Prediction of compliance with preventive measures among teachers in the context of the COVID-19 pandemic

Abstract

The objective of this study is to examine, in primary and high schools, teachers' compliance with preventive infection control measures (in the context of the COVID-19 pandemic). Inspired by the technology acceptance model (TAM) and occupational health and safety (OHS) literature on personal protective equipment (PPE) use, we propose a model of compliance with preventive measures among teachers. Data were collected following an observational, cross-sectional design. The data for the study were collected via a questionnaire survey of teachers working in the province of Quebec, Canada. To study the impact of the explanatory variables on the dependent variable, we developed a multiple linear regression model. This model was estimated to assess the preventive measures as a whole (six items). Results show that having tested positive for a COVID test in the last year, judging that the situation does not require the use of the mask or the protective glasses, training received on preventive measures, factors related to comfort and use of protective eyewear, as well as age influence teacher compliance with COVID-19 preventive measures.

1. Introduction

In 2019, the first cases of coronavirus disease 2019 (COVID-19)—caused by severe acute respiratory syndrome (SARS-CoV-2)—were detected in Wuhan, China. On March 11, 2020, the World Health Organization declared COVID-19 a pandemic (WHO, 2020b). In the following days, many governments worldwide deployed various measures to slow down the spread of the virus among their populations. Important measures to prevent this disease, caught when in contact with body fluids, include the use of personal protective equipment (PPE; Verbeek et al., 2020) and other preventive measures (hand washing, physical distancing, surface cleaning). Preventive measures, compliance, and PPE use depend on various individual and organizational factors, including the availability of clear and complete information on the risks at hand and on the proper use of protective equipment (Savoia et al., 2020).

In school systems, such measures took different forms over time. For example, the Government of Québec (Canada) announced the closure of its entire educational system, childcare, and universities (Government of Quebec, 2020b). In the fall of 2020, following several months of closures, these schools were reopened (Government of Quebec, 2020a). Research has found that schools are at low risk of COVID-19 transmission (El Jaouhari et al., 2021) and has shown a low prevalence of COVID-19 among school-aged youth (Macartney et al., 2020).

Low prevalence in schools does not mean that there is no risk. For example, Macartney et al. (2020) reported that one infected staff member would have contaminated six of his colleagues and six to seven children in a childcare center in New South Wales, Australia. A similar situation was reported in France, where Fontanet et al. (2021) reported a cluster of COVID-19 in one school with a 40.9 % infection rate. Although knowledge of the transmission chains of SARS-CoV-2 in schools is at its beginning (El Jaouhari et al., 2021), we can see that education personnel are exposed to COVID-19 risks (Macartney et al., 2020; Xu et al., 2020). While most studies have focused on the use of PPE in healthcare services, schoolteachers represent a population at risk, and little is known about the factors affecting the use of health and safety measures to control infections in classrooms (Lambert et al., 2020). Although COVID-19 is not recognized as an occupational disease in schools, teachers are at risk of contracting it and developing severe forms (Gaffney et al., 2020) and occupational health and safety (OHS) prevention are necessary for staff and students. Like many other jurisdictions, the Quebec government deployed a safety protocol proposing preventive measures such as physical distancing, frequent hand washing and disinfection of installations, as well as wearing a protective mask for staff and students (over 5th grade; Government of Quebec, 2020c). These preventive measures were consistent with existing knowledge on the control of infectious diseases in the workplace, particularly when wearing PPE (Reddy et al., 2019; Verbeek et al., 2020).

If PPE such as gloves, gowns, masks or face covers, visors, or protective eye wear help protect workers from body fluid exposure (Verbeek et al., 2020), proper use is essential for effective preventive measures (Savoia et al., 2020). Several studies have reported inappropriate PPE use in occupational settings (Currat et al., 2022; Dionne et al., 2021; Savoia et al., 2020; Shelus et al., 2020). While COVID-19 has highlighted the importance of PPE and preventive measures to protect workers' health, it has also revealed failures in their effective implementation. To protect workers effectively, it is important to understand the underlying causes of PPE use and compliance with preventive measures. However, most studies on this topic have focused on health services, despite the fact that teachers are also a vulnerable population. Consequently, little is known about the factors that influence the use of health and safety measures to control infections in classrooms.

Unlike healthcare professionals who are regularly exposed to infectious diseases and for whom the use of PPE and the application of preventive measures are an integral part of their jobs, the above COVID-19 practices are fairly new to teaching professionals and might constitute an innovation in the workplace. In this changing context, recent research has focused on organizational factors (Man et al., 2021; Michie et al., 2014; Savoia et al., 2020) and sociodemographic factors (Sapbamrer and Thammachai, 2020; Wong et al., 2021) as the main determinants for the adoption of safety behaviors and proper PPE use at work.

