2-Mohammadfam

JRHS 2009; 9(1): 7-12

Copyright © Journal of Research in Health Sciences

Evaluation of Injuries among a Manufacturing Industry Staff in Iran

Mohammadfam   I(PhD)a, Moghimbeigi A( PhD)b

aDepartment of Occupational Health, School of Public Health and Center for Health Research, Hamadan University of Medical Sciences, Iran

bDepartment of Biostatistics and Epidemiology, School of Public Health and Center for Health Research, Hamadan University of Medical Sciences, Iran

*Corresponding Author: A Moghimbeigi, , E-mail: moghimb@yahoo.com,

Received: 7 November 2008; Accepted: 19 March 2009

Abstract

Background: Occupational injury is related to personal characteristics. This phenomenon is a controversial issue. This paper presents the relationships of certain occupational and individual characteristics with frequency of occupational injuries.

Methods: A standardized injury questionnaire was completed for 199 employees in a big Iranian industrial company (MAPNA Group) by the researcher in the presence of the subjects. The data were analyzed using zero-inflated Poisson regression with random effects.

Results: We demonstrated a significant relation between the marital status (P< 0.001) and score of injures (P < 0.001) with number of injuries by employees. Technicians and supervisors have high chance of "not to be injured at all" relative to workers (P< 0.05).  Technicians and supervisors have less number of injuring than workers have (P< 0.05). In addition, increasing assessment score decreases the number of injuring of employees (P< 0.001).

Conclusion: Due to being aware of the risks and remedial measures, married employees and workers should be assisted by occupational specialists.

Keywords: Occupational injury, Industry, Iran

Introduction

Injuries and non-communicable diseases are significant contributors to childhood, adoles­cent and adulthood deaths and morbidity in developing countries (1, 2). Despite their high prevalence, injuries remain a neglected public health problem in developing coun­tries (3, 4). It is important to identify risk factors in order to inform, plan, implement and evaluate health promotion policies and strategies that reduce the occurrence of injuries (5, 6). Risk factors associated with injuries include socioeconomic, demo­graphic, psychosocial and behavioral factors (7). Oc­cupational injuries result in severe socio-economical consequences (8, 9). The annual incidence of injuries workers from the Ira­nian Ministry of Labor (IML) is 43%. Oc­cupational injuries are mainly caused by work conditions (10), but some individual characteristics also increase the risk of accidents (11, 12). Some workers have more than one injury during a short period. There­fore, identification of risk factors involving in injury management should be taken into consideration (13, 14).

The concept of injury is used to indicate that many individuals have more accident-related health problems than others (15, 16). Green (15) was the first to observe that a relatively small proportion of workers in a British munitions factory had most of the injuries. He suggested that the explanation for this clustering of injuries in certain persons was to be found in their personality make up.

Several researches were conducted in differ­ent areas of Iran. All of these tried to expect prevalence of work accident or to find rela­tion between accident and many factors (17-22).

The objective of this research was to investi­gate the individual and occupational factors in the frequency of occupational injuries in a manufacturing industry in Iran. In terms of basic personal characteristics, we regard num­ber of injury as the tendency of an indi­vidual to experience more injury than others. We did not include exposure to risk as part of the definition itself, because the extent to which individuals expose themselves to risk may be largely determined by personality characteristics.

Methods

This study was carried out in a manufactur­ing industry (MAPNA Group) with eight branches that are spread out in the many provinces of Iran. Then we selected propor­tional to size of each branch's employers who worked in duration of 2007 in this manufacturing industry. An injury proneness questionnaire was used to evaluate the injury proneness (23). The questionnaire has 65 questions that questions have five-scale re­sponding alternative, “never=0”, “rarely=1”, “sometimes=2”, “often=3”, “most times=4”. The score of each employer's injuring is sum­mation of 65 questions scores. Then Persian version of this questionnaire was provided. After questionnaire validity checking, it had high reliability (Cronbachs alpha: 0.726). Furthermore, the number of occupational in­jury in the 2007, work experience (years), job (0: Workers, 1: Technician, 2: Supervi­sors 3: Managers), marital status (0: Single 1: Married) and workplace (company) were included.

