 2022. We studied all physicians in Ontario (including postgraduate trainees) who were registered with the College of Physicians and Surgeons of Ontario (CPSO) as of 14 December 2020. To practise medicine in Ontario, physicians must register with the CPSO, thus by using registration data, we aimed to limit non-response bias. We excluded out-of-province physicians using data (place of residence was obtained from the Registered Persons Database (RPDB)), physicians who had not accessed Ontario Health Insurance Plan (OHIP)-insured services for their own care for ≥5 years and physicians who had missing identifiers (eg, age or sex). Missing data are presented and analysed independently using the same methods as other categories. We structured our report based on the Strengthening the Reporting of Observational Studies in Epidemiology guidelines (please see online supplemental appendix 1).23

Supplemental material

Patient or public involvement

Patients or the public were not involved in the design, or conduct, or reporting, or dissemination plans of our research.

Data sources

ICES is an independent, non-profit research institute whose data repository consists of administrative health services records of Ontarians eligible for universal health coverage (ie, OHIP) since 1986. The records are organised into various databases, which can be used for research to guide health system evaluation and improvement. Due to data-sharing agreements, datasets within ICES are not publicly available. As part of privacy protection, with ICES datasets, direct personal identifiers (ie, name and health card number) are removed and replaced with a unique encoded identifier. This unique encoded identifier is used to link across the different ICES databases, so that relevant information regarding the individual can be gathered from one database to another. In brief, through ICES, demographics and vital status were obtained from the RPDB, and data on physician service claims were obtained from the OHIP Claims Database. Physicians’ registration information and practice specialty were obtained from the CPSO via an existing data-sharing agreement with ICES. On the other hand, data on COVID-19 vaccine administration, including dates and vaccine type, were collected from the Ontario Ministry of Health’s province-wide COVID-19 Vaccination Database (COVaxON). All analyses of the linked datasets were completed at ICES. The authors did not have access to information that could identify individual participants during or after data collection.

Variables

Physicians’ residential locations were reported as the Local Health Integration Network regions, as defined by the Ontario Ministry of Health,24 that are associated with the postal codes of their home addresses as recorded in RPDB. Using the same postal code data from RPDB, we also classified whether physicians reside in an urban or rural area: if the postal code was linked to a geographical location with a population less than 10 000, the residential setting was classified as rural (if data were missing, the physician was categorised as such). This definition corresponds to that of Statistics Canada, which considers rural as the population living in areas outside the commuting zones of large urban centres (populations of 10 000 or more).25 We categorised physicians by specialty based on the ‘primary specialty’ data from the CPSO database, which also contained data on subspecialty training. Lastly, we determined whether physicians provided care in a long-term care home by the OHIP claim location associated with their billings. Details on the ICES databases and study variables used in this study are provided in online supplemental appendices 2–4.

COVID-19 vaccine uptake

We examined uptake of Health Canada-approved vaccines among Ontario physicians between 14 December 2020 and 13 February 2022. To limit self-report bias, data on vaccine uptake were obtained directly from the Ontario Ministry of Health’s COVID-19 Vaccination Database (where the vaccination record is entered in real time by the personnel administering the vaccine). We report the number and proportion of physicians who had received at least one, at least two, and three doses of a COVID-19 vaccine, as well as those who had received none, by the study end date. In physicians who had received at least two, or three doses, we report the number of days elapsed between each dose.

Statistical analyses

Study variables are presented as frequencies and proportions, means and SDs, or medians and IQRs, as appropriate. Cumulative probabilities of physicians receiving their first, second and third doses were estimated with Kaplan-Meier curves and were plotted over time from 14 December 2020 until 13 February 2022. No formal statistical tests were conducted, and no CIs were calculated. All statistical analyses were performed using SAS V.9.4 (SAS Institute).

