The Study of the Impact Factors of Behaviors of Nursing Intern Students: PRECEDE-Based Structural Equation Modeling Analysis

Background: Tobacco hazard is known as one of the most critical public health problems. Nurses could and should help patients to quit smoking. However, the lack of tobacco cessation education of nurses at medical school leads to the absence of behaviors helping smokers to quit. The analysis of the factors that inuencing the nursing interns to help smokers to quit provide evidence for nursing school to administer tobacco cessation education to facilitate more behaviors of smoking cessation in future nurses. Method: To explore the interaction and the effect coecient among the inuencing factors of the 5As behaviors of nursing interns in helping smokers to quit provide and to provide evidences for the tobacco cessation education, a questionnaire based on the PRECEDE theory was developed with consideration of predisposing factor, enabling factor and reinforcing factor. Random Cluster Sampling was used to conduct a survey in 13 teaching hospitals selected from 29 in Chongqing to assess the 5As behaviors of all nursing interns. A structural equation model was established to test the path and effect coecient of their 5As behaviors. Results: The survey investigated 1358 nurses, and the model tted well with the data. The tobacco cessation education, one of the enabling factors, imposed max effect on the 5As of nursing interns directly (cid:0) β=0.542 (cid:0) p<0.001 (cid:0) and indirectly (cid:0) β=0.38 (cid:0) p<0.001 (cid:0) through self-ecacy. The reinforcing factors ( smoking cessation environment) imposed direct effect (β=0.305 (cid:0) p<0.001 (cid:0) and indirect effect on the 5As of nursing interns through attitude changes. Predisposing factors including self-ecacy inuenced the 5As directly, while attitude (cid:0) β=0.001 (cid:0) p<0.05 (cid:0) and knowledge (cid:0) β=0.008 (cid:0) p<0.05 (cid:0) of tobacco cessation implied indirect inuence on the 5As through self-ecacy change.


Background
Tobacco hazard is known the most critical public health problems. A total of 8 millions died from tobacco use annually [1] . China is the largest producer and consumer of tobacco products, accounting for 40% of the global cigarettes [2] . China has 316 million current smokers and over 1 million died from diseases because of tobacco use every year [3] .
The Framework Convention on Tobacco Control of WHO stresses the important function of health professionals in smoking cessation and prevention, and puts request on them to perform brief consultation and offer simple advises for smokers to quit smoking [4] . Nurses are the largest group of health professionals. They offer medical and health services in hospitals, schools, communities and other places. They are extraordinary important for health promotion and playing signi cant roles in public health across the world [5] , and are also critical power to in uence the social smoking pattern. Therefore, nurses play pivotal role in smoking cessation [6] .
Nursing students are potential nurses and are supposed to learn the updated knowledge of tobacco cessation. They should assume good self-control in smoking meanwhile help others quit smoking. A few factors that in uence the involvement of students in tobacco cessation include the knowledge of and attitudes towards smoking cessation, self-smokers or not [7] . Smoker students presented negative belief in smoking cessation, and they lack identity with professional roles in helping smokers quit [8] and hold incorrect conceptions of tobacco dependence treatment with a consideration of ineffective short consultation and treatment for the enhancement of smoking cessation rate [9] . Those factors exert negative effect on the implementation of tobacco cessation, so they provide much lower rate of smoking cessation consultation than non-smoker students [10] . The absence of relative education induces inappropriate knowledge of tobacco and smoking. Global Healthcare Professional Students Survey (GHPSS) found that merely 22.6% respondents had ever received formal training about tobacco cessation [11] . Sarna [12] proposed to conduct tobacco use control education in nursing students.
Afterwards, lots of schools launched relative education among students including nursing and medical students [13][14][15] . Then some studies began to explore the effect of characteristics of courses on the conception and knowledge of students in helping smokers to quit [16] .
Except the above factors, none is known to in uence nursing students to help clients to quit smoking, and it is not clear how those factors interact with each other. Behavior presence is characterized with complexity and variety. Many theories have studied the causes of the occurrence of behaviors. The mode of PRECEDE-PROCEED, proposed by Greens, et al. [17] , is a commonly used way of health promotion, which is framed theoretically with PRECEDE(Predisposing, Reinforcing, and Enabling Causes in Educational Diagnosis and Evaluation) and PROCEED (Policy, Regulatory, and Organizational Constructs in Educational and Environmental Development). PRECEDE summarizes the factors that in uence the desired behaviors as predisposing factors, enabling factors and reinforcing factors when education evaluation is performed. PROCEED provides support for policy, regulation, and organizational construction when education intervention is implemented, evaluating the effect of the intervention on the quality of life and health of the targeted population. Compared to traditional health education, the Green mode is multidisciplinary based and well speci c, universally applied in the eld of health education [18][19] . Therefore, based on the theory of PRECEDE, this study is to evaluate the in uencing factors of the 5As in smoking cessation from predisposing factor, reinforcing factor and enabling factor. It is to be tested and veri ed by structural equation model and to analyze the degree of in uence and interrelation of the factors.