Statistical analysis

The number of injured for employees is count data and Poisson regression is appropriate model for analyzing such data. In many situa­tions where there is over-dispersion or excess zero, other distribution such as zero-inflated Poisson (ZIP) or negative binomial (NB) may be more appropriate for fitting data. The ZIP distribution regards as mixture of Poisson distribution and degenerate com­ponent placing all its mass at zeros. For counts, ZIP regression model is to examine the effects of risk factors or confounders by allowing both log-linear and the logistic regression to be linear functions of some covariates. In this context, there are score tests for extra zeros where the zero-inflation probability depends on covariates (24), and correlated count data (25). Xiang et al. (26) presented a score test for over-dispersion in a zip mixed regression. We used these score tests for model selection.

In this cross-sectional study, because em­ployees were nested within company, zero-inflated mixed Poisson regression (27) ap­plied for analysis the data. We did analysis of data with programs that are written with S-plus and Wald score test to justify about parameters significant level.  

Results

The mean SD of age of the people under study was 30.4 (6.4) yr. The work experi­ences mean was 7.0 (4.99) yr. About 48% were high school educated and 34% were college-educated. About 73% of samples were married and 24% had experienced an injury or more at their work.

The mean score in assessment of injury was 60.3 (10.18). The minimum and maximum achieved scores were 27 and 82, respec­tively. The assessment score of injury of 14.0 % of the employees was less than 50, 30.5% between 50 and 60, 39.5% between 60 and 70 and 16.0% of the rest was more than 70. In other words, in the companies under study, about 40% of employees are working in critical occupations. Table 1 shows that the overall prevalence of current injury of samples was 23.5% (95% CI 17.6%, 29.4%). The mean of number of injury by employee is 0.435 (95% CI 0.266, 0.604).

Table 1: Occupational injury frequency distribution of employees in one year (2007)

Number of injury

Frequancy

%

0

153

77.0

1

33

16.6

2

4

2.0

3

4

2.0

4

2

1.0

5

0

0.0

6

1

0.5

7

0

0.0

8

0

0.0

9

1

0.5

10

1

0.5

Assuming a Poisson distribution for number of injured for sample employees, the ex­pected number of zeros is 129. Therefore compared with 153 observed, 24 extra zero are observed relative to those expected under the Poisson assumption. Comparing NB and Poisson regressions shows that for these data NB is more appropriate (P<0.001). with com­pany correlated random effect, Xiang et al (25) score test value shows significantly zero-in­flation against Poisson distribution (P<0.001). In addition, the score test of Jansakul and Hinde (24) shows that there was extra zeros against NB regression (P<0.001), however with company correlated random effect, Xiang et al (26) score test value shows that there is not significantly over-dispersion against  zero-inflated Poisson regression.

Table 2 shows the result of fitting mixed ZIP regression model to number of injury by employees. At first, we considered full model with interactions between covariates. How­ever, since interaction terms and other factors were not significant, Table 2 shows the model with main effects for significant factors. All factors were candidate to enter to this model. Hence age, education and work experience, are not included in the model, i.e. they are not related to number of injury by employ­ments. Zero-inflated part of this model shows that supervisors (Adj. OR=0.004; P <0.001) and technicians (Adj. OR=0.141; P =0.020) have high chance of not to be injured at all relative to workers. In addition, this part of model indicates that employees with high score have less chance not to be injured at all than high score are (Adj. OR=0.925; P =0.039).

Under Poisson part of this model, married employers (Adj. RR= 4.735; P= 0.001) have high risk relative to singles to have more number of injured. The number of injured by supervisors (Adj. RR= 0.335; P = 0.036) and technicians (Adj. RR= 0.442; P= 0.020) are less than workers are. In addition, increasing assessment score decrease number of injuring of employees (Adj. RR=0.938; P <0.001). 

Furthermore a Pearson statistic for mixed effect of ZIP yields 187.049 on 186 degree of freedom (P= 0.54). Again, there is no evidence of lack of fit for the fitted model.