Results

43 630 physicians were registered with the CPSO as of 14 December 2020. 1962 physicians were excluded due to data cleaning exclusions (ie, missing or invalid age or sex, permanent residence outside Ontario, or record of death on or before 14 December 2020) and 401 physicians were excluded due to OHIP ineligibility. Thus, our final study cohort consisted of 41 267 physicians (figure 1). The mean age was 47 years, 56.0% were male, 46.1% practised family medicine and 6% provided care in a long-term care home (table 1). Between 14 December 2020 and 13 February 2022, 39 359 (95.4%) physicians received at least one dose of a Health Canada-approved COVID-19 vaccine, 39 148 (94.9%) received at least two doses and 35 834 (86.8%) received three doses; 1908 (4.6%) received no doses. The cumulative probabilities of physicians receiving their first, second and third doses over time are shown in figure 2.

Figure 1
Figure 1

Flow diagram outlining how the cohort was established. *Registered with the College of Physicians and Surgeons of Ontario. †In the Registered Persons Database, the province where a person’s permanent address is located is encoded by the first two digits of the variable ‘prcddablk’ and 35=Ontario. If data for this variable were missing or if the first two digits were not 35, the physician was considered to be a non-Ontario resident. In addition, physicians with no registered healthcare services in the prior 5 years were considered to be non-Ontario residents. ‡Record of death in the Registered Persons Database. OHIP, Ontario Health Insurance Plan.

Figure 2
Figure 2

Kaplan-Meier curves estimating the cumulative probabilities of physicians in Ontario receiving their first, second and third doses of a COVID-19 vaccine during our study period. HCWs, healthcare workers; LTC, long-term care; NACI, National Advisory Committee on Immunization.

Table 1

COVID-19 vaccination status in 41 267 physicians in Ontario as of 13 February 2022

Regarding physician characteristics, the proportion of physicians who received three doses was 87.2% among females and 86.6% among males. By age, the proportion was 90.4% among those aged ≥60 years and 81.2–89.5% among the other age groups. By specialty, the proportion that received three doses was 88.7% among family physicians, 89% among specialists and 88.2% among trainees. By Local Health Integration Network (LHIN) regions, the proportion that received three doses was 85% in Toronto Central (largest by population), 88.1% in North West (smallest by population) and 82.6–91.1% among other regions. As for residential setting, the proportion that received three doses was 90.1% among physicians residing in a rural area, and 86.7% among those residing in an urban centre. Of 2480 physicians who worked in long-term care, 2319 (93.5%) had received three doses.

In terms of the time elapsed between vaccine doses, the median number of days between the first and second doses was 39 (IQR 28–88) and it was 253 (IQR 183–285) between the second and third doses.

Discussion

In this study, we found that by 13 February 2022, 94.9% of Ontario physicians had received at least two doses (primary series) of a Health Canada-approved COVID-19 vaccine, and 86.8% of Ontario physicians had received three doses (primary two-dose series plus a booster). These figures are comparatively higher than the general population in Ontario, in which 78.2% of those eligible had received two doses and only 45.6% had received three doses by our study end date.26 Of physicians who had received three doses, the proportions were comparable between family physicians (88.7%), specialists (89%) and trainees (88.2%). Those aged 60 years or older had the highest proportion (90.4%), while those aged 30–39 years had the lowest (81.2%). Early in the pandemic, the evidence suggested higher infection severity and mortality with older age, and this could have influenced these vaccination trends observed.27 28