Page 4/19
2 Methods 2.1 theories and indexes PRECEDE Predisposing, Reinforcing, and Enabling Causes in Educational Diagnosis and Evaluation) is the evaluation phase of Green mode, also known as ecological mode of health promotion. Green mode stresses the conduction of ample evaluation of factors which in uence behaviors: predisposing factor, enabling factor and reinforcing factor when developing health education plan. predisposing factor: the internal basis for behavior occurrence, including knowledge, belief, attitude, self-e cacy, etc. of the individual; enabling factor: the premise of the realization of behavior motivation and willingness, that is, the necessary skills, resources and social conditions for the realization or formation of behavior changes; reinforcing factor: the factor that provide continuous incentives for long-term maintenance or repetition of behaviors after occurrence. The predisposing factor is the internal motivation, and the enabling factor and reinforcing factor are the external conditions.
(1) predisposing factor The theory believes "predisposing factor" would in uence behaviors. This study, considering its characteristics, de ned the core variables: Predisposing factors are students' knowledge, attitudes, selfe cacy et, al. related to smoking cessation.
Among the questions about the knowledge of smoking cessation: 6 about tobacco epidemic, 10 about tobacco and disease, 5 about the harm of second-smoking on families and 9 about the regulations of smoking cessation.
Attitudes towards the participation in smoking cessation: Medical professionals should routinely implement the 5As to help smokers to quit; It is the responsibility of medical professional to help smokers to quit. Responses ranked from strongly disagree to strongly agree, scored on a scale of 1-5.
Self-e cacy: It is the pro ciency of 5As in helping patients to quit smoking. Responses were from fully unskilled to very skilled, scored on a scale of 1-4.
(2) Enabling factor The available tobacco cessation instruction and resources include the tobacco cessation education and guidance to students. Resources cover the familiarity with China Clinical Smoking Cessation Guidelines, tobacco cessation websites, smoking control drugs and consulting-phone number. Responses were from completely unknown to well known, scored on a scale of 1-5. The duration of tobacco cessation education for nursing students at school was: 0 minute, 30 minutes, 31-60 minutes, 1-2 hours, 2-3 hours, 3-4 hours, 4-5 hours or more than 5 hours. The number of 5As instructions from clinical tutors was also considered as one of the resources.
(3) Reinforcing factors They were the factors that reinforce or interfere behaviors, such as the observations of 5As performed by clinical nurses in internship. Responses were scored 1 when never seen 5As performed by tutors in clinic, 2 when twice seen, 3 when 4-9 times seen, 4 when 10-25 times seen and 5 when over 25 times seen.
When nursing students implement smoking cessation interventions, if they felt their behaviors were consistent with the setting, as well as the practice and expectation of tutors, they would reinforce and maintain those interventions.