Table 2: Result of fitting ZIP regression random effects for occupational injury

Variable

Poisson part

Zero-inflation part


Adj. RR* (95% CI)

P-value

Adj. OR** (95% CI)

P-value

Marital status (reference: Single)

4.735 (1.915, 11.711)

0.001

2.351 (0.31, 23.942)

0.470

Job (reference:  Worker)





Technician

0.442 (0.222, 0.880)

0.020

0.141 (0.027, 0.734)

0.020

Supervisor

0.335 (0.120, 0.931)

0.036

0.004 (0.003, 0.005)

0.000

Manager

0.531 (0.204, 1.379)

0.194

0.137 (0.013, 1.415)

0.095

Score

0.938 (0.914, 0.962)

0.000

0.925 (0.859, 0.996)

0.039

δ2 (Company)

0.877

0.004

deviance

159.623

Pearson  statistic (DF)

177.674 ( 185)

-2log-likelihhood

268.444

*Adjusted Relative Risk , **Adjusted Relative Risk

Discussion

This research is one of the few studies conducted of its kind in the Iran. A number of studies investigated some aspects of accidental behavior among industrial em­ployees mostly on prevalence rates as well as prevention, control and determinants of accident without considering accident prone­ness (17-22). However, the aim of this study was to investigate the association between demographic factors and accident proneness scores with number of injures by employees with using ZIP regression model. It was found that marital status, kind of jobs and score of employees are played an important role on number of injuries.

Workers in these companies were involved in critical occupations, for instance operators of crane, pressure welder and the roof plumber. They have high assessment score of accident proneness over 60. This mentioned group is more likely to be involved in injuries and have the high rate of unsafe behavior ratio. Concerning the nature of their work, the employment of this group of the individuals in critical occupations will be very much dangerous (28). In the companies under study, there were six instances of death, 5 of which included the after-mentioned critical occupations.

In this study, the highest score of accident proneness is considered for the occupational groups, which are more dangerous for first accident, for instance the groups of construc­tion workers putting up scaffolding and drivers. Making a mistake can be very catas­trophic in these jobs. This result indicates the more necessity of implementing control­ling measures. However, there is indirect relation between increasing assessment score and number of injures and it is conflict with other studies (29, 30).

The study re-emphasizes that it may be there are problems in married persons that these problems are risk factors for increasing injuries in married persons. Married persons with these problems need to be identified early to reduce the impact of these problems in the short and long-term. This will reduce the burden of injuries in employees, and may influence injuries that happen later.

The results of the study emphasize the need to screen the accident-prone individuals in the course of inspections and recruitment, assign them to less critical tasks, design and implement regular training and retraining sessions.

Acknowledgements

The authors would like to thank Hamadan University of Medical Sciences for support­ing this research. The authors declare that there is no conflict of interests.

