Applying a health action model to predict and improve healthy behaviors in coal miners

Abstract

Introduction:

One of the most important ways to prevent work-related diseases in occupations such as mining is to promote healthy behaviors among miners. This study aimed to predict and promote healthy behaviors among coal miners by using a health action model (HAM).

Method:

The study was conducted on 200 coal miners in Iran in two steps. In the first step, a descriptive study was implemented to determine predictive constructs and effectiveness of HAM on behavioral intention. The second step involved a quasi-experimental study to determine the effect of an HAM-based education intervention. This intervention was implemented by the researcher and the head of the safety unit based on the predictive construct specified in the first step over 12 sessions of 60 min. The data was collected using an HAM questionnaire and a checklist of healthy behavior.

Results:

The results of the first step of the study showed that attitude, belief, and normative constructs were meaningful predictors of behavioral intention. Also, the results of the second step revealed that the mean score of attitude and behavioral intention increased significantly after conducting the intervention in the experimental group, while the mean score of these constructs decreased significantly in the control group.

Conclusion:

The findings of this study showed that HAM-based educational intervention could improve the healthy behaviors of mine workers. Therefore, it is recommended to extend the application of this model to other working groups to improve healthy behaviors.

Keywords

health action model coal mining health education healthy behavior occupational health promotion

Introduction

According to the National Institute for Occupational Safety and Health, an estimated 5400 deaths from occupational accidents occur annually in the United States, with a large share of this mortality occurring among miners (1). Although mining is usually recognized as a local or regional profession, this industry has seen dramatic growth over the past decade, and today large multinational companies in the United States of America, South America, Africa, Europe, Asia and Australia have actively engaged in this area by investing in modern equipment and adopting new methods of excavation (2).

In Iran, there are vast and rich mines, and a large workforce is employed in this industry (3). Coal is a valuable product in Iran, but, despite its importance in economic development, little attention has been paid to associated health problems among miners (4).

The recognition of risk factors in workplaces such as mines, and improvement of working environments to prevent work-related diseases is chiefly the responsibility of industry authorities and health practitioners (5,6). Coal mine safety in Iran is a serious problem, as occupational injuries are common in the coal mining industry. Statistics provided by the Mine Safety and Health Administration (MSHA) show that death rates in coal mines are higher than in other mines and private industries. On the other hand, compared with United States mortality rates, in the 10-year period up to 2012, death rates in Iran’s coal mines were, on average, more than 100 times those of the USA (4).

Education in healthy behaviors is an effective step in the prevention of work-related diseases and education, and plays a pivotal role in progress and development in various fields related to accident prevention (7,8). Research has shown that the most effective training programs are theory based and rooted in the behavior change model. Therefore, in all studies intended to raise knowledge and change the attitudes and behaviors of a target population, it is essential to explore views and models of behavior change to choose an appropriate method (9,10). Therefore, a health action model (HAM) was adopted as a model for provision of factors supporting behavioral intention because intention was introduced as one of the factors underlying successful intervention (11). In a study in Spain, this model was used as a framework for workers’ safety training programs, with the findings revealing the effectiveness of HAM in promoting healthy behaviors among workers (12).

The HAM is a subcategory of health promotion models designed to provide a comprehensive framework of the major factors influencing the adoption and implementation of healthy behaviors. This model adopts a new approach to health promotion objectives, and explores various aspects associated with the formation of intention and behavior of individuals, such as human role (beliefs, attitudes and normative systems) and environmental factors (13 –15).

Rennie described the constructs of this model to be applied in workplaces as follows (16): 1) Cognitive system: basic knowledge of workers about safety, occupational health and safety behaviors. 2) Normative system: workplace norms and regulations regarding safety, by-laws and guidelines. 3) Motivational system: intrinsic motive of individuals to engage in safe actions, or motivating factors of safe behaviors in an organization. 4) Belief system: values and beliefs of workers about effectiveness and benefits of safety and safe practice. 5) Facilitators of safety practice: conditions and factors that contribute to safe behaviors (Figure 1).

Figure 1.

Health action model.

