Article Text

Cohort profile
Cohort profile: recruitment and retention in a prospective cohort of Canadian healthcare workers during the COVID-19 pandemic
  1. Nicola Cherry1,
  2. Anil Adisesh2,
  3. Igor Burstyn3,
  4. Quentin Durand-Moreau1,
  5. Jean-Michel Galarneau1,4,
  6. France Labrèche5,
  7. Shannon M Ruzycki6,7,
  8. Tanis Zadunayski1
  1. 1Department of Medicine, University of Alberta Faculty of Medicine & Dentistry, Edmonton, Alberta, Canada
  2. 2Division of Occupational Medicine, University of Toronto Temerty Faculty of Medicine, Toronto, Ontario, Canada
  3. 3Department of Environmental and Occupational Health, Drexel University, Philadelphia, Pennsylvania, USA
  4. 4University of Calgary, Calgary, Alberta, Canada
  5. 5IRSST, Montreal, Québec, Canada
  6. 6Department of Medicine, University of Calgary, Calgary, Alberta, Canada
  7. 7W21C, University of Calgary Cumming School of Medicine, Calgary, Alberta, Canada
  1. Correspondence to Dr Nicola Cherry; ncherry{at}ualberta.ca

Abstract

Purpose Healthcare workers were recruited early in 2020 to chart effects on their health as the COVID-19 pandemic evolved. The aim was to identify modifiable workplace risk factors for infection and mental ill health.

Participants Participants were recruited from four Canadian provinces, physicians (medical doctors, MDs) in Alberta, British Columbia, Ontario and Quebec, registered nurses (RNs), licensed practical nurses (LPNs) and healthcare aides (HCAs) in Alberta and personal support workers (PSWs) in Ontario. Volunteers gave blood for serology testing before and after vaccination. Cases with COVID-19 were matched with up to four referents in a nested case-referent study.

Findings to date Overall, 4964/5130 (97%) of those recruited joined the longitudinal cohort: 1442 MDs, 3136 RNs, 71 LPNs, 235 PSWs, 80 HCAs. Overall, 3812 (77%) were from Alberta. Prepandemic risk factors for mental ill health and respiratory illness differed markedly by occupation. Participants completed questionnaires at recruitment, fall 2020, spring 2021, spring 2022. By 2022, 4837 remained in the cohort (127 had retired, moved away or died), for a response rate of 89% (4299/4837). 4567/4964 (92%) received at least one vaccine shot: 2752/4567 (60%) gave postvaccine blood samples. Ease of accessing blood collection sites was a strong determinant of participation. Among 533 cases and 1697 referents recruited to the nested case-referent study, risk of infection at work decreased with widespread vaccination.

Future plans Serology results (concentration of IgG) together with demographic data will be entered into the publicly accessible database compiled by the Canadian Immunology Task Force. Linkage with provincial administrative health databases will permit case validation, investigation of longer-term sequelae of infection and comparison with community controls. Analysis of the existing dataset will concentrate on effects on IgG of medical condition, medications and stage of pregnancy, and the role of occupational exposures and supports on mental health during the pandemic.

  • COVID-19
  • EPIDEMIOLOGY
  • OCCUPATIONAL & INDUSTRIAL MEDICINE
  • PREVENTIVE MEDICINE
  • Health & safety

Data availability statement

Data are available on reasonable request.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

STRENGTHS AND LIMITATIONS OF THIS STUDY

  • Recruitment of a broad spectrum of healthcare workers close to the start of the COVID-19 pandemic through their professional organisations.

  • Consent to link to records held by public health departments allows for validation of self-reports of vaccinations and episodes of COVID-19 infection.

  • Repeated contacts permit charting the evolution of anxiety, depression and sources of stress through the course of the pandemic.

  • The inclusion of a nested case-referent study allows exposure reporting in near real time.

  • The absence of good denominator data limits the ability to examine recruitment bias.

