Professional characteristics, attitudes, and practices associated with stress and quality of life among Canadian animal health workers

Abstract

Objective

To describe the knowledge, attitudes, and practices (KAP) towards COVID-19 of Canadian companion animal health workers (AHW); to measure their perceived stress and quality of life (QoL); and to explore professional risk factors associated with stress and QoL.

Sample

We sampled 436 companion animal veterinarians and technicians.

Procedure

The study had cross-sectional and cohort components. It was conducted online in August to December 2020, and repeated in May to July 2021, using a questionnaire assessing the respondents’ professional characteristics, COVID-19 KAP, perceived stress, and QoL.

Results

Overall, AHW had sufficient knowledge of COVID-19 transmission, and reported having adopted good preventive practices. Since the beginning of the pandemic, participants reported increases in new clients (76%), in refusal of new clients (53%), and in pet euthanasia (24%). Increased client refusal and pet euthanasia were associated with greater stress and poorer professional QoL, whereas perceived susceptibility to and adoption of measures against COVID-19 were associated with lower stress and better QoL.

Conclusion and clinical relevance

For AHW, professional characteristics were associated with stress and professional QoL. This information is important for developing strategies to cope with the ongoing shortage of AHW and with future public health crises.

Introduction

Following the COVID-19 pandemic declaration by the World Health Organization in March 2020, the government of Canada declared a state of public health emergency. This was soon followed by recommendations for the general population to shelter at home and to limit travel outside of their homes except for essential work and acquiring basic necessities. Veterinary clinics were listed among essential services and recommendations were made by veterinary regulator entities (e.g., Ordre des Médecins vétérinaires du Québec) to treat only emergency cases and postpone all nonessential appointments and surgeries (1). Although this recommendation may seem relatively straightforward, exactly what constituted an “emergency” was left to the judgment of veterinarians (2). Furthermore, conditions not initially considered an emergency may become urgent when postponed. Implementation of infection control measures was also recommended at all veterinary clinics to help prevent the spread of COVID-19 (1).

In knowledge, attitude, and practice (KAP) studies conducted in China, Vietnam, and Italy, although > 80% of human health workers had a sufficient level of knowledge about the SARS-CoV-2 virus, they also expressed high levels of concern regarding the disease (3–6). Despite this, that health workers in Chinese health centers still expressed willingness to care for patients infected with COVID-19 was attributed to training and previous experience caring for such patients (4). In another study in China, frontline health workers practiced strict preventive behaviors, which were associated with higher levels of education and < 8 h of work per day (3). In a study from Vietnam, high knowledge levels among human health workers were associated with positive attitudes, including willingness to self-isolate if infected with COVID-19 (5).

Evidence of SARS-CoV-2 transmission to companion animals was limited at the beginning of the pandemic (7), but was suspected, especially for cats and ferrets (8). To our knowledge, only 1 study, conducted in Nigeria, assessed SARS-CoV-2-related KAP of animal health workers (AHW) (9). In that study, 67% of participants had satisfactory knowledge scores for COVID-19, and 98% reported practicing mitigation measures while working. Investigations into workplace practices before the current pandemic revealed that use of personal protective equipment (PPE) by veterinarians was generally poor (10).

Before the pandemic, a cross-sectional Canadian study of veterinarians reported higher mean scores than the general population for burnout, anxiety, and depression, with female veterinarians having significantly higher scores than their male counterparts (11). This study used multiple validated tools to assess mental health and quality of life (QoL) of participants, including the perceived stress scale (PSS) and the ProQoL, which is used to assess professional QoL. This recent study mainly assessed gender differences for mental health and QoL indicators, but in previous research, work-related factors such as workload and euthanasia were associated with compassion fatigue, burnout, and suicide (12,13). These work-related factors seemed to change during the pandemic, but also before and likely after (14), contributing to poor mental health and QoL of veterinarians.

Although the COVID-19 pandemic generated stress and decreased QoL in the general population (15), it is unclear how AHW were affected. We hypothesized that strict population-level containment strategies and uncertainties around risks of exposure to SARS-CoV-2 via infected companion animals generated severe stress and decreased QoL among AHW.

Our objectives were to describe the KAP for COVID-19 of Canadian companion animal veterinarians and animal health technicians; to measure their perceived stress and QoL; to describe changes in KAP, stress, and QoL between 2 periods of the pandemic; and to explore professional risk factors associated with stress and QoL during the early stages of the COVID-19 pandemic.