Since health education models have been developed to help researchers and specialists gain deeper insight about the nature of an intended behavior and determine its effective factors, it is essential to evaluate the strength of model constructs in predicting behavioral intention so that the most appropriate program goals as well as methods of undertaking change and measuring results are specified. Therefore, the goals of this study were to predict health behaviors among coal miners by using HAM, and to promote healthy behaviors among coal miners by using HAM in the Tabas coal mine, Iran.

Methods

The study was conducted on 200 coal miners in Iran from 20 March 2016 to 15 October 2016 in two steps. The first step involved a descriptive study to identify predictor and influence constructs (attitudes, beliefs and norms) of HAM on behavioral intention. For this purpose, 200 miners were selected using a random sampling method. For a sampling of each of the five independent broadcast mines, 40 miners were randomly selected.

Considering a type 1 error of 0.05 and power of 80%, the 200 miners were calculated using the following formula

n = \frac{z_{1 - \frac{α}{2}}^{2} P (1 - P)}{d^{2}}

Having at least 1 year’s experience in the coal mine and lack of disability were considered as inclusion criteria. Missing data concerning any item within the questionnaires was also excluded from the final analysis.

In the first step, data was collected using the HAM questionnaire and the healthy behavior checklist. After analyzing data by linear regression and identifying effective and predictor constructs of HAM, the second step was implemented.

In the second step, a quasi-experimental study was undertaken to determine the effect of HAM-based educational programs. Thus, from the 200 workers, 80 were randomly selected and assigned to experimental (n = 40) and control (n = 40) groups. The pre-test was administered to both groups using the HAM questionnaire and healthy behavior checklist. Then, the intervention program based on predictor constructs of HAM in the first step was carried out in the experimental group.

To evaluate the effectiveness of interventions, after 3 months follow-up data was collected again using the HAM questionnaire and healthy behavior checklist. Finally, the data were analyzed by SPSS 20 software using statistical tests such as Chi-Square, Mann-Whitney, Fisher, Kruskal-Wallis, Wilcoxon, linear regression, paired t-test and independent t-test at a significance level of p < 0.05. Parametric tests were used for data with a normal distribution and non-parametric tests were used for data that were not distributed normally.

Educational intervention

The educational intervention was carried out by the researcher and the head of the safety unit based on predictor constructs determined in the first step of the study (attitudes, beliefs and norms) over 12 sessions of 60 min. The training sessions were in the form of focus groups and role playing. At the beginning, workers were divided into five groups of eight members. Initially, the issues of safety, accidents and the role of human error in accidents, belief in fate, and the role of supervisors and people in the prevention of accidents was discussed. At the end of the session, workers shared problems and hazards associated with their job, and were asked to offer solutions for the issues discussed. The latter, as stipulated in the HAM model, is mainly because adult learners prefer to be involved in learning. Adult learners often bring their experiences to the learning environment, and these can be utilized as a source of learning so that instead of opposing these experiences, adult learners should be incorporated in teaching to shape the learning experience.

In keeping with the goals of study, other measures were adopted, such as distributing educational pamphlets among participants, the presence of the head of security unit in sessions, screening photos and films on safety training, encouraging the participation of workers in reporting quasi-accidents and provision of feedbacks to improve safety, inspections by the deputy of safety, and listening to workers, serious handling of safety violations, provision of personal protective gadgets for more delicate work, and distribution of personal protective equipment.

Data collection instrument

In this study, the main data collection instrument was the HAM questionnaire and healthy behavior checklist. The latter consisted of two parts: the use of personal protective gadgets and safe behavior. The HAM questionnaire contained questions related to demographics and belief system constructs (14 items) (e.g. I believe accidents can be prevented), normative systems (8 items) (e.g. the senior management of this organization takes health and safety issues seriously), attitude system (4 items) (e.g. I think some of the rules and safety procedures are inoperable), knowledge (3 items) (e.g. what personal protective gadgets are required in your workplace) and behavioral intention (7 items) (e.g. I plan to use proper and safe tools at work). The items related to facilitators (5 questions) were designed by the researcher.

A 5-point Likert scale was used to score items related to belief, attitude, normative and behavioral intention constructs. Accordingly, the score ranges for constructs of belief system (14–70), normative system (8–40), attitude system (4–20) and behavioral intention (7–35) were determined. With regard to facilitators, the options ‘No’, ‘Somewhat’, ‘Yes’ were assigned a score of zero to two, respectively. To measure participants’ knowledge, three open questions were asked about personal protective gadgets required at the workplace as well as risk factors and safety suggestions. The score range for knowledge construct was in the range of 1 to 18.

