Healthcare workers (HCWs) play a critical role in providing quality healthcare for patients infected with SARS-CoV-2, the virus that causes COVID-19. Critical care and emergency department nurses, as well as other healthcare workers, must ensure effective infection prevention and control procedures (IPC) are appropriately used to prevent nosocomial spread and disease acquisition.1 Simultaneously, there are competing pressures and additional challenges in providing care, including higher patient volumes and acuity, a higher personal risk, limited and more judicious resource use, and increased rates of burnout.2,3,4,5 There is increased anxiety and fear among front-line HCWs such as nurses for their own safety and that of their families.6 Organizational and social factors, including HCW confidence in their ability to effectively deliver IPC, are important to protect physical and psychological health.7 Further, workplace culture, as well as an increased support by local institutions and management, are key factors in adherence to IPC guidelines.8,9,10

During the COVID-19 pandemic, use of personal protective equipment (PPE) and trust in institutions’ differing PPE recommendations and requirements has become a significant issue. While PPE recommendations in hospital settings have been largely consistent, use and availability of equipment has been variable.11 During previous outbreaks, willingness to work decreased when there was a lack of PPE and when there was a perception of putting oneself or one’s family at risk.6,12 Further, trust in government has historically been considered an important determinant of citizens’ adherence to public health policies and guidelines. Belief in recommendations from government and public health officials leading to the adoption of necessary behaviours to decrease risk, as well as a decrease in anxieties, stems from effective communication from officials along with actions that follow-through, and satisfaction with their performance.13,14 A better understanding of factors associated with institutional trust among HCWs is needed to inform policy and practices to support staff health and wellbeing. In this study, our primary objective was to evaluate ICU and emergency department nurses’ barriers and facilitators of IPC guideline adherence, along with their level of trust in organizational and government readiness. We sought to characterize their perceptions of IPC guidelines they have received, adherence to IPC protocols, readiness to adhere to IPC guidelines to prevent SARS-CoV-2 infection in their place of work, and their perceived level of personal risk.

Methods

We conducted a cross-sectional online survey to rapidly assess the views and readiness of Canadian HCWs who were providing care in community and hospital settings (Electronic Supplementary Material [ESM]). We sought to follow previously published survey reporting guidelines.15,16 We used a survey that was developed by the World Health Organization COVID-19 Research Roadmap Social Science and IPC working groups at the onset of the COVID-19 pandemic, because no previously validated measure of HCW perceptions was suitable.17,18,19 Internationally, the survey comprised basic respondent demographic and pandemic work experience characteristics as well as a series of closed-ended questions eliciting beliefs and practices of HCWs in the context of the COVID-19 outbreak. Once these questions were developed, the Theoretical Domains Framework (TDF) was used to evaluate the completeness of included items.20,21 The TDF, which had previously been applied to understanding clinicians’ behaviour, provides a framework that captures core constructs from multiple behavioural theories into 14 domains.22 Questions for this survey addressed the following TDF domains: knowledge; skills; social/professional role and identity; beliefs about capabilities; beliefs about consequences; environmental context and resources; and intentions, social influences and emotions. Additional items in the survey, not included in the TDF framework, assessed three dimensions of institutional trust (competence, honesty, acting in best interests of HCWs) and were based on a previously validated measure.23 Items were created on a seven-point Likert scale, ranging from “strongly disagree” to “strongly agree”. A total of 41 Likert scale questions were asked as well as 13 targeted questions regarding basic demographics and work context. Our survey was adapted to fit the Canadian context; additions and changes to the international version included non-binary gender options and rewording of clinical roles, practice setting, and employment type for consistency with Canadian terminology.

