Hazard Communication

DEPARTMENT OF LABOR

Occupational Safety and Health Administration

29 CFR Parts 1910, 1915, and 1926

[Docket No. OSHA-H022K-2006-0062 (formerly Docket No. H022K)]

RIN 1218-AC20

Hazard Communication

AGENCY:

Occupational Safety and Health Administration (OSHA), Department of Labor.

ACTION:

Proposed rule; request for comments.

SUMMARY:

OSHA is proposing to modify its existing Hazard Communication Standard (HCS) to conform with the United Nations' (UN) Globally Harmonized System of Classification and Labelling of Chemicals (GHS). OSHA has made a preliminary determination that the proposed modifications will improve the quality and consistency of information provided to employers and employees regarding chemical hazards and associated protective measures. The Agency anticipates this improved information will enhance the effectiveness of the HCS in ensuring that employees are apprised of the chemical hazards to which they may be exposed, and in reducing the incidence of chemical-related occupational illnesses and injuries.

The proposed modifications to the standard include revised criteria for classification of chemical hazards; revised labeling provisions that include requirements for use of standardized signal words, pictograms, hazard statements, and precautionary statements; a specified format for safety data sheets; and related revisions to definitions of terms used in the standard, requirements for employee training on labels and safety data sheets. OSHA is also proposing to modify provisions of a number of other standards, including standards for flammable and combustible liquids, process safety management, and most substance-specific health standards, to ensure consistency with the modified HCS requirements.

DATES:

Written comments.

Written comments, including comments on the information collection determination described in Section VIII of the preamble (OMB Review under the Paperwork Reduction Act of 1995), must be submitted by the following dates:

Hard copy:

Comments must be submitted (postmarked or sent) by December 29, 2009.

Facsimile and electronic transmissions:

Comments must be sent by December 29, 2009.

Informal public hearings.

The Agency will schedule an informal public hearing on the proposed rule. The location and date of the hearing, procedures for interested parties to notify the Agency of their intention to participate, and procedures for participants to submit their testimony and documentary evidence will be announced in the

Federal Register.

ADDRESSES:

Written comments.

You may submit comments by any of the following methods:

Electronically:

You may submit comments electronically at

http://www.regulations.gov,

which is the Federal e-Rulemaking Portal. Follow the instructions on-line for making electronic submissions.

Fax:

If your submissions, including attachments, are not longer than 10 pages, you may fax them to the OSHA Docket Office at (202) 693-1648.

Mail, hand delivery, express mail, messenger or courier service:

You must submit three copies of your comments to the OSHA Docket Office, Docket No. OSHA-H022K-2006-0062, U.S. Department of Labor, Room N-2625, 200 Constitution Avenue, NW., Washington, DC 20210. Deliveries (hand, express mail, messenger and courier service) are accepted during the Department of Labor's and Docket Office's normal business hours, 8:15 a.m.-4:45 p.m., E.T.

Instructions:

All submissions must include the Agency name and the docket number for this rulemaking (Docket No. OSHA-H022K-2006-0062). All comments, including any personal information you provide, are placed in the public docket without change and may be made available online at

http://www.regulations.gov.

Therefore, OSHA cautions you about submitting personal information such as social security numbers and birthdates.

Docket:

To read or download comments submitted in response to this

Federal Register

notice, go to Docket No. OSHA-H022K-2006-0062 at

http://www.regulations.gov

or to the OSHA Docket Office at the address above. All comments are listed in the

http://www.regulations.gov

index; however, some information (

e.g.,

copyrighted material) is not publicly available to read or download through that Web page. All comments, including copyrighted material, are available for inspection and copying at the OSHA Docket Office.

Electronic copies of this

Federal Register

document are available at

http://regulations.gov.

Copies also are available from the OSHA Office of Publications, Room N-3101, U.S. Department of Labor, 200 Constitution Avenue, NW., Washington, DC 20210; telephone (202) 693-1888. This document, as well as news releases and other relevant information, are also available at OSHA's Web page at

http://www.osha.gov.

FOR FURTHER INFORMATION CONTACT:

For general information and press inquiries, contact Jennifer Ashley, Office of Communications, Room N-3647, OSHA, U.S. Department of Labor, 200 Constitution Avenue, NW., Washington, DC 20210; telephone (202) 693-1999. For technical information, contact Maureen O'Donnell, Directorate of Standards and Guidance, Room N-3718, OSHA, U.S. Department of Labor, 200 Constitution Avenue, NW., Washington, DC 20210; telephone (202) 693-1950.

SUPPLEMENTARY INFORMATION:

I. Introduction

The preamble to the proposal to modify the Hazard Communication Standard includes a review of the events leading to the proposal, a discussion of the reasons why OSHA believes these modifications are necessary, the preliminary economic and regulatory flexibility analysis for the proposal, and an explanation of the specific provisions set forth in the proposed standard. The discussion follows this outline:

I. Introduction

II. Issues

III. Events Leading to the Proposed Modifications to the Hazard Communication Standard

IV. Overview and Purpose of the Proposed Modifications to the Hazard Communication Standard

V. Need and Support for the Proposed Modifications to the Hazard Communication Standard

VI. Pertinent Legal Authority

VII. Preliminary Economic Analysis and Initial Regulatory Flexibility Analysis

VIII. OMB Review Under the Paperwork Reduction Act of 1995

IX. Federalism

X. State Plans

XI. Unfunded Mandates

XII. Protecting Children From Environmental Health and Safety Risks

XIII. Environmental Impacts

XIV. Public Participation

XV. Summary and Explanation of the Proposed Modifications to the Hazard Communication Standard

(a) Purpose

(b) Scope

(c) Definitions

(d) Hazard Classification

(e) Written Hazard Communication Program

(f) Labels and Other Forms of Warning

(g) Safety Data Sheets

(h) Employee Information and Training

(i) Trade Secrets

(j) Effective Dates

(k) Other Standards Affected

(l) Appendices

XVI. References

XVII. Authority and Signature

XVIII. Proposed Amendments

In the preamble, OSHA references a number of supporting materials. References to these materials are given as “Document ID#” followed by the last four digits of the document number. The referenced materials are posted in Docket No. OSHA-H022K-2006-0062 (which is available at

http://www.regulations.osha.gov

). The documents are also available at the OSHA Docket Office (see

ADDRESSES

section above). For further information about accessing documents referenced in this

Federal Register

notice,

see

Section XIV (Public Participation—Notice of Hearing).

II. Issues

OSHA requests comment on all relevant issues, including economic impact and feasibility, environmental impact, effects on small entities, proposed revisions to the HCS, and subsequent modifications to other standards. OSHA has received many comments on the issues raised in the Advance Notice of Proposed Rulemaking (ANPR) (71 FR 53617, September 12, 2006), and the Agency has considered those comments in the development of this proposal. This section identifies issues on which the Agency seeks additional information and comment to supplement that received in response to the ANPR, as well as new topics related to this proposal. While new comments are welcome, OSHA requests that comments submitted in response to the ANPR not be resubmitted as they are retained in the rulemaking record and reconsidered throughout the process.

OSHA is including these issues at the beginning of the document to assist readers as they consider the comments they plan to submit. However, to fully understand the questions and provide substantive input in response to them, the parts of the preamble that address these issues in detail should be read and reviewed. These include Section VII, which addresses the impacts of the NPRM, and thus provides the background related questions 2 through 5. Section XV provides the Summary and Explanation of the proposed regulatory text, and Section XVII is the text itself. These are key to understanding questions 6 through 26. It should be noted that the Federal Register's required format for a modification of an existing standard does not allow the Agency to provide the full text of the rule,

i.e.,

the regulatory text in this document only addresses those paragraphs that OSHA is proposing to change. Therefore, the Agency is putting a marked up version of the text of the current rule on its web page to help readers understand the proposed changes in context. The marked up text will be found on

www.osha.gov

under Hazard Communication in the subject index.

OSHA requests that comments be organized, to the extent possible, around the following issues and numbered questions. Submitting comments in an organized manner and with clear reference to the issue raised will enable all participants to easily see what issues the commenter addressed and how they were addressed. This is particularly important in a rulemaking such as GHS which affects many diverse industries. Many commenters, especially small businesses, are likely to confine their interest (and comment) to the issues that affect them, and they will benefit from being able to quickly identify comment on their issues in others' submissions. Of course, OSHA also welcomes relevant comments concerning the proposal that fall outside the issue questions raised in this section. However, the Agency is particularly interested in receiving public responses, supported by evidence and reasons, to the following questions:

Need and Support for the Standard

1. OSHA has made a preliminary determination that the proposed modifications to the HCS would increase the quality and consistency of information provided to employers and employees. Specifically, OSHA believes that standardized label elements would be more effective in communicating hazard information; standardized headings and a consistent order of information would improve the utility of SDSs; and training would support and enhance the effectiveness of the new label and SDS requirements. Is this assessment correct? OSHA requests information that reflects on the effectiveness of the proposed modifications to the HCS in protecting employees from chemical hazards in the workplace.

Economic Impacts and Economic Feasibility

2. The preliminary economic analysis in Section VII raises a variety of specific questions and issues with respect to the preliminary economic analysis. OSHA would appreciate it if you could place answers to these issues as heading 2 in your comments and further organize comments on the preliminary economic analysis (PEA) as follows:

a.

Industrial profiles.

This covers issues concerning how many employees, establishments and products would be affected by the proposed standard. OSHA welcomes comments on all aspects of the industrial profile and is particularly interested in comments on the number of affected employees, and the number of SDSs that would need revision, by industry.

b.

Issues with respect to estimated benefits of the proposed standard.

OSHA considers three kinds of benefits in this preliminary analysis: Benefits associated with preventing injuries, illnesses, and fatalities through clearer and more accessible information; benefits associated with reducing the time that safety and health managers and logistics and emergency response personnel spend on hazardous chemicals through clearer and easier-to-find information; and benefits associated with reducing the time needed to develop and review SDSs because of international harmonization. OSHA is particularly interested in comments on the scope of these benefits; the extent to which they are already being achieved by existing practices; and the extent to which they depend on other countries following the harmonization effort.

c.

Issues with respect to the costs and range of costs of the proposed standard.

OSHA preliminarily estimated the principal costs of the standard to chemical producers for reclassification of chemicals; remaking SDS's; and redoing labels; and to chemical users for familiarization and program changes for managers and for training exposed employees. OSHA welcomes comments on all aspects of the costs, and is particularly interested in comments on the extent to which chemical producers may have already met some of the requirements of the standard and the time and professional skills needed for the activities the standard would require.

d.

Issues with respect to economic impacts and feasibility of the proposed standard, including the sensitivity of OSHA's economic feasibility determination with respect to various assumptions.

OSHA welcomes comments on all aspects of the economic impact and economic feasibility analyses.

e.

All other issues with respect to the PEA.

Effects on Small Entities

3. OSHA has certified that the proposed standard will not have a

significant impact on a substantial number of small entities. Nevertheless, because of the number of small entities affected, OSHA has prepared a voluntary initial regulatory flexibility analysis, the results of which are described in Section VII of the proposed rule. Do you consider the estimated costs and impacts on small entities presented there to be reasonable? Why or why not?

4. Are there alternatives to the rule as a whole or specific requirements of the rule that reduce impacts on small entities while still protecting the health of employees and meeting the broad goal of a globally harmonized system?

Environmental Impacts

5. OSHA has preliminarily determined that the proposed standard will not have any adverse effects on the environment, and may have positive effects on the environment. OSHA welcomes comments on this determination.

Hazard Classification

6. OSHA is proposing to adopt all of the physical and health hazard classes in the GHS. Among the physical and health hazard classes, OSHA is proposing to include all hazard categories in the GHS except Acute Toxicity Category 5 for oral, dermal, or inhalation exposures; Skin Corrosion/Irritation Category 3; and Aspiration Hazard Category 2. If you believe that the exclusion of these hazard categories is not consistent with the scope and/or level of protection provided by the current HCS, please describe any recommended changes to this proposal and the reasons you think these changes are necessary.

7. OSHA has proposed a definition for unclassified hazards be added to the HCS to ensure that all hazards currently covered by the HCS—or new hazards that are identified in the future—are included in the scope of the revised standard until such time as specific criteria for the effect are added to the GHS and subsequently adopted by OSHA. Will this approach provide sufficient interim coverage for hazards such as combustible dust? Are there other hazards for which criteria should be developed and added to the GHS? Please provide information regarding these hazards, and the information available to characterize them.

8. OSHA believes it may be more appropriate to add specific coverage for simple asphyxiants to the standard in the final rule to ensure everyone properly addresses their coverage rather than addressing them under the unclassified hazard definition. This effect is simple and straightforward, and could be addressed in a definition that does not involve extensive criteria. OSHA is requesting comment on this approach. A possible definition would be as follows:

“Simple asphyxiants” are substances that displace oxygen in the ambient atmosphere, and can thus cause oxygen deprivation in exposed workers that leads to unconsciousness and death. They are of particular concern in confined spaces. Examples of asphyxiants include: nitrogen, helium, argon, propane, neon, carbon dioxide, and methane.

OSHA would also like to solicit comments on specific label elements for simple asphyxiants. No symbol would be required, but the signal word “warning” would be used, with the hazard statement “may be harmful if inhaled”. In addition, a precautionary statement such as the following would be required: May displace oxygen in breathing air and lead to suffocation and death, particularly in confined spaces.

All other requirements of the standard that apply to hazardous chemicals would also apply to chemicals that meet this definition. These substances would generally be covered already under the proposed rule as compressed gases, and may also pose other effects such as flammability that would have to be addressed as well. They are also already covered under the existing HCS. Is the definition suggested by OSHA sufficient to cover this effect? Do you have suggestions for modifying this definition? Are the label elements suggested appropriate?

9. In order to help to ensure that health hazard determinations are properly conducted under a performance-oriented approach, the HCS includes a “floor” of chemicals that are to be considered hazardous based on several cited reference lists. In addition, the existence of one toxicological study indicating a possible adverse effect is considered sufficient for a finding of hazard for any health effect. Under the GHS, there is no floor of chemicals cited, nor is there an across-the-board provision such as the one-study criterion. Instead, specific, detailed criteria are provided for each type of health hazard to guide the evaluation of relevant data and subsequent classification of the chemical. The proposed modifications to the HCS would align the standard to the GHS approach, and thus do not include the floor of chemicals nor the universal one-study rule. Would the proposed detailed criteria provide sufficient guidance for a thorough hazard evaluation?

10. OSHA has edited the chapters in the GHS for classification of physical and health hazards to remove material not directly related to classification and to otherwise streamline the text. OSHA anticipates providing the decision logics separately to serve as guidance, but has not included them in the regulatory text. Are there any additions, subtractions, or clarifications of the classification criteria from the GHS that OSHA needs to consider?

