Abstract for Poster 52

 

 

Exposure, communication and hazard: A case control study of skin sensitizers and worker sensitization

T. Keegel*¹, A.S. Fletcher¹, J. Cahill², R.L. Nixon², S. Sakata³, M. Moyle²,
A.D. LaMontagne^1
1University of Melbourne, Melbourne, Australia
²Occupational Dermatology Research and Education Centre, Melbourne, Australia
³Monash University, Melbourne, Australia

Background

Effective communication regarding hazardous substances is vital for the adequate protection of human health and the environment¹. The importance of international consistency with information provision has been recognised through the development of the Globally Harmonised System for the Classification and Labelling of Chemicals (GHS)². With the advent of the GHS, under which Safety Data Sheets (SDS) will supersede Material Safety Data Sheets (MSDS), it is timely to consider both how the model of information delivery operates and its effectiveness across all stages of the hazard communication pathway.

Most studies of MSDS examine readability^3,4 and uptake by workers^5,6, or accuracy of reporting specified health effects^7-13. Previous studies have not specifically tested the association between the presence of MSDS warning information and prevention of exposure or associated disease. Hazard communication designed to inform employers and workers has been legislated in many industrialised countries^14-16. Legislation generally consists of criteria for the classification of hazard and for the provision of communication elements such as MSDS or SDS, appropriate labelling and placarding.

In Australia, manufacturers are required to provide MSDS for all hazardous materials ¹⁷. These materials are required to be classified as hazardous and labeled with a general statement of hazard such as the warning “Hazardous according to the criteria of the National Occupational Health and Safety Commission (NOHSC)”. Recognized skin sensitizers present at a concentration >= 1% must be listed in the ingredients and identified with the designated risk phrases, R43 “May cause sensitization by skin contact”. Australian Occupational Health and Safety (OHS) legislation also specifies that manufacturers must disclose all ingredients of a product at the request of a treating registered medical practitioner, for use in the diagnosis and management of a worker’s skin condition ¹⁸. We hypothesise that accurate reporting of a skin sensitizer should serve to minimize exposure and therefore reduce the probability of sensitization. This study tests the hypothesis amongst a population of workers with suspected occupational skin disease.

Methods

Setting and population

Between 1/1/01-30/9/03, N=705 workers with suspected occupational contact dermatitis attended the Occupational Dermatology Clinic, Melbourne, Australia. At the time of making their initial appointment, workers are routinely requested to bring along available MSDS for all products to which they come into contact as part of their usual work. A clinician will contact the product manufacturers and request verification of listed ingredients, as well as details of unlisted ingredients. If requested by the manufacturer, the relevant section of the OHS Act ¹⁸ is provided in writing. Workers who presented an MSDS for a product, which contained a sensitizer, were included in the study. The sensitizer may have been listed on the MSDS, or may have been identified when a clinician followed up ingredients with the manufacturer. Patch testing was utilized to determine sensitization. Workers were patch tested to sensitizers which were contained in the products, and/or the product itself, if the product was suitable for testing. A worker who had a positive reaction to a patch test for a sensitizer contained in the product was defined as a case. Workers who did not have a positive reaction to a sensitizer contained in the product were defined as controls.

Hazard communication variables

MSDS accuracy was assessed in terms of:

1.       presence of the general statement of hazard “Hazardous according to the criteria of NOHSC Australia”

2.       listing of sensitizer names, and

3.       presence of the risk phrase R43.

Clinician’s assessment of working conditions

At the time of diagnosis workers are routinely questioned by a clinician about occupational exposure. Appropriate provision and worker uptake of personal protective equipment is also assessed. Clinician determined assessment of exposure based on worker report was built into the model as a proxy measure for workplace exposure.

Sensitizer potency rating

A measure of hazard was built into the model through an estimation of sensitizer potential.

Relative potency of the sensitizer was determined by local lymph node assay tests¹⁹, from published data²⁰ and personal communication ²¹.

Analysis

Required sample size to detect differences between MSDS accuracy between the workers who were sensitized (cases) and non-sensitized (controls) was calculated using EpiInfo ²². We report numbers of workers who brought MSDS to the clinic and MSDS accuracy in terms of the three criteria.  MSDS accuracy was compared between cases and controls and odds ratios were calculated to describe the likelihood of worker sensitization. Multiple logistic regression methods were used to model for worker sensitization adjusted for confounders. Statistical analyses were performed using Stata 7 ²³.

Results

127/705 (18%) workers brought MSDS for products that contained sensitizers to their appointments and were included in the study. Of these, 46/127 (36.2%) were sensitized to a sensitizer contained in the product and classified as cases, with the remaining 85 workers classified as controls.

60 (47%) of the MSDS correctly stated that the product was hazardous and were labelled with the general warning “Hazardous according to the criteria of NOHSC”. 101 (80%) of the MSDS listed sensitizers, 50 (39%) had the appropriate R43 risk phrase “May cause sensitization”. 33(25%) of the MSDS were accurate for all three criteria.

Information regarding adequate exposure control, as determined by clinician assessment, was available for 105 (83%) workers. Adequate exposure control was significantly protective against sensitization (odds ratio [OR] 0.08, (95% confidence interval [95%CI]: 0.03-0.3). Sensitization potential, used as a measure of hazard, was available for 94 (74%) of the sensitizers. Increased hazard was significantly predictive of worker sensitization, with workers using highly sensitizing substances twice as likely to be sensitized (OR 2.0, 95%CI: 1.3-3.0). Information regarding combined exposure and hazard was only available for 78 (62%) workers, thus impacting on power when built into the predictive models. 