We employed convenience sampling through email and Slack channels via the distribution lists of the Canadian Association of Critical Care Nurses (CACCN) (1,100 list serve members) and the Canadian Association of Emergency Physicians (CAEP) (1,857 listserve members). Both organizations approved access to their listserves for this purpose. These organizations were chosen as they were felt to be representative of HCWs in a variety of Canadian critical care and emergency settings; they are large, broad networks with membership across Canada and representation from both academic and community hospitals. The survey was distributed between 16 March 2020 and 25 May 2020, over the course of the initial surge of COVID-19 in Canada (ESM, eFigure).24 The survey was sent out weekly for three consecutive weeks. The survey remained open for an additional two weeks following the final reminder. Study data were collected and managed using REDCap electronic data capture tools hosted at BC Children’s Hospital.25 All questionnaires were self-administered digitally, with voluntary participation, and consent was implied by survey completion. We did not collect re-identifiable information, including any linked personal identifiable information. The protocol was approved by The University of British Columbia/Children’s and Women’s Health Centre of British Columbia Research Ethics Board (Vancouver, BC, Canada) (Reference number: H20-00803). Research Ethics Board approval was not required or sought from investigators’ other affiliations.

Although distributed across multi-disciplinary networks, the majority of respondents were nurses. As a result of the poor response rate from HCWs in non-nursing categories limiting our ability to make valid comparisons across groups, we limited our analysis to nurses.

Statistical analysis

Our analysis was primarily descriptive. Incomplete surveys were included for the questions to which they contributed. We present descriptive statistics as proportions and mean (standard deviation [SD]) for continuous variables or median [interquartile range (IQR)] for discrete scales. Seven-point Likert scales were used to assess agreement with statements of belief, with a seven indicating “strongly agree” and one indicating “strongly disagree”. Our survey aimed to align analytical approaches with other adaptations of this survey being conducted in other international settings.

A series of composite scores was created as a summation of the Likert responses in the category. To assess trust in health facilities and government, the survey tool contained a validated trust measure, comprising three statements graded on a seven-point Likert scale that capture three different dimensions of trust: perceptions of competence, honesty, and actions that are in the employees’ best interests. These levels of trust were each asked of three levels of governance: local health facility, regional government, and national government (ESM).23 These nine questions were combined into a single “Trust score” out of a maximum of 63 points. Furthermore, we combined the scored Likert responses under the following thematic groups to further interpret the data: emotions, service demand, environment, skills and intentions, beliefs, and social influences/professional role. The Emotions score was based on responses to questions regarding perceived personal risk and fear on the job. The Service Demand score reflected perceptions of whether the health system can handle current and future patient demands. The Environment score reflected the clarity of reporting measures of exposures, guidance materials, and ease of access to infection control practices. The Skills and Intentions combined score reflected training, confidence, and use of PPE. The Professional and Social Role score reflected how perceptions of IPC measures are reflected in one’s role and the influence on others’ IPC adherence. Finally, the Beliefs score was calculated from answers regarding their beliefs in the effectiveness of PPE and IPC procedures, and the amount of strain these procedures create. For the belief category, where two questions were framed in the positive and two in the negative, negative scores were reversed and added to positive ones to develop a composite score.

We stratified number of children into any vs no children, and stratified years since completion of training into < 5 years, 5–10 years, 11–20 years, and > 20 years. We additionally stratified marital status into no partner, partnered, and prefer not to say, and practice setting into community hospital, academic hospital, and other. We created an aggregate COVID Care score based on whether respondents reported any suspected or confirmed COVID-19 cases in their hospital (yes or no/unsure), and whether they cared for a suspected or confirmed COVID-19 patient (yes or no/unsure), with “yes” responses corresponding to one point and “no” or “unsure” responses contributing to no points, with the total score ranging from 0 to 2. Similarly, we created an aggregate epidemic score, based on whether they’ve worked in a clinical setting (yes or no/unsure) in a previous epidemic and cared for a patient in that setting (yes or no/unsure), with “yes” responses corresponding to one point and “no” or “unsure” responses contributing to no points, with the total score ranging from 0 to 2.

We compared responses to individual questions using Kruskal–Wallis or Chi square tests for categorical variables and compared continuous variables using univariate linear regressions and Student’s t tests. We assessed responses to composite score measures using one-way analysis of variance (ANOVA) and assessed categorical and continuous variables using linear regressions and student t tests, respectively. We reported the T-statistic from Student’s t tests as T, the beta from linear regressions as B, and the F-statistic from one-way ANOVAs as F. To control for potential type 1 errors from multiple tests, a new threshold of significance was calculated using the Bonferroni correction for each model. The adapted level of significance for these correlations was P = 0.0055 (0.05/9).