The objective of this study is to examine, in primary and high schools, teachers' compliance with preventive infection control measures (in the context of the COVID-19 pandemic). Despite guidelines issued by the relevant authorities for the use of protective measures (procedure masks, eye protection, hand washing, etc.), several observations have emerged that the wearing of PPE and compliance with other preventive measures are not always applied in the population (Dionne et al., 2021). Therefore, this study aims to examine the frequency of compliance with preventive measures in the education sector and to identify the factors that influence and predict teachers' behaviors.

1.1. Technology acceptance model (TAM)

The technology acceptance model (TAM) has been widely used to explain and predict individual and collective behaviors underpinning the acceptance of various technologies (Dünnebeil et al., 2012; Paré et al., 2014). Proposed by Davis (1989) and based on the theory of reasoned actions (Fishbein and Ajzen, 1977), the TAM suggests that perceived ease of use and perceived usefulness will influence the intention to use a technology and, ultimately, the use of such technology. Perceived usefulness refers to an individual's perception that using a particular technology would improve performance, while perceived ease of use refers to an individual's perception that using a device would be easy and effortless (Davis, 1989). Venkatesh and Davis (2000) later proposed an extended model (TAM2) by adding antecedents to perceived usefulness. These antecedents refer to social influence processes (e.g., subjective norms, voluntariness, and compliance) and cognitive instrumental processes (e.g., job relevance, output quality, and perceived ease of use; Venkatesh and Davis, 2000).

Considering that our study focuses on preventive measures and PPE use and not new information technologies, we considered the specifics of OHS. Recent studies on the adoption and use of preventive measures—including PPE in healthcare services (Carr et al., 2010; Dünnebeil et al., 2012; Paré et al., 2014) and in OHS (Choi et al., 2017; Man et al., 2021; Wong et al., 2021)—show commonalities with the TAM. For example, Man et al. (2021) identified the factors influencing PPE use among construction workers. Their results highlight the positive influence of perceived ease of use and perceived usefulness on attitudes toward PPE. They concluded that workers with high levels of perceived usefulness and ease of use were more likely to use PPE. They also added a risk perception variable to their model and found that it had a direct impact on PPE use. Choi et al. (2017) also provided empirical support for the TAM through their results, which indicate a positive effect of perceived usefulness and perceived ease of use on workers' acceptance of an OHS device. They also introduced new perspectives to the concepts of perceived usefulness and perceived ease by adding a social influence variable. While they considered perceived usefulness through the perception that the use of the device would imply benefits for OHS, they conceptualized perceived ease of use through the comfort that using the device would bring. Furthermore, the TAM's theoretical contributions to PPE use are reinforced by the findings of Wong et al. (2021), which are fully consistent with the theory initially developed (Davis, 1989). Indeed, these findings support a positive influence of perceived usefulness and perceived ease of use on a “favorable” attitude toward PPE and, consequently, the intention to use it. Taken together, the studies above suggest that the TAM is relevant for identifying the factors predictive of PPE use and teacher health and safety.

In this context, our study adapted the definitions of perceived usefulness and perceived ease of use for them to better reflect OHS specifically. On the one hand, perceived usefulness can be thought of as the perceived benefits of PPE use and sanitation enforcement in relation to COVID-19 prevention. Thus, it can refer to the extent to which teachers believe that PPE use and compliance with sanitary measures can protect them efficiently. On the other hand, perceived ease of use can be seen as the perception of little or no constraints related to PPE use and compliance with sanitary measures. As such, it can refer to the extent to which teachers perceive little or no constraint or discomfort in relation to the use of PPE and application of sanitary measures in classrooms. As defined, perceived usefulness and perceived ease of use may be associated with different factors that have been identified as determinants of worker health and safety behaviors.

1.1.1. Perceived risk and usefulness

Risk perception and perceived benefits of prevention measures are considered important predictors of PPE use and workers' adoption of safe behaviors (Sapbamrer and Thammachai, 2020). Risk perception has a positive influence on attitudes and behaviors toward safety (Arezes and Miguel, 2008; Man et al., 2021). Man et al. (2021) found that workers are more likely to use PPE when they perceive a high level of risk. Tinoco et al. (2019) reported similar findings regarding hearing protection. Studies indicate that risk perception drives workers to use PPE even when it is not mandatory (Lombardi et al., 2009). Some authors have also shown that healthcare professionals do not use recommended PPE when they perceive health and safety risks to be minimal (Harrod et al., 2020) or when working with a patient not known to be infected (Çiriş Yildiz et al., 2022). On the other hand, medical doctors who perceive a high level of risk are keener to adequately protect themselves (Savoia et al., 2020; Tan et al., 2006). From a health and safety perspective, risk perception would be a strong determinant for using PPE or following preventive measures and can clearly be associated with perceived usefulness among education professionals.

