Factors influencing risk perception and nosocomial infection prevention practices of frontline nurses during the COVID-19 pandemic

Risk perception refers to an individual’s objective perception of potential harm or loss, i.e., his or her judgment of the attributes and severity of a particular risk [34], which is an important reference for accepting and coping with risk and is affected by various factors, such as personality, knowledge, education, and emotion. According to risk compensation theory and protective motivation theory, risk perception directly or indirectly affects an individual’s behavior. When an individual perceives a high risk, he or she will constantly adjust his or her cognitive processes and choose more protective behavior to avert the risk. In the traffic safety education and public health fields, risk perception has been extensively studied, and risk perceptions have been found to lead to changes in protective measures [35‐37].

After preliminary investigations, we found that nurses in China fully appreciated the danger of COVID-19, a novel infectious disease similar to the highly infectious and pathogenic SARS, and have acquired various degrees of risk perception. In addition, the nurses generally adopted necessary COVID-19 prevention practices, such as practicing hand hygiene, correctly selecting and using PPE, avoiding crowds, and implementing quarantine management for COVID-19 patients. According to risk compensation theory and protective theory, the nurses’ level of risk perception may have affected their nosocomial infection prevention practices and may therefore indirectly affect the rate of nosocomial infection with COVID-19.

Based on the above discussion, we propose the following hypothesis:

The broken window effect is a criminological theory, which holds that if a bad behavior in an environment is overlooked, people may feel encouraged to imitate the behavior or even an aggravated version of the behavior. Broken window theory has been fully applied in crime prevention, environmental protection, and education, among others. For example, environmental design has been applied to urban crime prevention in the UK [38]. Tourists’ perceived environment quality will affect their intention to engage in environmentally responsible behaviors [39]. Educators also discovered that if a school is in disorder and has physical risks, its basic educational goals and processes may be jeopardized [40].

The workplace safety perceptions of frontline nurses during the COVID-19 pandemic mainly included reasonable ward zoning, PPE supply, and quarantine management of COVID-19 patients. First, based on previous SARS prevention experience, a modified form of traffic control bundling was adopted to reduce nurses’ exposure to the disease, which is conducive to reducing the probability of nosocomial infection [41]. Second, during the epidemic, PPE was provided to nurses in the ward. PPE, an important aspect of workplace safety, includes masks, protective clothing, face screens, hats, etc. PPE has been confirmed to effectively reduce the nosocomial infection rate by not only isolating the virus and protecting the PPE user but also decreasing the viral load and preventing secondary transmission [42]. Surgical masks played an important role in preventing SARS infection [43]; N95 masks are more effective than surgical masks in reducing the spread of droplets. Medical protective clothing has a strong protective effect on blocking liquid, microorganisms, and aerosols. Third, the management of COVID-19 patients in the ward is an important part of nurses’ workplace safety. When COVID-19 patients walked in the ward, the airflow was found to be significantly contaminated [44], and COVID-19 is spread through droplets and aerosols. Therefore, practicing good quarantine management of COVID-19 patients and limiting their scope of activities reduce the risk of nosocomial spread of the disease.

According to our survey, nurses working on the frontline clearly perceived the basic workplace conditions mentioned above, which affected their risk perception at work. In addition, according to broken window theory, workplace safety perceptions also affect nurses’ work performance and practices for preventing nosocomial infection.

Based on the above discussion, we propose the following hypotheses:

Studies have shown that employees’ understanding of prevention directly affects their occupational safety practices and ultimately has an impact on occupational injuries [45]. Therefore, training nurses on the prevention of nosocomial infection will affect their nosocomial infection prevention practice and thus their nosocomial infection rate.

During the COVID-19 epidemic, hospitals in China conducted timely and comprehensive training on preventing infection that focused on two aspects: (1) common knowledge, such as hand hygiene, proper PPE selection and use, and occupational exposure management; and (2) prevention education focusing on the pathological and infectious characteristics of COVID-19, such as clinical diagnosis, diagnosis and treatment, transmission routes, and patient isolation management [46]. In the preliminary investigation, we found that nurses have strictly complied with the preventive requirements and have constantly adjusted their preventive practices according to their own circumstances and actual care needs.