(4) Result variables
The 5As interventions performed by students on patients in the period of internship: asking patients the status of smoking, advising smokers to quit, assessing the tobacco dependence and willingness to quit, assisting patients to quit and arranging following up services for those who were trying to quit. The number of patients helped by students with 5As was taken as result variable: 1= none patients, 2=1-3 patients, 3=4-9 patients, 4=10-25 patients and 5=>25 patients. H13: Attitudes towards tobacco cessation change the acquisition of relevant knowledge and indirectly in uences 5As performance H14: Self-e cacy of tobacco cessation indirectly in uences 5As performance through attitudes H15: Knowledge of tobacco cessation facilitate self-e cacy and indirectly in uences 5As performance See pathway in gure 1

Research objects and data collection
The design and method of this cross-sectional study in Chongqing has been previously reported [20] . To get to know the 5As behaviors and in uencing factors of nursing students in helping smokers to quit in clinic work, a web-based survey was conducted among nursing students in 13 teaching hospitals in Chongqing in January 2019. The study was approved by the ethics committee of the First A liated Hospital of Chongqing Medical University (2019-157). The participants were nursing students who were taking clinical practicum before graduation. A total of 1522 students participated in the study, with a response rate of 62.4% (1522 / 2400), excluding questionnaires ful lled less than 180 seconds and addressed with non Chongqing areas. The effective questionnaires were 1358 with an effective rate of 89.2%.

Statistics and analysis
SPSS 20.0 software IBM Corporation Armonk NY US was used to analyze the data. The characteristics of participants were descriptively analyzed by mean and standard deviation or constituent ratio. A structural equation model was developed by the software of Amos24 through path analysis. Firstly, con rmatory factor analysis was used to test the measurement variables whose effectiveness was evaluated by the measurement results of SEM. Factor load indicated how much they contributed to latent variables, and aggregation validity evaluated the internal consistency of the questionnaire through composite reliability (CR) and average variance extracted (AVE). This study described the differential validity through the comparison of the square value of the correlation coe cient between AVE and latent variables. When AVE was higher than the square value of the correlation coe cient among latent variables, the differential validity between latent variables was better. The latent variables included smoking cessation knowledge, attitude, self-e cacy, education, resources, environment and 5As behaviors. Then, a structural equation model was used to test the relationship among target variables. This model made an assessment by generalized least square (GLS) instead of maximum likelihood (ML), because the data went beyond the hypothesis of multivariate normality and the sample size was quite large. The theoretical mode was tested and revised until it was theoretically and statistically acceptable.
Generally, the ratio of chi square to degree of freedom was used to assess the model t, and the relative chi square should be less than 2.0. But it was sensitive to the sample size and a large size might induce a signi cant chi square. Thus, other variable measures had also been adopted, including Root Mean Square Error of Approximation RMSEA < or =0.05 Goodness-of-t Index GFI and Adjusted Goodness-of-t Index AGFI > or = 0.9, Parsimony Normed Fit Index PNFI and Parsimony Goodness-of-t Index PGFI and Parsimony Comparative Fit Index PCFI >0.5. As a result, the nal t model indicated the correlation and the estimates and levels of regression parameters.

Table1
The frequency of 5As of students in assisting patients N=1358 The load factor of this study was from 0.436 to 0.952. All CRS were higher than the threshold value 0.6, indicating good reliability [21] . Except the dimension of knowledge, the mean variance of each structure was more than 0.5, suggesting that the factors were fully integrated [21][22] . Table 3 displayed the statistical outcomes of con rmatory factor analysis.  Table 4 showed that the AVE values on the diagonal were higher than the square values of the correlation coe cients among other latent variables in the same row and column, indicating that the model was well discriminated.