Results

Respondent characteristics

We analyzed responses from 319 nurses. The general survey completion rate for all respondents was 78%. The total response rate was incalculable because of the nature of the distribution network across more than one listserv.

The majority of respondents were bedside ICU and emergency nurses (n = 256), with the remainder working in a variety of nursing roles including nurse specialists, nurse educators, and charge nurses. The majority were female (83%) and partnered (71%), with a mean (SD) age of 41.7 (11.5) yr (Table 1). Almost two-thirds of respondents worked in academic hospitals (63%), while 35% worked in community settings and the remainder worked in other settings. On average, respondents had been in practice for 15 (11) yr. Over half of respondents had previously worked in a clinical setting during an epidemic (60%), while 63% had personal experience caring for patients with a novel respiratory pathogen (e.g., SARS, MERS Co-V, H1N1) (Table 2). Over 90% respondents were working in settings that had provided care for patients with suspected and/or confirmed COVID-19, while two-thirds (66%) had personally cared for COVID-19 patients.

TABLE 1 Respondent characteristics
Full size table
TABLE 2 Respondent experiences related to COVID-19 epidemic and previous outbreaks
Full size table

Access to and perception of personal protective equipment

Respondents reported a high rate of PPE use for suspected or confirmed cases of COVID-19, with universal use of hand hygiene (100%), and very high rates of use of face masks (97%) and disposable gloves (99%) (n = 229) (Table 2). Respondents also reported high use of gowns and face shields (86% each), while isolation was only used in 84% of cases. Overall, respondents reported moderate ease in accessing PPE, with a median [IQR] score of 5 [3–6] out of 7. Bedside nurses reported higher scores on the combined Environment measure (T = 2.36; 95% confidence interval [CI], 0.53 to 6.08; P = 0.02) compared with all other nurses, indicating they had easier access to PPE and isolation facilities and had clear IPC policies that were well supported by public health authorities; however, this difference was not significant after adjustment for multiple comparisons (Figure; Table 3). Older respondents and those with more years of experience also reported higher scores compared with younger respondents and those earlier in their independent career (B = 0.15, P = 0.02 and F = 15.77, P < 0.001, respectively), with only years in independent practice being significant after adjustment for multiple comparisons (Table 3). Over 95% of respondents indicated positively (agree or strongly agree) that they intend to always use a mask, eye protection, gown and gloves when taking care of suspected or confirmed COVID-19 patients. There was no observed difference among respondents based in academic hospitals compared with other settings.

Figure
figure 1

Responses to individual Likert-scale questions by thematic category. Figure description: Responses to nine questions on a Likert scale of 1–7, with 1 representing strongly disagree and 7 representing strongly agree. Percentages on each question, from left to right, represent the percent who disagreed with the question (strongly disagree, disagree, somewhat disagree), those who neither agreed nor disagreed (in grey bar), and those who affirmed the question (somewhat agree, agree, strongly agree).

Full size image
TABLE 3 Associations between respondent characteristics and aggregated perspective scores*†
Full size table

Perceived risk of COVID-19 infection

Respondents were most concerned about the risk to their families and their own risk of becoming ill (Figure). The majority of respondents (59%) disagreed with the statement that they have little control over whether or not they become infected with COVID-19, with only 27% affirming the statement. Older respondents disagreed more strongly with questions related to fears, with a lower Emotions score showing overall less fear for themselves and their families (B = -0.16, P < 0.001) (Table 3). Those with previous experience working during outbreaks scored highest on fear of becoming ill and fear of looking after patients with COVID-19 (χ2 = 21.48, P = 0.002 and χ2 = 12.61, P = 0.05, respectively). Partnered status did not significantly influence respondents’ Emotions score, nor did whether they had children (F = 3.03, P = 0.08 and T = -1.60, P = 0.11, respectively) (Table 3).