1.1.2. Perceived ease of use

Perceived ease of use can include comfort when using PPE. For example, agricultural workers do not wear respiratory masks for comfort reasons even when dangerous pesticides are handled (Sapbamrer and Thammachai, 2020). During COVID-19, physical and psychological concerns have been raised regarding the long-term use of protective masks among medical doctors (Savoia et al., 2020).

Physically, prolonged use of masks (> 4 h) can cause headaches, skin lesions, voice disorders, difficulty breathing, heat stress, and discomfort (Ribeiro et al., 2020; Shelus et al., 2020; Unoki et al., 2021). Psychologically, the use of such masks, especially in a professional setting, is associated with an increase in the perception of vocal effort and communication difficulties (intelligibility and coordination of speech and sound attenuation; Atcherson et al., 2020; Caniato et al., 2021; Corey et al., 2020; Ribeiro et al., 2020). Specific visual-protection equipment (glasses or visors) is associated with impediment during precision tasks, decrease in visual acuity (vision distorted by glasses, fog, scratches), discomfort (Lombardi et al., 2009), and visual limitations (Ruskin et al., 2021). Such perceived ease-of-use factors could very well apply to teaching professionals.

1.1.3. Organizational factors

Consistent with previous research, this study classified organizational factors into three main groups: PPE availability (Lombardi et al., 2009; Savoia et al., 2020), training (Harrod et al., 2020; MacFarlane et al., 2008), and risk awareness measures (Geana, 2020; Shelus et al., 2020; Tan et al., 2006).

1.1.3.1. PPE availability

According to Lombardi et al. (2009), the availability and accessibility of PPE are determining factors for its use. In the first few months of the COVID-19 pandemic, the availability and accessibility of PPE was a critical issue across different countries (WHO, 2020a). Shortages of PPE forced jurisdictions to adopt rationalization strategies, which involved prolonged mask use (beyond the recommended duration) and reuse of single-use equipment (INSPQ, 2021; Savoia et al., 2020). These strategies could have sent mixed messages regarding the importance of properly wearing protective equipment.

1.1.3.2. Training

Training is positively associated with PPE use and conformity to preventive measures. For example, agricultural workers are more likely to use PPE when handling pesticides when properly trained (MacFarlane et al., 2008; Sapbamrer and Thammachai, 2020). Training would not only predict usage and frequency of usage but also conformity to applying preventive measures (Avory and Coggon, 1994). Similar observations have been made in other sectors, such as healthcare (Harrod et al., 2020). During COVID-19, Morioka et al. (2020) noted that knowledge gaps led some healthcare professionals to deviate from the use of PPE and application of preventive measures. Training was also associated with the proper use of PPE. Among healthcare professionals, training can reduce self-contamination when putting on or removing PPE (Currat et al., 2022; Harrod et al., 2020). In this context, training is a predictive variable for the use of PPE and conformity to preventive measures among education professionals.

1.1.3.3. Risk awareness measures

A positive association can be established between awareness, access to information, and safety behaviors. Tan et al. (2006) found that access to proper information and awareness had a positive effect on perceived effectiveness and PPE use during the SARS pandemic of 2003. Recently, Wong et al. (2021) found a positive relationship between safety awareness and PPE use. On the other hand, inadequate information on the need to wear a mask and lack of information on the usefulness of a mask to protect against COVID-19 were found to be associated with a low proportion of mask use among the US population (Shelus et al., 2020). Savoia et al. (2020) reported similar observations among healthcare professionals in Italy during the pandemic.

1.1.4. Socio-demographic factors

Sociodemographic factors, such as age and level of education, are known to be predictive variables of PPE use and conformity to preventive measures (Lombardi et al., 2009; Sapbamrer and Thammachai, 2020; Tinoco et al., 2019). Lombardi et al. (2009) reported a relationship between age and risk perception. Experience is associated with better risk perception and compliance with safety measures. Education level is positively associated with PPE use and compliance with safety measures (MacFarlane et al., 2008; Sapbamrer and Thammachai, 2020; Tinoco et al., 2019).