To evaluate the validity of the questionnaire, it was sent to 10 experts in health education, occupational health and occupational medicine as well as three specialists working in Tabas Coal Mine. The views and comments of these individuals were incorporated in determining the content and face validity, and to fine-tune the study instrument. To assess reliability, Cronbach’s alpha was calculated for belief system (0.72) attitude system (0.74) normative system (0.83) facilitators (0.75) and behavioral intention (0.90).

Ethical considerations

This study was approved by the Ethics Committee of Mashhad University of Medical Sciences (registration No.: IR.MUMS.REC.1394.52). All participants signed a written informed consent form and were assured about the confidential nature of data analysis and the voluntary basis of participation, which allowed them to withdraw the study at any time.

Results

Based on the results of the first step of the study, the mean (SD) for age (30.22 ± 4.68) and years of work experience (5.20 ± 3.96) of participants was computed. The latter was in the range of 1–17 years. As for marital status, 10.5% of participants were single and 89.5% were married. Further, 12.5% had a history of work-related accident and the remaining 87.5% had never experienced such incidents. The demographic variables and their relationship with HAM constructs are shown in Table 1.

Table 1.

The relationship between demographic characteristics (age, work experience, job title, marital status, education level, history of work accident) and HAM constructs in subjects under study.

Model constructs	Mean ± SD	Belief	Normative	Attitude	Safety attitude	Knowledge	Facilitators	Self-efficacy	Behavior
Age^a	22.4 ± 68.30	r = 0.06 p = 0.396	r = 0.008 p = 0.914	r = 0.09 p = 0.180	r = 0.09 p = 0.163	r = 0.13 p = 0.067	r = 0.05 p = 0.477	r = 0.09 p = 0.200	r = 0.01 p = 0.820
Years of work experience^a	5.20 ± 3.96	r = 0.08 p = 0.265	r = 0.11 p = 0.126	r = 0.03 p = 0.634	r = 0.03 p = 0.609	r = 0.08 p = 0.262	r = 0.10 p = 0.142	r = 0.06 p = 0.390	r = 0.10 p = 0.131
Job title	No. (%)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)
Mechanic	38 (19.1)	55.34 ± 6.31	28.39 ± 5.31	13.82 ± 3.34	29.16 ± 5.19	9.08 ± 2.88	6.66 ± 2.78	27.84 ± 4.91	6.58 ± 0.79
Worker	113 (56.5)	55.80 ± 5.85	28.91 ± 5.32	13.82 ± 3.65	29.63 ± 4.38	9.37 ± 2.25	6.94 ± 2.20	27.43 ± 4.21	8.85 ± 0.85
Electrician	20 (10.0)	53.70 ± 5.33	26.95 ± 4.32	14.20 ± 2.48	29.10 ± 4.82	9.37 ± 2.24	7.05 ± 2.23	28.20 ± 3.90	8.15 ± 0.87
Operator	15 (7.5)	58.20 ± 3.36	28.47 ± 6.13	14.80 ± 2.08	28.73 ± 2.63	11.27 ± 2.09	6.53 ± 2.10	26.27 ± 4.99	8.87 ± 1.25
Workman	14 (7.0)	55.43 ± 4.55	26.07 ± 4.21	14.14 ± 1.92	28.86 ± 2.41	11.36 ± 2.10	6.57 ± 1.95	25.64 ± 3.99	8.07 ± 0.73
Kruskal–Wallis test results		χ² = 6.60 p = 0.15	χ² = 5.25 p = 0.26	χ² = 2.52 p = 0.64	χ² = 3.54 p = 0.47	χ² = 16.5 p = 0.002	χ² = 1.48 p = 0.83	χ² = 5.40 p = 0.24	χ² = 9.31 p = 0.54
Marital status	No (%)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)
Single	21 (10.5)	55.05 ± 5.83	28.29 ± 5.88	13.90 ± 3.60	29.0 ± 4.91	9.95 ± 2.88	2.86 ± 2.64	26.4 ± 4.91	8.52 ± 0.75