Introduction

This paper reports on the establishment of a longitudinal cohort of healthcare workers (HCWs) from four Canadian provinces, working with patients potentially infected with the SARS-CoV-2 virus in a range of healthcare settings. It was set up at the very start of the pandemic in Canada, when little was known about the likely extent or length of demands on the healthcare workforce. Experience from the first SARS in 2003 suggested that HCWs would be at increased risk of both infection1 and adverse effects on mental health.2 Very early reports on HCWs exposed to the SARS-CoV-2 virus confirmed risks for both infection3 and psychological outcomes.4

The initial objective of establishing the cohort was to determine risk factors that could be mitigated to reduce risk of infection and mental distress in HCW, but the width of data collected laid the groundwork for studies of infection, the sequelae of infection and the impact of the pandemic on mental health. Two additional study groups were later identified within the cohort, a serology subcohort to address immune response to vaccination and infection and a nested case-referent study to obtain near real-time data on occupational exposures related to infection. There was a clear need for such a comprehensive approach, to clarify understanding of factors other than the standard demographic ones (age, gender and job) on serological response, and the impact of infection and vaccination on resiliency to the psychological demands of work. Early in the pandemic many short-term studies of HCWs were set up to establish risk in particular workplaces but few had the funding, resources or established protocol to follow workers through the length of the pandemic. Other cohorts of similar size and longevity as the one described here have been established either as seroprevalence studies5–7 or studies of mental health.8 9

A major challenge for longitudinal studies is attrition between contacts.10 This is particularly of concern if the reason for non-response is closely associated with the study outcome and risk factors of interest.11 In the cohort reported here, early attrition of those with prepandemic asthma or mental ill health could bias estimates of the prevalence of post-COVID-19 syndromes.12 13 An additional objective was to keep attrition to a minimum and to report on factors systematically related to non-response.

Cohort description

Recruitment procedures

Cohort recruitment began in Alberta in March 2020 and was extended to include HCWs in the four largest provinces of Canada (Alberta, British Columbia (BC), Ontario and Quebec). It comprised physicians (medical doctors, MDs), registered nurses and psychiatric nurses (RNs) licensed practical nurses (LPNs), personal support workers (PSW) and healthcare aides (HCAs). Potential participants were contacted, where possible, through their professional or workplace organisation.

MDs in Alberta, BC and Ontario were approached through their provincial college, the licensing authority for physicians. In Alberta and BC, the colleges forwarded recruitment materials, explaining the aims of the study, to physicians who contacted the research team if they wished to take part. In Ontario, the college provided emails to the study team to contact physicians directly but requested these not be used for further follow-up: physicians who had responded to the initial mailing were retained. The Collège des Médecins du Québec provided links to the study on their website and requests were sent to the presidents of medical specialty federations to invite participation and posted on social media addressed to MDs. Only specialists in internal medicine, infectious disease, medical microbiology, respiratory medicine, critical care, anaesthesiology, emergency medicine, geriatrics/care of the elderly, occupational medicine, public health and preventive medicine were approached together with physicians in family medicine and general practice. In BC and Ontario, to keep numbers manageable, community-based physicians were approached only from selected administrative areas (Hamilton, Peel. York and Ottawa in Ontario: Abbotsford, Burnaby, Chilliwack, Coquitlam, Delta, New Westminster, Port Coquitlam, Port Moody, Surrey and White Rock in BC).

RNs, LPNs and HCAs were recruited only in Alberta. RNs were recruited using email addresses provided by the College and Association of Registered Nurses of Alberta: the list of addresses only included those who had previously given approval to be approached about future (unspecified) research. Registered psychiatric nurses and LPNs were approached through their colleges. The Alberta Register of Health Care Aides was held by the College of Licensed Practical Nurses of Alberta, which agreed to contact this group also. PSWs in Ontario were approached through the Ontario Personal Support Workers Association and through one of the larger groups providing home care outside the public system.

The recruitment information sent to potential participants included an introduction to the purpose of the study, together with links to a more detailed information sheet, to an online consent module and, if consent given, to the phase 1 questionnaire (online supplemental materials 1). The consent module included space for contact details (email and phone number). In Alberta, physicians and nurses who completed the consent but not the full questionnaire were invited to complete a short questionnaire to establish their demographic details. Participants in Alberta were also asked for consent to link their identity to the provincial administrative health database.