Professional and social role

Respondents reported very high rates of agreement with statements regarding expectations to follow IPC guidelines (99% agreement), the importance of following IPC guidelines to prevent COVID-19 spread (98% agreement), and that they encourage junior colleagues to adhere to IPC guidelines (98% agreement) (Figure). Those with more years of experience since completing training scored higher on the combined measure (F = 5.50, P = 0.02) indicating they felt more strongly that there is an expectation to follow IPC guidelines as part of their role, and that there is a culture of support from senior staff to junior staff to adhere to IPC protocols; however, this was not significant after adjustment for multiple comparisons (Table 3).

Beliefs

Respondents felt strongly that, while IPC procedures add to their workload, they are a necessary additional strain; with only 14% reporting that IPC procedures are an unnecessary strain (Figure). Respondents were more ambivalent as to whether IPC measures would effectively prevent COVID-19 spread. Bedside nurses had higher overall Beliefs score (T = 2.93; 95% CI, 0.44 to 2.30; P = 0.004) (Table 3) than all other nurses, showing confidence in PPE and IPC procedures. Experience working in previous epidemics or in centres with COVID-19 did not influence respondents’ Beliefs score.

Skills and Intentions

Overall, respondents reported a very high affirmation of always using PPE when caring for COVID-19 patients (Figure). While only 7% of respondents felt they had insufficient training in IPC for general communicable diseases, 19% felt they had insufficient IPC training specifically for COVID-19. Older respondents and those further from training felt more strongly that they had sufficient skills (B = 0.05, P = 0.05 and F = 8.58, P = 0.004, respectively), with only years in independent practice being significantly associated with higher confidence in their IPC skills after adjustment for multiple comparisons. Those with experience working in previous pandemics or with novel respiratory pathogens felt more confident regarding their skills compared with those without previous exposure (F = 5.33, P = 0.02), which was not significant after adjustment for multiple comparisons. Bedside nurses, compared with all other nurses, felt more confident in their IPC skills (T = 2.60; 95% CI, 0.31 to 2.33; P = 0.01), which was not significant after adjustment for multiple comparisons.

Perceived organizational preparedness

Respondents felt equivocal as to whether their work setting is able to manage current and future demand, responding less positively than on most other measures. Overall, 71% of respondents affirmed that current COVID-19 patient demand is being adequately managed, while only 33% affirmed that they believe the system will be able to manage demand over the next three months (Figure).

Trust in organizational preparedness to prevent infection

Bedside nurses had more trust in the organizational, provincial, and national preparedness than other workers did (T = 3.74; 95% CI, 3.26 to 10.64; P < 0.001) (Table 3). Older age was associated with higher trust in the system (B = 0.31, P < 0.001). Similarly, there was greater trust in the system among those further out from training (F = 18.09, P < 0.001), with mean trust scores (out of a maximum trust score of 63) ranging from 33.3 (12.1) among those in practice for less than 5 years to 42.8 (14.0) among those in practice for more than 20 years (ESM). Those with experience during previous outbreaks also had more trust (F = 7.87, P = 0.005) (Table 3). Our survey was distributed during the peak of Canada’s first wave; however, whether respondents were personally caring for suspected/confirmed COVID-19 patients or working in hospitals that cared for suspected/confirmed COVID-19 patients did not change respondents’ trust levels.

Discussion

Our survey of Canadian emergency and critical care nurses, early after the onset of the COVID-19 outbreak, shows high levels of trust in the healthcare system. While respondents felt comfortable with IPC procedures for general communicable diseases, they felt less comfortable with IPC procedures surrounding COVID-19. This difference potentially speaks to the emerging level of evidence surrounding COVID-19 and the early period of the pandemic captured by this study.

Nurses practicing in settings caring for COVID-19 patients felt confident in their IPC skills, potentially reflecting the emergence of IPC education and COVID-19 policies early in the pandemic in centres expecting to see COVID-19 cases. Overall, respondents showed strong beliefs regarding the importance of IPC and reported very strong social and professional expectations to follow IPC procedures, and perceived that following IPC procedures is part of their job. At the same time, respondents showed beliefs that they have some autonomy and control over whether they become infected.