Married	179 (89.5)	55.73 ± 5.70	28.40 ± 5.17	13.96 ± 3.05	29.4 ± 4.30	9.62 ± 2.39	6.84 ± 2.25	27.4 ± 4.30	8.53 ± 0.90
Mann–Whitney test result		Z = 0.30 p = 0.76	Z = 0.20 p = 0.84	Z = 0.28 p = 0.77	Z = 0.25 p = 0.80	Z = 0.64 p = 0.51	Z = 0.26 p = 0.79	Z = 0.90 p = 0.36	Z = 0.20 p = 0.83
Level of education	No. (%)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)
Junior high	40 (20.0)	55.5 ± 55.62	27.68 ± 4.99	14.07 ± 2.86	28.78 ± 4.6	9.13 ± 2.65	6.55 ± 1.82	26.28 ± 4.4	8.48 ± 0.64
High school diploma	122 (61.0)	55.79 ± 5.60	28.76 ± 5.26	13.93 ± 3.07	29.53 ± 4.3	9.61 ± 2.29	6.92 ± 2.28	27.41 ± 4.4	8.55 ± 0.90
Associate degree	21 (10.5)	55.43 ± 5.17	27.81 ± 5.25	13.57 ± 3.72	28.86 ± 3.6	10.52 ± 2.0	7.10 ± 2.42	28.48 ± 3.9	8.52 ± 0.92
Bachelor’s degree and higher	17 (8.5)	55.24 ± 7.51	28.06 ± 5.77	14.29 ± 3.31	30.18 ± 4.6	10.06 ± 3.1	6.65 ± 3.10	28.3 ± 4.18	8.47 ± 1.23
Kruskal–Wallis test result		χ² = 0.005 p = 0.941	χ² = 1.385 p = 0.239	χ² = 0.054 p = 0.816	χ² = 0.83 p = 0.361	χ² = 0.73 p = 0.392	χ² = 1.93 p = 0.164	χ² = 1.61 p = 0.203	χ² = 0.28 p = 0.595
History of work accident	No. (%)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)	(Mean ± SD)
Yes	25 (12.5)	54.96 ± 6.46	27.60 ± 4.45	12.76 ± 2.63	29.12 ± 4.6	9.52 ± 2.52	6.32 ± 2.63	26.88 ± 4.5	8.60 ± 0.76
No	175 (87.5)	55.75 ± 5.60	28.5 ± 5.34	14.13 ± 3.13	29.4 ± 4.33	9.67 ± 2.44	6.91 ± 2.23	27.45 ± 4.3	8.51 ± 0.90
Mann–Whitney test result		Z = 0.85 p = 0.39	Z = 0.81 p = 0.41	Z = 2.19 p = 0.02	Z = 0.23 p = 0.816	Z = 0.45 p = 0.64	Z = −0.94 p = 0.34	Z = 0.70 p = 0.47	Z = 0.47 p = 0.63

Spearman correlation test.

HAM: Health action model.

As shown in Table 1, there was a significant relationship between job title and knowledge constructs. Also, the constructs of attitude and history of work accidents were significantly correlated. Type of employment was also correlated with constructs of attitude and facilitators. However, no significant relationship was found between demographic characteristics and other HAM constructs.

At this step, predictive analysis was implemented using linear regression. The results of linear regression test between behavioral intention and HAM constructs of the subjects under study indicated that attitudes, beliefs and normative constructs were significant predictors of behavioral intention. As shown in the column of standardized coefficients, ‘attitude’ construct (B = 0.314) was the most powerful predictor of behavioral intention, followed by ‘belief’ and ‘normative’ constructs (B = 0.282 and B = 0.213 respectively). This model could explain 37% of the variation in behavioral intention (R² = 0.37). In other words, HAM constructs could significantly predict 37% of variation in behavioral intention (Table 2).

Table 2.

Linear regression analysis by considering behavioral intention as the response variable in the absence of demographic variables statistics.

Predictor	Stat
	Beta regression coefficients	Standard error	Standardized regression coefficients (B)	T	p-value	Descriptions
Belief	0.215	0.049	0.282	4.401	0.001	Final model
Normative	0.177	0.056	0.213	3.714	0.002
Attitude	0.440	0.087	0.314	5.067	0.001

Adjusted R² = 0.37.