Structure and timelines

The full cohort was contacted to complete questionnaires at each of the four phases of the study. Phase 1 (baseline) from April 2020, phase 2 in October 2020, phase 3 in May 2021 and phase 4 in May 2022. Cohort members were asked to give a prevaccine blood sample for serology testing in October 2020. Those who were willing and able to give serial blood samples postvaccine gave such samples from March 2021 to July 2022. A nested case-referent study recruited cases and matched referents from October 2020 to March 2022.

Newsletters were sent to update cohort members on study progress and preliminary findings in March 2021, November 2021 and March 2022.

Collection of information by questionnaire

Information was collected online using the Qualtrics survey platform, offered in both English and French. Reminders to complete were sent by email, with the alternative of a telephone interview offered to those not initially responding online. Where the participant had provided a phone number, this was used to obtain information or as a reminder that a questionnaire had not yet been completed. At each phase subsequent to the baseline contact, a short questionnaire was offered in the weeks immediately prior to closing down on-line access to the questionnaire, allowing those busy or reluctant to contribute core information. A payment of $C50 (Canadian) was offered to those initially unwilling to complete the phase 4, questionnaire. No other financial incentive was offered.

The questionnaires at each phase had much information in common, including questions about work with COVID-19 patients, availability of personal protective equipment (PPE), infection with the COVID-19 virus and, after December 2020, vaccination. The Hospital Anxiety and Depression scale (HADS)14 was included at each phase, together with a community acquired pneumonia questionnaire15 and substance use (alcohol, tobacco, cannabis and medications for anxiety and sleep). All four questionnaires asked participants to rate statements about changes at work, confidence in working with infected patients, sources of worries and support and also asked an open-ended question on stressful events.The baseline (phase 1) questionnaire included also questions on heath in the 12 months before the start of the pandemic (treatment for anxiety or depression and use of medication for asthma), and histories of smoking and chronic lung disease (bronchitis, emphysema or chronic obstructive lung disease). These were added to the questionnaire after the early recruitment of MDs who completed these questions at phase 2. Age, gender, marital status and children under 18 years in the home were asked only at baseline: ethnicity was collected at phase 2. The contact at phases 2–4 contained questions on institutional supports for mental health and, at phase 4, health sequelae of COVID-19 infection.

The questionnaires were essentially the same for all with only slight changes in wording to address each occupational group. Phase 2–4 questionnaires were sent to everyone who had consented to follow up unless they indicated they were unwilling to take part further (refused) had moved away from the participating provinces with no intent to return (moved away), had left healthcare work (either from choice or retirement) or were lost to trace (ie, neither email nor phone number were in service). Where a respondent indicated that they had some personal difficulty (such as illness or marital breakdown) that might not exclude them permanently, one or more follow-up contacts were omitted.

Collection of prevaccine serology samples

At the phase 2 questionnaire (October 2020), participants were asked for permission to approach them to give a blood sample for analysis of serological markers of infection and immune response to vaccination and infection. It was not possible for the team to collect blood samples ‘in house’ but arrangements were made in each province for one or more clinical laboratories to take the sample, when approached by the participant with a study requisition. In Alberta, the clinical laboratories covered the whole of the province but this was not the case elsewhere. In BC, the provincial healthcare laboratories provided back-up in areas not covered by the commissioned commercial laboratory. In Ontario and Quebec, only commercial laboratories with limited coverage were available to collect samples: in those provinces participation in the serology study was effectively impossible for those living at a great distance from a commercial collection centre. Attempts to collect samples more informally and to transport them by courier to the commercial laboratory were not always successful, but a small number of participants in Ontario and Quebec were able to use this route.