11. Certain physical hazard classification criteria (

i.e.,

for self-reactive chemicals, organic peroxides, self-heating chemicals, explosives) either directly reference packaging or quantity, or rely on test methods that reference packaging or quantity. The criteria were developed for transport concerns. Clearly, quantity and packaging can greatly affect safe transport of chemicals that pose hazards such as those listed above. However, OSHA seeks comments on whether the criteria as stated in the GHS are appropriate for the workplace. Does use of these criteria present any obstacles to classification or create any difficulties for suppliers or users of chemicals? Describe any difficulties these criteria may present and any suggestions for addressing these issues, particularly recommendations that would be consistent with the GHS and maintain the GHS level of safety for these chemicals.

12. The GHS gives countries guidance on a cut-off or concentration limit for chemical mixtures containing target organ toxicity hazards. In Appendix A, Section A.8.3, OSHA is proposing to make the suggested 20% concentration limit mandatory so that label preparers are clear on what needs to be done. Please comment on whether this mandatory concentration limit is appropriate. If you have an alternative, please provide it along with the rationale.

Labels

13. The proposal would require pictograms to have a red frame. As discussed in Section V, OSHA believes that use of the color red will make warnings more noticeable and will aid in communicating the presence of a hazard. However, the GHS gives competent authorities such as OSHA the discretion to allow use of a black frame when the pictogram appears on a label for a package which will not be exported. For packages that will not be exported, should the modified standard allow black frames on pictograms, or should the pictogram frame be required to be presented in red?

14. In addition to the pictograms, signal word and hazard statements, GHS labels must include precautionary statements. OSHA is proposing to require the text in the precautionary statements in the GHS to be on HCS labels. As discussed in Section XV Summary and Explanation of the Proposed Standard, these statements are codified under the GHS, meaning that numbers have been assigned to them. In addition, the appropriate statements to use for each hazard class and category have been indicated in the GHS annexes. This means that label preparers will know exactly what precautionary statements to apply once they complete their hazard classification, and chemical users will see consistent language on labels to indicate the necessary precautionary measures. However, the statements are not yet considered to be part of the harmonized text like hazard statements are; rather they are included in the GHS as an suggested language. OSHA expects that other countries may adopt the codified precautionary statements when they put GHS in place. For example the EU has required that labels use the GHS codified precautionary statement text in adapting the GHS. Since OSHA did not previously require the use of precautionary statements, and had no such recommended statements to provide, the Agency is proposing to use those currently in the GHS as the mandatory requirements with the option of consolidating statements where appropriate (

See

Appendix C). OSHA anticipates this approach will provide the maximum benefit. OSHA is also seeking comment on whether any of these statements should be modified or if other precautionary statements should be included.

In addition, as discussed in Section IV, OSHA has presented other alternatives with regards to precautionary statements, and OSHA is soliciting comment on these options as well. Specifically, OSHA is seeking feedback on whether the Agency should include the GHS precautionary statements as nonbinding examples, through a non mandatory appendix or guidance, rather than as required statements, or whether OSHA should allow label preparers to develop their own precautionary statements rather than specifying the text to be used.

15. OSHA has not proposed to require the exploding bomb pictogram or specific precautionary statements for Division 1.4S ammunition and ammunition components because the specified GHS label elements may not accurately reflect the hazards of these materials. Is this sufficiently protective? Are any adjustments to the label elements for Division 1.4S ammunition and ammunition components necessary? Describe any requested changes and explain why such revisions are necessary.

16. In the current HCS, OSHA has a provision that requires labels to be updated within three months of obtaining new and significant information about the hazards. The Agency has not been enforcing this provision for many years, and there has been an administrative stay on enforcement. OSHA is including the provision in this proposal, and inviting comment on it with the intention of including it in the final rule and lifting the stay. Is three months the appropriate time interval for updating? Are there any practical accommodations that need to accompany this limit (for example, related to stockpiles of chemicals)? Provide any alternatives you consider appropriate, as well as documentation to support them.

Safety Data Sheets (SDSs)

17. As discussed in Section XV, the Agency is proposing to require that OSHA permissible exposure limits (PELs) be included on the SDS, as well as any other exposure limit used or recommended by the chemical manufacturer, importer, or employer preparing the safety data sheet. OSHA welcomes comments on this approach, along with an explanation of the basis for your position.

18. OSHA is proposing that Section 15 of the SDS be non-mandatory. As indicated in Appendix D, Section 15 addresses regulatory information concerning the chemical. OSHA is considering requiring the substance specific standards be referenced in this section, which would make Section 15 mandatory. Would employers and employees benefit from having this information in this section of the SDS?

Other Standards Affected

19. OSHA is proposing to align the definitions of the physical hazards to the requirements of the GHS categories in safety standards for general industry, construction, and maritime standards, which either directly reference the HCS or provide information pertinent to the Safety Data Sheets (SDSs). In most cases OSHA has modified the standards to maintain scope and protection. However, the changes in definitions for flammable liquids Category 1 and 2 and flammable aerosols appear to be more than simply rounding to the nearest significant number.

○

Flammable liquids Category 1 and 2:

The boiling point cut-off for Category 1 is reduced from 100 deg F (37.8 deg C) or less to 95 deg F (35 deg C) or less, which could shift some liquids from Category 1 to Category 2.

○

Flammable aerosols:

OSHA is proposing to adopt the GHS method to determine flammability rather than the method defined by the Consumer Product Safety Commission (CPSC).

OSHA's decision to change these definitions to be consistent with the GHS is based not only upon harmonizing its standards with those of other countries that have adopted or may adopt the GHS, but OSHA is also concerned with making its standards internally consistent. OSHA believes the methods used to classify these physical hazards are similar enough so that substances that are currently regulated by OSHA would continue to be regulated and that few, if any, changes would result in a shift in regulatory coverage. Would the proposed changes have any impact on your operations? If so, describe the anticipated effects.

20. OSHA is proposing to eliminate the term “combustible liquid” in 29 CFR 1910.106. 1910.107, 1910.123, 1910.124, 1910.125, and 1926.155 for liquids with a flashpoint above 100 °F. To reflect consistency with the revised HCS where appropriate, OSHA is proposing to add the specific flashpoint criteria. This will maintain equivalent protection. Are there other standards that OSHA should update with the new terminology?

21. OSHA is proposing to modify the language required on signs in substance-specific health standards. The Agency developed the proposed language to reflect the terminology of the revised HCS while, at the same time, providing adequate warning through language that is consistent with the current sign requirements for these chemicals. An added benefit is the hazard warnings on signs specified for these standards will now be consistent throughout OSHA standards. For example, all carcinogens will now bear the hazard statement “MAY CAUSE CANCER”. OSHA believes that providing language that is consistent on both signs and labels will improve comprehension for employees. Does the proposed language on signs accurately convey the hazards?

22. OSHA is proposing to revise the substance-specific health standards' provisions on labeling for producers and importers of chemicals and substances. Currently in the substance-specific standards OSHA requires specific language on labels for certain chemicals. OSHA is proposing to change these labeling requirements by referring those responsible for labeling to the modified HCS and including in each substance-

specific standard a list of health effects that must be considered for hazard classification. The modified HCS will dictate the specific language (

i.e.,

signal word, hazard statement(s), and precautionary statement(s)) that is required on labels through the classification process. However, OSHA is proposing to maintain specific language for labels on contaminated clothing and waste/debris containers to ensure adequate hazard communication for the downstream recipients. How would the removal of required language for labels from substance-specific standards affect your work place? Are there hazard warnings that will be lost that do not have an equivalent hazard or precautionary statement? Are there alternatives to OSHA's approach for the substance-specific standards that will assure information is disseminated in a manner that is consistent with the modified HCS labeling requirements?

23. In determining the health hazards that need to be considered by manufacturers, importers and distributors when classifying chemicals regulated by the substance-specific standards, OSHA is proposing to primarily rely on the determinations made by the Agency in each rulemaking, the

NIOSH Pocket Guide to Chemical Hazards

(2005) and the International Chemical Safety Cards, and use as a secondary source the health effects identified by the European Commission (2007). OSHA is proposing to include a health hazard only if it is identified as such by two or more of these organizations. Are there other sources of information that OSHA should consult?

24. As detailed in the Summary and Explanation section of this document, OSHA is not proposing in this rulemaking to update the electrical standards (general industry 1910 subpart S and construction 1926 subpart K) or Explosives and blasting agents (general industry 1910.109 and construction 1926.914). These subparts are “self-contained” in that they do not rely on other OSHA standards for regulatory scope or definitions, but reference external organizations (such as the National Fire Protection Association [NFPA]). OSHA believes that these standards could be updated when the referenced external organizations adopt applicable GHS elements. If OSHA were to change these standards to comply with the GHS, how would this impact your operations?

Effective Dates

25. OSHA has proposed to require that employers train employees regarding the new labels and safety data sheets within two years after publication of the final rule to ensure they are familiar with the new approach when they begin to see new labels and SDSs in their workplaces. Is the proposed time appropriate?

26. OSHA has proposed that chemical manufacturers, importers, distributors, and employers be required to comply with all provisions of the modified final rule within three years after its publication. Does this allow adequate time to review hazard classifications and amend them as necessary, and to revise labels and safety data sheets to reflect the new requirements? Would a shorter time frame be sufficient?

27. Are there any other factors that should be considered in establishing the phase-in period?

Compliance Assistance and Outreach

28. OSHA received many comments in response to the questions in the ANPR regarding compliance assistance and outreach and is seeking additional comment in this proposal. However, comments already submitted need not be resubmitted. Please refer to the discussion in Section XV. Specifically, OSHA is interested in your responses to the following: What types of materials or products would best assist employers in understanding and complying with the modified HCS? OSHA seeks input to identify the tools that would be most useful to employers and employees, the subjects of greatest interest (

e.g.,

classification criteria, labels, safety data sheets), and the best means of distributing these materials.

29. OSHA received a number of comments that suggested that a data base of chemical classifications should be developed and maintained to assist chemical manufacturers and importers in performing hazard classifications. This approach has been adopted in some other countries. Would such a data base be helpful? Who would be responsible for doing the classifications and maintaining them? How would the data base be kept aligned with other countries' classifications?

Alternative Approaches

30. OSHA has described alternatives to the scope and application of the proposed rule in the preamble, Section IV. These include consideration of allowing voluntary implementation of the GHS; exemptions based on size of the business; adopting some components of the GHS but not others; and not adopting all of the required label elements. The Agency requests comments on these alternatives, with data to support the views expressed. Suggestions and support for other alternatives are requested as well.

III. Events Leading to the Proposed Modifications to the Hazard Communication Standard

OSHA's Hazard Communication Standard (HCS) (29 CFR 1910.1200; 1915.1200; 1917.28; 1918.90; and 1926.59) was first issued in 1983 and covered the manufacturing sector of industry (48 FR 53280, November 25, 1983). In 1987, the Agency expanded the scope of coverage to all industries where employees are potentially exposed to hazardous chemicals (52 FR 31852, August 24, 1987). Although full implementation in the non-manufacturing sector was delayed by various court and administrative actions, the rule has been fully enforced in all industries covered by OSHA since March 17, 1989 (54 FR 6886, February 15, 1989). In 1994, OSHA made a number of minor changes and technical amendments to the HCS to help ensure full compliance and achieve better protection of employees (59 FR 6126, February 9, 1994). The development of the HCS is discussed in detail in the preambles to the original and revised final rules (see 48 FR 53280-53281; 52 FR 31852-31854; and 59 FR 6127-6131). This discussion will focus on the sequence of events leading to the development of the Globally Harmonized System of Classification and Labelling of Chemicals (GHS) and the modifications to the HCS included in this proposed rule.

The HCS requires chemical manufacturers and importers to evaluate the chemicals they produce or import to determine if they are hazardous. The rule provides definitions of health and physical hazards to use as the criteria for determining hazards in the evaluation process. Information about hazards and protective measures is then required to be conveyed to downstream employers and employees through labels on containers and safety data sheets. All employers with hazardous chemicals in their workplaces are required to have a hazard communication program, including container labels, safety data sheets, and employee training. (

Note:

The HCS uses the term “material safety data sheet” or “MSDS”, while the GHS uses “safety data sheet” or “SDS”. For convenience and for consistency with the GHS, safety data sheet or SDS is being used throughout this document and that term would replace MSDS in the modified HCS.)

To protect employees and members of the public who are potentially exposed to chemicals during their production,

transportation, use, and disposal, a number of countries have developed laws that require information about those chemicals to be prepared and transmitted to affected parties. These laws vary with regard to the scope of chemicals covered, definitions of hazards, the specificity of requirements (

e.g.,

specification of a format for safety data sheets), and the use of symbols and pictograms. The inconsistencies between the various laws are substantial enough that different labels and safety data sheets must often be developed for the same product when it is marketed in different nations.

Within the U.S., several regulatory authorities exercise jurisdiction over chemical hazard communication. In addition to OSHA's HCS, the Department of Transportation (DOT) regulates chemicals in transport, the Consumer Product Safety Commission (CPSC) regulates consumer products, and the Environmental Protection Agency (EPA) regulates pesticides, as well as having other authority over labeling under the Toxic Substances Control Act. Each of these regulatory authorities operates under different statutory mandates, and has adopted distinct hazard communication requirements.

Tracking the hazard communication requirements of different regulatory authorities is a burden for manufacturers, importers, distributors, and transporters engaged in commerce in the domestic arena. This burden is magnified by the need to develop multiple sets of labels and safety data sheets for each product in international trade. Small businesses may have particular difficulty in coping with the complexities and costs involved. The problems associated with differing national and international requirements were recognized and discussed when the HCS was first issued in 1983. The preamble to the final rule included a commitment by OSHA to review the standard regularly to address international harmonization of hazard communication requirements. OSHA was asked to include this commitment in recognition of an interagency trade policy that supported the U.S. pursuing international harmonization of requirements for chemical classification and labeling. The potential benefits of harmonization were noted in the preamble:

* * * [O]SHA acknowledges the long-term benefit of maximum recognition of hazard warnings, especially in the case of containers leaving the workplace which go into interstate and international commerce. The development of internationally agreed standards would make possible the broadest recognition of the identified hazards while avoiding the creation of technical barriers to trade and reducing the costs of dissemination of hazard information by elimination of duplicative requirements which could otherwise apply to a chemical in commerce. As noted previously, these regulations will be reviewed on a regular basis with regard to similar requirements which may be evolving in the United States and in foreign countries. (48 FR 53287)

OSHA has actively participated in a number of such efforts in the years since that commitment was made, including trade-related discussions on the need for harmonization with major U.S. trading partners. The Agency also issued a Request for Information (RFI) in the

Federal Register

in January 1990 to obtain input regarding international harmonization efforts, and on work being done at that time by the International Labor Organization (ILO) to develop a convention and recommendations on safety in the use of chemicals at work (55 FR 2166, January 22, 1990). On a closely related matter, OSHA published an RFI in May 1990 requesting comments and information on improving the effectiveness of information transmitted under the HCS (55 FR 20580, May 17, 1990). Possible development of a standardized format or order of information was raised as an issue in the RFI. Nearly 600 comments were received in response to this request. The majority of responses expressed support for a standard SDS format, and the majority of responses that expressed an opinion on the topic favored a standardized format for labels as well.