Model t (con rmatory analysis of the t)
Theoretical model was revised with t index, correction index and theoretical experience after the path was analyzed by the least square method with AMOS 24.0. The absolute t index (= 3.395, GFI = 0.937; AGFI = 0.922) and the reduced t index (PCFI = 0.678, PGFI= 0.756, PNFI= 0.682) of the model showed a good t between the data and the model. See Table 5.
In sum, the measurement model showed good reliability, su cient aggregation and discrimination validity. The t of the data and the model was acceptable.
3.4 Analysis of the knot path of in uencing nursing interns' 5As in helping patients Structural equation model estimated that path H1 (t = 0.422, P = 0.673), H2 (t = 1.930, P = 0.053), H5 (t = 1.166, P = 0.244) did not work. That is, the direct in uence of smoking cessation knowledge, attitude and resources of students' 5As did not present. Path H9 (t = 1.599, P = 0.110) tobacco cessation education played no effect on the related knowledge, and H10 (t = 1.301, P = 0.193) smoking cessation environment had no effect on self-e cacy, while other paths were signi cant at the P 0.05 level. Tobacco cessation education directly in uenced the nursing interns' 5As (β = 0.542, P < 0.001) in helping patients quit smoking, and had positive effect on self-e cacy (β = 0.380, P < 0.001) and tobacco cessation attitude (β = 0.13, P < 0.05), thus indirectly in uencing 5As behaviors. Smoking cessation environment (reinforcing factor) (β = 0.305, P < 0.001) had direct effect on students' behavior, and had a positive effect on smoking cessation attitude (β = 0.201, P < 0.001). Tobacco cessation resources (β = 0.305, P < 0.001) had active effect on students' self-e cacy, indirectly in uencing students' 5As behaviors. The pro ciency of 5As (self-e cacy) (β = 0.099, P < 0.05) had direct effect on 5As behaviors. The knowledge of smoking cessation (β = 0.08, P < 0.05) had positive effect on self-e cacy, indirectly in uencing 5As behaviors. The attitudes towards tobacco cessation (β = 0.132, P < 0.05) had positive effect on knowledge and indirectly in uenced the behaviors of 5As. See Figure 2 for the path and coe cient of 5As behavior effect of nursing students.

Discussion
This study, based on the theory of PREEDE, developed a structural equation model and analyzed the 5As behaviors that in uenced the participation of nursing students in the tobacco cessation. It found that students scarcely performed smoking cessation behaviors [20] , and the enabling factor exerted the greatest effect on the implementation of tobacco cessation education; the next one was the reinforcing factor (the smoking cessation environment involving the smoking cessation behavior of the clinic tutor in the hospital); while the predisposing factor (knowledge, attitude and self-e cacy) had little effect on the implementation of 5As.
The research path demonstrated that knowledge and attitude did not directly but through self-e cacy in uenced the 5As behaviors, which was consistent with Bandura's interpretation of self-e cacy, that is, the judgment of knowledge and ability determined the motivation of behavior [23] . H5 tobacco cessation resources did not directly in uence the 5Asbehavior, because people could change the 5As behaviors only after having mastered the knowledge and technology of the used resources and improved the selfe cacy of behavior implementation. Teachers' instruction had no effect on the knowledge of tobacco cessation, which might be attributed to that the knowledge was mainly about tobacco hazards, tobaccorelated disease and regulations, while most of which was relatively common and students had mastered when they were learning the fundamental theories of health. Smoking cessation environment had no effect on self-e cacy. As mentioned above, it needed the corresponding resources and technologies to enhance the self-e cacy of 5As behaviors. According to the study outcoming and the current situation of students' involvement in tobacco cessation, we suggested: 4.1 Stressthe responsibility and behaviors of clinical tutors of instructing students to help patients to quit smoking The study found that tutors' instruction, though not much, was the major factor that in uenced students' 5As behaviors to help smokers to quit [20] . Overseas schools had implemented theoretical education on tobacco cessation among medical students on campus and clinical tutors had administered intensive instruction on smoking cessation in practicum, which played signi cant roles in instructing students to help smokers to quit [15] . Previous studies [24] reported that clinical nurses in China performed few smoking cessation behaviors on patients, which could be attributed to the lack of tobacco cessation education of knowledge and skills in their medical college study [25] and the lack of relevant on-the-job training. In addition, the absence of the philosophy of disease prevention and health promotion and the responsibility to help patients to set up healthy lifestyles resulted in the lack of active study of the skills of disease prevention and health promotion. Therefore, along the call of Global Health Action, medical professionals should take a proactive leading role in tobacco cessation, stop smoking themselves rstly, help patients to quit smoking, and instruct students the skills of smoking cessation.