Collection of postvaccine serology samples

Those who had agreed to give a prevaccine serology sample were approached again in early 2021 to ask if they would be willing to take part in a serology substudy giving samples at 4, 7 and 13 months after their first dose of vaccine: a 10-month sample was added subsequently, following the early observation of a rapid decline in antibody concentration between 4 and 7 months. Participants were asked to consider whether such a regimen of repeated samples would be feasible for them, in light of their distance from a collection centre and their personal commitments. Those who agreed were sent a requisition for each sample as it became due and asked to arrange for sample collection locally.

Nested case-referent study

A nested case-referent study was established within the cohort, starting in October 2020 and continuing to 31 March 2022. It considered exposures in the 21 days prior to a positive PCR test. Those with a positive PCR test were matched, as referents, with up to four people in the same province, on job classification, self-reported gender and first vaccination date. Both cases and referents were asked, in the days immediately following case infection, to complete a detailed exposure questionnaire about experiences and exposures in the days leading up to the case’s positive test. More details are given in Cherry et al.16

Updating of data on infection and vaccination

Each of the four questionnaires sent to the full cohort asked about positive tests for the SARS-CoV-2 virus and questionnaires 2–4 about vaccination shots received, once this became available to HCWs in mid-December 2020. In addition, participants were regularly reminded to contact the research team if they had a positive COVID-19 test or received a vaccine. Those in the postvaccine subcohort completed a brief questionnaire at each sample, updating infection and vaccination status. All participants were asked for consent to approach the public health authorities in their province to obtain records of positive PCR tests and vaccinations.

Protocol for reminders

At recruitment potential participants who were approached through their professional associations in Alberta and BC received two email reminders following the initial invitation. At each questionnaire postrecruitment non-responders received two email reminders (three if there was no telephone contact) and, if they had volunteered a telephone contact, up to two phone reminders. The same protocol was used for the case-referent study. Contact with the serology cohort was two-way with participants requesting help in arranging blood sampling and providing information on when and where samples had been collected. Those in this subcohort received reminders to book an appointment before a sample was due, the requisition before the due date and reminders to give a sample once the due date had passed. Because of the high engagement of participants, the overlapping of subcohorts and a policy of rapid personalised response to email or telephone queries, interaction between survey staff and cohort members became more informal and frequent than is implied by the reminder protocol outlined.

Numbers recruited

A total of 5130 HCWs gave consent for completion of the baseline questionnaire. Of these 4964 (96.8%) agreed to follow up. The follow-up cohort comprised, by order of recruitment, 1442/1490 (96.8%) MDs, 3136/3227 (97.2%) RNs, 71/74 (95.9%) LPNs, 235/257 (91.4%) PSWs and 80/82 (97.6%) HCAs. Those who declined follow-up were more likely to be younger and male.

The estimated proportion accepting the invitation to join the study ranged from 18% of Alberta nurses to less than 4% of Ontario physicians, where recruitment could not continue beyond an initial email. In Alberta and BC, about 7% of physicians approached joined the study, based on the number of emails sent. Estimates of the proportion participating were unsatisfactory for LPNs, HCAs and PSWs: where the numbers approached by email could not be ascertained and where, for HCAs, emails were not actively maintained. In Quebec, recruitment of physicians depended on their response to a link on the college of physician’s website. It seems likely that response to individual emails (had this been possible) would have been higher. We know nothing, other than professional role and province, about any of those not accepting the invitation to join and cannot investigate response bias from low response at recruitment.

Retention in the main cohort

By the time of the phase 4 questionnaire in spring 2022, 127 had become ineligible (2 were known to have died, 36 had moved from the study provinces and 89 had retired/left healthcare work), leaving a denominator of 4837. Among the 538 who remained in the cohort but who did not respond, 100 refused and contact details failed for 18. The research team made some contact with 61 of 420 remaining participants but the reason for eventual non-response was not known. The response rate at phase 4 was thus 88.9% (4299/4837) with 4731 completing a questionnaire at phase 2 and 4519 at phase 3.

Figure 1 gives the flow chart for recruitment and retention for the four phases of the main cohort.

Figure 1

Flow chart of recruitment and retention for all phases of the main cohort. HC, healthcare.