In June 1992, the United Nations Conference on Environment and Development issued a mandate (Chapter 19 of Agenda 21), supported by the U.S., calling for development of a globally harmonized chemical classification and labeling system:

A globally harmonized hazard classification and compatible labelling system, including material safety data sheets and easily understandable symbols, should be available, if feasible, by the year 2000.

This international mandate initiated a substantial effort to develop the GHS, involving numerous international organizations, many countries, and extensive stakeholder representation.

A coordinating group comprised of countries, stakeholder representatives, and international organizations was established to manage the work. This group, the Inter-Organization Programme for the Sound Management of Chemicals Coordinating Group for the Harmonization of Chemical Classification Systems, established overall policy for the work and assigned tasks to other organizations to complete. The Coordinating Group then took the work of these organizations and integrated it to form the GHS. OSHA served as chair of the Coordinating Group.

The work was divided into three main parts: Classification criteria for physical hazards; classification criteria for health and environmental hazards (including criteria for mixtures); and hazard communication elements, including requirements for labels and safety data sheets. The criteria for physical hazards were developed by a United Nations Subcommittee of Experts on the Transport of Dangerous Goods/International Labour Organization working group and were based on the already harmonized criteria for the transport sector. The criteria for classification of health and environmental hazards were developed under the auspices of the Organization for Economic Cooperation and Development. The ILO developed the hazard communication elements. OSHA participated in all of this work, and served as U.S. lead on classification of mixtures and hazard communication.

Four major existing systems served as the primary basis for development of the GHS. These systems were the requirements in the U.S. for the workplace, consumers and pesticides; the requirements of Canada for the workplace, consumers and pesticides; European Union directives for classification and labeling of substances and preparations; and the United Nations Recommendations on the Transport of Dangerous Goods. The requirements of other systems were also examined as appropriate, and taken into account as the GHS was developed. The primary approach to reconciling these systems involved identifying the relevant provisions in each system; developing background documents that compared, contrasted, and explained the rationale for the provisions; and undertaking negotiations to find an agreed approach that addressed the needs of the countries and stakeholders involved. Principles to guide the work were established, including an agreement that protections of the existing systems would not be reduced as a result of harmonization. Thus countries could be assured that the existing protections of their systems would be maintained or enhanced in the GHS.

An interagency committee under the auspices of the Department of State coordinated U.S. involvement in the development of the GHS. In addition to OSHA, DOT, CPSC, and EPA, there were a number of other agencies

involved that had interests related to trade or other aspects of the GHS process. Different agencies took the lead in various parts of the discussions. Positions for the U.S. in these negotiations were coordinated through the interagency committee. Interested stakeholders were kept informed through e-mail dissemination of information, as well as periodic public meetings. In addition, the Department of State published a notice in the

Federal Register

that described the harmonization activities, the agencies involved, the principles of harmonization, and other information, as well as invited public comment on these issues (62 FR 15951, April 3, 1997). Stakeholders also actively participated in the discussions at the international level and were able to present their views directly in the negotiating process.

The GHS was formally adopted by the new United Nations Committee of Experts on the Transport of Dangerous Goods and the Globally Harmonized System of Classification and Labelling of Chemicals in December 2002. In 2003, the adoption was endorsed by the Economic and Social Council of the United Nations. The GHS will be updated as necessary to reflect new technology and scientific developments, or provide additional explanatory text. This proposed rule is based on Revision 3 of the GHS, published in 2009.

Countries have been encouraged to implement the GHS as soon as possible, and established a goal to have fully operational systems by 2008. This goal was adopted by countries in the Intergovernmental Forum on Chemical Safety, and was endorsed by the World Summit on Sustainable Development. The U.S. participated in these groups, and agreed to work toward achieving these goals. While much progress was made by the U.S. and other countries by the end of 2008, most are still in the process of implementing the GHS.

OSHA published an Advance Notice of Proposed Rulemaking (ANPR) on the GHS in September of 2006 (71 FR 53617, September 12, 2006). The ANPR provided information about the GHS and its potential impact on the HCS, and sought input from the public on issues related to GHS implementation. Over 100 responses were received, and the comments and information provided were taken into account in the development of the modifications to the HCS included in this proposed rule. At the same time the ANPR was published, OSHA made a document summarizing the GHS available on its Web site (

http://www.osha.gov).

OSHA remains engaged in a number of activities related to the GHS. The U.S. is a member of both the United Nations Committee of Experts on the Transport of Dangerous Goods and the Globally Harmonized System of Classification and Labeling of Chemicals, as well as the Subcommittee of Experts on the Globally Harmonized System of Classification and Labeling of Chemicals. These permanent UN bodies have international responsibility for maintaining, updating as necessary, and overseeing the implementation of the GHS. OSHA and other affected Federal agencies actively participate in these UN groups. In addition, OSHA and EPA also participate in the GHS Programme Advisory Group under the United Nations Institute for Training and Research (UNITAR). UNITAR is responsible for helping countries implement the GHS, and has ongoing programs to prepare guidance documents, conduct regional workshops, and implement pilot projects in a number of nations. OSHA also continues to be involved in interagency discussions related to coordination of domestic implementation of the GHS, and in discussions related to international work to implement and maintain the GHS.

IV. Overview and Purpose of the Proposed Modifications to the Hazard Communication Standard

The intent of the HCS is to ensure that the hazards of all chemicals are evaluated, and that information concerning chemical hazards and associated protective measures is transmitted to employers and employees. The standard achieves this goal by requiring chemical manufacturers and importers to review available scientific evidence concerning the physical and health hazards of the chemicals they produce or import to determine if they are hazardous. For every chemical found to be hazardous, the chemical manufacturer or importer must develop a container label and an SDS and provide both documents to downstream users of the chemical. All employers with employees exposed to hazardous chemicals must develop a hazard communication program, and ensure that exposed employees are provided with labels, access to SDSs, and training on the hazardous chemicals in their workplace.

The three information components in this system—labels, SDSs, and employee training—are all essential to the effective functioning of the program. Labels provide a brief, but immediate and conspicuous summary of hazard information at the site where the chemical is used. SDSs provide detailed technical information and serve as a reference source for exposed employees, industrial hygienists, safety professionals, emergency responders, health care professionals, and other interested parties. Training is designed to ensure that employees understand the chemical hazards in their workplace and are aware of protective measures to follow. Labels, SDSs, and training are complementary parts of a comprehensive hazard communication program—each element reinforces the knowledge necessary for effective protection of employees.

Information required by the HCS reduces the incidence of chemical-related illnesses and injuries by enabling employers and employees to implement protective measures in the workplace. Employers can select less hazardous chemical alternatives and ensure that appropriate engineering controls, work practices, and personal protective equipment are in place. Improved understanding of chemical hazards by supervisory personnel results in safer handling of hazardous substances, as well as proper storage and housekeeping measures.

Employees provided with information and training on chemical hazards are able to fully participate in the protective measures instituted in their workplaces. Knowledgeable employees can take the steps required to work safely with chemicals, and are able to determine what actions are necessary if an emergency occurs. Information on chronic effects of exposure to hazardous chemicals helps employees recognize signs and symptoms of chronic disease and seek early treatment. Information provided under the HCS also enables health and safety professionals to provide better services to exposed employees. Medical surveillance, exposure monitoring, and other services are enhanced by the ready availability of health and safety information.

OSHA believes that the comprehensive approach adopted in the HCS—requiring evaluation of chemicals and the transmittal of information through labels, SDSs, and training—is sound. The proposed modifications to the rule do not alter that approach. Rather, the proposed modifications to the rule are intended to improve the effectiveness of the HCS by enhancing the quality and consistency of the information provided to employers and employees. OSHA believes this can be accomplished by modifying the requirements of the standard to conform with the more specific and detailed provisions of the GHS for classification,

labeling, and SDSs. OSHA's rationale for this belief is summarized below. The evidence supporting this preliminary conclusion is presented in Section V of this preamble, and the proposed revisions to the HCS are discussed in detail in Section XV.

HCS Provisions for Classification, Labeling, and SDSs

The HCS covers a broad range of health and physical hazards. The standard is performance-oriented, providing definitions of hazards and parameters for evaluating the evidence to determine whether a chemical is considered hazardous. The evaluation is based upon evidence that is currently available, and no testing of chemicals is required.

The standard covers every type of health effect that may occur, including both acute and chronic effects. Definitions of a number of adverse health effects are provided in the standard. These definitions are indicative of the wide range of coverage, but are not exclusive. Any adverse health effect that is substantiated by a study conducted according to established scientific principles, and reporting a statistically significant outcome, is sufficient for determining that a chemical is hazardous under the rule.

Most chemicals in commerce are not present in the pure state (

i.e.,

as individual elements or compounds), but are provided as mixtures of chemicals. Evaluation of the health hazards of mixtures is based on data for the mixture as a whole when such data are available. When data on the mixture as a whole are not available, the mixture is considered to present the same health hazards as any ingredients present at a concentration of 1% or greater, or, in the case of carcinogens, concentrations of 0.1% or greater. The HCS also recognizes that risk may remain at concentrations below these cut-offs, and where there is evidence that is the case, the mixtures are considered hazardous under the standard.

The current definitions of physical hazards in the HCS were derived from other OSHA standards that address such chemicals (

e.g.,

flammable chemicals), or from the DOT criteria for physical hazards at the time OSHA promulgated the HCS. DOT subsequently changed their criteria to be consistent with the internationally harmonized transport requirements, and the HCS criteria for classification of physical hazards are generally not consistent with current DOT requirements.

The HCS establishes requirements for minimum information that must be included on labels and SDSs, but does not provide specific language to convey the information or a format in which to provide it. When the HCS was issued in 1983, the public record strongly supported this performance-oriented approach (

see

48 FR 53300-53310). Many chemical manufacturers and importers were already providing information voluntarily, and in the absence of specific requirements had developed their own formats and approaches. The record indicated that a performance-oriented approach would reduce the need for chemical manufacturers and importers to revise these existing documents to comply with the HCS, thus reducing the cost impact of the standard. In recognition of the work that had been voluntarily completed, OSHA decided to allow labels and SDSs to be presented in any format desired, as long as the minimum information requirements of the standard were met.

GHS Provisions for Classification, Labeling, and SDSs

The GHS is an internationally harmonized system for classifying chemical hazards and developing labels and safety data sheets. However, the GHS is not a model standard that can be adopted verbatim. Rather, it is a set of criteria and provisions that regulatory authorities can incorporate into existing systems, or use to develop a new system.

The GHS is designed to allow regulatory authorities to choose provisions that are appropriate to their particular sphere of regulation. This is referred to as the “building block approach.” The GHS includes all of the regulatory components, or building blocks, that might be needed for classification and labeling requirements for chemicals in the workplace, transport, pesticides, and consumer products.

Regulatory authorities such as OSHA adopt the provisions of the GHS that are appropriate for their particular regulatory sector, but do not need to adopt all of the criteria and provisions of the GHS. For example, the GHS includes criteria for classifying chemicals for aquatic toxicity. Since OSHA does not have the regulatory authority to address environmental concerns, OSHA would not adopt the GHS criteria for aquatic toxicity. The building block approach may also be applied to the criteria for defining hazards. For example, the acute toxicity criteria in the GHS are much broader than those currently found in the HCS. This is to allow consumer product authorities the ability to address the protection of children and other vulnerable populations. OSHA would not need to adopt all of the acute toxicity categories to maintain protection of employees in the workplace.

The building block approach can also be applied when a regulatory authority decides which parts of the system to adopt. For example, the GHS includes classification criteria and provisions for labels and SDSs. While OSHA is proposing to adopt all of these elements because the current HCS cover labels and SDSs, consumer product and transportation authorities are not expected to require SDSs.

Under the GHS, each hazard or endpoint (

e.g.,

Explosives, Carcinogenicity) is considered to be a hazard class. The classes are generally sub-divided into categories of hazard. The definitions of hazards are more specific and detailed than those currently in the HCS. For example, under the HCS, a chemical is either an explosive or it is not. Under the GHS, there are seven categories of explosives, and assignment to these categories is based on the classification criteria provided.

The GHS generally applies a tiered approach to evaluation of mixtures. The first step is consideration of data on the mixture as a whole. The second step allows the use of “bridging principles” to estimate the hazards of the mixture based on information about its components. The third step of the tiered approach involves use of cut-off values based on the composition of the mixture, or for acute toxicity, a formula which is used for classification. The approach is generally consistent with the current requirements of the HCS, but provides more detail and specification and allows for extrapolation of data available on the components of a mixture to a greater extent—particularly for acute effects.

Hazard communication requirements under the GHS are directly linked to the hazard classification. For each class and category of hazard, a harmonized signal word (

e.g.,

Danger), pictogram (

e.g.,

skull and crossbones), and hazard statement (

e.g.,

Fatal if Swallowed) are specified. These specified elements are referred to as the core information for a chemical. Thus, once a chemical is classified, the GHS provides the specific core information to convey to users of that chemical. The core information allocated to each category generally reflects the degree of severity of the hazard. Precautionary statements are also required on GHS labels. The GHS provides example precautionary statements, but they are not yet

considered formally harmonized. In other words, it would be possible for regulatory authorities to use different language for the precautionary statements. However, it appears likely that the language in the examples will become the harmonized text of the GHS on precautionary statements in the near future. The most recent revision to the GHS has codified these statements (

i.e.,

assigned numbers to them) as well as aligned them with the hazard classes and categories. Codification allows reference to them in a shorthand form, and makes it easier for authorities using them in regulatory text to organize them. In addition, there are provisions to allow supplementary information so that chemical manufacturers can provide data in addition to the specified core information.

The GHS establishes a standardized 16-section format for SDSs to provide a consistent sequence for presentation of information to SDS users. Items of primary interest to exposed employees and emergency responders are presented at the beginning of the document, while more technical information is presented later. Headings for the sections (

e.g.,

First Aid Measures, Handling and Storage) are standardized to facilitate locating information of interest. The harmonized data sheets are consistent with the order of information included in the voluntary industry consensus standard for safety data sheets (ANSI Z400.1).