Materials and methods

Study design

This study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Ethics Committee of the Université de Montréal (CERSES-20-097-D; approved on July 22, 2020). We first designed this study as a cohort study, whereby participants of a first online survey were to be followed in time to measure changes in stress, QoL, and KAP about COVID-19 during the pandemic. The questionnaire was first shared from August to December 2020 (T1), and participants who gave their consent for a follow-up questionnaire were contacted again from May to July 2021 (T2). We explored associations between work-related and KAP factors and stress and QoL outcomes, with a cross-sectional design, during the period for which veterinary clinics had the greatest stringency (T1).

Source population and recruitment strategy

The target-source population was a convenience sample. We contacted the Canadian Veterinary Medical Association (CVMA) and the provincial veterinary and animal health technician associations by email and asked them to share an invitation to participate in the study with their members (open voluntary survey). These associations shared the invitation to their members through newsletters and listserv communication. Participants were included if they were veterinarians or animal health technicians, 18 y or older, residents of Canada, and currently practicing in companion animal health clinics in Canada (including general, emergency, referral, and mixed practice).

Measurement of variables of interest

The LimeSurvey platform (https://www.limesurvey.org/) was used to administer an online electronic questionnaire. We developed the questionnaire in English (see Appendix I, available online from: www.canadianveterinarians.net) and translated it into French. During the development phase, the questionnaire was tested by 8 collaborators, and 37 questions were modified following this testing phase to improve clarity. The final questionnaire was presented on 13 pages with 1 to 3 questions per page, or a table with multiple statements. Participants were allowed to go back through the questionnaire and change their answers before submitting it.

Information about the survey was provided to participants when they accessed the survey online, at which point they were asked for their informed consent (whether they agreed to the conditions of the survey) (Appendix I, available online from: www.canadianveterinarians.net). Participants were asked to answer sociodemographic and work-related questions, including about their gender, age, and province; occupation (veterinarian or animal health technician); years of experience; and work schedule.

Questions about knowledge, risk perception, adoption of preventive measures at work, and perception of the effectiveness of such measures were developed following the Health Belief Model (16). These questions were asked only in the first questionnaire (T1). Sufficient knowledge about modes of transmission of COVID-19 was defined as correctly classifying 7 of 8 choices (modes of transmission: droplets, aerosols, surfaces, food, humans, and humans to animals; not modes of transmission: mosquitoes and water), following information available from the public health expertise and reference center at the time of the survey (17).

We summarized various KAP elements in indices averaging multiple questionnaire elements selected using a factor analysis (Table 1). Briefly, this factor analysis was done using a principal factor method and an oblique rotation transformation. Variables with an item-total correlation > 0.5 were included (18), and the internal consistency reliability of each index was assessed using Cronbach’s alpha (19).

Table 1.

Information related to indices describing diverse characteristics of animal health workers in Canada, including tested and included questions, and Cronbach’s alpha for each index.

Index	Cronbach’s alpha	Questions	Included
Perceived	0.77	If other Canadians do not follow protective measures against COVID-19, MY risk of susceptibility becoming infected is [very low to very high]	Yes
		If other Canadians do not adopt protective measures against COVID-19, their risk of becoming infected is [very low to very high]	Yes
Worrying	0.56	COVID-19 is a severe disease [strongly disagree to strongly agree]	Yes
		It is easy to protect myself against COVID-19 [strongly disagree to strongly agree]	No
		I believe there is great scientific uncertainty around COVID-19 [strongly disagree to strongly agree]	No
		I am worried about becoming infected with COVID-19 [strongly disagree to strongly agree]	Yes
Perception of measures’ efficiency	0.85	Wearing a mask while at work [strongly disagree to strongly agree]	Yes
		Limit the number of clients in the clinic [strongly disagree to strongly agree]	Yes
		Use of personal protective equipment during consultations [strongly disagree to strongly agree]	Yes
		Regular disinfection of surfaces [strongly disagree to strongly agree]	Yes
		Installation and use of a plexiglass barrier [strongly disagree to strongly agree]	Yes
		Have clients leave their pets at the entrance [strongly disagree to strongly agree]	Yes
		Conduct triage by phone before scheduling appointments with clients [strongly disagree to strongly agree]	Yes
		Advise workers to stay home if they have any symptoms compatible with COVID-19 [strongly disagree to strongly agree]	No
Adopted measures	0.58	Wearing a mask while at work [never to always]	No
		Limit the number of clients in the clinic [never to always]	Yes
		Use of personal protective equipment during consultations [never to always]	Yes
		Regular disinfection of surfaces [never to always]	Yes
		Installation and use of a plexiglass barrier [never to always]	No
		Have clients leave their pets at the entrance [never to always]	No
		Conduct triage by phone before scheduling appointments with clients [never to always]	No
		Advise workers to stay home if they have any symptoms compatible with COVID-19 [never to always]	No