1.2. Predictive model on compliance with preventive measures among teachers

Inspired by the TAM and OHS literature on PPE use (Choi et al., 2017; Man et al., 2021; Wong et al., 2021), we propose a model of compliance with preventive measures among teachers. Consistent with previous research that proposed TAM adaptations (Carr et al., 2010; Venkatesh and Davis, 2000) while integrating a specific OHS approach (Man et al., 2021), our model includes factors affecting PPE use and conformity with preventive measures. The model proposes additional variables, such as organizational and socio-demographic variables, in addition to the original TAM. The model suggests that risk perception and PPE usefulness (individual factors), PPE-related factors, organizational factors, and socio-demographic factors are antecedents of PPE use and compliance with preventive measures (Fig. 1 ).

Fig. 1.

Predictive model on compliance with preventive measures among teachers.

2. Methods and measures

Data were collected following an observational, cross-sectional design. The data for the study were collected via a questionnaire survey of teachers working in the province of Quebec, Canada. Given the nature of the position held by the respondents, the questionnaire was administered online (via LimeSurvey software). A link to the questionnaire was relayed to potential participants by union and employer associations as well as through social media such as Facebook and LinkedIn. This mode of data collection also ensured anonymity and confidentiality and thus minimized social desirability bias. The survey was conducted between May 20 and July 5, 2021, with 252 teachers. The data were automatically aggregated in the software and subsequently exported in SPSS format. The teachers answered the questionnaire in French. Before sending, the questionnaire was pretested.

This study evaluates the factors that may influence teachers' behaviors from a TAM perspective. Since, to our knowledge, there is no validated model for the study of this phenomenon, the proposed model is exploratory and aims first to identify the factors that influence teachers' behaviors. Thus, the analyses carried out were single-level. Multi-level analyses could be conducted in future research, particularly to add control or intermediate variables or to control for respondents' social desirability.

Based on the model suggested above, we classified, in four categories, the factors that are likely to influence the dependent variable on preventive measures compliance: (i) organizational factors, (ii) PPE-related factors, (iii) individual factors, and (iv) socio-demographic variables. The measurements of the dependent and independent variables are detailed in the next section.

2.1. The dependent variable on preventive measures compliance (PMC)

In the context of the COVID-19 pandemic, countries have adopted guidelines to oversee preventive measures in the workplace. In Quebec, it is the Occupational Health and Safety Commission (CNESST) that has produced guidelines, including health standard guides for the school network. Therefore, our study was based on CNESST (2021) guidelines to define preventive measures required in the education sector at the time of our investigation.

We measured compliance with preventive measures through a series of statements about wearing protective equipment (procedural masks, glasses, or visors), washing hands, and cleaning work surfaces and common areas. Respondents were asked to assess how frequently they engaged in each activity using a five-point scale ranging from 1 (never) to 5 (always).

The scale used in this study included six activities of preventive measure compliance, as presented in Table 1 . Hence, the degree of compliance with preventive measures was measured using the sum of the scores of items corresponding to responses to these six assertions. Scores of the respondents, which initially ranged from 6 to 30, were weighted to consider “does not apply” answers. Thus, for each respondent, the sum of their scores was divided by the number of applicable items. Even though the initial index has integer values from 1 to 5, once weighted, it can take non-integer values. Both unidimensionality and internal consistency of the global index were assessed using principal component factor analysis and Cronbach's alpha (see the Results section).

Table 1.

Preventive measures among teachers in the context of the COVID-19 pandemic.

Activity 1 When required, I wear a procedure mask.

Activity 2 I change my procedure mask at least once a day.

Activity 3 I wear my glasses or my protective visor when a distance of 2 m cannot be respected.

Activity 4 I clean my hands regularly.

Activity 5 I clean the work surfaces (desk, chair, board, etc.) after my visit.

Activity 6 I clean the common areas (table and chair used for breaks, meals) after my visit.

2.2. Organizational factors

Three variables were assessed as organizational factors that could impact teachers' compliance with preventive measures. The first is the availability of material and protective equipment, which can have an impact on contracting COVID-19 (Ahmad et al., 2022). Accessibility and availability are well-known organizational factors that impact PPE use (Lombardi et al., 2009). In this study, the level of material and protective equipment availability was measured using an index of two items assessing whether (i) the protective equipment (masks, glasses, etc.) is available and provided in sufficient quantity by the establishment and whether (ii) equipment for hand washing and surface disinfection is available and placed in strategic places in the establishment. For each item, respondents were asked to assess whether they agreed or disagreed using a five-point scale ranging from 1 (strongly disagree) to 5 (strongly agree). The index was thus the sum of the scores of the items corresponding to the teacher's responses. Respondents' scores, which initially ranged from 2 to 10, were weighted to consider “does not apply” answers. Thus, for each respondent, the sum of their scores was divided by the number of applicable items. Even though the initial index ranges from 1 to 5, once weighted, it can take non-integer values. As for the dependent variable, both unidimensionality and internal consistency of the global index were assessed (see the Results section).