Advantages of the Proposed Modifications to the Standard

OSHA believes that the detailed and specific classification requirements of the GHS would result in better, more consistent information being provided to employers and employees. Classification under the revised criteria would not only indicate the type of hazard, but would generally give an indication of the degree of severity of the hazard as well. This information would be helpful to both employers and employees in understanding chemical hazards and identifying and implementing protective measures. The detailed criteria for classification are also expected to result in greater accuracy in hazard classification and more consistency among classifiers. By following the detailed criteria, classifiers are less likely to reach different interpretations of the same data.

OSHA also believes that standardized presentation of information on labels and safety data sheets would improve the comprehensibility of chemical hazard information. Employers and employees would be given the same core information on a chemical regardless of the supplier. Use of standardized pictograms would complement and reinforce the information provided through signal words and hazard statements. Pictograms are also anticipated to improve communication for those who are not functionally literate, or who are not literate in the language used on the label. The standardized format for SDSs is expected to make the information easier for users to find, with the information employees and emergency responders need most appearing in the beginning of the document for easy identification and reference.

Standardized requirements for labels and SDSs are also expected to increase the accuracy of chemical hazard information. With consistent presentation of information, the task of reviewing SDSs and labels to assure accuracy would be simplified. Individuals preparing and reviewing these documents should find it easier to identify any missing elements, and OSHA enforcement personnel should be able to more efficiently examine SDSs and labels when conducting inspections.

Another advantage that will result from adopting a system that has harmonized hazard statements in it relates to the use of “control banding,” a guidance approach to recommending control measures for chemical exposures. The approach uses information that is readily available to small and medium-sized employers with chemicals in their workplaces to provide them with workplace-specific control recommendations. Basically, the system uses such information to estimate the degree of severity of the hazard and the amount of chemical present, and relates that to the degree of control needed. The control banding approach relies on harmonized hazard statements to allow the system to estimate the degree of severity of the hazard. Initially based on the European hazard classification system, it has now been converted to the GHS phrases. The use of control banding to provide guidance for chemical safety and health approaches in U.S. workplaces cannot be accomplished until harmonized hazard statements are readily available. Adoption of the GHS and its phrases would open up the possibility that control banding guidance can be used in the U.S. to help small and medium-sized employers select and implement appropriate control measures. For more information on control banding, please

see http://www.cdc.gov/niosh/topics/ctrlbanding/.

OSHA is proposing modifications to the HCS that are necessary for consistency with the GHS. The GHS does not include requirements for a written hazard communication program or for employee training. OSHA is not proposing any substantive changes to the requirements for a written hazard communication program. However, OSHA believes that additional training would be necessary to ensure that employees understand some elements of the new system. In particular, some training and familiarization would be needed for pictograms to be effective. The Agency is therefore proposing modified training requirements to address the new label elements and SDS format that would be required under the revised standard.

The GHS leaves certain matters to the competent authority (

i.e.,

the regulatory authority with jurisdiction over that sector) to determine. OSHA would maintain its current approaches in these situations. For example, the scope and application provisions in the HCS address the interface of the OSHA requirements with requirements of other agencies. These scope provisions would remain unchanged under the proposed rule.

The proposed modifications to the HCS primarily affect manufacturers and importers of hazardous chemicals. Chemical manufacturers and importers would be required to re-evaluate chemicals according to the new criteria in order to ensure they are classified appropriately. For health hazards, this will necessitate placing the chemical in the appropriate hazard category as well as the hazard class. For physical hazards, however, the new criteria are generally consistent with current DOT requirements for transport. Therefore, if the chemicals are transported (

i.e.,

they are not produced and used in the same workplace), this classification should already be done for physical hazards for purposes of complying with DOT's transport requirements. This should minimize the additional work required for classification of physical hazards. Preparation and distribution of modified labels and safety data sheets by chemical manufacturers and importers would also be required. Those chemical manufacturers and importers already following the ANSI Z400.1 standard for safety data sheets should already have the appropriate format, and would only be required to make some small modifications to the content of the sheets to be in compliance.

Compliance requirements for chemical users would be limited. Workplaces where chemicals are used would need to integrate the new

approach into their hazard communication program, assuring that employees understand the pictograms and other information provided on labels and SDSs. Employers who use chemicals, and exposed employees, would benefit from receiving labels and safety data sheets presented in a consistent format. The information should be easier to find and comprehend, allowing it to be used more effectively for the protection of employees.

Changing the HCS to make it conform to the GHS will also make it necessary to modify a number of other OSHA standards. Modifications are proposed to the standards for Flammable and Combustible Liquids in general industry (29 CFR 1910.106) and construction (29 CFR 1926.152) to align the requirements of the standards with the GHS hazard categories for flammable liquids. A modification to the Process Safety Management standard (29 CFR 1910.119) is proposed to ensure that the scope of the standard is not changed by the proposed modifications to the HCS. In addition, modifications to most of OSHA's substance-specific health standards are proposed to ensure that requirements for signs and labels are consistent with the modified HCS.

OSHA's preliminary determination to modify the HCS is based on its assessment of the potential to improve employee safety and health by adopting the GHS approach to hazard communication. However, GHS implementation is also expected to accomplish a number of other objectives, and produce additional benefits. By providing an internationally comprehensible system for hazard communication, the GHS is anticipated to enhance the protection of the environment and of human health in all sectors, not only the workplace. The GHS provides a framework for developing a hazard communication system for those countries without an existing system, thus protecting employees around the world and helping to ensure that the appropriate information is received with chemicals imported into American workplaces. Implementation of the GHS is also expected to reduce the need for testing and evaluation of chemicals, since classification would be based on existing data and would only need to be performed once for each substance. In addition, implementation of the GHS is expected to facilitate international trade in chemicals, as the need to identify and comply with diverse and complex hazard communication requirements in different countries would be reduced or eliminated.

Alternative Approaches

In this section OSHA presents several alternatives to the proposed GHS modification to the HCS to respond to concerns raised by commenters through the ANPR. OSHA provides the following discussion of these alternatives and their potential impacts and requests comments regarding their relative costs, benefits, feasibility, impact on small businesses, impact on worker safety and health, and any other issues on which commenters may wish to provide feedback.

This rulemaking seeks to improve employee protections by adopting an internationally harmonized approach to hazard communication issues. While the current HCS provides protections for exposed workers by disseminating information about chemicals in their workplaces, OSHA believes, as discussed in Section V, that the adoption of GHS strengthens and refines the system, and gives OSHA the opportunity to improve worker safety by improving hazard communications. The GHS has the same general concept of an integrated, comprehensive process of identifying and communicating hazards, but provides more extensive criteria to define the hazards in a consistent manner, as well as standardizes label elements and SDS formats to help to ensure that the information is conveyed consistently.

Additionally, the Agency believes that adoption of the GHS as proposed will simplify implementation insofar as OSHA's preferred alternative would clearly be considered “harmonized” with other regulatory authorities in the world, and thereby acquire the full benefits of harmonization.

This is in line with the GHS, which anticipates that countries will adopt the hazard classification criteria and required label elements, as well as SDS requirements in workplaces. As stated in the introduction to the GHS (3rd revision):

1.1.3.1.3 In the workplace, it is expected that all of the GHS elements will be adopted, including labels that have the harmonized core information under the GHS, and safety data sheets. It is also anticipated that this will be supplemented by employee training to help ensure effective communication.

As addressed in Section XV, many commenters supported the concept of OSHA moving forward to adopt the GHS (Document ID #s 0003, 0007, 0047, 0050, 0052, 0062, 0106, 0011, 0033, 0038, 0123, 0130, 0151, 0163, and 0171). While others objected to adoption, OSHA has identified and responded to their concerns in Section XV as well. In addition, there were several commenters who noted that small chemical manufacturers that are not in international trade of chemicals would have a large burden associated with adopting the GHS, and questionable benefits due to their lack of international trade. (Document ID # 0022). Others simply noted that they believed there would be high costs and limited benefits for such employers, or that it would be costly and difficult to adopt (

Document ID #s

0015, 0026, 0178, and 0144). There was no discussion in any of these comments about potential alternatives.

It should be noted that it appears that all of these commenters assumed the primary benefits of adopting the GHS would be in facilitating international trade. As has been addressed in Section VII, OSHA has based the benefits of this action on improved communication to workers and has provided initial estimates of a range of benefits that would be achieved in this area; trade benefits which, while recognized, have not been quantified. Therefore, grandfathering or other exemptions related to this rule might result in workers in those facilities receiving lower benefits of increased comprehensibility relative to workers in other types and sizes of workplaces; OSHA considers this a serious concern that could potentially exclude a group of workers exposed to hazardous chemicals from the increased benefits associated with clearer and more specific classification criteria, as well as standardized label elements.

Alternatives:

In order to respond to the concerns raised in these comments, OSHA solicits comment on several options:

1. The first option is designed to facilitate voluntary adoption of GHS within the existing HCS framework. Specifically, this approach would involve recognition and adoption of the GHS, with minimal changes to the current HCS. Under this approach, entities could opt to adopt GHS or continue to follow their current practice under HCS.

Therefore, companies would decide whether they would continue complying with the existing standard, or comply with the GHS. This would reduce the costs for those companies that choose to remain in compliance with the existing HCS, and allow those companies that foresee the benefits of GHS compliance from a trade perspective to adopt its provisions. Another version of this option would be to exempt small chemical producers from complying.

2. A second option that OSHA is seeking to solicit comment on would

make modifications to the current HCS in order to improve hazard communication through adoption of components of the GHS. Under this option OSHA would add requirements for standardized hazard statements, signal words, and precautionary statements being added to the current HCS, but otherwise would follow the approach outlined in Alternative 1 above.

Since the standardized labels are relatively inexpensive to implement, while reviewing classifications is more costly, this has the potential to reduce the overall cost of implementation of the revised rule.

A variation on this alternative would entail incorporation of some, but not all, of the label elements. In particular, the Agency would not adopt the precautionary statements since these are not yet considered to be “harmonized” under the GHS—they are provided for guidance and reference, but competent authorities may choose to implement other statements. The precautionary statements could be adopted later when they are harmonized under the GHS. Or, alternatively, OSHA could either allow label preparers to use whatever precautionary statements they deem appropriate or develop its own set of statements to require.

From OSHA's perspective, a key issue regarding the alternative approaches presented is that the classification criteria in the GHS are different from the hazard definitions in the current HCS. In general, as discussed in Section XV, they cover the same scope of hazard so these differences do not result in significant differences in the chemicals covered. But the GHS criteria divide most of the hazard classes into hazard categories that convey the severity of the effect, while few of the hazard classes in the current HCS take this approach. The standardized label elements are associated with these specific hazard categories,

i.e.,

the harmonized pictograms, signal words, and hazard statements are assigned by hazard category and reflect the degree of hazard it presents to those exposed. Likewise, the precautionary statements assigned are also reflective of the degree of hazard, with responses related to these presumed hazard levels.

Additionally, with regard to the first alternative, there will be chemicals that will be classified in different hazards classes under the GHS classification scheme versus the HCS hazard determination step. In addition, these chemicals will also be assigned to hazard categories under GHS where there are none now. This is particularly true for the classification of mixtures for all hazards, except the chronic health hazards, since the hazard determination scheme in the current HCS is based solely on concentration limits and the GHS classification scheme is based on bridging principles. Under the alternatives presented workers might be given different hazard information when exposed to a chemical purchased from two different suppliers. OSHA notes that this would be similar to the situation under the current performance-oriented HCS, but this approach may forego an opportunity to make the system more consistent.

OSHA is interested in comments related to the alternatives addressing the extent to which differences in classification between the GHS and HCS might create confusion or otherwise result in problems. OSHA is further interested in comments addressing the classification of mixtures under the alternatives discussed, given the differences in classification under HCS and GHS applicable to mixtures.

Given the current variability in MSDS and labels under the performance based HCS, OSHA believes that this approach might not have a negative impact on safety and health relative to our current HCS. However, the Agency anticipates that components of the GHS would confer benefits external to producers (

e.g.,

the benefits associated with clearer and more specific classification criteria, as well as labels or other changes that could potentially make easier for users to locate and understand the information they are seeking), adoption of this alternative could result in foregone benefits. In addition, a small number of chemicals or mixtures might be labeled differently due to differing categorization results between the existing HCS and GHS.

OSHA is generally seeking comment on the possible cost impacts associated with the alternatives on the chain of chemical suppliers. OSHA notes that large and small producers are not mutually exclusive so that a large business or distributers engaged in international trade cannot simply and straightforwardly choose to implement the GHS regardless of their suppliers. Small businesses sell to large businesses. If small businesses do not adopt the GHS, then the large businesses or the distributor would either have to generate GHS classifications for chemicals they buy from them or request that small businesses supply data and labels using GHS classifications. Likewise, chemical producers often provide their products to distributors who then sell them to customers unknown to the original producer. Thus knowing whether or not a product will wind up in international trade may be questionable in some situations. A producer may provide a substance to another company, who then formulates it into a product that is sold internationally—thus the original producer is involved in international trade without necessarily realizing it. In theses cases, costs would be incurred for the conversion to GHS. This issue was raised in comments regarding the effective dates for the rule, when many suggested it was not appropriate to differentiate dates based on the size of the business. For example, ORC Worldwide, Inc. stated (Document ID # 0123):

OSHA should consider a company's place in the manufacturing supply chain, not size, in determining how the phase-in is implemented. It would be sensible to start with producers of raw materials and basic chemicals. The technical information, classification and categorization they perform will be useful downstream for the intermediate chemical producers and specialty chemical manufacturers. Lastly, the end user will benefit from the influx of information developed by the upstream professionals.

OSHA solicits comment on whether a voluntary system, or a system based on business size, could be successfully implemented given the structure of the supply system.

OSHA seeks comment on how companies that use chemicals, but don't produce them, would be affected under an alternative approach. Rather than potentially simplifying compliance and improving comprehensibility, the user of chemicals would continue to see variation in labels on purchased chemicals. This would be further complicated by the fact that the underlying criteria for these labels may be different as well, and thus the warnings would be too. If there is no requirement for such employers to be familiar with the new system, and train their employees, then there will be new pictograms and signal words with no structure for ensuring they are understood and the appropriate precautions are implemented.

Regarding Alternative 2, under OSHA's proposed approach the label provisions are relatively cost-efficient to adopt given that the GHS assigns the various required elements by hazard class and category and once the classification or re-classification has been accomplished, the GHS provides the specific information for the label.