Characteristics of participants

At baseline, the majority of participants were working in Alberta (77%) and were aged 40 years or older (62%): PSWs (76%) and HCAs (73%) had a greater proportion over 40 years. Most were married or living as married (75%): fewer than half (43%) had a child aged <18 years living in their home. The majority of the participants were female (4111) with 848 male and 5 who reported their gender as non-binary. Overall, 83% reported their ethnicity as white, with the highest proportion in RNs (87%). Only 2% reported themselves as Indigenous and 1% as black (table 1).

Table 1

Demographics of participants by work role

Exposures and potential confounders

Overall, 88% (4184/4760) of those known to be working at baseline were in direct contact with patients (either ‘hands on’ or virtually), and of these 58% had all or part of their work in a hospital setting, 35% in the community and 12% in a residential setting (online supplemental materials 2). A small proportion (7%) worked only in another setting, including roles related to COVID-19 (screening, contact tracing). Physicians (65%) and nurses (61%) were most likely to be working in a hospital setting, whereas PSWs and HCAs were more likely to be working in the community (including the client’s home) or in a residential community such as a care home.

Smoking habit and health in the 12 months before the start of the pandemic differed markedly with work role. Physicians (1%), RNs (6%) and LPNs (13%) reported less often than PSWs (32%) or HCAs (25%) that they had smoked tobacco in the 12 months before March 2020 and were less likely to report chronic lung disease other than asthma. Overall, 24% reported having been treated for anxiety or depression in the 12 months before the pandemic. MDs were least (17%) and LPNs most (48%) likely to report treatment for these conditions (table 2).

Table 2

Smoking and health in the 12 months before the start of the pandemic by work role

Characteristics of those lost to follow-up

Factors associated with attrition between baseline and phase 4 are shown in table 3, with completion tabulated by data collected at baseline. Partial completion of the baseline questionnaire was a strong predictor of attrition, but the phase 4 completion rate was 79% even in this group. As seen in the bivariate logistic regression in table 3, completion was higher with greater age and in those married or living as married and lower in those smoking in the year before the start of the pandemic. The missing data on smoking, lung disease, use of asthma medication, anxiety and depression seen in the bivariate analysis arose from a failure to complete the full baseline questionnaire and was strongly related to attrition at phase 4. The multivariable analysis presented in table 3 is limited to those fully completing the baseline questionnaire. It supported the bivariate results for all factors other than smoking, where the effect was less marked. Treatment for anxiety or depression in the 12 months before the start of pandemic showed only a slight association with attrition in the bivariate analysis: this was reduced in the multivariable model.

Table 3

Completion of the phase 4 questionnaire in spring/summer 2022

Participation in the serology study

The numbers giving serology samples prevaccine and postvaccine are given in table 4: 59% (2940/4964) of the full cohort gave a blood sample for serology before vaccination and 60% (2752/4567) at least one sample postvaccine (excluding the 397, 8.0%, not known to have been vaccinated). Participation was much lower in PSWs than in other HCWs with only 16% giving a prevaccine sample and 12% postvaccine. Factors associated with giving prevaccine and postvaccine serology samples are shown in online supplemental materials 2. Women, older HCWs, nurses, those who were married and those working directly with COVID-19 patients were more likely to give a sample. Smokers and those having a child in the home were less likely to give a prevaccine sample (online supplemental table 2). MDs and RNs were most likely to join the postvaccine serology subcohort, as were those who were older and married (online supplemental table 3). Having a child at home and smoking were again associated with a lower participation.