As a second organizational factor variable, we assessed training, which is likely to be an important intervention for reducing risk exposure (Bhamra et al., 2021; MacFarlane et al., 2008). Risk perception plays an important role in decisions to apply safety measures, and training is an essential element (Lombardi et al., 2009). Training was measured using a three-item index regarding whether teachers reported having received, since the beginning of the COVID-19 pandemic, face-to-face or on-line training on (i) the use of procedure masks (wearing, adjustment, handling), (ii) the use of protective glasses or visors (wearing, cleaning), and (iii) hand washing. For each statement, teachers were asked to assess whether they agreed or disagreed using a five-point scale ranging from 1 (strongly disagree) to 5 (strongly agree). The statistics of the weighted index are presented below (see the Results section).

We also measured the degree of awareness and information provided in the workplace (Legault Faucher and Mélançon, 2003). This organizational variable was assessed using a five-item index assessing whether teachers agreed or disagreed with the following items: (i) awareness posters are placed in the workplace to remind people of preventive measures, (ii) I receive emails to remind me to respect preventive measures, (iii) my colleagues remind me to respect the preventive measures, (iv) the school management reminds me to respect the preventive measures, and (v) a resource person is present in the school to encourage us to respect the preventive measures and to answer our questions, if necessary. Thus, for each teacher, a weighted index of awareness and information was calculated. Both the unidimensionality and internal consistency of the global index were assessed (see the Results section).

2.3. PPE-related factors

Personal protection equipment can lead to discomfort or even injury to those who wear it. In a recent scoping review, Unoki et al. (2021) found adverse effects of wearing PPE during COVID-19, such as headaches, voice disorders, skin injuries, and other manifestations. These adverse effects could influence the way teachers respect (or not) preventive measures (Legault Faucher and Mélançon, 2003). We measured factors related to personal protective using two variables about the characteristics of procedure masks and of glasses or protective visors. Accordingly, we asked teachers to assess, when they did not use a procedure mask, whether it was influenced by (i) the discomfort (e.g., heat, difficult difficulty), (ii) the injuries that it generated (e.g., skin problem), or (iii) the fact that it interfered with the tasks (e.g., difficult communication). Similar questions were asked about eye protection, discomfort, injuries (e.g., headache), and interference with tasks (e.g., poor vision, fogging in glasses). For each statement, teachers were asked to assess whether they agreed or disagreed using a five-point scale ranging from 1 (strongly disagree) to 5 (strongly agree). From the weighted index, we created binary variables coded 1 (yes) if the teacher-weighted index was >3.5, and 0 (no) otherwise.

2.4. Individual factors

Three variables assessed whether individual factors influenced compliance with preventive measures. We first estimated the teachers' attitude toward the need (or not) to wear a mask. We also estimated their attitudes toward wearing (or not) eye protection. Risk perception and exposure are associated with a positive impact on attitudes toward using PPE (Man et al., 2021). For these two variables, we asked teachers whether they agreed (5) or not (1) with the fact that when they do not wear masks or eye protection, it is because they judge that the situation does not require it. For analysis purposes, binary variables were created from these questions. Thus, the answers of agree (4) or strongly agree (5) were categorized as 1 (yes); other responses or “not applicable” responses were categorized as 0 (no). A third individual factor concerns whether respondents had (1 = yes) or had not (0 = no) received a positive test result for COVID-19 in the last year. In other sectors, having experienced an illness is associated with the use of PPE in the workplace (Sapbamrer and Thammachai, 2020). Statistics on these variables are presented in the Results section.

2.5. Socio-demographic variables

The socio-demographic variables were gender, school level, COVID-19 outbreak in the workplace, and age. Socio-demographic factors are known to affect PPE use. Age is associated with a wider use of PPE; young workers are less likely to perceive the risks and less likely to use protective equipment (Lombardi et al., 2009). Education level and literacy level are also associated with higher use of PPE (Sapbamrer and Thammachai, 2020).