 

 

An increased risk of sensitization was observed for each of the measures of MSDS accuracy (Table 1). This was significant for MSDS that correctly had the R43 risk phrase (OR 4.2, 95%CI 1.9-8.9) and for MSDS that correctly had all three measures of accuracy (OR 4.0, 95%CI 1.8-9.2). When the model was adjusted for confounding by exposure and hazard, neither the R43 risk phrase, nor MSDS accurate for all three measures, were significant for the prediction of worker sensitization (Table 2). However, as discussed above, this may be caused in part by reduced sample size. 

 

  

Discussion

Protective effect for the general warning, sensitizer listing and R43 risk phrase was not observed. Rather an R43 and MSDS accurate for all three measures were significantly associated with worker sensitization. Confounding factors were suspected and accordingly built into the model. Once hazard and exposure were factored into the analysis none of the measures on the MSDS designed to provide protection for workers were associated with decreased worker sensitization.

This study is limited in a number of ways. The sample is based on tertiary referral occupational contact dermatitis clinic data and as such there may be selection bias. There may be some measurement error caused by the clinician assessment of workplace exposure.  Although the study sample size was adequate according to our calculations, information regarding confounders were only available for a proportion of workers. Finally, the possibility of uncontrolled confounding cannot be rejected.

Information regarding hazardous substances must be suited to the context of the workplace where it is to be implemented. The hazard communication system assumes that MSDS and SDS will smoothly disseminate accurate and comprehensible information from the manufacturer/supplier through to the employer and the employee. Workers need guaranteed access to MSDS/SDS in the work place, in addition to training regarding the interpretation of MSDS/SDS. This study suggests that a system of hazard communication, which is reliant on the dissemination of information through MSDS/SDS, is inadequate for the protection of workers.

References

1.        Hadden SG. Providing citizens with information about health effects of hazardous chemicals. Journal of Occupational Medicine 1989; 31: 528-34.

2.        United Nations. Globally harmonised system of classification and labelling of chemicals (GHS). New York and Geneva: United Nations, 2003.

3.        Kolp P, Sattler B, Blayney M et al. Comprehensibility of material safety data sheets. Am J Ind Med 1993; 23: 135-41.

4.        Sadhra S, Petts J, McAlpine S et al. Workers' understanding of chemical risks: electroplating case study. Occupational & Environmental Medicine 2002; 59: 689-95.

5.        Pearson C, Game W, Corbett C et al. Hazardous substances in the construction industry. Journal of Occupational Health Safety- Australia New Zealand 1995; 11: 510-2.

6.        Phillips CC, Wallace BC, Hamilton CB et al. The efficacy of material safety data sheets and worker acceptability. Journal of Safety Research 1999; 30: 113-22.

7.        Bernstein JA. Material safety data sheets: are they reliable in identifying human hazards? Journal of Allergy & Clinical Immunology. 2002; 110: 35-8.

8.        Frazier LM, Beasley BW, Sharma GK et al. Health information in material safety data sheets for a chemical which causes asthma. J Gen Intern Med 2001; 16: 89-93.

9.        Kanerva L, Henriks-Eckerman ML, Jolanki R et al. Plastics/acrylics: material safety data sheets need to be improved. Clin Dermatol 1997; 15: 533-46.

10.     Kolp PW, Williams PL, Burtan RC. Assessment of the Accuracy of Material Safety Data Sheets. American Industrial Hygiene Association Journal 1995; 56: 178-83.

11.     Lerman SE, Kipen HM. Material safety data sheets. Caveat emptor. Archives of Internal Medicine 1990; 150: 981-4.

12.     Cote R, Davis H, Dimock C et al. The evaluation and hazard classification of toxicological information for workplace hazardous materials information system material safety data sheets. Regulatory Toxicology and Pharmacology 1998; 27: 61-74.

13.     Welsh MS, Lamesse M, Karpinski E. The verification of hazardous ingredients disclosures in selected material safety data sheets. Appl Occup Environ Hyg 2000; 15: 409-20.

14.     Nielsen JB, Grandjean P. Criteria for skin notation in different countries. American Journal of Industrial Medicine 2004; 45: 275-80.

15.     Boeniger MF, Ahlers HW. Federal government regulation of occupational skin exposure in the USA. International Archives of Occupational & Environmental Health 2003; 76: 387-99.

16.     Roggeband R, Basketter DA, De Groot AC et al. Labelling of skin sensitizers: the new European Dangerous Preparations Directive. Contact Dermatitis 2001; 44: 321-4.

17.     National Occupational Health & Safety Commission. National Code of Practice for the Preparation of Material Safety Data Sheets [NOHSC:2011(2003)]. Canberra, 2003.

18.     Occupational Health and Safety (Hazardous Substances) Regulations. In: 211. Manufacturer's and importer's duty to disclose commercially confidential information to a registered medical practitioner, Vol. S.R. No 143/1999, 1999.

19.     Basketter DA, Balikie L, Dearman RJ et al. Use of the local lymph node assay for the estimation of relative contact allergenic potency. Contact Dermatitis 2000; 42: 344-8.

20.     Basketter DA, Kimber I. Predictive testing in contact allergy: facts and future. Allergy 2001; 56: 937-43.

21.     Basketter D. In., 2004.

22.     Dean A, Arner T, Sangam S et al. Epi Info 2000, a database and statistics program for public health professionals for use on Windows 95, 98, NT, and 2000 computers.  Atlanta, Georgia, USA: Centers for Disease Control and Prevention, 2000.

23.     StataCorp. Stata Statistical Software: Release 7.0. In. College Station, TX: Stata Corporation, 2001.

 

Content last modified: 24 May 2005