OSHA solicits comment on whether requiring this standardized approach to labeling under the HCS, without the infrastructure of the GHS will be burdensome for the chemical

manufacturer to accomplish OSHA further solicits comment on whether confusion may result from labels that may look the same but which actually reflect different classification criteria. Under this approach, chemical producers will have to assess their current determinations and attempt to relate them to the established hazard classes and categories. Alternatively, OSHA could create a regulatory system assigning HCS categories to each GHS label elements; comments are welcomed on the impact on benefits and costs, and the feasibility of such an approach. OSHA believes it is unlikely that this component of Alternative 2 would provide significant savings over reviewing classifications for purposes of putting the chemicals into GHS classes and categories.

OSHA is concerned that chemical producers following this approach might not be able to use their labels in other countries where the GHS has been adopted. OSHA is further concerned that adopting only some elements of the GHS label may be confusing and may fail to provide useful information regarding the possible hazardous effects of exposure. Delaying adoption of the precautionary statements may also reduce the effectiveness of the labels significantly, and reduce the appropriate information on the SDSs as well. A variation on this alternative—to simply require precautionary statements, but not to specify what they are, may generate significant variation due to the performance-oriented approach that allows the label preparer to determine what they are or if they are included. One communication advantage of providing the information in the same language from label-to-label is that workers and other users can be assured that the same action is required. If you take a simple preventive measure such as “wash your hands,” but convey it in several different ways, the reader of the label will think you mean something different. This is one of the advantages of providing the text for these statements in the revised HCS. In addition, since these precautionary statements will be translated, this should make it easier for those participating in international trade to produce and use labels.

Thus, OSHA solicits comment on a range of alternative approaches to regulatory adoption of GHS and welcomes comments on these options. The costs and benefits are further addressed in Section VII.

V. Need and Support for the Proposed Modifications to the Hazard Communication Standard

Chemical exposure can cause or contribute to many serious adverse health effects such as cancer, sterility, heart disease, lung damage, and burns. Some chemicals are also physical hazards and have the potential to cause fires, explosions, and other dangerous incidents. It is critically important that employees and employers are apprised of the hazards of chemicals that are used in the workplace, as well as associated protective measures. This knowledge is needed to understand the precautions necessary for safe handling and use, to recognize signs and symptoms of adverse health effects related to exposure when they do occur, and to identify appropriate measures to be taken in an emergency.

OSHA established the need for disclosure of chemical hazard information when the HCS was issued in 1983 (48 FR 53282-53284). This need continues to exist. The Agency estimates that 880,000 hazardous chemicals are currently used in the U.S., and over 40 million employees are now potentially exposed to hazardous chemicals in over 5 million workplaces.

Chemical exposures result in a substantial number of serious injuries and illnesses among exposed employees. The Bureau of Labor Statistics estimates that employees suffered 55,400 illnesses that could be attributed to chemical exposures in 2007, the latest year for which data are available (BLS, 2008). In that same year, 17,340 chemical-source injuries and illnesses involved days away from work (BLS, 2009).

The BLS data, however, do not indicate the full extent of the problem, particularly with regard to illnesses. As noted in the preamble to the HCS in 1983, BLS figures probably only reflect a small percentage of the incidents occurring in exposed employees (48 FR 53284). Many occupational illnesses are not reported because they are not recognized as being related to workplace exposures, are subject to long latency periods between exposure and the manifestation of disease, and other factors (

e.g.,

Herbert and Landrigan, 2000; Leigh

et al.,

1997; Landrigan and Markowitz, 1989).

The HCS currently serves to ensure that information concerning chemical hazards and associated protective measures is provided to employers and employees. However, OSHA's experience, along with information acquired since the HCS was issued, indicates that modifications to the standard may be appropriate. The Agency believes that the proposed changes, based on the GHS, will substantially improve the quality and consistency of the information provided to employers and employees. OSHA further believes the proposed revisions to the HCS will enhance workplace protections, because better information will enable employers and employees to take measures that would result in a reduction in the number and severity of chemical-related injuries and illnesses.

A key foundation underlying this belief relates to the comprehensibility of information conveyed under the GHS. All hazard communication systems deal with complicated scientific information being transmitted to largely non-technical audiences. During the development of the GHS, in order to construct the most effective hazard communication system, information about and experiences with existing systems were sought to help ensure that the best approaches would be used. Ensuring the comprehensibility of the GHS was a key issue during its development. As noted in a

Federal Register

notice published by the U.S. Department of State (62 FR 15956, April 3, 1997): “A major concern is to ensure that the requirements of the globally harmonized system address issues related to the comprehensibility of the information conveyed.” This concern is also reflected in the principles of harmonization that were used to guide the negotiations and discussions during the development of the GHS. As described in Section 1.1.1.6(g) of the GHS, the principles included the following: “[T]he comprehension of chemical hazard information, by the target audience,

e.g.,

workers, consumers and the general public should be addressed.”

To help in the development of the GHS, OSHA had a review of the literature conducted to identify studies on effective hazard communication, and made the review and the analysis of the studies available to other participants in the GHS process. Prepared by researchers at the University of Maryland, the document entitled “Hazard Communication: A Review of the Science Underpinning the Art of Communication for Health and Safety” (Sattler

et al.,

1997) has also long been available to the public on OSHA's Hazard Communication web page. More recently, OSHA conducted an updated review of the literature published since the 1997 review. This updated review examined the literature relevant to specific hazard communication provisions of the GHS (ERG, 2007).

Further work related to comprehensibility was conducted during the GHS negotiations by researchers in South Africa at the University of Cape Town—the result is an annex to the GHS related to comprehensibility testing (see GHS Annex 6, Comprehensibility Testing Methodology) (United Nations, 2009). Such testing has been conducted in some of the developing countries preparing to implement the GHS, and has provided these countries with information about which areas in the GHS will require more training in their programs to ensure people understand the information. The primary purpose of these activities was to ensure that the system developed was designed in such a way that the messages would be effectively conveyed to the target audiences, with the knowledge that the system would be implemented internationally in different cultures with varying interests and concerns.

Also among the agreed principles that were established to guide development of the GHS was that the level of protection offered by an existing hazard communication system should not be reduced. Following these principles, the best aspects of existing systems were identified and included in a single, harmonized approach to classification, labeling, and development of SDSs.

The GHS was developed by a large group of experts representing a variety of perspectives. Over 200 experts provided technical input on the project. The United Nations Sub-Committee of Experts on the GHS, the body that formally adopted the GHS and is now responsible for its maintenance, includes 32 member nations as well as 17 observer nations. Authorities from these member states are able to convey the insight and understanding acquired by regulatory authorities in different sectors, and to relate their own experiences in implementation of hazard communication requirements. In addition, over two dozen international and intergovernmental organizations, trade associations, and unions are represented, and their expertise serves to inform the member nations. The GHS consequently represents a consensus recommendation of experts with regard to best practices for effective chemical hazard communication, reflecting the collective knowledge and experience of regulatory authorities in many nations and in different regulatory sectors, as well as other organizations that have expertise in this area. A number of United States-based scientific and professional associations have endorsed adoption of the GHS. The American Chemical Society indicated its support for the GHS, stating: “The American Chemical Society (ACS) strongly supports the adoption of the GHS for hazard communication in general and specifically as outlined in the ANPR” adding that “* * * ACS anticipates that OSHA implementation of GHS in the U.S. will enhance protection of human health and the environment through warnings and precautionary language that are consistent across different products and materials as well as across all workplaces” (Document ID #0165). In comments submitted in response to the ANPR, the American Industrial Hygiene Association (AIHA) affirmed its support for modification of the HCS to adopt the GHS. AIHA maintained that standardized labels and safety data sheets will make hazard information easier to use, thereby improving protection of employees (Document ID #0034). The American Society of Safety Engineers also indicated its support for the GHS rulemaking (Document ID #0139). While acknowledging that the GHS presents a number of concerns and challenges, the Society of Toxicology has also expressed its support for the GHS, stating that “a globally harmonized system for the classification of chemicals is an important step toward creating consistent communications about the hazards of chemicals used around the world” (SOT, 2007). The American Association of Occupational Health Nurses joined these organizations in advocating adoption of the GHS, arguing that standardization of chemical hazard information is critical to protecting the safety and health of employees (Document ID #0099). The positions taken by these organizations point to wide support for the GHS among the scientific and professional communities.

In addition to the endorsement of the GHS by a group of experts with extensive knowledge and experience in chemical hazard communication and support from scientific and professional associations with expertise in this area, a substantial body of evidence indicates that the proposed modifications to the HCS will better protect employees. Specifically, this evidence supports OSHA's belief that: (1) Standardized label elements—signal words, pictograms, hazard statements and precautionary statements—would be more effective in communicating hazard information; (2) standardized headings and a consistent order of information would improve the utility of SDSs; and (3) training would support and enhance the effectiveness of the new label and SDS requirements.

This evidence was obtained from a number of sources. OSHA has commissioned several studies to examine the quality of information on SDSs (Karstadt, 1988; Kearney/Centaur 1991a, 1991b; Lexington Group, 1999); the General Accounting Office (GAO) has issued two reports based on its evaluation of certain aspects of the HCS (GAO 1991, 1992); a National Advisory Committee on Occupational Safety and Health (NACOSH) workgroup conducted a review of hazard communication and published a report of its findings (NACOSH, 1996); and a substantial amount of scientific literature relating to hazard communication has been published. As mentioned previously, OSHA commissioned a review of the literature, and a report based on that review was published in 1997 (Sattler

et al.,

1997). An updated review was published in 2007 (ERG, 2007). In addition, OSHA conducted a review of the requirements of the HCS and published its findings in March of 2004 (OSHA, 2004). Key findings derived from these sources are discussed below.

OSHA's rationale for adopting the GHS is tied to anticipated improvements in the quality and consistency of the information that would be provided to employers and employees. Hazard classification is the foundation for development of this improved information. Indeed, hazard classification is the procedure of identifying and evaluating available scientific evidence in order to determine if a chemical is hazardous, and the degree of hazard, pursuant to the criteria for health and physical hazards set forth in the standard. Hazard classification provides the basis for the hazard information that is provided in labels, SDSs, and employee training. As such, it is critically important that classification be performed accurately and consistently.

The GHS provides detailed scientific criteria to direct the evaluation process. The specificity and detail provided help ensure that different evaluators would reach the same conclusions when evaluating the same chemical. Moreover, the GHS refines that classification process by establishing categories of hazard within most hazard classes. These categories indicate the relative degree of hazard, and thereby provide a basis for determining precise hazard information that is tailored to the level of hazard posed by the chemical. The classification criteria established in the GHS thus provide the necessary basis for development of the specific, detailed hazard information that would enhance the protection of employees.

Labels

Labels provide a brief, conspicuous hazard summary at the work site where a chemical is used. Labels serve as an immediate visual reminder of chemical hazards, and complement the information presented in training and on SDSs.

The HCS currently requires that labels on hazardous chemical containers include the identity of the hazardous chemical; appropriate hazard warnings that convey the specific physical and health hazards, including target organ effects; and the name and address of the chemical manufacturer, importer, or other responsible party. The HCS does not specify a standard format or design elements for labels.

OSHA is proposing a requirement that labels include four new, standardized elements:

a signal word; hazard statement(s); pictogram(s); and precautionary statement(s) (see Section XV for a detailed discussion of the proposed requirements). The appropriate label elements for a chemical would be determined by the hazard classification. OSHA believes that these standardized label elements would better convey critically important hazard warnings, and provide useful information regarding precautionary measures that would serve to better protect employees.

A great deal of literature has been developed that examines the effectiveness of warnings on labels. However, some important limitations must be recognized in applying this information to workplace labels for hazardous chemical products. Most studies have examined labels for prescription and non-prescription medications, alcoholic beverages, or consumer products. Relatively few studies pertain specifically to labels for hazardous chemicals in the workplace. Much of the literature is also characterized by the use of research subjects such as college students or consumers. Such subjects may not be representative of workplace populations, as these subjects may differ from typical employees in terms of product knowledge, hazard perception, perceptual abilities, and safety motivation. In addition, some studies involve non-U.S. populations that may not be representative of the U.S. workforce.

Nevertheless, the literature provides a substantial body of information applicable to workplace chemical labels. In spite of the differences in affected populations, workplace chemical labels have many characteristics that are comparable to those found in other sectors. Pharmaceutical labels, for example, are similar to chemical labels in that they often have explicit instructions for use which, if not followed, can cause adverse health effects or death. Designers of pharmaceutical labels also encounter many of the same challenges faced by those who design chemical labels, such as container space limitations and the need to convey information to low-literate or non-English literate users. In addition, some of the research is not directly related to any particular sector or type of product. Some findings related to use of color, for example, could reasonably be applied to a wide variety of label applications. Relevant finding from the literature are presented in the sections that follow.

Signal Words

A signal word is a word that typically appears near the top of a warning, sometimes in all capital letters. Common examples include DANGER, WARNING, CAUTION, and NOTICE. The signal word is generally understood to serve a dual purpose: alerting the user to a hazard and indicating a particular level of hazard. For example, users generally perceive the word DEADLY to indicate a far greater degree of hazard than a term like NOTICE.

The proposal prescribes one of two signal words for labels—DANGER or WARNING—depending on the hazard classification of the substance in question. These are the same two signal words used in the GHS. DANGER is used for the more severe hazard categories, while WARNING denotes a less serious hazard. These signal words are similar to those in other established hazard communication systems, except that some other systems have three or more tiers. For example, ANSI Z129.1 (the American National Standard for Hazardous Industrial Chemicals—Precautionary Labeling) uses DANGER, WARNING, and CAUTION, in order of descending severity (ANSI, 2006).

A number of recent studies have examined how people perceive signal words and, in particular, how they perceive signal words to be different from one another. Overall, this research supports the use of signal words in labels, demonstrating that they can attract attention and help people clearly distinguish between levels of hazard. The research also supports the decision to use only two tiers, as many recent studies have found clear differences between DANGER and WARNING but little perceived difference between WARNING and CAUTION.

Wogalter

et al.

investigated the influence of signal words on perceptions of hazard for consumer products (Wogalter

et al.,

1992). Under the pretext of a marketing research study, 90 high school and college students rated product labels on variables such as product familiarity, frequency of use, and perceived hazard. Results showed that the presence of a signal word increased perceived hazard compared to its absence. Between extreme terms (

e.g.,

NOTE and DANGER), significant differences were noted.

Seeking to test warning signs in realistic settings, Adams

et al.

tested five industrial warning signs on a group of 40 blue-collar workers employed in heavy industry, as well as a group of students (Adams

et al.,

1998). Signs were manipulated to include four key elements (signal word, hazard statement, consequences statement, and instructions statement) or a subset of those elements. Participants were asked questions to gauge their reaction and behavioral intentions. Overall, 77 percent (66 percent of the worker group) recognized DANGER as the key word when it appeared, and more than 80 percent recognized BEWARE and CAUTION, suggesting that the signal word was generally noticed, and it was recognized as the key alerting element. DANGER was significantly more likely than other words to influence behavioral intentions.