At the beginning of the second step (intervention step), following the selection of experimental and control groups the homogeneity of two groups was evaluated in terms of demographic variables. The demographic variables of the two groups are shown in Table 3. Statistical analysis showed that none of the demographic variables in the experimental and control groups were significantly different.

Table 3.

Homogeneity of the experimental and control groups in terms of demographic variables (age, work experience, education level and marital status, job title and work accident experience, etc.).

Demographic variable		Group
		Control	Experiment	Test results
		SD ± mean	SD ± mean	Test results
Age (year)		28.87 ± 4.23	29.70 ± 4.34	Z (Mann–Whitney) = 0.89 p = 0.327
Work experience (year)		4.38 ± 3.10	4.40 ± 3.40	Z (Mann–Whitney) = 0.07 p = 0.945
		No. (%)	No. (%)
Level of education	Junior high	5 (12.5)	5 (12.5)	Z (Mann–Whitney) = −0.22p = 0.823
	High school diploma	27 (67.5)	26 (65)
	Associate degree	4 (10)	4 (10)
	Bachelor’s degree and higher	4 (10)	5 (12.5)
Marital status	Married	37 (92.5)	35 (87.5)	Fisher’s Exact Test = 0.11p = 0.712
Marital status	Single	3 (7.5)	5 (12.5)	Fisher’s Exact Test = 0.11p = 0.712
Job title	Mechanic	7 (17.5)	8 (20)	Pearson Chi-Square = 2.23p = 0.676
	Worker	7 (17.5)	12 (30)
	Electrician	9 (22.5)	6 (15)
	Operator	9 (22.5)	8 (20)
	Workman	8 (20)	6 (15)
Experience of work accident	Yes	6 (15)	5 (12.5)	Pearson Chi-Square = 0.11p = 0.754
Experience of work accident	No	34 (85)	35 (87.5)	Pearson Chi-Square = 0.11p = 0.754

Table 4 shows the mean and SD of HAM constructs before and after the intervention in both experimental and control groups. As shown in the table, with regard to the constructs of belief system, normative systems and knowledge, there was no significant difference between the mean scores of the control group before and after the intervention, but in the experimental group, the mean scores of belief system, normative systems and knowledge constructs had increased significantly. Moreover, prior to the intervention, the constructs of belief system, normative systems and knowledge were not significantly different in control and experimental groups, but a significant difference was observed between the two groups after the intervention (Table 4).

Table 4.

Distribution of means and standard deviation of HAM constructs before and after intervention.