Based on the above discussion, we propose the following hypotheses:

Anxiety refers to psychological characteristics represented by nervousness and worry and accompanying physiological changes, such as increased blood pressure [47]. In the field of cognitive psychology, Spielberger proposed the theory of state-trait anxiety [48], where state anxiety refers to an individual’s nervous and anxious emotions in the face of specific dangers or stress events, and trait anxiety is a specific personality or trait that manifests as constant uneasiness and apprehension. Previous studies have confirmed that both state anxiety and trait anxiety have an impact on learning ability, professional behavior, safety and accidents, and risk perception [49‐52]; furthermore, studies have also revealed that anxiety is correlated with risk perception and that individuals with high levels of anxiety may perceive an exaggerated risk [53], which is likely to lead to fear or withdrawal behaviors.

During the COVID-19 epidemic, anxiety was an important psychological characteristic of frontline nurses and likely resulted from the following causes: (1) the epidemic outbreak was a major public health emergency event that had an enormous psychological impact on the public; (2) the work intensity of frontline nurses was generally high; (3) the epidemic lasted a long time, and the nurses worked in high-risk posts for long periods, imposing tremendous and continuous psychological pressure on them. This situation persisted for a long time without proper intervention and therefore caused anxiety for the nurses. The nurses’ anxiety can be divided into two categories: (1) state anxiety triggered by high-risk care work in a hospital during the epidemic; and (2) trait anxiety, which is a stable personality characteristic. In this study, we examined the state-trait anxiety of frontline nurses and its impact on their risk perception and nosocomial infection prevention practices. Therefore, we propose the following hypotheses:

According to the World Health Organization (WHO) definition, job burnout refers to the occupational phenomenon caused by chronic work stress that is poorly alleviated. Job burnout is an important factor affecting employees’ health and their pursuit of medical care; however, this condition does not fall within the category of disease [54]. Job burnout mainly manifests as emotional exhaustion, depersonalization, and a low sense of personal accomplishment. Previous studies have confirmed that job burnout may affect employees’ occupational safety practices and can become a work safety hazard. A study of firefighters suggested that those experiencing job burnout show a decreased rate of reporting safety hazards, decreased PPE usage, and a decreased probability of compliance with safety rules [55]. Job burnout was found to be correlated with anxiety, which may be a new mechanism through which job burnout affects occupational safety and job performance. In an investigation of nurses in neonatal intensive care units, job burnout was found to be significantly correlated with anxiety levels [56]. Through a study on police officers, Maria showed that under the intermediary variable of emotional exhaustion, a high-intensity workload can predict anxiety levels [57]. In addition, depersonalization was found to be closely associated with anxiety disorders and depression [58]. To date, a correlation between the low personal accomplishment variable and anxiety has not been reported. According to the survey, when faced with substantial psychological pressure and a high-intensity workload, some frontline nurses did not receive proper psychological intervention, which led to job burnout phenomena such as depersonalization and emotional exhaustion. According to the theoretical bases described above, the job burnout of nurses may indirectly affect their nosocomial infection prevention practices by affecting their state anxiety. Therefore, we propose the following hypotheses:

According to modern stress effect theory, a stressor is a stimulus that an individual perceives and that generates a positive or negative stress response in a certain job or in certain internal and external environments; when it exceeds a certain limit, the stressor will cause psychological and physiological reactions in the individual, such as insomnia, fear, anxiety, and depression [59, 60]. During the COVID-19 epidemic, public health policies and the work environment changed profoundly and changed the work, family, and personal lives of medical personnel accordingly, which led to changes in nurses’ mental status and thus affected their perceptions of the workplace during the pandemic and the actions that they took in response.

Changes in the stress of nurses during a pandemic have been previously studied. During the SARS and MERS outbreaks, more than 50% of health care workers felt increased work stress [61], with increased negative emotions and concern about their personal safety and their families [62]. During the COVID-19 outbreak and treatment, similar stressors and their effects were also present among nurses. Therefore, we propose the following hypothesis:

In this study, we referenced the items of scales used in previous studies and applied appropriate modifications according to the characteristics of the actual work that the nurses performed during the epidemic, ultimately generating the items needed for the questionnaire used in this study. Additionally, we invited 20 experts in the field to evaluate and conduct a pilot test of the draft questionnaire; based on the experts’ feedback, we repeatedly revised and tested the questionnaire.

The questionnaire used in this study included two parts: demographics and scaled questions. First, the nurses’ basic information, including gender, age, education, marital status, job title, department, workplace, etc., was collected. Second, the actual performance of the nurses during the epidemic regarding the factors related to risk perception and preventive practices was examined. For each variable, questions were scored using a 5-level Likert scale (Strongly disagree = 1; Disagree = 2; No opinion = 3; Agree = 4; Strongly agree = 5) to facilitate calculation using structural equation modeling (SEM). The responses to the questions were not scored as right or wrong but aimed to reflect the respondent’s attitude or tendency regarding an item; therefore, the respondents were instructed to answer according to their true perception during the epidemic. The contents of the scaled questionnaire are listed in Table 1.