We first asked participants what gender they identified with. The response “man” was categorized as 0, and “women” 1. At the school level, we asked them to specify whether they had worked in a primary school (0) or a high school (1). They were also asked to signal whether there had been a COVID-19 outbreak in their workplace in the previous year. A “no” response was coded as 0, and “yes” was coded as 1.

Finally, the variable age was measured as follows: 29 years old and under was a binary variable coded 1 if the teacher reported this age, and 0 otherwise; 30 to 39 years old was a binary variable coded 1 if the teacher reported this age, and 0 otherwise; 40 to 49 years old was a binary variable coded 1 if the teacher reported this age, and 0 otherwise; and 50 years old and over was a binary variable coded 1 if the teacher reported this age, and 0 otherwise.

3. Results

For the dependent and independent variables, based on multiple item scales included in the econometric model, we conducted principal component factor analysis (PCFA) on the construct scales to assess their unidimensionality. Assessments of statistical reliability were performed by computing Cronbach's alpha.

For the dependent variable, preventive measures compliance, the result of the PCFA indicates that one factor explains 40.10 % of the original variance of the phenomenon studied with an initial eigenvalue of 2.406. For the three independent variables based on multiple-item scales, namely, the index referring to material protective equipment availability (two items), training (three items), and awareness and information (five items), the results of the PCFA indicate that, in all cases, one factor explains, respectively, 25.16 %, 36.23 %, and 19.82 % of the original variance of these three constructs. Once the unidimensionality of the additive scales measuring the three independent variables based on multiple-item scales was established, an assessment of statistical reliability was necessary. To make such an assessment, item analysis of the components of these additive scales was performed by computing Cronbach's alpha. This coefficient provides a reliability coefficient for multiple-item scales, such as those included in the scales of the four variables. Cronbach's alpha was 0.649 for the six items of preventive measure compliance, 0.656 for the two items of the material and protective equipment availability index, 0.888 for the three items of the training index, and 0.629 for the five items of the awareness and information index. The statistics for the variables used in this study are presented in Table 2 .

Table 2.

Descriptive statistics.

Variables	Type of variables	Min.	Max.	Mean	Standard deviation	Cronbach's ∝
Continuous variables:
Preventive measures compliance	Index: six items	1	5	4.24	0.61	0.649
Material and protective equipment availability	Index: two items	1	5	4.31	0.93	0.656
Training	Index: three items	1	5	2.30	1.21	0.888
Awareness and information	Index: five items	1	5	3.38	0.72	0.629
Categorical variables
Not judging the need to wear a mask	•No •Yes					96 % 4 %
Not judging the need to wear eye protection	•No •Yes					63.9 % 36.1 %
Procedure mask characteristics	•No •Yes					96 % 4 %
Eye protection characteristics	•No •Yes					61.9 % 38.1 %
Received a positive test result for COVID in the last year	•No •Yes					96 % 4 %
Had a COVID outbreak in the workplace in the last year	•No •Yes					30.4 % 69.6 %
Gender	•Man •Woman					10.4 % 89.6 %
Age	•29 years old and under •30 to 39 years old •40 to 49 years old •50 years old and over					10.4 % 30.7 % 36.2 % 22.7 %
School level	•Primary school •High school					56.2 % 43.8 %

Number of teachers = 252.

3.1. Descriptive analysis

In this study, compliance with preventive measures is operationally defined as including six activities (see Table 1). Computation of the answers to the questions on preventive measures compliance shows a generally high compliance. More precisely, about 99 %, 83 %, and 97 % of the teachers surveyed often or very often, respectively, wore a procedure mask (activity 1), changed the procedure mask at least once a day (activity 2), and washed hands regularly (activity 4). Conversely, 48.8 % of the respondents often or very often wore eye protection when a distance of 2 m could not be respected (activity 3). Lastly, about 75 % of teachers reported cleaning often or very often work surfaces and common areas (activities 5 and 6). Globally, the average index (six items) of preventive measures compliance indicates a mean of 4.24 (on a 1–5 scale), thus indicating a preventive measure compliance value among teachers of >4 (often). The frequency distribution of compliance with the preventive measures is shown in Table 3 .

Table 3.

Frequency distribution of preventive measures among teachers in the context of the COVID-19 pandemic and average of the global index.