		Difference	After intervention	Before intervention	Intra group comparison (Wilcoxon nonparametric test)
		Mean ± SD	Mean ± SD	Mean ± SD	Intra group comparison (Wilcoxon nonparametric test)
Self-efficacy	Experiment group	2.70 ± 2.87	30.27 ± 2.97	27.58 ± 4.24	Z = 4.65p < 0.001*
	Control group	0.20 ± 1.62	27.80 ± 2.99	27.60 ± 3.30	Z = 0.95p = 0.340
	Inter-group comparison (Mann–Whitney test)	Z = 4.19p < 0.001*	Z = 3.71p < 0.001*	Z = 0.42p = 0.677
Behavior	Experiment group	0.88 ± 0.89	9.60 ± 0.71	8.73 ± 0.82	Z = 4.31p < 0.001*
	Control group	0.33 ± 0.94	8.70 ± 0.69	8.38 ± 0.95	Z = 0.21p = 0.033*
	Inter-group comparison (Mann–Whitney test)	Z = 2.45p = 0.014*	Z = 5.08p < 0.001*	Z = 1.60p = 0.11
Belief	Experiment group	3.53 ± 2.75	59.33 ± 4.18	55.8 ± 3.83	Z = 4.79p < 0.001*
	Control group	−0.08 ± 1.93	55.5 ± 5.71	55.58 ± 6.19	Z = 0.22p = 0.829
	Inter-group comparison (Mann–Whitney test)	Z = 5.34p < 0.001*	Z = 2.79p = 0.005*	Z = 0.69p = 0.478
Normative	Experiment group	2.90 ± 2.63	31.28 ± 3.96	28.38 ± 4.79	t = 6.98p < 0.001*
	Control group	−0.03 ± 1.96	29.05 ± 4.44	29.08 ± 5.32	t = 0.10p = 0.922
	Inter-group comparison (Independent t-tests)	t = 6.00p < 0.001*	t = 2.37p = 0.020*	t = 0.62p = 0.538	–
Attitude	Experiment group	1.30 ± 1.40	15.05 ± 2.44	13.75 ± 2.75	Z = 4.27p < 0.001*
	Control group	−0.45 ± 1.11	13.98 ± 2.78	14.43 ± 3.16	Z = 2.40p = 0.017*
	Inter-group comparison (Mann–Whitney test)	Z = 5.23p < 0.001*	Z = 2.04p = 0.041*	Z = 0.53p = 0.569	–
Behavioral intention	Experimental group	1.98 ± 2.71	31.83 ± 3.01	29.85 ± 3.28	Z = 4.22p < 0.001*
	Control group	−0.70 ± 1.83	28.05 ± 4.65	28.75 ± 5.55	Z = 2.34p = 0.019*
	Inter-group comparison (Mann–Whitney test)	Z = 4.75p < 0.001*	Z = 3.86p < 0.001*	Z = 0.22p = 0.828	–
Facilitators	Experimental group	1.08 ± 1.21	7.33 ± 1.84	6.25 ± 2.14	Z = 4.37p < 0.001*
	Control group	0.40 ± 1.45	6.7 ± 1.60	6.30 ± 2.26	Z = 1.44p = 0.149
	Inter-group comparison (Mann–Whitney test)	Z = 1.99p = 0.046*	Z = 1.85p = 1.066	Z = 0.05p = 0.977	–
Knowledge	Experimental group	3.08 ± 2.38	12.3 ± 3.24	9.23 ± 2.72	Z = 4.82p < 0.001*
	Control group	−0.11 ± 0.81	8.85 ± 2.26	8.96 ± 2.63	Z = 0.29p = 0.773
	Inter-group comparison (Mann–Whitney test)	Z = 5.45p < 0.001*	Z = 4.97p < 0.001*	Z = 0.412p = 0.681	–

: Statistically significant.

The mean score of attitude and behavioral intention rose significantly in the experiment group after the intervention, whereas this figure dropped significantly in the control group. Further, a comparison of the two groups before and after the intervention showed that the mean score of attitude and behavioral intention was not significantly different between the two groups before the intervention, but a significant difference was found after the intervention.

In determining the effect of education on facilitators, the findings indicated a significant increase in the mean score of the experimental group after intervention, whereas the mean score of the control group was not significantly different before and after intervention. A comparison of the two groups before and after the intervention revealed that the two groups were not significantly different, but the two groups were significantly different in terms of mean scores after intervention.

Discussion

In the present study, HAM structures had a predictive power of 37% for behavioral intention, which in general demonstrates the usefulness of HAM for predicting behavioral intention. Among HAM constructs, attitudes, beliefs and normative systems possessed the highest predictive power. The study of Mazaheri et al., which applied HAM model to workers in Isfahan Steel Company, Iran, showed that belief and attitude systems were major predictors of behavioral intention (17).

In this study, belief system was found to be the most powerful construct associated with behavioral intention. People’s attitudes about a subject are the outcome of their perspectives regarding that subject. Attitudes can be considered as a valid and reliable predictor of occupational accidents. With many safety studies, the focus is on the attitudes of people, and changing attitudes towards safety has been the major factor affecting a person’s behavior (18 –20).

Belief system also played a key role in behavioral intention, as it was the second major predictor of behavioral intention. Williamson et al. used a safety atmosphere assessment questionnaire to compare motivational and belief questions; their findings showed that motivational questions had greater bias whereas belief questions had higher reliability and validity (21).

Tones et al. emphasize the importance of belief system in this model, arguing that belief system is a cognitive structure in the HAM model, before focusing on Fisherbin’s definition of belief, according to which belief is a social judgment that associates a concept or subject to a particular trait (22). In planning interventions, the importance of these beliefs should be appreciated, and systematic methods need to be adopted for their modification.