Laughery

et al.

also demonstrated the usefulness of signal words. The authors tested the warnings on alcoholic beverage containers in the U.S., and found that a signal word (WARNING) was one of several factors that decreased the amount of time it took for participants to locate the warning. (Laughery

et al.,

1993).

Several studies have tested the arousal strength or perceived hazard of different signal words.

Arousal strength

is a term used to indicate the overall importance of the warning, and incorporates both the likelihood and severity of the potential threat. Silver and Wogalter tested the arousal strength of signal words on college students and found that DANGER connoted greater strength than WARNING and CAUTION (Silver and Wogalter, 1993). The results failed to show a difference between WARNING and CAUTION. Among other words tested, DEADLY was seen as having the strongest arousal connotation, and NOTE the least.

Griffith and Leonard asked 80 female undergraduates (who were unlikely to have already received industrial safety training) to rate signal words. Results included a list of terms in order of “meaningfulness,” representing conceptual “distance” from the neutral

term NOTICE (Griffith and Leonard, 1997). From most to least meaningful, these terms were reported to be DANGER, URGENT, BEWARE, WARNING, STOP, CAUTION, and IMPORTANT.

Wogalter

et al.

asked over 100 undergraduates and community volunteers to rank signal words (Wogalter

et al.,

1998). DEADLY was perceived as most hazardous, followed by DANGER, WARNING, and CAUTION. All differences were statistically significant. In a follow-up experiment using labels produced in the ANSI Z535.2 (American National Standard for Environmental and Facility Safety Signs), ANSI Z535.4 (American National Standard for Product Safety Signs and Labels), and alternative formats, the authors found a similar rank order for signal words with all labeling systems. Finally, the authors tested the same terms on employees from manufacturing and assembly plants and found the same general order: DEADLY, then DANGER, then WARNING and CAUTION with no significant difference between the last two terms.

In more of a free-form experiment, Young asked 30 subjects to produce warning signs for a set of scenarios, using different sign components available on a computer screen (Young, 1998). In roughly 80 percent of the signs, the participant chose to use a signal word. DANGER, DEADLY, and LETHAL were more likely to be used for scenarios with severe hazards; CAUTION and NOTICE for non-severe scenarios. WARNING was used equally in both types of scenarios. The author suggests that these results support a two-tiered system of signal words. In a separate task, users ranked the perceived hazard of signal words, resulting in the following list from most to least severe: DEADLY, LETHAL, DANGER, WARNING, CAUTION, and NOTICE.

While these studies have focused on the relative perceptions of signal words, others have sought to evaluate how the absolute meaning of common signal words is perceived. Drake

et al.

asked a group of students and community volunteers to match signal words with definitions borrowed from consensus standards and other sources (Drake

et al.,

1998). Participants matched DANGER to a correct definition 64 percent of the time, while NOTICE was matched correctly 68 percent of the time. WARNING and CAUTION were matched correctly less than half of the time, suggesting confusion. The authors recommended using WARNING and CAUTION interchangeably. The authors also suggested that a standard set of signal words (but not synonyms) is helpful for users with limited English skills, who can be trained to recognize a few key words.

Signal word perceptions are reported to be consistent among some non-U.S. populations, as well. Hellier

et al.

asked 984 adults in the UK to rate DANGER, WARNING, and CAUTION on a hazard scale from 1 (low) to 10 (high) (Hellier

et al.,

2000a). DANGER was ranked as 8.5, WARNING was ranked as 7.8, while CAUTION was rated as 7.25. These results are consistent with the findings of studies on subjects in the U.S. In a second study published in 2000, Hellier

et al.

asked a mixed-age group of participants in the UK to rate the arousal strength of 84 signal words commonly used in the U.S. (Hellier

et al.,

2000b).The authors found that DANGER is stronger than WARNING, while WARNING and CAUTION are not significantly different from each other.

Similar results were found among workers in Zambia. Banda and Sichilongo tested GHS-style labels using four different signal words (as well as other variables) (Banda and Sichilongo, 2006). Among workers in the industrial and transport sectors, DANGER was generally perceived as the most hazardous signal word. WARNING was one of a group of terms that were largely indistinguishable from one another, but distinct from DANGER. The authors support adoption of the GHS, suggesting that having just two possible signal words will lead to “more impact and less confusion about the extent of hazard.”

In addition, comparable results were found in South Africa (London, 2003). In a large study on SDS and label comprehensibility conducted for South Africa's National Economic Development and Labour Council (NEDLAC), DANGER was generally ranked as more hazardous than WARNING by participants in the four sectors tested: industry, transport, agriculture, and consumers.

Cumulatively, these studies provide a clear indication that signal words are effective in alerting readers that a hazard exists, and in conveying the existence of a particular level of hazard. The studies have found a generally consistent hierarchy of signal words with respect to perceived hazard. DANGER and WARNING appear to connote different levels of hazard, while the perceived difference between WARNING and CAUTION is often insignificant.

Pictograms

A pictogram is a graphical composition that may include a symbol along with other graphical elements, such as a border or background color. A pictogram is a communication tool and is intended to convey specific information.

The proposed rule includes requirements for use of eight different pictograms. Each of these pictograms consists of a different symbol in black on a white background within a red square frame set on a point (

i.e.,

a red diamond). The specific pictograms that are required on a label would be determined based on the hazard classification of the substance in question.

OSHA believes that the proposed pictograms would make warnings on labels more noticeable and easier for employees to understand. In particular, symbols are expected to improve comprehension among people with low literacy and those who are not literate in the English language. It should be remembered that pictograms would be used not only in conjunction with other label elements, but in the context of the hazard communication program as a whole. Training that includes an explanation of labels (included in the proposed rule) would ensure that pictograms are understood by employees.

A considerable amount of evidence supports the belief that pictograms can serve as useful and effective communication tools. In reviewing this evidence, it should be noted that some sources offer distinct definitions for “pictogram,” “pictorial,” “symbol,” and other terms describing graphical elements. For example, Rogers

et al.

state that: “Pictorials refer to pictures that represent the concept of interest (

e.g.,

a picture of a fire extinguisher). Symbols are more abstract representations of a concept, the meaning of which must be learned (

e.g.,

the use of a skull and crossbones to denote poison)” (Rogers

et al.,

2000). ANSI and others combine these terms in the definition of “symbol,” however, and for the purposes of discussing the literature on this subject, these terms are used interchangeably.

Symbols serve several important functions in warning labels. As Wogalter

et al.

explain, symbols may alert the user to a hazard more effectively than text alone:

Symbols may be more salient than text because of visual differentiations of shape, size, and color. Usually symbols have unique details and possess more differences in appearance than do the letters of the

alphabet. Letters are highly familiar and are more similar to one another than most graphical symbols (Wogalter

et al.,

2006).

Symbols also can bolster a text message and improve label comprehension among individuals with low literacy, and those who do not understand the language in which the label text is written (Parsons

et al.,

1999).

Several researchers have sought to evaluate how people comprehend symbols, including those symbols that are incorporated in the proposed rule. Some studies have found that the skull and crossbones icon—one of the symbols included in the proposed rule—is among the most recognizable safety symbols. For example, Wogalter

et al.

asked 112 undergraduates and community volunteers to rank various label elements (Wogalter

et al.,

1998). Among shapes and icons, the skull symbol (in this case, without the crossbones) was rated most hazardous and most noticeable. The skull connoted the greatest hazard among industrial employees as well. Smith-Jackson and Wogalter asked 48 English-speaking workers to rate the perceived hazards of six alerting symbols (Smith-Jackson and Wogalter, 2000). The skull was rated significantly higher than all other symbols.

Some research has examined other pictograms included in the proposed rule. As part of an experiment to see how individuals comprehend warnings on household chemical labels, Akerboom and Trommelen asked 60 university students whether they understood the meaning of several pictograms, including four that are included in the proposed rule (Akerboom and Trommelen, 1998). The authors reported the following levels of comprehension for these pictograms:

• Flame: 93 percent comprehension;

• Skull and crossbones: 85 percent comprehension;

• Corrosion: 20 percent comprehension; and

• Flame over circle: 13 percent comprehension.

Only the flame and skull and crossbones pictograms met the 85 percent comprehension criteria suggested by ANSI Z535.3 (the American National Standard Criteria for Safety Symbols) (ANSI, 2002a). The authors recommend that labels present the hazard phrase [statement] and symbol together, along with corresponding precautions, as would be required under the proposed rule.

Banda and Sichilongo tested comprehension of labels that included the proposed pictograms among 364 workers in four sectors in Zambia (transport, agriculture, industrial, and household consumers) (Banda and Sichilongo, 2006). Within this population, the skull and crossbones symbol was widely understood, as was the “flame” symbol. Based on these results, the authors suggest a preference for symbols that depict familiar, meaningful, and recognizable images.

London performed a similar study among the same four sectors in South Africa, finding that the skull and crossbones was understood by at least 96 percent of each sector and “flame” by at least 89 percent (London, 2003). “Exploding bomb” was correctly comprehended by 44 to 71 percent of each sector. Many health-related symbols did not fare well, and six symbols had less than 50 percent comprehension across all four sectors. Outside the transport sector, “Gas cylinder” was the least well comprehended symbol.

These findings indicate that some of the pictograms included in the proposed rule are already widely recognized by a general audience. Others, however, are not commonly understood. Therefore, simply adding some of the proposed pictograms on labels will not provide useful information unless efforts are also undertaken to ensure that employees understand the meaning of the pictograms. As Wogalter

et al.

noted, some studies have found slower processing, poorer recognition, and greater learning difficulties with symbols versus with text—particularly if the symbols are complex or non-intuitive (Wogalter

et al.,

2006). These results emphasize the need to train employees on the meaning of the pictograms that would be included on chemical labels.

Where pictograms are used and understood, communication of hazards can be improved. Houts

et al.

studied long-term recall of spoken medical instructions when accompanied by a handout with pictograms (Houts

et al.,

2001). Nearly 200 pictograms were tested with 21 low-literate adults (less than grade 5 reading level). Immediately after training, participants recalled the meaning of 85 percent of the pictograms, and they recalled 71 percent after 4 weeks. This study found that recall was better for simple pictograms where there is a direct relationship between the image and its meaning—that is, where no inference is required.

Another body of literature focuses on the utility of symbols in general. Ganier found that people generally construct mental representations faster with pictures than they do with text, supporting earlier findings on the usefulness of symbols (Ganier, 2001). Evans

et al.

found similar results with a task in which undergraduates were asked to sort items into categories using either text clues, visual clues, or a combination of pictures and text (Evans

et al.,

2002). When categories were fixed (

i.e.,

sorting instructions were specific), people sorted the cards more consistently with one another when presented with pictures than when presented with text alone.

In a follow-up article on the South African study mentioned previously, Dowse and Ehlers found that patients receiving antibiotics adhered to instructions much better when the instructions included pictograms (54 percent with high adherence, versus 2 percent when given text-only instructions) (Dowse and Ehlers, 2005).

Pictograms also serve to attract attention to the hazard warnings on a label. To examine factors that influence the effectiveness of pharmaceutical labels, Kalsher

et al.

asked subjects to rate the noticeability, ease of reading, and overall appeal of labels with or without pictorials (Kalsher

et al.,

1996). A group of 84 undergraduates gave consistently higher ratings to labels with pictorials. A group of elderly subjects had similar preferences, rating labels with pictorials as significantly more noticeable and likely to be read.

Laughery

et al.

found similar results with a timed test on alcoholic beverage labels (Laughery

et al.,

1993). When a pictorial was present to the left of the warning showing what not to do when drinking, the amount of time it took to find the label was significantly reduced. An icon consisting of the alert symbol (an exclamation mark set within a triangle) and the signal word WARNING also decreased response time. The fastest response time came when four different enhancements (including the pictorial and the icon) were included. In a follow-up exercise, an eye scan test found that the pictorial had a particularly strong influence on reaction time, compared with other enhancements.

As far as chemical labels are concerned, London found that symbols tend to be the most easily recalled label elements (London, 2003). In the comprehensibility test of labels among South African workers mentioned previously, symbols were the most commonly recalled elements—particularly the skull and crossbones—and people recalled looking at symbols first. Symbols were also cited as by far the most important factor in determining hazard perception. Overall, the author concludes that “Symbols are therefore key to attracting attention, and informing risk perception regarding a chemical.”

Wogalter

et al.

found less encouraging evidence on pictorials, however (Wogalter

et al.,

1993). The authors tested the influence of various warning variables on whether subjects wore proper protective equipment during a task involving measuring and mixing chemicals. Warning location and the amount of clutter around the warning had significant effects on compliance, but the presence or absence of pictorials did not.

Meingast asked subjects to recall warning content after viewing labels that were considered either high quality (with color signal icons, pictorials, and organized text conforming to ANSI Z535.4, the American National Standard for Product Safety Signs and Labels) or low quality (text only) (Meingast, 2001). Pictorials were the items remembered most often, accounting for 48 percent of what viewers of high quality labels recalled. The author suggests that these pictorials also served the role of dual coding, meaning that they help to improve the retention of corresponding text.

Other recent studies support this dual-coding function of pictorials, finding that symbols tend to be most effective when paired with redundant or reinforcing text. For example, Sojourner and Wogalter asked 35 participants to rate several prescription label formats in terms of ease of reading, ease of understanding, overall effectiveness, likelihood of reading, overall preference, pictorial understanding, and how helpful pictorials are in helping to remember the instructions (Sojourner and Wogalter, 1997). The authors found that people prefer fully redundant text and pictorials, which they judged easiest to read, most effective, and preferred overall. Dual-coded pictorials aided understanding and memory more than labels with pictorials only (no text). In a follow-up study, Sojourner and Wogalter gave undergraduates, young adults, and older adults a free recall test after viewing medication labels (Sojourner and Wogalter, 1998). Fully redundant text and pictorials led to significantly greater recall than other formats, and were rated most effective by all age groups.

Similarly, Sansgiry

et al.

found that pictograms on over-the-counter drug labels improved comprehension, but only when they were congruent with the corresponding text (Sansgiry

et al.,

1997). A group of 96 adults were less confused, more satisfied, more certain about their knowledge, and understood more when shown labels that contained congruent pictures and verbal instructions, versus verbal instructions alone. The results were significantly better with congruent pictures and text than with either pictures alone or incongruent pictures and text.