How frequently do you apply the following preventive measures?	Scale measurement (in % of researchers) (N = 252)						Median (mode)
	Not applicable/Missing data	Never (1)	Rarely (2)	Sometimes (3)	Often (4)	Always (5)
Activity 1	0	0	0	0.8	7.9	91.3	5.00 (5)
Activity 2	0	2.4	6.7	7.5	24.2	59.1	5.00 (5)
Activity 3	2.8	22.2	13.9	12.3	20.6	28.2	4.00 (5)
Activity 4	0.4	0	0.4	1.6	19.4	78.2	5.00 (5)
Activity 5	0.8	4.4	6.3	13.1	34.9	40.5	4.00 (5)
Activity 6	6	4.4	6.3	7.5	24.2	51.6	5.00 (5)
Global preventive measures compliance index				Average 1–5 scale (SD)			4.24 (0.61)

SD: Standard deviation.

3.2. Regression results for the prediction of preventive measures compliance

To study the impact of the explanatory variables on the dependent variable, we developed a multiple linear regression model. This model was estimated to assess the preventive measures as a whole (six items). To check for multicollinearity between the variables, we observed the correlation matrices and found that the highest correlation was 0.687. Similarly, the smallest value observed in the tolerance test was 0.489, suggesting the absence of multicollinearity (Menard, 1995). After verifying collinearity and model linearity assumptions, we obtained the results presented in Table 4 .

Table 4.

Regression equation predicting degree of compliance with preventive measures.

Dependent variable: Preventive measures compliance (index of six items)
Independent variables	Standardized coefficients (β)	P value
Organizational factors:
Material and protective equipment availability (two items)	−0.050	0.320
Training (three items)	0.116	0.031**
Awareness and information (five items)	0.055	0.302
Personal protective equipment–related factors
Procedure mask characteristics	0.023	0.739
Eye protection characteristics	−0.287	<0.001***
Individual factors (perceived risk and usefulness):
Not judging the need to wear a mask	−0.375	<0.001***
Not judging the need to wear eye protection	−0.259	<0.001***
Received a positive test result for COVID in the last year	0.109	0.030**
Socio-demographic variables:
Gender (being a woman)	0.118	0.022**
School level (working in a high school)	0.076	0.137
Had a COVID outbreak in the workplace in the last year	0.041	0.410
29 years old and undera	−0.130	0.033**
30 to 39 years olda	−0.151	0.030**
40 to 49 years olda	−0.102	0.136
Number of cases: 238
Adjusted R2: 0.438

*, **, and *** indicate that variable is significant at 10 %, 5 %, and 1 %, respectively.

Two-tailed test.

^aWith “50 years old and over” as the reference category.

As we can observe, some factors influenced and predicted teachers' compliance with preventive measures. These factors were the training received by teachers, eye protection characteristics, not judging the need to wear a mask and eye protection, received a positive test result for COVID in the last year, being a woman, and the age of teachers. For this last variable, the results show that being “29 years old and under” and “30 to 39 years old” negatively predicted compliance with preventive measures, in comparison with being “50 years old and over.”

The value of the adjusted R² was 0.438, which is the degree of variance in the magnitude of respect for preventive measures explained by this model. Furthermore, the degree of meaning of the model was <0.001, which suggests that the null hypothesis that all the parameter coefficients are zero is strongly rejected. Consequently, the model was significant at the 1 % level.

4. Discussion and conclusion

Inspired by the TAM (Davis, 1989) and the OHS literature on PPE use (Choi et al., 2017; Man et al., 2021; Wong et al., 2021; Man et al., 2021), this study presents a predictive model of how teachers applied preventive measures in the context of the COVID-19 pandemic. Evaluating the application of preventive measures with concrete activities, this study thus adds to the evidence about transfer practices by examining them for teachers in a very peculiar context.

The findings of this study show that teachers exhibit very satisfactory compliance with preventive measures. Of the six activities explored, wearing eye protection was the one with the lowest compliance (48.8 %). The result is not surprising when we look at the fact that teachers reported that characteristics of eye protection (discomfort, injuries, headache, poor vision, and fogging) are strongly negatively related to compliance with preventive measures. These results must be put in perspective that teachers reported that protection glasses were either mandatory (58.9 %) or recommended (34 %) in their workplace. Despite this, wearing eye protection was not very high.

Regarding training, most teachers stated that they had not received, since the start of the COVID-19 pandemic, face-to-face or online training on the use of procedural masks (wearing, adjusting, handling), use of protective glasses or visors (wearing, cleaning) and on hand washing. Accordingly, the results show an average global index of 2.30, which means that, on average, the teachers did not agree that they had followed a training course. Face-to-face and online training can improve risk perception and practices (Bhamra et al., 2021). Indeed, in light of our results, teacher training could have improved compliance with preventive measures directly, but also indirectly, through its impact on the judgment of the need to wear eye protection, while risk perception and awareness are important in wearing protection (Bhamra et al., 2021; Simard et al., 1991; Tinoco et al., 2019). Not only could this training focus on PPE use and the application of other preventive measures, but also on the notion of risk to ensure a common understanding of the need to apply preventive measures and the perception of the effectiveness of these measures to limit the spread of infections.