Perceived susceptibility to COVID-19 was assessed with 2 Likert scale questions (5 points, very low to very high) and worrying about COVID-19 was assessed with 2 Likert scale questions (5 points, strongly disagree to strongly agree). Perception of the efficiency of various measures to prevent the spread of COVID-19 was assessed with 7 Likert scale questions (5 points, strongly disagree to strongly agree), whereas adoption of measures to prevent the spread of COVID-19 was assessed with 3 Likert scale questions (4 points, never to always). Participants were also asked to answer “yes” or “no” to questions asking whether they perceived that there had been increases in euthanasia, treatment refusals from their clients, new clients, and refusal of new clients since the beginning of the pandemic.

We used standardized questions to assess perceived stress. The PSS is a validated, 10-item, self-reported instrument that scores participants’ responses regarding the frequency of positive and negative feelings over the previous month (20). This tool estimates a stress level, with 40 being the highest level of perceived stress, and 0 the lowest. Participants’ QoL was assessed for their personal and professional lives. The EQ-5D is a validated, 5-item instrument that has been used to assess personal QoL in research and clinical trials (21). The ProQoL is a 30-item instrument used to assess professional QoL of those working in caregiving domains (22). The EQ-5D tool includes 5 questions regarding mobility, self-care, usual activities, pain and discomfort, and anxiety and depression, as well as 1 question on overall health. The combined results across the 5 questions were translated on a standardized scale validated for Canadians (utility score) (23). The ProQoL tool gives 3 scores, ranging from 10 to 50, for compassion satisfaction, burnout, and secondary traumatic stress. The 3 scores can be categorized as low, middle, or high, following a normative benchmark (24).

Statistical analyses

We conducted statistical analyses using R software (version 4.0.5) with the R Studio interface (Version 1.3.1093) (25). Multiple entries from the same IP address and incomplete questionnaires were manually removed. The loss-to-follow-up bias was assessed by comparing participants who completed both questionnaires to participants who did not for sociodemographic variables, KAP, and stress and QoL variables at T1. Proportions and 95% confidence intervals (CI) were calculated for categorical variables, and medians and ranges were obtained for continuous variables. The change in KAP, stress, and QoL between T1 and T2 was assessed graphically, when available (knowledge about COVID-19, increase in new clients, new client refusal, treatment refusal, euthanasia, PSS, QoL, and ProQoL). In addition, agreement between T1 and T2 was assessed using the Cohen kappa statistics for dichotomized variables (increase in new clients, new client refusal, treatment refusal, and euthanasia).

Complete-case models to explore the association between work-related variables and stress and QoL were built using data from the first survey (T1) with a regressive elimination approach, following the causal diagram developed by the research team (available at dagitty.net/mhNYbCj) and keeping confounders with P < 0.16 (26). The PSS and QoL utility scores were assessed with linear regression models. The ProQoL scores were coded as binary outcomes and assessed with logistic regression models. Categorical variables selected by < 1% of respondents were considered as missing values in the models. Linearity and homoscedasticity of the residuals were verified for the linear models, and the fit of the logistic regression models was assessed using the Hosmer and Lemeshow goodness-of-fit test.

Results

Totals of 607 and 340 respondents started the first (T1) and second (T2) questionnaires, and totals of 519 and 302 completed them, respectively, representing completion proportions of 85.5 and 88.8%. After considering inclusion criteria, 436 and 189 participants were eligible at T1 and T2, respectively, representing a loss to follow-up of 56.5%. Participants who completed both surveys were generally similar to participants who did not, but were older and had more work experience (Table 2). Almost all participants were women, with a majority aged between 25 and 44 y, and ~2/3 were animal health technicians (Table 2). A total of 20.4% of participants were veterinary clinic owners.

Table 2.