The findings about normative system revealed the lower predictive power of this construct compared to attitudes and beliefs constructs in predicting behavioral intention. Obviously, all structures of a model do not have the same effect in predicting behavior (23). One reason for the low predictive power of a normative system is the dominant culture of the society, and the conservative responses of individuals in fear of perceived adverse consequences that may follow a written comment on laws and regulations or performance of supervisors and managers.

Based on the intervention group the results revealed the positive effect of an HAM-based education intervention on improving healthy behaviors of workers. According to the results, the education intervention improved the belief system of workers. This significant change in the belief system of participants in the experimental group was related mainly to their increased knowledge of the desired behavior and the positive experiences of individuals after performing healthy behaviors. Naturally, training classes and discussion groups helped change the beliefs of workers, as the supportive atmosphere created in such classes could yield psychological outcomes for participating individuals (24,25).

With regard to the normative system, this study implemented educational action for support of supervisors and training managers, with the results of intervention indicating its effectiveness on enhancing normative condition. In their studies, Cook and McSween (26) and Geller et al. (27) show that direct observation of workers’ behaviors by supervisors or managers is effective in improving attitudes towards safe behavior.

Studies have shown that a strong link between managers and members of workgroup fosters healthy actions and practices. Also, management plays a key role in the effectiveness of health programs. Managers should actively put into practice the health-related ideas of people, allocate adequate resources to health issues, address complaints and suggestions related to health promptly, attend safety meetings, pay attention to safety and health training, visit workplaces on a regular basis and pledge their commitment to safety regulations (28).

As for interventions related to attitude construct, results reflected the positive impact of these interventions on improving workers’ attitudes. Lund argued that attitude change could affect health actions significantly by manipulating social norms (29). Furthermore, researchers have shown that failure to deal with attitudes related to safety and health interventions would retain the current rate of occupational accidents, with equipment and engineering controls being unable to exert any marked effect (30).

The findings of this research on health actions showed that a training intervention enhanced healthy behavior in the experimental group. This could be attributed to the impact of the intervention on predictor constructs of behavioral intention (attitude, belief and normative systems), as these interventions can foster self-efficacy in workers and consequently improve behavioral intention, which in turn encourages healthy actions and behaviors.

This study had a number of limitations, the most important of which was its restriction to male workers.

Conclusion

The results of this study supported the usefulness of an HAM model in affecting attitudes, beliefs, norms, behavioral intention, knowledge and facilitators of healthy actions and behaviors in workers. Overall, the findings revealed the effect of an HAM-based educational intervention on promoting healthy behaviors in workers. Future studies are recommended to assess this model in female workers as well. The use of instructional strategies to raise workers knowledge about hazards at work and provide feedback to workers’ safety are suggested as practical actions.

Footnotes

Authors’ contribution

Design and implementation of the project: Hadi Tehrani

Scientific monitoring of the project: Nooshin Peyman

Analysis of data: Mohammad Vahedian-Shahroodi

Involvement in implementation of the project: Faeze Mohammadi

Participation in writing article: Mahdi Gholian-Aval, Hadi Tehrani

ORCID iD

Hadi Tehrani

Conflict of interest

The authors declare that there is no conflict of interest.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Page

. Blood on the coal: the effect of organizational size and differentiation on coal mine accidents. J Safety Res. 2009; 40: 85–95.

Banerjee

Cukic

(eds) On the cost of mining very large open source repositories. Proceedings of the First International Workshop on BIG Data Software Engineering. Piscataway, NJ: IEEE Press, 2015.

Pazand

. Characterization of REE geochemistry of coals from the Sangrud coal mine, Alborz coalfield, Iran. Arab J Geosci. 2015; 8: 8277–8282.

Mahdevari

Shahriar

Esfahanipour

. Human health and safety risks management in underground coal mines using fuzzy TOPSIS. Sci Total Environ. 2014; 488: 85–99.

Yorio

Wachter

. The impact of human performance focused safety and health management practices on injury and illness rates: do size and industry matter? Saf Sci. 2014; 62: 157–167.

Haas

Yorio

. Exploring the state of health and safety management system performance measurement in mining organizations. Saf Sci. 2016; 83: 48–58.

