Determinants of the intention of elementary school nurses to adopt a redefined role in health promotion at school

Through a systematic review, Godin and colleagues [20] concluded that psychosocial theories are effective in understanding the cognitive mechanisms leading to the adoption of professional behaviour in healthcare. Eccles and colleagues [21] and Godin and colleagues [20] concluded that intention is a valid proxy measure for behaviour among clinicians, and the best prediction of intention was observed among samples of nurses. Nonetheless, few studies have focused on understanding the psychosocial determinants of healthcare professionals' intentions and behaviour. Among theory-based studies of healthcare professionals' behaviour, the TPB [22] and the theory of interpersonal behaviour (TIB) [23] have been used most often to date, and these theories have outperformed other psychosocial theories in the prediction of healthcare professionals' intentions and behaviour [20]. The TPB was preferred because it contains most of the variables recognised for their predictive capacity, such as control beliefs, social influences, normative beliefs, and consequences related to the adoption of a particular behaviour [20]. Moreover, TPB constructs are clearly defined and easy to operationalise and measure [24]. In addition, the TPB was formulated to take into consideration behaviour not always under volitional control [22, 24]. ESNs are subject to the rules of their health unit.

According to Ajzen, individual behaviour is directly defined by an intention to adopt a particular behaviour. Intention refers to an individual's motivation regarding the performance of a given behaviour. The attitude towards the behaviour, the subjective norm, and perceived behavioural control are direct determinants of intention. As far as we know, this model has not been applied yet to the study of the redefinition of the role of school nurses. However, the TPB was successfully used to better understand the intention of public health nurses to adopt clinical behaviours [25‐30]. Attitudinal factors and perceived behavioural control were the most important predictors of intention.

Following the recommendation of Perkins and colleagues [31], additional psychosocial constructs were added to the TPB model in order to gain a global understanding of the studied behaviour. With respect to factors explaining health professionals' intentions, Godin and colleagues [20] found that the most significant cognitive factors were beliefs about capabilities, beliefs about consequences, social influences, social/professional role and self-identity, and moral norm. In the literature, it is generally acknowledged that the TPB explains up to 40% of the variance in the prediction of intention [32]. The higher value reported in Godin and colleagues' [20] systematic review (59% explained variance) could possibly be related to the addition of other variables, such as Triandis's self-identity and moral norm. Self-identity helps to identify characteristics of the participants having the intention to adopt the proposed role. As suggested by Godin and colleagues [20], moral norm takes into consideration the ethical dimension of healthcare professionals' behaviour, and, as a single construct, it was found to be a frequent significant determinant of intention. This variable provides information on the moral obligation felt by the participants towards the proposed role. Finally, when a behaviour is performed in unstable or difficult contexts, conscious decision making is likely to be necessary to initiate and carry out the behaviour. Under these conditions, past behaviour (along with attitude and subjective norms) may contribute to intention [33].

Health and social services centers (Centres de santé et de services sociaux [CSSSs]) that had a minimum of five ESN positions were recruited. This criterion is necessary to conduct a multilevel analysis that could be further realised. From the 50 CSSSs that met this criterion, 42 agreed to participate. The study population held part- and full-time ESN positions and included those on call for ESN position replacement. These on-call nurses had a minimum of six months of experience in the past year in school healthcare services in the province of Québec, Canada. School nurses work under the jurisdiction of local health and social services organisations, known as CSSSs. Data to estimate the number of positions per CSSS were obtained at the CSSS level, the only place where such data are recorded. Of the 358 mailed questionnaires, 256 were received. Among those received, three were returned uncompleted and two were completed by high school nurses. High school nurses have different mandates and different working conditions. Thus, their opinions would not reflect ESN realities. Consequently, 251 questionnaires were completed satisfactorily. Respondents were mailed Can$10 for compensation upon receipt of the completed questionnaire. The variation in response rate was generally homogenous across CSSSs, with a slightly higher proportion in smaller CSSSs. The average response rate of 70% served to meet the recommendations of Rashidian and colleagues [34] regarding sample size needed to predict intention based on the TPB.

An authorisation to proceed with the survey was requested in a letter sent to the head office of every CSSS, also asking that a local contact person be identified. This person was usually the assistant to the school health program coordinator and was responsible for explaining the project to the ESN and distributing the consent form, questionnaire (see additional file 1), and the preaddressed and prestamped envelope to them. The prestamped envelope was to be mailed directly to the researcher with the completed questionnaire and signed consent form. One week later, each contact person distributed the recall letter sent by the researcher to all ESNs. This study was approved by all 21 ethical committees, including the local university, participating CSSSs, the Health and Social Ministry's Central Committee, and the Montréal Regional Public Health Unit.

Firstly, descriptive analyses of the sample were performed to better describe the variables. A correlational analysis with Pearson coefficients was also carried out between studied variables. A multiple hierarchical linear regression analysis was performed to identify the determinants of intention. This was done as follows: first, past behaviour was entered; second, the proximal variables of the TPB were entered; third, Triandis's variables of moral norm and self-identity were entered; fourth, perceived barriers, facilitating factors, and normative beliefs were entered; and last, sociodemographic variables were added. After each step, variables not reaching p < .05 were eliminated from subsequent steps. Finally, a discriminant analysis of the beliefs was performed in order to identify the main targets of action. All analyses were performed using SAS software, Version 9.1 (SAS Institute, Cary, NC, USA) [44].