Descriptive statistics for sociodemographic; work-related; and knowledge, attitude, practice variables for animal health workers in Canada from August to December 2020 (T1) who completed only the first survey (T1) or both surveys [T1 and T2 (May to July 2021)] as part of a cohort study on stress and quality of life of animal health workers during the early stage of the COVID-19 pandemic. Categorical variables are presented as the proportion per category with 95% confidence intervals (CI), or as the proportion with 95% CI for the “yes” category for “yes/no” variables. Continuous variables are presented as the mean with 95% CI.

Variables	Missing values	Proportion or mean (95% CI)
		Completed T1 only (n = 247)	Completed T1 and T2 (n = 189)
Gender	0
Woman		94.0% (90.0; 96.5)	94.2% (89.6; 96.9)
Man		5.2% (2.9; 9.0)	4.2% (2.0; 8.5)
Other		0	0.5% (0; 3.4)
Prefer not to answer		0.8% (0.1; 3.2)	1.1% (0.2; 4.2)
Age (y)	0
18 to 24		10.1% (6.8; 14.7)	6.9% (3.9; 11.7)
25 to 34		46.0% (39.7; 52.4)	38.6% (31.7; 46.0)
35 to 44		27.0% (21.7; 33.1)	26.5% (20.4; 33.4)
45 to 54		10.5% (7.1; 15.1)	18.0% (12.9; 24.4)
55+		6.5% (3.9; 10.5)	10.1% (6.3; 15.5)
Region	0
Western provinces and Territoriesa		21.8% (16.9; 27.5)	23.8% (18.1; 30.6)
Ontario		25.4% (20.2; 31.4)	30.7% (24.3; 37.9)
Quebec		36.3% (30.4; 42.6)	30.7% (24.3; 37.9)
Atlantic provincesb		16.5% (12.3; 21.9)	14.8% (10.2; 20.9)
Occupation	0
Veterinarian		35.1% (29.2; 41.4)	32.8% (26.3; 40.1)
Animal health technician		64.9% (58.6; 70.8)	67.2% (59.9; 73.7)
Experience (y)	0
< 5		35.5% (29.6; 41.8)	23.8% (18.1; 30.6)
5 to 10		21.0% (16.2; 26.7)	24.9% (19.0; 31.8)
11 to 15		17.7% (13.3; 23.2)	15.3% (10.7; 21.5)
16 to 20		13.7% (9.8; 18.8)	11.1% (7.2; 16.7)
≥ 21		12.1% (8.4; 17.0)	24.9% (19.0; 31.8)
Working schedule	36
Did not work		4.8% (2.6; 8.7)	3.5% (1.4; 7.7)
Worked part-time		15.4% (11.1; 20.8)	17.3% (12.2; 24.0)
Worked < 40 h/wk		21.9% (16.9; 28.0)	23.1% (17.2; 30.3)
Worked 40 to 50 h/wk		48.3% (41.6; 54.9)	48.0% (40.4; 55.7)
Worked > 50 h/wk		9.6% (6.3; 14.4)	8.1% (4.7; 13.5)
Sufficient knowledge about COVID-19	18	53.4% (46.8; 59.9)	65.8% (58.4; 72.5)
Perceived susceptibility index	18	4.3 (4.0; 4.3)	4.3 (4.2; 4.5)
Worrying about COVID-19 index	18	3.7 (3.7; 3.8)	4.0 (3.8; 4.0)
Perception of measures’ efficiency index	18	4.4 (4.4; 4.5)	4.4 (4.4; 4.5)
Adopted measures index	98	3.7 (3.5; 3.7)	3.7 (3.5; 3.8)
Increase in new clients	18	76.9% (70.9; 82.1)	75.5% (68.6; 81.4)
Increase in new client refusal	18	52.1% (45.5; 58.7)	53.3% (45.8; 60.6)
Increase in treatment refusal	18	28.2% (22.6; 34.5)	23.9% (18.1; 30.9)
Increase of euthanasia	18	23.1% (17.9; 29.1)	24.5% (18.6; 31.4)

^aBritish Columbia, Alberta, Saskatchewan, Manitoba, Yukon, and Northwest Territories.

^bNew Brunswick, Nova Scotia, Prince Edward Island, and Newfoundland and Labrador.