Some evidence links use of pictograms directly to safer behavior. Jaynes and Boles investigated whether different warning designs, specifically those with symbols, affect compliance rates (Jaynes and Boles, 1993). Five conditions were tested: a verbal warning, a pictograph warning with a circle enclosing each graphic, a pictograph warning with a triangle on its vertex enclosing each graphic, a warning with both words and pictographs, and a control (no warning). Participants performed a chemistry laboratory task using a set of instructions that contained one of the five conditions. The warnings instructed them to wear safety goggles, mask and gloves. All four warning conditions had significantly greater compliance than the no-warning condition. A significant effect was also found for the “presence of pictographs” variable, suggesting that the addition of pictographs will increase compliance rates.

In addition to the evidence pertaining to the other graphical elements in pictograms, research indicates that the use of the color red in pictograms will serve to make warnings more noticeable. Red is also generally perceived to reflect the greatest degree of hazard, and is thus well-suited to identifying serious chemical hazards in the workplace.

In their review of the literature on warning effectiveness on behavioral compliance, Kalsher and Williams summarize several studies that examined the effects of adding color to warnings (Kalsher and Williams, 2006). Overall, Kalsher and Williams suggest that adding color can influence both the noticeability and effectiveness of warnings.

In a test on the noticeability of warnings, Swindell measured the amount of time it took subjects to locate warning text that had been embedded in medication instructions (Swindell, 1999). Warnings were found significantly faster when the icon and signal word were presented in either red or blue, causing the warning to stand out from the black text. Swindell's findings echo the results reported by Laughery

et al.,

who found that alcoholic beverage labels were located significantly faster when the text was red instead of black (Laughery

et al.,

1993). While these studies involve color on label elements other than the pictogram border, they provide a general indication that color attracts the attention of label users.

A number of researchers have investigated the hazard connotations of different colors. These investigations indicate that red is generally perceived to reflect the greatest degree of hazard. Yellow, orange, and black reflect a lesser degree of hazard. In a review of the literature, Parsons

et al.

suggest that the red-orange-yellow hierarchy generally matches people's perceptions of risk, including perceptions among native Spanish speakers (Parsons

et al.

, 1999). Experimental results that support the conclusion that red generally connotes the highest degree of hazard include:

■ Smith-Jackson and Wogalter asked English-speaking community members to rate the perceived hazard of ten ANSI safety colors (Smith-Jackson and Wogalter, 2000). Red, yellow, black, and orange were rated the highest (in descending order). Differences were statistically significant except the difference between yellow and black.

■ Among 80 college students asked to rate colors by Griffith and Leonard, red was rated the most “meaningful” color (

i.e.,

most distinct in meaning from neutral gray), followed by green, orange, black, white, blue, and yellow (Griffith and Leonard, 1997).

■ Wogalter

et al.

asked Spanish speakers to rank the perceived hazard of ANSI safety colors (Wogalter

et al.,

1997b). Red was ranked highest, followed by orange, black, and yellow.

■ Dunlap

et al.

surveyed 1169 subjects across several different language groups including English, German, and Spanish speakers (Dunlap

et al.,

1986). Subjects rated the color words red, orange, yellow, blue, green, and white according to the level of perceived hazard. The results demonstrated that the hazard information communicated by different colors followed a consistent pattern across language groups, with red having the highest hazard ratings.

■ Wogalter

et al.

asked undergraduates and community volunteers to rank various warning components (Wogalter

et al.

, 1998). Red connoted a significantly greater hazard than other colors, followed by yellow, orange, and black (in that order). A group of industrial workers ranked the colors from greatest to least hazard as follows: red, yellow, black, orange.

■ London asked workers in four sectors in South Africa to rank the colors red, yellow, green, and blue in terns of perceived hazard; 95 percent said red represents the greatest hazard, and 58 percent said yellow is the second greatest hazard (London, 2003).

■ Banda and Sichilongo asked workers in Zambia to rate the perceived hazard of various colors used in chemical labels (Banda and Sichilongo,

2006). Red was associated with the greatest hazard, followed by yellow.

■ Among a sample of 30 undergraduates who rated the perceived hazard of 105 signal word/color combinations, Braun

et al.

reported that red conveyed the highest level of perceived hazard followed by orange, black, green, and blue (Braun

et al.,

1994).

These reports are consistent in indicating that red is commonly understood to be associated with a high level of hazard—the highest of any color. OSHA anticipates that by using the color red on labels for hazardous chemicals, labels will be more effective in communicating hazards to employees—both by drawing the attention of employees and indicating the presence of a hazard through non-verbal means.

Hazard and Precautionary Statements

Hazard statements describe the hazards associated with a chemical. Precautionary statements describe recommended measures that should be taken to protect against hazardous exposures, or improper storage or handling of a chemical. The HCS currently includes a performance-oriented requirement for “appropriate hazard warnings” on labels. The proposed rule would require specific hazard statements and precautionary statements on labels. The statements would be determined based on the hazard classification of the chemical.

Standardized requirements for hazard and precautionary statements would provide a degree of consistency that is currently lacking among chemical labels. This lack of consistency makes it difficult in some instances for users to understand the nature and degree of hazard associated with a chemical, and to compare chemical hazards. For example, Beach relates experiences from the perspective of a doctor treating occupationally exposed patients (Beach, 2002). The author noted that different suppliers use different risk phrases for the same chemical, making it difficult for users to compare relative risks.

ANSI standard Z129.1 was developed to provide a consistent approach to labeling of hazardous chemicals. This standard gives manufacturers and importers guidance on how to provide information on a label, including standardized phrases and other information that can improve the quality of labels. Because it is a voluntary standard, however, the ANSI approach has not been adopted by all chemical manufacturers and importers. As a result of the diverse formats and language used, consistent and understandable presentation of information has not been fully achieved.

A preference for hazard statements was shown in EPA's Consumer Labeling Initiative (Abt Associates, 1999). This study asked consumers about their attitudes toward labels on household chemical products. Overall, consumers indicated that they like to have information that clearly connects consequences with actions, and they prefer to know why they are being instructed to take a particular precaution. A clear hazard statement can provide this information.

In some cases, clear and concise precautionary information is necessary to enable employees to identify appropriate protective measures. For example, Frantz

et al.

examined the impact of flame and poison warning symbols prescribed in certain regulations by the Canadian government (Frantz

et al.,

1994). The results suggest that although the generic meanings of these two symbols are well understood, people may have difficulty inferring the specific safety precautions necessary for a particular product.

Other reports have indicated that users prefer information that includes both an indication of the hazard and the recommended action (

i.e.

, the precautionary statement). Braun

et al.

examined statements in product instructions for a pool treatment chemical and a polyvinyl chloride (PVC) adhesive, asking subjects to rate the injury risk posed by each product (Braun

et al.

, 1995). The experimenters manipulated the instructions to include either recommended actions only, actions followed by consequences, consequences followed by actions, or a simple restatement of the product label. The authors found that actions paired with consequences led to significantly higher risk perception than a restatement of the label or actions alone. Although the preferred wording was longer than the alternatives, subjects did not feel that the instructions were too complex, suggesting that they appreciate having actions and consequences paired together. Freeman echoed these findings in a discussion on communicating health risks to fishermen and farmers, noting that to be useful, risk statements should be balanced with equally strong statements of ways to reduce or avoid the risk (Freeman, 2001).

Explicit precautionary statements may make it more likely that employees will take appropriate precautions. Bowles

et al.

asked subjects to review product warnings, then either decide what actions they should take or evaluate whether someone else's actions were safe, based on the warning (Bowles

et al.

, 2002). In general, situations that required the user to make inferences about a hazard—particularly when they had to come up with their own ideas for protective actions—led to decreased intent to comply. By providing clear precautionary instructions on the label, the proposed rule would eliminate the need for users to infer protective actions.

Some evidence indicates that using key label elements together can improve warning performance, compared with labels that only contain a subset of these elements. This is the approach taken in the proposed rule, which would require the signal word, pictogram(s), hazard statement(s), and precautionary statement(s) together on the label. In one study, Meingast asked students to recall information from two variations of warning labels: enhanced warnings with color, signal icons, pictorials, and organized text (following the ANSI Z535.4 standard); and warnings with text only (Meingast, 2001). The authors reported that the enhanced warnings were more noticeable, led to significantly greater recall, and made people report a higher likelihood of compliance.

Other findings agree that improving all label elements can improve warning performance. For example, Lehto tested information retrieval from three chemical label formats and found that subjects generally did best with an “extensive” format that included pictograms, paragraphs, and horizontal bars indicating the degree of hazard (Lehto, 1998). Subjects were able to answer more questions correctly when the label included a range of content—particularly information on first aid and spill procedures.

Wogalter

et al.

reported similar results in a test of four different signs that discouraged people from using an elevator for short trips (Wogalter

et al.

, 1997a). Three signs were text-only. The fourth sign had a signal word panel, icons, a pictorial, and more explicit wording indicating the desired behavior (

i.e.

, “use the stairs”). Subjects rated the enhanced sign as more understandable, and a field test found that it significantly increased compliance over the other options.

The effectiveness of a combination of elements was also investigated in a study of warnings on alcoholic beverage containers (Laughery

et al.

, 1993). Laughery

et al.

tested warnings to determine which elements influenced noticeability. The authors manipulated labels by adding a pictorial, adding an alert symbol with a signal word, making the text red, and/or adding a border around the warning. The warning was

located fastest when all four of these modifications were present, suggesting that the best designs include a combination of enhancements.

These findings support the belief that the proposed label elements, in combination, would likely be more effective in communicating hazard information than the individual elements would be if presented alone. Although the warnings examined in these studies are different than those included in the proposed rule, they indicate that enhancements such as color and symbols can increase the effectiveness of a label, and that presenting hazard information and corresponding precautions together may improve understanding. OSHA therefore believes that this evidence substantiates its belief that the proposed labeling requirements will result in more effective transmittal of information to employees.

Overall, the presentation of information on labels through standardized signal words, hazard statements, pictograms, and precautionary statements would provide clearer, more consistent, and more complete information to chemical users. Comments received in response to the ANPR support this view (

e.g.

, Document ID #s 0054, 0032, 0124, 0124, and 0158). For example, the Refractory Ceramic Fibers Coalition (Document ID #0030) pointed to the benefits of this approach, stating:

Employers and employees would be given the same information on a chemical regardless of the supplier. This consistency should improve communication of the hazards. It may also improve communication for those who are not functionally literate, or who are not literate in the language written on the label. In addition, having the core information developed already, translated into multiple languages, and readily available to whomever wishes to access it, should eliminate the burden on manufacturers and users to develop and maintain their own such systems. Thus the specification approach should be beneficial both to the producers and the users of chemicals.

Labels are intended to provide an immediate visual reminder of chemical hazards. Whereas labels currently may be presented in a variety of formats using inconsistent terminology and visual elements, labels prepared in accordance with the proposed requirements would be consistent. Standardized signal words and hazard statements would attract attention and communicate the degree of hazard. Pictograms would reinforce the message presented in text and enhance communication for low-literacy populations. Precautionary statements would provide useful instructions for protecting against chemical-source injuries and illnesses.

Safety Data Sheets

The HCS requires chemical manufacturers and importers to develop an SDS for each hazardous chemical they produce or import. SDSs serve as a source of detailed information on chemical hazards and protective measures. Each SDS must indicate the identity of the chemical used on the label; the chemical and common name(s) of hazardous ingredients; physical and chemical characteristics; physical and health hazards; the primary route(s) of entry; exposure limits; generally applicable precautions for safe handling and use; generally applicable control measures; emergency and first aid procedures; the date of preparation of the SDS; and the name, address and telephone number of the party preparing or distributing the SDS. The HCS does not require this information to be presented in any particular order or to follow a specific format.

Since the HCS was adopted in 1983, access to chemical information has improved dramatically due to the availability of SDSs. While the effectiveness of SDSs is evident, there are concerns regarding the quality of information provided. In particular, concerns have been raised regarding the accuracy (

i.e.

, the correctness and completeness of the information provided) and comprehensibility (

i.e.

, the ability of users to understand the information presented) of information provided on SDSs.

OSHA is proposing a requirement that the information on SDSs be presented using consistent headings in the sequence specified in the GHS (see Section XV for a detailed discussion of the proposed requirements). The Agency believes that a standardized order of information would improve the utility of SDSs by making it easier for users to locate and understand the information they are seeking. A standardized format would also be expected to improve the accuracy of the information presented on SDSs.

A number of studies have demonstrated the benefits provided by SDSs. In May 1992, the General Accounting Office (GAO) issued a report presenting the findings of an examination of difficulties small employers were said to experience in complying with the HCS, as well as issues relating to the costs of compliance (GAO, 1992). The findings were based on the results of a national survey of construction, manufacturing, and personal services providers. A total of 1,120 responses were received from employers.

One very important finding of the GAO survey was that almost 30% of employers reported that they had replaced a hazardous chemical with a less hazardous substitute because of information presented on an SDS. With regard to the HCS as a whole, GAO found that over 56% of employers reported “great” or “very great” improvement in the availability of hazard information in the workplace and in management's awareness of workplace hazards. Forty-five percent of those in compliance with the HCS considered the standard to have a positive effect on employees, compared with only 9% who viewed the effect as negative. The results indicate that when chemical hazard information is provided, the result is generally recognized as beneficial to employees.

A number of other studies support this conclusion. For example, in a survey of 160 workers at a large national laboratory, more than 90 percent of respondents said that SDSs are satisfactory or very satisfactory in providing protective information and answering questions (Phillips

et al.,

1999).

Conklin demonstrated the utility of SDSs among employees of a multinational petrochemical company (Conklin, 2003). Across three countries (the U.S., Canada, and the United Kingdom), 98 percent felt that the SDS is a satisfactory information source (the percentage was similar across all three countries). Seventy-two percent said they would request an SDS all or most of the time when introduced to a new chemical, although 46 percent of workers said that SDSs are too long. The author notes, however, that this sample did not include any workers with low literacy.

A number of investigations have raised concerns that, in some cases, the information on SDSs is not comprehensible to employees. In 1991, OSHA commissioned a study that evaluated the comprehensibility of SDSs by a group of unionized employees in manufacturing industries located in the State of Maryland (Kearney/Centaur, 1991). The study assessed the ability of these employees to understand information regarding the route of entry of the substance, the type of health hazard present, appropriate protective measures, and sources of additional help.

Each of the 91 participating workers was provided with and tested on four different SDSs. The workers answered the test questions based on information

supplied on each of the SDSs. It should be noted that the employees who volunteered for this study understood that it relied on reading comprehension. This created a selection bias, as employees with reading difficulties would not be likely to volunteer for the study.

The results of the tests indicated that workers on average understood about two-thirds of the health and safety information on the SDSs. The best comprehension was associated with information providing straightforward procedures to follow (

e.g.,

in furnishing first aid, dealing with a fire, or in using personal protective equipment) or descriptions of how a chemical substance can enter the body. Workers had greater difficulty understanding health information addressing different target organs, particularly when more technical language was used. Workers also reportedly had difficulty distinguishing acute from chronic effects based on information presented in the SDSs.