The questionnaire survey was conducted from February 1 to April 1, 2020, over a time span of 2 months at hospitals involved in epidemic work, with those in Wuhan as the epicenter. Through the Wenjuanxing online platform, 2797 copies of the questionnaire were collected; 2546 valid copies of the questionnaire were retained after excluding 251 copies with incomplete answers or highly random answers, resulting in a valid questionnaire rate of 91.03%.

The study was approved by the Ethical Community of Wuhan University. The respondents of the online survey consented to participate in the study.

(1) Workplace safety perceptions have the most significant positive impact on nurses’ nosocomial infection prevention practices. Consistent with broken window theory, a good working environment and safety management have a positive impact on nurses’ work quality. Manapragada et al. [70] showed that the safety climate of medical staff is highly positively correlated with their safety performance, which supports our findings. (2) The behavior model constructed in this study suggests that nurses’ grasp of nosocomial infection prevention knowledge has a significantly positive impact on their adoption of nosocomial infection prevention practices. Previous studies have suggested that nurses’ knowledge of nosocomial infection prevention may affect their attitude and thus may ultimately indirectly affect their nosocomial infection prevention practice [71]. Therefore, providing training for COVID-19 frontline nurses regarding nosocomial infection prevention will help them improve their basic prevention skills. Judith showed that when medical personnel have full knowledge of nosocomial infection prevention and hand hygiene supplies are adequately available in the workplace, the hand hygiene compliance of personnel significantly improves [72]. A study on the relationship between the knowledge and practice of preventing infection through quarantine conducted by Suliman et al. [73] showed that although medical staff had good knowledge of isolation management, they exhibited poor implementation compliance. Therefore, we speculate that the nosocomial infection safety practices of nurses are jointly affected by their knowledge of nosocomial infection prevention and workplace safety perceptions, and the single factor of knowledge of nosocomial infection prevention does not directly lead to good safety practices.

(1) A moderate level of state anxiety can positively affect nurses’ nosocomial infection prevention practice, i.e., the quality of nosocomial infection prevention practices among nurses with relatively higher state anxiety is better than that among nurses with low state anxiety. Previous studies on workplace anxiety have indicated that anxiety has both positive and negative aspects. Excessive anxiety affects employees’ work performance, but moderate levels of anxiety have a stimulating effect on employees, increasing their receptivity to feedback regarding their work performance and thus their vigilance in terms of the supervision and management that they receive, which leads to better performance [74, 75]. Based on the behavioral model proposed in this study, the nurses’ state anxiety during the epidemic was at an appropriate level and had a positive effect on their nosocomial infection prevention practices. Therefore, heeding employees’ state anxiety levels, implementing effective measures to reduce state anxiety (e.g., promoting successful COVID-19 prevention and treatment, strengthening organizational support and psychological counseling interventions for nurses), and maintaining nurses’ state anxiety levels within a moderate range are issues warranting the attention of hospital managers. (2) Interestingly, trait anxiety negatively predicts nurses’ nosocomial infection prevention practices, i.e., higher trait anxiety levels correspond to lower execution of nosocomial infection prevention practices. According to state-trait anxiety theory, trait anxiety is a longer-lasting and more stable personality trait. Under the same stress scenario, individuals with high trait anxiety are more prone to experiencing more anxious emotions with high stability [48]. Previous studies have suggested that employees with high trait anxiety have poor work performance [75] and that the level of state anxiety can predict employees’ inattention in dangerous situations [76]. These studies suggest that frontline nurses with high levels of trait anxiety should concentrate on their work when working in high-risk areas to reduce work errors and occupational exposure risks. Hospital administrators must direct attention towards employees with high levels of trait anxiety, e.g., by strengthening training on the prevention of nosocomial infections, through psychological intervention, etc., to reduce their risk of nosocomial infection. (3) During the COVID-19 epidemic, nurses’ long-term work stress and work intensity significantly increased, resulting in varying degrees of job burnout. Emotional exhaustion and depersonalization are two important dimensions of job burnout. Emotional exhaustion manifests as an employee’s loss of enthusiasm and work motivation, while depersonalization manifests as indifference and negligence towards work objects and environments. Manomenidis et al. [77] found that job burnout can reduce nurses’ hand hygiene compliance, which also verifies our findings. In the proposed model of factors affecting nurses’ nosocomial infection prevention practices, we found that the depersonalization and emotional exhaustion associated with job burnout can indirectly affect nurses’ nosocomial infection prevention practice by affecting their state anxiety.