Reported knowledge, attitudes, and practices (KAP)

Just over 1/2 of the participants had sufficient knowledge about COVID-19 modes of transmission at T1 (58.9%, 95% CI: 54.0 to 63.6). At T2, this proportion did not increase (63.3%, 95% CI: 55.9 to 70.1), but some participants had improved knowledge (of participants who had insufficient knowledge at T1, 45.2% had sufficient knowledge at T2) or deteriorated knowledge (of participants who had sufficient knowledge at T1, 26.7% had insufficient knowledge at T2; Figure 1). Insufficient knowledge was largely driven by respondents who did not know that SARS-CoV-2 can be transmitted from humans to companion animals, which was selected by 42% and 52% of the participants at T1 and T2, respectively (see Appendix II, available online from: www.canadianveterinarians.net). In both surveys, < 1/5 of participants thought that contact with an infected animal was a mode of transmission (T1: 19.1%, 95% CI: 15.5 to 23.3; T2: 20.0%, 95% CI: 14.7 to 26.5).

Figure 1.

Distribution of the change in number of correct answers to identifying COVID-19 modes of transmission at 2 times during the COVID-19 pandemic (T1: August to December 2020, T2: May to July 2021) for animal health workers classified as having sufficient knowledge^a at T1 and/or at T2.

^a “Sufficient knowledge” was defined as correctly classifying 7 of 8 modes of transmission.

Participants’ index for perceived susceptibility to COVID-19 at T1 varied from low to very high (median: 4, i.e., high), whereas the range of the index for worrying about COVID-19 varied from strong disagreement to strong agreement (median: 4, i.e., agreement). The index for the perception of effectiveness of the different measures varied from strong disagreement to strong agreement of effectiveness (median: 4.4, i.e., agreement to strong agreement), and the index for reported adoption of various measures varied from never to always (median: 3.7, i.e., sometimes to always). Based on detailed frequency of adoption of the 8 measures, almost all measures were generally reported to be adopted by the participants (Appendix II, available online from: www.canadianveterinarians.net). However, conducting triage by phone was always done by 45.1% of respondents, regular disinfection was always done by 60.3% of respondents, and plexiglass was never used by 18.3% of respondents (but was always used by 65.3%).

At T1, 76.3% of the participants perceived that there was an increase in new clients since the beginning of the pandemic, and 52.6% refused new clients. In addition, 23.7% of participants perceived there was an increase in the number of euthanasia procedures they had to perform, and 26.3% perceived there was an increase in treatment refusal from their clients. Proportions of respondents who perceived increases in new clients, refusal of new clients, euthanasia, and treatment refusal were similar at T2, but there was poor agreement between perception of T1 and T2 (Appendix II, available online from: www.canadianveterinarians.net).

Stress and quality of life (QoL)

Differences in PSS and QoL scores between T1 and T2 for each participant were distributed around 0 (no difference; Appendix II, available online from: www.canadianveterinarians.net). Due to the important loss to follow-up and limited differences in outcomes between T1 and T2, only data from T1 were analyzed in the following steps. The median PSS score was 22 (range: 3 to 36) and the median QoL utility score was 0.83 (range: 0.17 to 0.95). At T1, 30.9% (95% CI: 26.6 to 35.5) of participants had a low compassion satisfaction score, 65.2% (95% CI: 60.5 to 69.6) had a high burnout score, and 83.1% (95% CI: 79.1 to 86.4) had a high secondary traumatic stress score.

For every 1-point increase in adoption of measures to limit the spread of COVID-19 index, the perceived stress of the participants decreased by 2.8 points on the PSS (Table 3). Participants who perceived there was an increase in client refusal and in euthanasia since the beginning of the pandemic had slightly greater PSS than participants who did not perceive these changes (1.6 and 1.4 points, respectively; Table 3). Participants who reported not working at the time of the survey had 4.8 and 8.2 × higher odds of low compassion satisfaction and high burnout scores, respectively, than participants who worked 40 to 50 h/wk (Table 4). Participants who reported working part-time also had 2.6 × greater odds of low compassion satisfaction than participants who worked 40 to 50 h/wk. In addition, for every 1-point increase in perceived susceptibility, perceived effectiveness of measures, and adoption of preventive measures indices, odds of having a high burnout score were 0.9, 0.5, and 0.3 × lower, respectively (Table 4). Participants who thought there was an increase in euthanasia in their practice had 1.9 × greater odds of having a high burnout score (Table 4). Finally, participants who reported an increase in clients in their practice had 2.4 × greater odds of having a high secondary traumatic stress score (Table 4). No professional risk factors were associated with a difference in QoL as measured by the EQ-5D (all models are presented in Appendix III, available online from: www.canadianveterinarians.net).