Develop more effective smoking cessation resources
Smoking cessation resources could provide assurance for students to implement 5As behaviors.
Literature reported that [26] : smoking cessation resources were important factors that in uenced medical professionals to offer smoking-quit assistance. Therefore, the establishment of professional tobacco cessation website, the construction of smoking cessation clinic, the research and invention of drugs to control smoking, the development and promotion of appropriate skills of smoking cessation, provided available resources for medical professionals and conditions for interns to implement smoking cessation behavior.

Set up smoking cessation environment and culture
The implementation of 5As behaviors by teachers had positive effect on students. China actively launched media publicity while consolidating regulations of smoking cessation, setting a good social environment [27] . Teaching hospitals should actively construct smoking-free hospitals, and develop a culture of actively helping smokers to quit, so that students would take behaviors consistent with the environment and felt that their behaviors were in line with social expectations and their behaviors would be reinforced, resulting in internal incentives to perform more behaviors to help others to quit smoking.

Promote the course education of tobacco cessation among nursing students
At present, more than 100 universities in China have established tobacco cessation courses [28] , but that for nursing students has not been reported. As the largest staff of medical professionals, nurses work the longer time with patients and most of them are female and non-smokers who are more active in smoking cessation [29] . It is of great signi cance to implement tobacco cessation education in nursing. Firstly, it can restrain the increase of the state of female smoking. The present female smoking rate in China is far lower than that of men. But in recent years, when the male smoking rate has reached saturation, tobacco producers and seller stake aim on women, depicting cigarettes as symbol of "independence", "freedom" and "equality" and they produce women cigarettes of "thin sticks" and "peppermint avor", and advertise smoking as losing weight, subsequently women smoking rate gradually increases [30] .Actually, women smoking is not only harmful to their own health, but also to the newborn and the health quality of the next generation [31] . In addition, Chinese family education responsibility are mainly on women. Present nursing students are also future family educators. Their cognition and behavior of tobacco use directly in uence the next generation's tobacco use behaviors. Therefore, the administration of tobacco cessation education among nursing students can achieve multiple effects such as reducing female smoking and the tobacco use of next generation.
This study is a con rmatory study based on the ecological theory of health education PREDEDE. A 5As behavior model that in uences the nursing interns to help smokers to quit was constructed, and the relationship between various factors and the coe cient were tested, which provided further direction for the implementation of tobacco cessation education in nursing students to improve the 5As skills. This study also has some limitations. Firstly, when designing the reinforcing factors, the main consideration was placed on the internal in uence of teachers' smoking cessation behaviors on the awareness of participation of students. The external evaluation of students after the implementation of 5As behavior was not included, and the internal motivation should also be taken seriously. Secondly, the dimension of knowledge mainly included the tobacco hazards and regulations related to tobacco cessation, little about the skills of 5As. This may be attributed to the indirect effect of knowledge on 5As behavior. There are various knowledge about tobacco cessation, and how to select to be included in the questionnaire needs to be further improved. Thirdly, the ratio of chi square and degree of freedom of the model was over 3. It suggested that the t degree of the model and the data needed further improvement although other tting indexes indicated that the data t well with the model. Fourthly, this survey was a self-report from students, which might be dishonestly over reported due to the bias of social expectations. Finally, although the relationship and the coe cient between the factors that in uence the 5As behavior of nursing students had been analyzed, a causal judgment could not be achieved because the data in this study was collected from a cross-sectional survey.

Conclusions
The instruction of smoking cessation by instructors produced the greatest effect on students' behaviors, so the clinical teachers' responsibility and skills training of tobacco cessation should attract more attention. They should actively help smokers themselves and instruct students to use the 5As to help smokers to quit in clinic. Hospitals keep going on the establishment of smoking-free hospitals, developing a good social culture of smoking cessation, and develop more effective tobacco cessation resources to urge medical professionals to help more smokers to quit. Students were informed the voluntary participation and signed informed consent.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used during the current study are available from the corresponding author on reasonable request.

Figure 1
In uencing factors and correlations of 5As in tobacco control among nursing students based on PRECEDE theory