An examination of the correlation matrix indicated that all psychosocial variables were correlated to intention. Attitude, perceived control, and subjective norm equally presented the greatest association with intention (r = .78; p < .0001). Tests for multicollinearity were performed and none was detected. Variance inflation factors (VIF) were well below 10, and the condition index was under 30. According to Kline [48], multicollinearity is present when the correlation between two independent variables is greater than .85; none of the coefficients of correlation between the independent variables reached that level. Furthermore, residuals must be normally distributed [49]. Indeed, residual distribution followed a normal curve. An analysis of proportion showed that 73.1% of ESNs indicated moderate to strong intentions to play the role, which means they scored between 5 and 7 on the seven-point Likert-type scale used. With regard to the prediction model, moral norm added to the TPB constructs to predict school nurses' intention. Table 3 shows the steps applied for the multiple linear regression analysis. The strongest determinant of intention was perceived behavioural control (β = 0.36), followed by moral norm (β = 0.27), attitude (β = 0.24), and subjective norm (β = 0.21). The final model explained 83% of the variance in the intention of ESNs to adopt the proposed redefined health-promotion role in the context of the HPS approach in Québec.

The variables retained for analyses were the salient underlying beliefs from proximal constructs for which a significant relation with intention was identified (e.g., barriers and facilitating factors for perceived behavioural control, behavioural beliefs for attitude, and normative beliefs for subjective norm). In order to identify the beliefs that will serve to guide proper actions, a discriminant analysis contrasting high and low intenders was performed. The results indicated that the item 'This role would allow me to be valued in the performance of my duties' explained the greatest portion of the variance (R ² = .17; p < .0001). Additional items that also contributed to this prediction were 'If I agreed to play the proposed role, school nurses would approve' (R ² = .07; p < .0001); 'I would be able to play this role despite the nursing shortage' (R ² = .04; p < .003); 'If I agreed to play the proposed role, the parents of the students would approve' (R ² = .03; p < .008); 'To accept to play the role proposed would require me to have leadership' (R ² = .02; p < .04); and 'This role would allow me to gain a better understanding of school needs' (R ² = .02; p < .04). These six items represented the underlying beliefs distinguishing those who had a high intention from those who had a low intention (Wilks's lambda = .83; F[1] 239 = 50.29; p < .0001).

For managers and administrators, it is valuable information to know that approval by parents and school nurses, increased feeling of valorisation, and increased knowledge of school needs would motivate school nurses to play such a role. These beliefs are factual information of worth to use to market the role. MacDonald and Schoenfeld [66] found that public health nurses' involvement in research and use of their input in planning, delivering, and evaluating health-promotion programs promoted a sense of achievement, increased feelings of being valued, and greater respect from other professionals. Others also demonstrated that positive outcomes for students contributed to feelings of worth [10, 67]. The role proposed implies that nurses become involved in research. Thus, their involvement in research linking their health-promoting actions to child health outcomes, child well-being, and academic outcomes would likely contribute to feelings of worth and produce evidence-based knowledge that can promote their role as decision makers [56].

Findings also indicate that a nursing shortage and leadership require that action be taken. High intenders feel that they could overcome barriers such as the nursing shortage. However, this finding needs to be interpreted carefully by health administrators in order to avoid burnout among these human resources. In fact, if supervisors want to increase the proportion of high intenders, they need to ensure adequate nurse staffing in schools, as a nursing shortage is recognised as a serious threat to the deployment of health-promotion practices by nurses [52]. Indeed, a shortage of nurses means understaffing and decreased presence in the milieu. The success of the HPS approach depends on the stability of networks over time [68], and networking requires a minimum of shared time among the different actors [13]. A presence at school is also associated with leadership. Leadership has been shown to be a key component to improving nurse staffing and retention, as well as health-promotion practices [57]. In the international context of financial constraints, where additional nursing staffing has proven difficult to obtain, investing in the development of delegation skills could help school nurses increase their presence where needed to advance a role in health promotion. It would also contribute to heightened feelings of worth [63]. Leadership is a major skill that should be addressed at the university level, preferably in the context of a school health specialisation, so that school nurses are socialised to be leaders and expected to act as leaders before they enter the school system. Leadership training could also be provided to school nurses in the form of workshops in collaboration with local universities. Leadership development covers topics such as management of resources, marketing, team motivation, negotiation, effective communication, organisational change, contribution to the development of policies, mentoring, and delegation [57].

This study also shows that ESNs do not form a homogeneous population and that individual considerations should be taken into account for the implementation of interventions planned for the traditional and the innovative type of ESN. The latter seems to rely more on individual resources, while the former tends to rely more on organisational resources [4].

This study presents some limitations. Even though the sample accounts for a large portion of the ESN population, only local health organisations with five ESNs or more were invited to participate in order to ensure the required number of participants for a multilevel analysis [69]. Thus, smaller regional sites were not included in the present study. Furthermore, the sample was composed of volunteers. Therefore, responses to this study are subject to self-selection biases. Also, it may have been difficult for participants to determine their intention regarding the hypothetic role, since none of them has played this precise role in the past. There is also a potential influence of the social desirability bias. Thus, some caution should be exercised before generalising the results.