Table 3.

Estimates and 95% confidence intervals (CI) from linear regression models for variables associated^a with perceived stress of animal health workers in Canada during the COVID-19 pandemic (August to December 2020).

Variables	Crude estimate (95% CI)	Adjusted estimate (95% CI)	Confounders included in models
Adoption of measures against COVID-19 index	−3.35 (−4.74; −1.96)	−2.81 (−4.18; −1.44)	Age
Increased new client refusal	1.90 (0.69; 3.12)	1.56 (0.36; 2.76)	Occupation and increase in new clients
Increased euthanasia	2.27 (0.84; 3.69)	1.44 (0.04; 2.83)	Age, occupation, and increase in new clients

^aOnly adjusted associations with P < 0.05 are presented (all models are available in Appendix III, available online from: www.canadianveterinarians.net).

Table 4.

Odds ratio (OR) and 95% confidence intervals (CI) from logistic regression models for variables associated^a with 3 tools measuring the professional quality of life (ProQoL) of animal health workers in Canada during the COVID-19 pandemic (August to December 2020).

Variables	Crude OR (95% CI)	Adjusted OR (95% CI)	Confounders included in models
Low compassion satisfaction
Work schedule
Worked 40 to 50 h/wk	Ref.	Ref.	Age and region
Did not work	4.20 (1.53; 12.14)	4.75 (1.69; 14.04)
Worked part-time	2.09 (1.15; 3.77)	2.58 (1.38; 4.84)
Worked < 40 h/wk	1.33 (0.76; 2.30)	1.60 (0.89; 2.86)
Worked > 50 h/wk	1.13 (0.49; 2.45)	1.39 (0.59; 3.09)
Perceived susceptibility to COVID-19 index	0.57 (0.40; 0.80)	0.88 (0.82; 0.95)	Age and region
Adoption of measures against COVID-19 index	0.45 (0.28; 0.74)	0.40 (0.24; 0.68)	Work schedule and region
High burnout
Work schedule
Worked 40 to 50 h/wk	Ref.	Ref.	Gender, experience, and region
Did not work	8.13 (1.60; 148.3)	8.20 (1.59; 150.6)
Worked part-time	0.93 (0.52; 1.69)	1.09 (0.59; 2.06)
Worked < 40 h/wk	0.92 (0.55; 1.56)	1.16 (0.67; 2.04)
Worked > 50 h.0/wk	0.63 (0.31; 1.32)	0.80 (0.37; 1.73)
Perceived susceptibility to COVID-19 index	0.64 (0.45; 0.90)	0.89 (0.82; 0.96)	Age, worrying about COVID-19, and region
Perceived effectiveness of measures against COVID-19 index	0.58 (0.38; 0.87)	0.52 (0.32; 0.84)	Work schedule, perceived susceptibility, worrying about COVID-19, and region
Adoption of measures against COVID-19 index	0.34 (0.19; 0.58)	0.31 (0.16; 0.57)	Age, work schedule, and region
Increase of euthanasia	2.10 (1.27; 3.59)	1.93 (1.16; 3.32)	Experience
High secondary traumatic stress
Increase in new clients	2.46 (1.42; 4.22)	2.40 (1.38; 4.12)	Gender

^aOnly adjusted associations with P < 0.05 are presented (all models are available in Appendix III, available online from: www.canadianveterinarians.net).

Ref. — Reference category.

Discussion

Our results described the KAP of Canadian AHW related to the COVID-19 pandemic, assessed their perceived stress and QoL, and identified meaningful associations among these characteristics. In summary, positive perception and adoption of measures against COVID-19 were associated with less stress and better QoL indicators. Moreover, impacts of the pandemic such as perceived increase in clientele, increase in new client refusal, and increase in euthanasia were associated with more stress and poorer QoL indicators.

Overall, AHW had sufficient knowledge of COVID-19 transmission, perceived they had a high susceptibility to COVID-19, agreed preventive measures were effective, and had adopted good preventive practices. Our study population were Canadian AHW and included mainly educated female respondents in a high-income country, 3 identified factors associated with good COVID-19 KAP in previous studies (27–29).