Risk perception during the COVID-19 epidemic is an important psychological feature of nurses working on the frontline in the fight against the epidemic and an important factor influencing their mental health, work performance, and job change intention. In this study, we focused on factors influencing nurses’ risk perception and found the following two important influence paths: (1) Nurses’ safety perception of the ward environment has a significant influence on their risk perception. Safety climate refers to an employee’s intuitive perception of the degree to which his or her company values safety and determines the employee’s occupational safety practices and initiative in participating in safety management [78, 79]. A survey of US construction workers showed that in a more active workplace safety climate, workers demonstrate a higher hazard identification capability and a higher degree of risk perception [14]. In the proposed SEM, nurses’ assessment of workplace safety includes the appropriateness of ward division, PPE availability, the quality of isolation and management of COVID-19 patients, etc., and the workplace safety variable represents the safety climate of the hospital to a certain extent. Therefore, that the workplace safety perceptions of the frontline nurses would have a significant influence on their risk perception is a reasonable finding. (2) Nurses’ state anxiety significantly affects risk perception. Previous studies have suggested that employees’ mental health is an important factor in risk perception. A survey of the relationship between mental state and risk perception showed that the anxiety and paranoid personality group was more likely than the control group to think that negative events occur with high probability [80]. Therefore, nurses may be overly concerned about adverse events during the COVID-19 epidemic while overestimating the risk of COVID-19, which may be the reason why state anxiety affects nurses’ risk perception.

In addition, in the proposed model, three invalid hypotheses are noteworthy. First, the hypothesis that risk perception affects the execution of nosocomial infection prevention practices is invalid. A study of nurses’ risk perception, knowledge, and prevention practices regarding occupational exposure to the Zika virus indicated that risk perception and knowledge of nosocomial infection prevention can ultimately affect nurses’ preventive practices by influencing their attitude towards prevention [71]. According to our hypothesis, under perceptions of high risk, nurses are expected to implement nosocomial infection prevention measures highly efficiently. The reason for this invalid influence path may be that (1) in the early stage of the epidemic, some nurses did not have a positive attitude regarding the prevention of nosocomial infection, (2) PPE was not readily available, and although the nurses had appropriate risk perceptions at the time, they were unable to adopt the corresponding high-quality nosocomial infection prevention practices because of a lack of appropriate materials. Second, the hypothesis of the influence path between knowledge of nosocomial infection prevention and risk perception was invalid, likely because training on the prevention of nosocomial infection was dominated by knowledge about how to effectively prevent COVID-19 and only minimally addressed the risk, prognosis, and other information; therefore, this factor was unable to influence nurses’ risk perception. Third, the influence path between trait anxiety and risk perceptions of COVID-19 was invalid, although the reasons for this result are unclear. We speculate that employees with trait anxiety may be distracted and thus unable to accurately assess workplace safety, which leads to the invalidity of the hypothesis of this influence path.

Based on the nurses’ education, job title, working years, and marital status, we performed a differential analysis of the influence paths and found that the effects of trait anxiety and workplace safety perceptions on risk perception and nosocomial infection prevention practices differed.

First, as the nurse’s education level and job title increased, the positive impact of workplace safety perceptions on risk perception and nosocomial infection prevention practices gradually increased for the following reasons: education level and job title reflect nurses’ knowledge level and work ability to a certain extent, and those with high academic qualifications and high professional titles may be better able to observe and make judgements regarding workplace safety based on their better knowledge of nosocomial infection. These more experienced nurses also have better clinical capabilities and a greater capability to execute nosocomial infection prevention practices. Therefore, among nurses with a high professional title and a higher education level, the risk perception and the broken window effect of workplace safety perceptions are more profound.

Second, risk perception and nosocomial infection prevention practices differed significantly between nurses who were single and those who were married with children because the latter group has more complex family relations, causing them to worry about the health of their families and children in addition to their own health. These factors increase the intensity of stressors in nurses with more work experience; additionally, the level of state anxiety increases accordingly but remains within an appropriate range, resulting in increased execution of nosocomial infection prevention practices.

Additionally, we found that with an increase in work years, the positive influence of state anxiety on the execution of nosocomial infection prevention practices increased significantly because nurses with more work experience are affected by other variables, such as marital status, and are most often married or married with children, which has a moderating effect similar to that of marital status.