XtremeSorb Spill control products
Get XtremeSorb spill materials here!

MSDS
Topics
Free Sites FAQ's Regulations Glossary Software Suppliers
Books Forum Poll Fun stuff Quiz Store
MSDS and safety supplies Search ALL our MSDS info
ILPI's Interactive Library of OSHA MSDS Regulations and Interpretations
previous topic OSHA Index next topic

Appendix C

Hazard Evaluation Procedures


The hazard evaluation procedures required by the standard are performance-oriented. Basically, OSHA's concern is that the information on labels and data sheets, and in the training program, is adequate and accurate. Although specific procedures to follow and sources of consultation cannot be definitively established, general guidance will be provided herein. The hazard evaluation process can be characterized as a "tiered" approach -- the extent to which a chemical must be evaluated depends to a large degree upon the common knowledge regarding the chemical, whether its health effects are under review, and how prevalent it is in the workplace.

  1. The first step for CSHOs [Compliance Safety and Health Officers] evaluating chemicals is to determine whether the chemical is part of the "floor" of chemicals to be considered hazardous in all situations.

    1. The floor of chemicals consists of three sources:

    2. Sources to generally establish hazards of the chemicals that are part of the floor of hazardous chemicals covered by the standard:

      In addition, the CSHOs [Compliance Safety and Health Officers] should check the NIOSH Registry of Toxic Effects of Chemical Substances (RTECS) to see if any hazards are indicated which do not appear in these sources. If there are, further study should be done to evaluate the hazards. RTECS should never be considered a definitive source for establishing a hazard since it consists of data that has not been evaluated. It is, however, a useful screening resource.

  2. The second step is to consult other generally available sources to see what has been published regarding the chemical. Patty's Industrial Hygiene and Toxicology would be one such source. OCIS contains a number of other chemical information sources. Material Safety Data Sheets available through information services would also be useful.

  3. The third step, for those chemicals where information is not readily available or where such available information is not complete, is to perform searches of bibliographic data bases. In general, the National Library of Medicine (NLM) services should be used. These include the Toxicology Data Bank (TDB), TOXLINE, and MEDLARS. The information generated by these data bases should be evaluated using the criteria in Appendix B of the HCS; i.e., to qualify as an acceptable study, it must be conducted according to scientific principles (e.g., in animal studies, the number of subjects is adequate to do statistical analyses of the results; a control group is used, and the study must show statistically significant results indicating an adverse health effect). This evaluation obviously requires a subjective, professional assessment. Any questions should be referred to the Directorate of Compliance Programs, Office of Health Compliance Assistance (through the Regional Office), for assistance. In general, uncorroborated case reports and in vitro studies, such as Ames tests, are useful pieces of information, but not definitive findings of hazards. Animal studies involving species other than those indicated in the acute hazard definitions must be evaluated as well. The acute hazard definitions are not included in the standard to "categorize" chemicals but rather to establish that chemicals meeting those definitions fall under the coverage of the standard.

  4. In some cases, the only information available on a substance may be employer-generated data. If the employer indicates that such information is the basis for the hazard evaluation, the CSHO [Compliance Safety and Health Officer] shall ask to see it in order to complete the OSHA evaluation.

  5. In cases where the employer denies the CSHO [Compliance Safety and Health Officer] access to its own hazard data and no published data on the chemical can be found to review the sufficiency of the hazard determination, the Regional Office shall be contacted for assistance in obtaining an administrative subpoena. The Directorate of Compliance Programs shall be contacted if assistance is required in order to obtain unpublished chemical hazard information available from other Federal agencies such as Environmental Protection Agency.

  6. If an employer has found any chemical to be non-hazardous , and the CSHO [Compliance Safety and Health Officer] has reason to believe it is hazardous, further investigation is required. The definitions of hazard in the standard are very broad, and it is not expected that many chemicals can be considered nonhazardous under this approach. Those most likely to be exempted would be chemicals that pose no physical hazards, and which have lethal dose findings above the limits found in the acute hazard definitions.

  7. In some cases, the employer may not have addressed in the Hazard Communication Program a specific chemical that the CSHO [Compliance Safety and Health Officer] knows to be present through knowledge of the process or through sampling or other investigation of the workplace. This situation should also be further investigated. If the CSHO has information to indicate that there is a hazard, the employer must be able to defend the finding of no hazard.

  8. Internet addresses for the above-mentioned organizations are:


Previous: Appendix B   |   Top of Page   |   Next: Appendix D


Entry last updated: Wednesday, June 17, 2009. This hypermarked and content-enhanced page is copyright 2001-2014 by ILPI, all rights reserved. Unauthorized duplication or posting on other web sites is expressly prohibited. For questions, comments and concerns, please contact us at our MSDS email address.

The official, public domain, OSHA version of this document is available at http://www.osha.gov/pls/oshaweb/owadisp.show_document?p_table=DIRECTIVES&p_id=1551&p_text_version=FALSE.