Stress and quality of life (QoL)

In a recent study in the Canadian veterinarian population that used a similar recruitment strategy to the one in our study, Canadian veterinarians had an average PSS score of 17 before the pandemic; and 28% of them had a low compassion satisfaction score, 42% had a high burnout score, and 65% had a high secondary traumatic stress score (11). The average PSS score of 21 in the present study was higher than that in the study conducted before the pandemic (11). Similarly, there were greater proportions of AHW who had high burnout (65%) and high secondary traumatic stress (83%) scores than in the study by Perret et al (2020) (11). The AHW in the present study also seemed to have a lower QoL (average utility score: 0.80) than the Canadian general population during the early stage of the COVID-19 pandemic (utility score: range = −0.01 to 0.95, median = 0.87) (30).

Compared to existing literature, we inferred that AHW have lower QoL than the general population, corroborating previous studies that veterinarians have increases in anxiety, burnout, depression, and suicide compared to the general population (11,31,32). The population we studied, however, also included animal health technicians. We inferred that these burdens were likely shared by all AHW, not only veterinarians. Indeed, there were no indications that the occupation (technician or veterinarian) of respondents was a risk factor for stress and professional QoL. A recent study from the USA reported that a greater proportion of veterinary support staff experienced serious psychological distress or suffered from burnout compared to veterinarians (14). The use of different tools to assess mental health and QoL and the different sociodemographic contexts in Canada and the USA could explain differences with our results. The effects of stress and poor QoL on professional activities during the pandemic has not been assessed here, but in a study from the USA involving a variety of AHW, 15% of respondents considered mental health a barrier to work during the COVID-19 pandemic (33).

In the present study, stress and professional QoL were worse than those reported by Perret et al (2020) for AHW before the pandemic (11). However, as our study did not assess stress and QoL before the pandemic, it is possible that participants in our study differed from those in the Perret et al (2020) study (11). In this case, discrepancies between studies must be interpreted with caution and remain to be validated. Nonetheless, in support of our observations, a longitudinal study from the USA reported that the pandemic had negative effects on the well-being of veterinarians and veterinary staff (14). Poorer mental health has also been reported in healthcare workers and in the general population, with some discrepancies based on location, time period, and sociodemographic characteristics (15,34).

Knowledge, attitudes, and practices (KAP)

Preventive measures were not all adopted despite an average perception that respondents were highly susceptible to COVID-19, worried about COVID-19, and strongly agreed that preventive measures were effective against COVID-19. For example, conducting phone triage was reported to be always done by < 1/2 of participants. This survey did not allow for exploration of reasons why this relatively simple measure was less commonly implemented, but leaving the definition of “emergency” cases to veterinarians’ judgement could have led veterinarians to regard phone triage as unnecessary (2). For example, castration of a young male dog may not seem to be an urgent operation, but delayed castration may cause aggression to develop in the dog and may lead to the eventual abandoning of the dog by the owner due to problematic behavior. In a study during the early stages of the pandemic, emergency cases were not managed uniformly by veterinary practices (35). Emergency case definition and management were not assessed in the present study, but we believe that the lack of a clear definition of what can be considered “emergency” cases can add stress to the daily work of veterinarians, even after the pandemic.

Measures such as wearing a mask, limiting the number of clients, using PPE, and regularly disinfecting surfaces were generally adopted. These 4 measures were in line with the general recommendations and requirements in most Canadian provinces during the study period, which can explain the high compliance compared to AHW-specific measures (e.g., phone triage and use of plexiglass). Adoption of various practices was also described by Muzzatti and Grieve (2022), but the qualitative approach of their research did not identify a gap in adoption of various practices (36). Telehealth services increased in veterinary practices in the USA in the early phase of the pandemic, but this increase only represented 38% of respondents, with discrepancies among regions (37). In clinics offering telehealth services, teletriage was commonly reported.

We summarized COVID-19 KAP questions into indices including perceived susceptibility, worrying about COVID-19, perceived effectiveness of preventive measures, and adoption of preventive measures. Both perceived susceptibility and perceived effectiveness of measures had a satisfactory Cronbach’s alpha value (> 0.7) (19), but worrying about COVID-19 and adoption of preventive measures indices had a poor alpha. These poor alpha values could have been due to the low number of questions, poor relatedness, or heterogenous constructs. Unfortunately, the factor analysis did not yield better constructs by adding factors to the indices, and these indices (with maximum alpha) were kept to simplify complex statistical models.