Our study showed a discord between intentions to use PPE and reported use: 96% of respondents agreed or strongly agreed that they intended to use PPE, while only 85% had actually used gowns and 86% had used a face shield when caring for suspected or confirmed COVID-19 patients. Thirty percent of respondents noted that they cannot easily access PPE. Combined, these findings speak to challenges accessing PPE early in the pandemic,26,27 particularly face shields.

Our finding that our nursing cohort felt a sense of control and noted high degrees of support from senior colleagues is in contrast to other literature showing that nurses are at high risk for burnout.2,3,4,5 In our study, which largely comprised front-line critical care and emergency nurses providing direct clinical care (though not exclusively), there was a strong sense of concern for personal health and risk to family; however, participants felt empowered and in control of their own situation. Having a sense of control has been correlated with lower rates of burnout,28,29 as has having strong support from managers among nurses.30 Recent studies conducted during the COVID-19 outbreak have shown high degrees of burnout, anxiety, and depression among similar populations2,3,31; while we did not specifically screen for these conditions, our findings do not appear to be in keeping with them, particularly as these feelings are often associated with a lack of a feeling of support and lack of feeling of control, which was not evident in our population.

Canadian nurses showed a high level of institutional trust in our survey. Intriguingly, older nurses, those more remote from their training, and those who had worked during previous epidemics, all had significantly higher trust in institutional systems, even though these nurses had higher levels of fear. While these findings could be due to statistical variation among the sub-group of older nurses, it may also reflect the high levels of societal fear of illness at the time of survey distribution, which was early in our experience and understanding of the disease. These HCWs, who are more experienced and perhaps more senior in the healthcare setting, trust the decision-making processes being made to prioritize their own safety. Nevertheless, their fear may stem from their perception of personal risk as a result of more advanced age and disease severity associated with older age.

As it relates to fear, it is notable that Canada’s COVID-19 outbreak did not reach the same levels of those in other settings (e.g., Italy) during the initial wave of the pandemic; ongoing fear may be related to fear of the outbreak worsening, as shown by respondents’ concern that the Canadian system will not have capacity to manage future caseloads. Fear in those with previous experience working during outbreaks, despite adequate training, might be related to having worked in epidemic experiences before and having lingering fears from recent outbreaks such as SARS.32,33 Ongoing monitoring for fear, as well as long-term mental health impacts, will be important to capture beyond the initial outbreaks.

Future directions

Future work should focus on evaluating change in trust, beliefs, and skills during evolving pandemics, as well as linking specific healthcare policies and geographies to respondent beliefs. A more in-depth understanding of the influence systems and communication strategies on perspectives, particularly assessing leadership styles, media impact, and information sharing, would be informative in developing potential interventions to support HCWs in future pandemics.

Limitations

Our study is limited as we did not collect geographical data and are therefore unable to assess how responses changed in relation to the different timing of outbreaks across the country, as well as different local and provincial policies. Overall, our study’s limited response rate may be related to the high volume of surveys and studies being initiated at the time of this study, with competing requests from other research groups and an emerging high clinical burden, as well as the listservs we were able to access for survey distribution. We sought to disseminate the survey quickly, early in the pandemic, to capture the rapidly changing healthcare landscape. Nevertheless, because of the resulting survey distribution, the majority of nursing respondents are based in critical care settings (ICU, emergency department); our findings therefore may not be generalizable to other nursing populations or other HCWs. Our gender skew, which likely reflects the high percentage of women in nursing, limited our ability to assess whether gender played a role in respondents’ perspectives, which has been shown in other studies.3,34 Finally, our findings might reflect stated rather than actual experiences of respondents, which is a recognized limitation of survey methodology.35

Conclusions

We conducted a survey of nurses in Canada during the early stages of the COVID-19 pandemic. Respondents expressed trust that the local, provincial, and national health system leadership will ensure protective measures are in place during the pandemic. Trust was highest among bedside nurses, older nurses, those with more years of independent practice, and those with experience during previous epidemics, despite the same population reporting higher levels of fear of becoming ill. Canadian nurses had strong beliefs in the utility of PPE and IPC measures, and self-reported high rates of adherence to both IPC measures as well as PPE use. Canadian nurses showed strong social and professional expectations to adhere to IPC measures.