Peyman

Rezai-Rad

Tehrani

Gholian-Aval

Vahedian-Shahroodi

Miri

. Digital media-based health intervention on the promotion of women’s physical activity: a quasi-experimental study. BMC Public Health. 2018; 18: 134.

Sorensen

Emmons

Hunt

Barbeau

Goldman

Peterson

, et al. Model for incorporating social context in health behavior interventions: Applications for cancer prevention for working-class, multiethnic populations. Prev Med. 2003; 37: 188–197.

Tehrani

Majlessi

Shojaeizadeh

Sadeghi

Kabootarkhani

. Applying socioecological model to improve women’s physical activity: a randomized control trial. Iran Red Crescent Med J. 2016; 18: e21072.

10.

Moos

. Theory-based active ingredients of effective treatments for substance use disorders. Drug Alcohol Depend. 2007; 88: 109–121.

11.

Racher

Annis

. Community health action model: health promotion by the community. Res Theory Nurs Pract. 2008; 22: 182–191.

12.

Nieto-Montenegro

Brown

LaBorde

. Using the health action model to plan food safety educational materials for Hispanic workers in the mushroom industry. Food Control. 2006; 17: 757–767.

13.

Moore

Buchanan

Fairley

Smith

. Public health action model for cancer survivorship. Am J Prev Med. 2015; 49: S470–S476.

14.

Seaman

. Food hygiene training: Introducing the food hygiene training model. Food Control. 2010; 21: 381–387.

15.

Whitehead

. Health education, behavioural change and social psychology: nursing’s contribution to health promotion? J Adv Nurs. 2001; 34: 822–832.

16.

Rennie

. Evaluation of food hygiene education. Br Food J. 1994; 96: 20–25.

17.

Amidi Mazaheri

Hidarnia

Ghofranipour

. Predictors of safety attitude among workers based on the health action model. J Health Sys Res. 2011; 7: 684–692.

18.

Glendon

Litherland

. Safety climate factors, group differences and safety behaviour in road construction. Saf Sci. 2001; 39: 157–188.

19.

Geller

. Behavior-based safety and occupational risk management. Behav Modif. 2005; 29: 539–561.

20.

Barling

Kelloway

Iverson

. Accidental outcomes: attitudinal consequences of workplace injuries. J Occup Health Psychol. 2003; 8: 74–85.

21.

Williamson

Feyer

Cairns

Biancotti

. The development of a measure of safety climate: the role of safety perceptions and attitudes. Saf Sci. 1997; 25: 15–27.

22.

Tones

Robinson

Tilford

. Health education: effectiveness and efficiency. New York: Springer; 2013.

23.

Kim

Ham

Yang

Choi

. The roles of attitude, subjective norm, and perceived behavioral control in the formation of consumers’ behavioral intentions to read menu labels in the restaurant industry. Int J Hosp Manag. 2013; 35: 203–213.

24.

Wheeler

Zuo

Bjornlund

. Farmers’ climate change beliefs and adaptation strategies for a water scarce future in Australia. Glob Environ Change. 2013; 23: 537–547.

25.

Hadiyono

JEP

Suryawati

Danu

Santoso

. Interactional group discussion: results of a controlled trial using a behavioral intervention to reduce the use of injections in public health facilities. Soc Sci Med. 1996; 42: 1177–1183.

26.

Cook

McSween

. The role of supervisors in behavioral safety observations. Prof Saf. 2000; 45: 33.

27.

Geller

Perdue

French

. Behavior-based safety coaching: 10 Guidelines for successful application. Prof Saf. 2004; 49: 42.

28.

Abudayyeh

Fredericks

Butt

Shaar

. An investigation of management’s commitment to construction safety. Int J Proj Manag. 2006; 24: 167–174.

29.

Lund

Lassen

Sandøe

. Public attitude formation regarding animal research. Anthrozoos. 2012; 25: 475–490.

30.

Haynes

Weiser

Berry

Lipsitz

Breizat

AHS

Dellinger

, et al. Changes in safety attitude and relationship to decreased postoperative morbidity and mortality following implementation of a checklist-based surgical safety intervention. BMJ Qual Saf. 2011; 20: 102–107.