Table 4

Blood samples for serology prevaccine and postvaccine by work role

Participation in the case-referent study

Of 542 cases eligible for the case-referent study, 534 (99%) completed the case-referent questionnaire, as did 1697 (93.5%) of 1815 selected referents.16

Findings to date

As reported here, occupational groups differed markedly on prepandemic health indicators including smoking, chronic lung disease and treatment for anxiety and depression. Retention in the main cohort and recruitment to the serology subcohort were higher in older and married HCWs. In the case-referent study, unvaccinated HCWs were at greater risk of infection if they worked hands‐on with patients with COVID‐19, on a ward designated for care of infected patients, or handled objects used by infected patients. Later in the pandemic, with almost universal HCW vaccination, risk from working with infected patients was much reduced.16

Discussion

The strengths of this cohort include baseline data collected in the first weeks of the pandemic from a spectrum of HCW roles and job settings. The great majority of those recruited have been retained, completing questionnaires throughout the pandemic. Participation in the serology subcohort was high (60%), given the time and organisation needed for up to four postvaccine samples. It is a strength to document factors associated with this participation. The case-referent study, nested within the cohort, allowed near ‘real-time’ reporting of exposures. The study depended on self-reporting of infection and vaccination and the ability to check these reports against public health records in at least two provinces (Alberta and BC) gives strength, particularly to the analysis of determinants of antibody response. This established cohort, with repeated measures over the course of the pandemic, provides sound data for analysis of workplace and personal factors related to vaccination, infection with the SARS-CoV-2 virus, and adverse health outcomes of COVID-19 illness, and also of the evolution of mental ill health during the pandemic.

The low response rate to the invitation to join the cohort was a weakness. Those recruited were potentially unrepresentative and this limits the generalisability of the results. Challenges in setting up the cohort included finding organisations with nominal lists and contact details who were willing and able, at the start of the pandemic, to supply those details or to act on behalf of the study to contact potential participants. Although considerable efforts were made to recruit and retain HCW other than physicians and nurses, workers in other roles were under-represented. A further weakness is that episodes of COVID-19 may have been under-reported and this cannot be checked in all provinces. Asymptomatic and unrecognised cases can be charted only for those giving serology samples (where nucleocapsid antibodies would result from infection rather than vaccination). Arrangements for collection of serology samples were not uniform across provinces. A high proportion of participants were from Alberta, resulting from the decision to limit recruitment of nurses to those working in that province: this reduces the capacity to generalise findings to nurses working across Canada. Only MDs contributed data from all four participating provinces.

The attrition rate (11%) in the study reported here is lower than other population-based studies of HCWs during the pandemic, which range from 45%17 and 57%18 to 68%–69%19 20 over shorter periods of follow-up. Low attrition rates over multiple contacts are rare in community-based studies, although de Graaf et al21 achieved a 20% attrition rate in the first follow-up in a prospective psychiatric epidemiological study. Nguyen et al22 report an attrition rate of 15% in a community-based cohort of COVID-19 with follow-up of 16 months. Both de Graaf and Nguyen attributed their successful retention to the large time-investment to encourage respondents, as was done here. Analysis of those not retained suggested that attrition was not random. As in other studies, age, marriage and higher socioeconomic status (here represented by MDs and RNs) were associated with higher retention and smoking with lower.23–25 Importantly, given that a main focus of the study was mental health, the lack of an association between attrition and treatment for anxiety or depression in the 12 months before the pandemic was encouraging. Previous studies have differed on the importance of baseline mental health as a predictor of attrition, with some21 26 27 reporting (as here) no relationship while others17 25 28 29 found those with mental ill health were less likely to respond at follow-up. Despite high participation in the cohort during the pandemic, it will be important to remain vigilant of the impact attrition may have on causal inference.

Recruitment to the serology subsample was not random, not least because of the less satisfactory arrangements for sample collection in Ontario and Quebec. The proportion of PSWs in Ontario (16%) giving a prevaccine sample was less than half that for HCAs in Alberta (44%) although these groups were similar in other ways. Inspection of prevaccine samples within physicians showed samples from only 40% of MDs in Ontario and Quebec but 78% for Alberta (where collection sites covered the whole province) and 62% for BC where black holes for collection were fewer than in Ontario and Quebec. Similar, though more marked, differences were evident for postvaccine samples, where commitment was asked for repeated blood samples over nine months: here rates for PSWs were only 12% compared with 45% for HCAs, and Ontario MDs with the lowest rate of participation among MDs at 44%. Bias in participation may be of limited importance for a serology study but biased estimates of IgG concentration in HCWs could result if rurality or remoteness was differentially associated both with infection or vaccination and difficulty in giving blood samples