With regard to geographical trends, 85% of physicians who resided in Toronto Central, the largest LHIN by population, had received three doses of a COVID-19 vaccine by our study end date. In comparison, most other LHINs had similar, if not greater, proportions of physicians who received three doses. One notable outlier was the Central West LHIN, which had the lowest proportion with three doses (82.6%). Of all the LHINs, Central West has the highest population growth rate and the lowest percentage of population who are seniors (age 65 years and older).29 Therefore, we speculate that physicians residing in this LHIN were likely to be younger and perhaps were attracted to this area of significant growth for job opportunities. Considering that the average age of resident physicians in Canada is just over 30 years,30 and that length of residency ranges from 2 to 6 years,31 newly certified physicians are most likely between the ages of 30 and 39 years. This age group, as shown in our data, had the lowest proportion with three doses of a COVID-19 vaccine. As for residential setting, physicians residing in a rural location (90.1%) seemingly had a higher proportion of having had three doses than urban physicians (86.7%). This is observed despite rural communities having less access to vaccines, exemplified in part by the lower availability of trained pharmacists to provide vaccinations.32 Canadian seniors residing in rural areas have been shown to have lower rates of vaccine intake.33 Given that seniors are at increased risk of infection, we speculate whether rural physicians had increased vaccine uptake to promote similar behaviours in their local communities. Media reports in rural communities have suggested that the presence of healthcare providers at pop-up vaccination clinics likely increased vaccine uptake.34

As shown in figure 2, nearly 94% of all Ontario physicians had received their first dose before the vaccine mandate for long-term care workers was announced on 31 May 2021, and 90% had completed their primary series before the implementation of mandates by Ontario hospitals that fall. Considering the geographical trends observed, altogether this would suggest that vaccine uptake in Ontario physicians was likely self-directed, and occurred irrespective of vaccination mandates, residential location or setting. However, as no formal statistical tests were performed in our study, significant differences between the compared groups cannot be determined, and comparisons are subjective.

The OMA survey, which reported that 98% of surveyed Ontario physicians self-reported to have already received both doses of a COVID-19 vaccine, was the first to report physician-specific information.15 However, the findings of the survey were presented with a single sentence through a news release by the OMA, and additional information regarding the survey is not available online. Comparing our findings on vaccination uptake with other Canadian provinces, data from a preprint study conducted between 1 March 2020 and 11 November 2021 in British Columbia, which had also used administrative healthcare databases, indicated that 98.3% of healthcare workers at Vancouver Coastal Health had received two doses of a COVID-19 vaccine.35 Among the cohort, 3326 individuals had their occupation listed as ‘medical staff’ (distinct from nurses and care aides), and within this group, 99% had received two doses of a COVID-19 vaccine, 0.2% had received just one and 0.8% had received none. However, it was unclear how many physicians were part of this group, as the authors had condensed more than 1000 occupational designations into seven categories. In addition, though vaccination rates reported in this study appear to be higher than Ontario, it did not capture a province-wide cohort, as Vancouver Coastal Health is one of five health authorities in British Columbia.32

With regard to studies in the UK, a single-centre study by Martin et al, which included all staff identified in the Electronic Staff Record at University Hospitals of Leicester National Health Service (NHS) Trust on 3 February 2021, found that of 19 044 healthcare workers, 64.5% had been vaccinated with one dose of a COVID-19 vaccine and 35.5% were unvaccinated.36 Physicians were reported to be the lowest proportion (57.4%) of vaccinated healthcare workers, though in a sensitivity analysis that excluded locum or bank contracts, vaccination rates in physicians were higher (69.4%). Martin et al reported that since their centre is one of the most ethnically diverse populations of healthcare workers in the UK, and vaccine uptake has been noted to be lower in ethnic minorities,37 38 this could have influenced the vaccination rates observed. On the other hand, a multicentre prospective cohort study by Hall et al of 23 324 healthcare workers working in publicly funded NHS hospitals across the UK, 89% had received at least one dose of a COVID-19 vaccine by 5 February 2021.38 Further, uptake by staff groups showed 92.5% of physicians were vaccinated with one dose, while 7.5% were unvaccinated. In Wales, a national-scale cohort study of 82 959 healthcare workers demonstrated that 89.7% had at least one dose, and 85.5% had two doses by 30 September 2021.39 Overall, the high rates of vaccine uptake in healthcare workers from larger multicentre studies in the UK appear to correlate with our findings.