Above all, all three variables measuring individual factors (perceived risk and having received a positive test result for COVID-19) were significant in our model. This adds to the evidence for the validation of the TAM, but in a particular sector and context, namely, teacher compliance with COVID-19 preventive measures in the OHS field.

Results also show that being a woman seems to have a positive impact on protective measures compliance. These results pertaining to gender are consistent with previous findings (Reed et al., 2006).

The study reveals that the age of teachers is related to compliance with preventive measures. Being “29 years old and under” and “30 to 39 years old” negatively predicted compliance with preventive measures, in comparison with being “50 years old and over.” This could be explained by the fact that COVID-19 has affected older people more, among whom there are more deaths and hospitalizations. This could explain why older individuals seem more aware of complying with the requirements regarding preventive measures. This result is consistent with other studies in different sectors that observed an association between age and PPE use and safety practices (Sapbamrer and Thammachai, 2020).

Inspired by recent OHS studies (Choi et al., 2017; Man et al., 2021; Wong et al., 2021), we used the TAM (Davis, 1989) while adding variables affecting attitudes toward preventive measures. This model allowed us to obtain a clear picture of teachers' compliance with preventive infection control measures during the COVID-19 pandemic. More recently, the TAM has evolved owing to the proposals of antecedent variables for perceived usefulness (Venkatesh and Davis, 2000). Future research could test the TAM2 by adding variables such as experience (Choi et al., 2017) or having experienced COVID-19 symptoms (Sapbamrer and Thammachai, 2020) as antecedents (intermediate variable) of the perceived usefulness of PPE use.

4.1. Research limitations

Because of the teachers' perspectives used in this study, we did not include other variables related to the context of the pandemic and local practices. Some other limitations could restrict the application of our study; notably, the number of respondents could limit the analysis and significance of the variables. In addition, our study is not a direct measure of practices but a report of the practices and perceptions of teachers.

4.2. Practical implications/applications to practice

It is important to understand prevention practices in the educational setting so that we can prevent the many outbreaks that were noted in this area during the COVID-19 pandemic in Canada. Thus, the identification of elements that can predict the behaviors of teachers would make it possible to better equip states and the education sector to lead them to act, allowing them to improve practices in schools and reduce outbreaks in this sector.

4.3. Originality/What is new about your research

To our knowledge, this study is the first to provide a model for predicting and explaining teachers' compliance with preventive measures related to infection control in school settings. Given the importance of this sector for the well-being of children and the economy of countries, improving practices and limiting the spread of infections in this sector must be part of comprehensive prevention strategies at the global level.

CRediT authorship contribution statement

Elena Laroche: Conceptualization, Methodology, Formal analysis, Investigation, Writing – Original, review and editing.

Pierre-Sébastien Fournier: Conceptualization, Writing, review and editing, Validation.

Nafissatou Cynthia Ouedraogo: Conceptualization, Writing, review and editing

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declaration of competing interest

Nothing to declare.

Biographies

Elena Laroche is a professor at the Department of Management at FSA Laval University in Canada. Her work focuses on improving occupational injury and organizational health through the application of preventive practices that promote risk prevention. She uses knowledge transfer models to study the processes that guide the application of preventive practices that promote occupational health and prevention. Elena Laroche is a researcher at the Centre d'expertise en gestion de la SST (Laval University).

Pierre-Sébastien Fournier is Full Professor and Director of the Department of Management at FSA Laval University in Canada. He holds a PhD degree in Industrial relations from Laval University. His research in Occupational Health and Safety Management and in Human Resources Management aims at understanding the daily reality of work to propose practical interventions to improve health, safety, and wellbeing in the workplace. His work also focuses on workload risks of managers and their employees.

Nafissatou Cynthia Ouedraogo is a PhD student in management at the Faculty of Administrative Sciences of Laval University. Affiliated with the Centre d'expertise en gestion de la santé et de la sécurité (Laval University), her research interests focus on the transfer of occupational health and safety knowledge, with a particular emphasis on identifying the factors contributing to the integration of new OHS knowledge in organizations. She is currently conducting research on the role of managers in the adoption of new OHS practices.

Data availability

Data will be made available on request.