References

  1. Bijur PE, Stewart-Brown S, Butler N. Child behavior and accidental injury in 11,966 preschool children. Am J Dis Child. 1986; 40:487-92.
  2. Bijur PE, Golding J, Haslum M.  Per¬sistence of occurrence of injury: can injuries of preschool children predict injuries of school-aged children? Pedi¬atrics. 1988; 82:707-12.
  3. Bijur PE, Golding J, Haslum M, Kurzon M. Behavioral predictors of injury in school-age children. Am J Dis Child. 1988; 142: 1307-12.
  4. Zwerling C, Whitten PS, Davis CS, Sprince NL. Occupational injuries among workers with disabilities: the national health interview survey 19851994. J Am Med Assoc. 1997; 278:2163-66.
  5. OConnor TG, Davies L, Dunn J, Golding J. Distribution of accidents, injuries, and illnesses by family type. ALSPAC Study Team. Avon Longitu¬dinal Study of Pregnancy and Child¬hood. Pediatrics. 106(5):1-6.
  6. Grossman, DC. The history of injury control and the epidemiology of child and adolescent injuries. Future Child. 10:23-52.
  7. Azadeh A, Mohammad Fam I. A frame¬work for development of intelli¬gent human engineering environment. Info Tech J. 5: 290-99.
  8. Bhattacherjee A, Chau N, Sierra CO, Legras B, Benamghar L, Michaely JP et al. Relationships of job and some individual characteristics with occupa¬tional injuries in employed people: a community based study. J Occup Health. 2003; 45:382-91.
  9. Bijttebier P, Vertommen H, Florentie K. Risk-taking behaviour as a mediator of the relationship between childrens temperament and injury liability. Psy¬chic Health. 2003; 18(5): 645-53.
  10. Azadeh A, Nouri J, Mohammad Fam I. The impacts of total system design fac¬tors on human performance in power plants. Am J Appl sci. 2005; 2:1301-04.
  11. Engel, HO. Accident proneness and illness proneness: a review. J R Soc Med. 1991; 84(3): 163-4.
  12. Dembe AE. The social consequences of occupational injuries and illnesses. Am J Ind Med.  2001; 40:40317.
  13. Gabel CL, Gerberich SG. Risk factors for injury among veterinarians. Epi¬demiology. 2002; 13:806.
  14. Gauchard GC, Chau N, Touron C , Benamghar L, Dehaene D, Perrin P et al. Role of certain individual charac¬teristics in occupational injuries due to disequilibrium: a case-control study in the employees of a railway company. Occup Env Med. 2003; 60:33035.
  15. Green, J. Accident proneness. J R Soc Med. 1991; 84(8): 510.
  16. Hindmarch, I. Accident proneness and illness proneness. J R Soc Med. 1991; 84(9): 570.
  17. Parvizpour, D. Epidemiology of work accident in Iran. Sing Med J. 1977; 18: 53-6.
  18. Parvizpour, D. An epidemiological study of industrial accidents among injured workers in Tehran, Iran. Iran J Pub Health,1979; 8: 51-60.
  19. Mohammad Fam I, Ghaziiadeh A. An epidemiological survey of load to death road accident in Tehran province in 1999. Sci J Kurd Uni Med Sci. 2002; 6(3): 35-40.
  20. Mohammadfam I, sadri G. The assess¬ment of job incidents in Iran from 1369-76. Tabib-e-shargh. 2000; 1:175-81.
  21. Mohammadfam I, Evaluation of oc¬cupational accidents and their related factors in Iranian Aluminum company in 1999. Sci J Kurd Uni Med Sci, 2001; 5(3) 18-23.
  22. Sanati KA, Yadegarfar G, Naghavi SHR, Sadr AH., Gholami M, Hadipour M et al. Occupational injuries in a synthetic fiber factory in Iran. Occup Med, 2009, 59: 62-5.
  23. Accident proneness test. Available from:  http://www.queendom.com/tests/index.htm
  24. Jansakul, N., Hinde, JP. Score tests for zero-inflated Poisson models. Comp Stat Data An, 2002; 40, 75 -96.
  25. Xiang, L, Lee, AH, Yau KKW, McL¬achlan GJ. A score test for zero-infla¬tion in correlated count data. Stat Med. 2006; 25, 1660-71.
  26. Xiang L, Lee AH, Yau KKW, McL¬achlan GJ. A score test for over-disper¬sion in zero inflated Poisson mixed regression model. Stat Med.  2007; 26, 160822.
  27. Yau, KKW, Lee, AH. Zero-inflated Poisson regression with random effects to evaluate an occupational injury pre¬vention programme. Stat Med. 2001; 20, 29072920.
  28. Marusic, A, Musek, J, Gudjonsson, G. Injury proneness and personality. Nor J psych. 2000; 155 (3): 157-161.
  29. Mandal, MK, Suar, D, Bhattacharya, T.  Side bias and accidents: are they re¬lated? Int J Neuro. 2001; .109: 139146.
  30. Kirschenbaum, A, Oigenblick, L, Gold¬berg, AI. Well being, work en¬vironment and work accidents.  Soc Sci Med. 2000; 50 (5): 631639.


JRHS Office:

School of Public Health, Hamadan University of Medical Sciences, Shaheed Fahmideh Ave. Hamadan, Islamic Republic of Iran

Postal code: 6517838695, PO box: 65175-4171

Tel: +98 81 38380292, Fax: +98 81 38380509

E-mail: jrhs@umsha.ac.ir