Changes in the workload since the beginning of the pandemic were observed by many AHW in this study. Three quarters of AHW perceived an increase in clientele since the beginning of the pandemic, 1/2 had increases in new client refusal, and 1/4 perceived increases in euthanasia and treatment refusal. It is not clear if there were real increases, but mixed methods research, anecdotal news, and internal reports identified an increase in adoptions of companion animals and increases in client numbers and euthanasia in veterinary practices in Canada and worldwide (36–38).

Risk factors

Associations between professional characteristics, COVID-19 KAP, stress, and professional QoL in AHW were, to our knowledge, not previously assessed. Indeed, research to identify risk factors for poor mental health and QoL during the pandemic have focused on activities and sociodemographic and health characteristics (15,39). Only 1 multinational study of the general population reported a weak correlation between mental health and knowledge and attitude scores, without adjustment for potential confounders such as gender and education level (28).

Differences in PSS according to the adoption of preventive measures score and the perceived increases in client refusal and in euthanasia varied between 1.4 and 2.8 points. These differences, although statistically significant, were small, considering that the PSS scale ranges from 0 to 40 (20) and, in a Canadian study on veterinarians, there was a difference of 3 to 4 points between males and females (11). As the literature assessing the association between professional characteristics or KAP and stress is limited, it is unclear how important these differences are. However, finding associations between the same characteristics or KAP and professional QoL indicators suggests these differences might be real, assuming stress and QoL are related. Work-related changes, however amplified by the pandemic, have likely continued since the end of the public health emergency (14). Our findings reiterated the importance of addressing work-related stress and its effects for animal health professionals.

Study limitations

The cross-sectional design used to evaluate associations limited our ability to suggest causation. For example, non-working AHW had greater odds of having low compassion satisfaction and high burnout scores than AHW who were working, regardless of their work schedule. We suppose, in this case, that high burnout scores would result in increased odds of not working, but our study design cannot assess directionality. As we did not assess them in time, all associations should be considered to have temporal ambiguity and causal interpretations should be avoided.

The longitudinal design that was planned for this study unfortunately resulted in poor follow-up, not allowing us to conduct repeated measures analyses. Based on descriptive analyses at T1 and T2, we inferred that sociodemographic characteristics of the population did not change. Moreover, the general absence of change in stress and QoL indicators at T2 was unexpected. When we planned this study, the progression of the pandemic was unknown and the timing of the 2 survey periods might have driven these results. Indeed, T1 was at the beginning of the second wave of cases in Canada (August to December 2020). At that time, no vaccines were available and the public health measures in place included case detection and isolation; contact tracing and quarantine; travel restrictions; restrictive closures (gathering restrictions, nonessential business and school closures); curfews; and personal measures. Conversely, T2 included the end of the third wave of cases (May to July 2021), but at that time, vaccination was available to the adult population and public health measures were generally lifted. These major changes could have had strong effects on the attitudes and practices and the stress and QoL of AHW, but we observed a lack of change between T1 and T2. Perhaps participants who had more stress and a worse QoL did not participate in T2, which would have driven the absence of changes between T1 and T2. Information about the participants’ stress and QoL before the pandemic would have been helpful for assessing how the pandemic affected them, and perhaps for confirming whether our results are comparable to those of similar studies (11), but such information was not assessed in this case.

Inclusion criteria in this study limited participants to companion AHW. Although 1 study reported minimal differences in COVID-19 KAP across job roles (33), it is not clear if findings would have been the same for AHW working in other settings (e.g., large animal clinics or government). This inclusion criterion could have influenced the high proportion of respondents who identify as female. Indeed, the proportion of women in the present study was higher than that in the membership of the CVMA in 2020 (62%) (40), which could have affected our findings. It is likely that our results reflected the female population of companion animal veterinarians, and the male population would need additional research.

Our findings highlighted that changes in companion animal clinics related to the pandemic; i.e., increases in clientele, new client refusal, and euthanasia, were associated with poorer stress and QoL scores in AHW. However, stress and QoL were improved in AHW who had a positive perception of and reported adopting measures against COVID-19. This information should be taken into serious consideration, as the Canadian animal care professions are facing an unprecedented shortage of AHW and the potential for future public health crises cannot be ignored.

Funding Statement

This study was funded by the Centre de Recherche en Santé Publique (CReSP).