As for time elapsed between each dose of a COVID-19 vaccination, Hall et al report a median length of 23 days between the first and second dose (IQR 21–26; range 19–28).38 In our study, this length was 39 days. Although the recommended interval between the first and second vaccine doses was 21 days,40 a shortage of vaccines in Canada occurred in late January 2021.41 42 To enable more Canadians to receive a first dose during the shortage, Canada’s National Advisory Committee on Immunization recommended that eligibility for second doses be delayed by up to 4 months.43 This policy change was announced on 3 March 2021, and its impact is observable on vaccine uptake trends in this study. As shown in figure 2, between March and June 2021, receipt of first doses significantly increased, while receipt of second doses plateaued. In addition, a longer than expected period between the second and third doses (median of 253 days) was likely due to healthcare workers only being eligible for the third dose beginning on 6 November 2021.44 The increased uptake of the third dose after this date was again observed in figure 2.

By the end of our study period, 4.6% of Ontario physicians were not recorded to have had any doses of a Health Canada-approved COVID-19 vaccine. Proportions by specialty showed 3.4% of family physicians and 4.1% of specialists had no record of vaccination. Certainly, refusal of vaccination could be one reason, and vaccine hesitancy in healthcare workers has been well described in the literature; major reasons include concerns about the safety and efficacy of COVID-19 vaccines given their relatively rapid development and approval.4–6 However, another reason could be that physicians were vaccinated outside Ontario and not captured by COVaxON data. Furthermore, of the 1851 physicians with missing data on practice specialty, 519 (28%) had no record of vaccination. We speculate that this group could have included physicians whose listed residence was in Ontario, but were not actively practising, or were practising/training outside of Ontario. Our hypothesis would explain the geographical differences observed, in which proportions of unvaccinated physicians by LHIN were the highest in Toronto Central (7.2%), followed by Champlain (6.5%). Toronto is the most populous city in Ontario, and Ottawa (represented by the Champlain LHIN) is the second.45 Therefore, we suspect that a considerable number of physicians with missing specialty data lived in either Toronto or Ottawa, the top-two urban centres, and influenced our findings. Moreover, both cities have large academic health centres46 47; thus, it is possible that a physician in-training was pursuing further education outside of Ontario. Overall, if we were to have excluded physicians with missing specialty data from our analysis, with a remaining cohort size of 39 416 physicians and 1389 of which had no record of vaccination, the proportion unvaccinated would decrease to 3.5%. Nonetheless, uptake of COVID-19 vaccines by Ontario physicians remains high, irrespective of physician characteristics, which may signify their confidence in the safety and efficacy of COVID-19 vaccines. This is especially important, as physician health practices can influence those of their patients.48

Strengths

Strengths of our study include the use of a population-based design that captured all physicians who were registered to practise in Ontario, which avoided non-response bias. We were also able to limit self-report bias by using data from provincial administrative databases that documented vaccination details. Lastly, our study not only captured vaccination characteristics of physicians across multiple specialties, but also provided data on the uptake of the third/booster dose in physicians.

Limitations

One major limitation was that no tests of difference were performed; thus, with absence of CIs, comparisons made between groups were subjective. The databases used in our study were limited to data from Ontario only; hence, physicians who received vaccinations outside of Ontario would be missed. As indicated previously, 4.5% of physicians in our cohort had missing specialty data and might have included physicians who were misclassified as living in Ontario, or not in active practice.

Conclusions

In our study, we found that by 13 February 2022, 14 months since vaccine availability, 86.8% of Ontario physicians had received three doses of a Health Canada-approved COVID-19 vaccine, compared with 45.6% of the general population. However, 4.6% of Ontario physicians had no record of vaccination; thus, a potential area of future research could include exploring vaccine hesitancy in physicians. Nonetheless, our findings suggest that Ontario physicians willingly received vaccinations prior to provincial mandates, which may signify their confidence in the safety and effectiveness of COVID-19 vaccines.

Supplemental material

This post was originally published on https://bmjopen.bmj.com