A similar result was reported by Conklin in a study involving employees of a multinational petrochemical company (Conklin, 2003). After viewing information on an unfamiliar chemical in a variety of SDS formats, a questionnaire was administered to workers to gauge their comprehension of the material presented. The workers reportedly answered 65 percent of the questions correctly.

A study that examined the comprehensibility of SDS to master printers was reported by the Printing Industries of America in 1990 (PIA, 1990). The subjects had an average of 13.9 years of formal education, or approximately two years beyond high school. In this study, 27 SDSs were selected and analyzed for reading levels using a software program, finding an average reading grade level of 14. The investigators found that employees with 15 years of education or more understood 66.2% of the information presented.

Some of the difficulty workers experience in understanding information presented on SDSs may be due to the vocabulary used in the document. Information presented at a reading level that exceeds the capability of the user is unlikely to be well understood. An example of this situation was reported by Frazier

et al.

(Frazier

et al.,

2001). The authors evaluated a sample of SDSs from 30 manufacturers of toluene diisocyanate, a chemical known to cause asthma. Half of the SDSs indicated that asthma was a potential health effect. One SDS made no mention of any respiratory effects, while others used language (

e.g.,

allergic respiratory sensitization) that the authors believed may not clearly communicate that asthma is a risk. However, the more technical language meets the requirements of the HCS.

Other reports substantiate the belief that many SDS users have difficulty understanding the information on the documents. For example, in a study evaluating the comprehensibility of SDSs at a large research laboratory, 39 percent of the workers found SDSs “difficult to understand” (Phillips, 1997). The study also indicated that a third of the information provided on SDSs was not understood. These results were obtained from a study population of literate, trained workers who spoke English as their first language.

Smith-Jackson and Wogalter corroborated this finding in a study involving 60 undergraduates and community volunteers (Smith-Jackson and Wogalter, 1998). The subjects were asked to sort SDS data into a logical order. After completing the task, subjects were asked for their opinions on the difficulty of the content. Overall, 43 percent found the information easy to understand, 42 percent said it was not easy, and the remaining 15 percent felt that only scientists, experts, or very experienced workers would be able to understand the information.

These studies are consistent in reporting that workers have difficulty understanding a substantial portion of the information presented on SDSs. This finding can be explained at least in part by the fact that not all of the information on SDSs is intended for workers. SDSs are intended to provide detailed technical information on a hazardous chemical. While they serve as a reference source for exposed employees, SDSs are also meant for other audiences as well. SDSs provide information for the benefit of emergency responders, industrial hygienists, safety professionals, and health care providers. Much of this information may be of a technical nature and would not be readily understood by individuals who do not have training or experience in these areas. For example, language that may be readily understood by a population of firefighters may be poorly understood by chemical workers.

In addition, Title III of the Superfund Amendments and Reauthorization Act (SARA, also known as the Emergency Response and Community Right-to-Know Act of 1986) mandated that SDSs be made available to State emergency response commissions, local emergency planning committees, and fire departments in order to assist in planning and response to emergencies, as well as to provide members of the general public with information about chemicals used in their communities. It is difficult, if not impossible, for a document to meet the informational needs of all of these audiences while being comprehensible to all as well.

Product liability concerns also play a role in the comprehensibility of SDSs. Producers of chemicals may be subject to “failure to warn” lawsuits that can have significant financial implications. Attempts to protect themselves against lawsuits can affect the length and complexity of SDSs, as well as the way in which information is presented.

In some cases the length and complexity of SDSs reportedly make it difficult to locate desired information on the documents. For example, in testimony before the U.S. Senate Subcommittee on Employment, Safety, and Training, one hospital safety director described a situation in which an employee was unable to find critical information on an SDS in an emergency situation:

* * * two gallons of the chemical xylene spilled in the lab of my hospital. By the time an employee had noticed the spill, the ventilation had already sucked most of the vapors into the HVAC. This, in turn, became suspended in the ceiling tile over our radiology department. Twelve employees were sent to the emergency room. To make the matter worse, the lab employee was frantically searching through the MSDS binder in her area for the xylene MSDS. Once she found it, she had difficulty locating the spill response section. After notifying our engineering department, she began to clean up the spill with solid waste rags, known for spontaneous combustion, and placing the rags into a clear plastic bag for disposal. She did not know that xylene has a flash point of 75 degrees Fahrenheit. She then walked the bag down to our incinerator room and left it there, basically creating a live bomb. Twelve people were treated from this exposure. The lab employee was very upset and concerned about the safety of the affected employees and visitors, and hysterically kept stating that she could not find the necessary spill response information (Hanson, 2004).

SDSs at this particular hospital were reported to range from one page to 65 pages in length.

To accommodate the needs of the diverse groups who rely on SDSs, a standardized format has been viewed as a way to make the information on SDSs easier for users to find, and to segregate technical sections of the document from more basic elements. A standardized format was also thought to facilitate computerized information retrieval systems and to simplify employee training.

OSHA established a voluntary format for SDSs in 1985 to assist manufacturers

and importers who desired some guidance in organizing SDS information. This 2-page form (OSHA Form 174) includes spaces for each of the items included in the SDS requirements of the standard, to be filled in with the appropriate information as determined by the manufacturer or importer. However, some members of the regulated community desired a more comprehensive, structured approach for developing clear, complete, and consistent SDSs.

In order to develop this structure, the Chemical Manufacturers Association (now known as the American Chemistry Council) formed a committee to establish guidelines for the preparation of SDSs. This effort resulted in the development of American National Standards Institute (ANSI) standard Z400.1, a voluntary consensus standard for the preparation of SDSs. Employers, workers, health care professionals, emergency responders, and other SDS users participated in the development process. The standard established a 16-section format for presenting information as well as standardized headings for sections of the SDS. An updated version of the ANSI standard published in 2004 is consistent with the GHS format that is included in the proposed rule.

By following the recommended format, the information of greatest concern to employees is featured at the beginning of the document, including information on ingredients and first aid measures. More technical information that addresses topics such as the physical and chemical properties of the material and toxicological data appears later in the document. The ANSI standard also includes guidance on the appearance and reading level of the text in order to provide a document that can be easily understood by readers.

OSHA currently allows the ANSI format to be used as long as the SDS includes all of the information required by the HCS. Because it is a voluntary standard, however, the ANSI format has not been adopted by all chemical manufacturers and importers. As a result, different formats are still used on many SDSs.

The International Organization for Standardization (ISO) has published its own standard for SDS preparation. This standard, ISO 11014-1, has been revised for consistency with the GHS (new version issued in 2009). The standard includes the same 16 sections as the GHS, as well as similar data requirements in each section. These two consensus standards, ANSI Z400.1-2004 and ISO 11014-1 (2009), have essentially the same provisions and are consistent with GHS. There are minor differences, such as units of measure recommended in the national ANSI standard versus the international ISO standard.

Another development has been the creation of International Chemical Safety Cards (ICSCs). The documents, developed by the International Programme on Chemical Safety, summarize essential health and safety information on chemicals for use at the “shop floor” level by workers and employers (Niemeier, 1997). ICSCs are intended to present information in a concise and simple manner, and they follow a standardized format that is shorter (one double-sided page) and less complex than the ANSI approach. The ICSCs were field tested in their initial stages of development, and new ICSCs are verified and peer reviewed by internationally recognized experts (Niemeier, 1997). ICSCs have been developed in English for 1,646 chemicals, and are also available in 16 other languages. The ICSCs are being updated to be consistent with the GHS.

A study by Phillips compared the effectiveness of different SDS formats as well as ICSCs among workers at a large national laboratory (Phillips, 1997). The employees represented a variety of trades, including painters, carpenters, truck drivers, and general laborers. Each worker was tested for knowledge regarding a hazardous chemical before and after viewing an SDS or ICSC. Three designs were tested: a 9-section OSHA form, the 16-section ANSI Z400.1 format (an earlier and slightly different version of the current ANSI Z400.1 format), and the 9-section ICSC. A subsequent paper described the final results of this study (Phillips, 1999). All three formats led to significant improvements in subjects' knowledge, and there was no statistically significant difference among the three formats in terms of total test score. However, there were a few significant differences in how well readers of each SDS format answered specific types of questions:

• The ICSC performed better than the OSHA form regarding chronic and immediate health effects.

• The other two formats performed better than the ANSI format on fire-related questions.

• The OSHA form performed better than the other two formats on spill response questions.

• The OSHA form was better than the ANSI format regarding carcinogenic potential.

In a separate comparison, Conklin also found similarities in the overall performance of several standard SDS formats (Conklin, 2003). In this study, employees of a multinational petrochemical company were given one of three versions of an SDS for an unfamiliar chemical: a U.S. version (OSHA's required content within an ANSI Z400.1-1998 16-part structure); a Canadian version following the 9-part structure prescribed by Canada's Workplace Hazardous Materials Information System (WHMIS); and a version following the European Union's content and 16-part structure. SDSs were controlled for font, layout, and reading level. Overall, Conklin found no statistically significant difference in mean post-test scores using the three different formats, although there were significant differences on 5 out of 10 questions (no one format was consistently better).

Because extensive searching can be a barrier to SDS use, researchers have examined whether there is a preferred order of information that more closely matches users' cognitive expectations. Smith-Jackson and Wogalter asked 60 undergraduates and community volunteers to arrange portions of six SDSs in the order they considered most usable (Smith-Jackson and Wogalter, 1998). The authors found a few consistent results:

• Information about health hazards, protective equipment, and fire and explosion data tended to be placed toward the beginning.

• Physical and reactivity data tended to be placed near the end.

• Spill or leak procedures were placed near the beginning or the middle, depending on the type of chemical.

A majority of subjects reported that they had attempted to prioritize the hazard information that needed to be communicated. The participants' suggested order of information generally did not match either the original SDS order or the order listed in the HCS—particularly the subjects' emphasis on health hazard information near the beginning.

In the previously discussed 1991 study that evaluated the comprehensibility of SDSs by a group of 91 unionized workers in manufacturing industries in the State of Maryland, a subset of the group (18 workers) was also tested on an ICSC (Kearney/Centaur 1991). While the results indicated that workers on average understood about two-thirds of the health and safety information on SDSs, ICSCs provided better results. The average ICSC test score ranged from 6% to 23% higher than the average test score on the four SDSs evaluated. This finding was considered by the authors to suggest that an improved format for SDSs may

serve to increase user comprehension of the information presented.

OSHA believes that a standardized format would improve the effectiveness of SDSs. The primary basis for this belief is very simple: A consistent format would make it easier for users to find information on an SDS. Headings for SDS sections would be standardized, so SDS users would know which section to consult for the information they desire. The sections would be presented in a consistent, logical sequence to further facilitate locating information of interest. Information commonly desired by exposed employees and of greatest interest to emergency responders (

e.g.,

Hazards Identification; First Aid Measures) would be presented in the beginning of the document for easy reference. More technical information (

e.g.,

Stability and Reactivity; Toxicological Information) would be presented later.

By segregating more complex information on an SDS from the information that is generally easier to understand, the standardized format included in the proposed rule has the potential to address many of the concerns that have been raised regarding the comprehensibility of information on SDSs. The standardized order of information will allow SDS users who desire only basic information about a hazardous chemical to find that information without having to sift through a great deal of technical information that may have little meaning to them. In emergency situations, rapid access to information such as first-aid measures, fire-fighting measures, and accidental release measures can be critically important.

A standardized format does not address all issues affecting SDS comprehensibility. Reading level and some design elements would continue to vary. In many respects, this is inevitable given the different target audiences that SDSs have, and the varying qualifications of those who prepare SDSs. Nevertheless, OSHA believes that the proposed revisions will result in a substantial improvement in the quality and ease of comprehension of information provided on SDSs.

In addition to the issues regarding comprehensibility, a number of researchers have raised concerns that some SDSs may be incomplete or contain erroneous information. The magnitude of the problem is unclear, because only very limited numbers of SDSs have been evaluated in these studies and in some cases the investigations were performed so long ago that the results may not reflect current practices. Nevertheless, the evidence appears to indicate that a substantial number of SDSs may not contain complete and correct information.

An initial examination of the accuracy of SDSs was commissioned by OSHA shortly after the scope of the rule was expanded to cover all industries in 1987 (Karstadt, 1988). The report, which analyzed the content of 196 SDSs for products used in auto repair and body shops, provided a general indication that the content and presentation of information was inconsistent on the SDSs examined. In 1991, OSHA commissioned an additional study that examined the accuracy of SDSs (Kearnet/Centaur, 1991). The study examined information presented in five areas considered crucial to the health of workers potentially exposed to hazardous substances. These five areas assessed were chemical identification of ingredients; reported health effects of ingredients; recommended first aid procedures; use of personal protective equipment; and exposure level regulations and guidelines. The evaluation indicated that 37% of the SDSs examined accurately identified health effects data, 76% provided complete and correct first aid procedures, 47% accurately identified proper personal protective equipment, and 47% correctly noted all relevant occupational exposure limits. Only 11% of the SDSs were accurate in all four information areas, but more (51%) were judged accurate, or considered to include both accurate and partially accurate information, than were judged inaccurate (10%). The study also concluded that the more recent SDSs examined (those prepared between 1988 and 1990) appeared to be more accurate than those prepared earlier.

This belief that some SDSs are not complete and correct was corroborated by an examination of SDSs for lead and ethylene glycol ethers (Paul and Kurtz, 1994). Although these substances are known reproductive and developmental toxicants, researchers found that 421 of 678 SDSs examined (62%) made no mention of effects on the reproductive system. OSHA also commissioned a study, completed in 1999, focusing specifically on the accuracy of first aid information provided on SDSs (Lexington Group, 1999). A total of 56 SDSs for seven chemicals were examined. First aid information on the SDSs was compared with information from established references. The researchers reported that nearly all of the SDSs reviewed had at least minor inaccuracies.

A standardized format does not directly address the concerns that have been raised regarding the accuracy of information present on SDSs. However, standardization would improve the accuracy of chemical hazard information indirectly. With consistent presentation of information, the task of reviewing SDSs and labels to assure accuracy would be simplified. Individuals preparing and reviewing these documents should find it easier to identify any missing elements, and compare information presented on an SDS to reference sources and other SDSs. OSHA enforcement personnel would be able to more efficiently examine SDSs when conducting inspections. The detailed entries proposed for the SDS are particularly noteworthy in this regard. The sub-headings would provide an organized and detailed list of pertinent information to be included under the headings on the SDS. For example, while the HCS currently requires physical and chemical characteristics

This text is long and has been trimmed here. Open the source document for the complete record.