Future plans

Results of the serology analysis (concentration of IgG against the receptor binding site of the spike protein of the SARS-CoV-2 virus) will be entered, together with demographic data, into the publicly accessible database being compiled by the Canadian Immunology Task Force (CITF).30 Further, we will take advantage of the data available through Canada’s public healthcare system to validate self-reports on infection and vaccination as these health reports become available from cooperating provinces. Using these administrative data, we will investigate longer-term sequelae of COVID-19 infection, and the influence of pre-existing health conditions by linking those in the Alberta cohort to administrative health records both before and since the start of the pandemic. We have in place the necessary consents and approvals to do this.

Within data on the present cohort, serology analysis concentrates on the extent to which medical conditions, medications and stage of pregnancy add to vaccination, infection and age in determining IgG. Mental health changes during the pandemic are documented through scores on the HADS completed at each of four contacts, with changes during the pandemic examined by work demands, availability of PPE, vaccination and infection, within risk groups defined by prepandemic health indicators. An analysis of stressful events reported by participants will describe the evolution of sources of stress since the start of the pandemic. The extent to which supports offered through work modify anxiety and depression scores is being assessed. The severity of response to infection is considered in linked analyses of symptom reporting, time off work and post-COVID-19 symptoms.

Collaboration

In Alberta, we have an agreement to link administrative health records of each HCW to the health records of five community referents, matched on age, sex and geographical region to allow comparison with infection rates in the cohort. We are also working with provincial public health authorities in Alberta, BC and Quebec to link records of infection and immunisation to the cohort database. As further collaborations we are exploring comparing data from this cohort with a parallel cohort of paramedics recruited from five Canadian provinces including Alberta, BC and Ontario.31 Such collaborations will be greatly facilitated by the databank being established by the CITF, which will contain data on all projects (including this and others on HCWs) funded as a national initiative early in the pandemic. Through this databank anonymised data will be made available to researchers internationally on request to the CITF. Collaborations involving comparative data are welcomed as are those wishing to access serum samples, within the constraints of consents given by participants.

Patient and public involvement

The first phase of the study was set up very rapidly after COVID-19 was reported in Canada. Discussions between the study team and leadership of HCW organisations in the four provinces were arranged to ensure the approach taken and study questions were acceptable to participants, and to inform recruitment strategies. We solicited feedback about the research topics throughout phase 1, and phase 2 was designed taking account of feedback on the initial questionnaire and the perceived need by the leaders of participant healthcare organisations for additional information on mental health supports. Ongoing communication towards participants was individual and frequent. Participants received three newsletters with updates on preliminary results and further explanatory background information, with options to comment on the study and to submit questions. Participants in the serology substudy received their individual postvaccine antibody results.

Data availability statement

Data are available on reasonable request.

Ethics statements

Patient consent for publication

Ethics approval

This study involves human participants and was approved by University of Alberta Health Ethics Board (Pr000099700)Unity Health Toronto Research Ethics Board (REB#20-298). Participants gave informed consent to participate in the study before taking part.

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • Twitter @doctorsdilemma, @ShannonRuzycki

  • Contributors NC, AA, IB, QD-M, J-MG, FL and SMR contributed to the initial design of the study. J-MG established the dataset. TZ coordinated the ongoing acquisition of data. NC analysed the data in discussion with IB and J-MG. NC drafted the manuscript and acts as guarantor. All authors contributed to the interpretation of the data and critically revised the draft. All read and approved the final manuscript.

  • Funding Seed funding was obtained from the College of Physicians and Surgeons of Alberta. Grant funding was obtained from the Canadian Institutes of Health Research (Funding Reference number 173209). This funding was extended by a grant from the Canadian Immunology Task Force.

  • Disclaimer No funder had any part in the design of the study or in the collection, analysis or interpretation of the data or in the writing of the manuscript.

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.