ADVANCEMENT OF THE SCIENCE
ously reported in the qualitative risk assess- ments. In these situations, responses should be interpreted with caution because of reli- ability and lack of knowledge on specific haz- ards or recall bias. Guidance via training for the interviews and questionnaires for better comprehen- sion with regard to specific hazards might improve responses and subsequently lower the degree of mischaracterization, under- estimation, or overestimation of the risks (Gallizzi & Tempesti, 2014). It is equally important that the sample size of work- ers and people in the community be large enough to avoid selection biases. If field evaluation methods used to assess specific exposures to specific hazards are implemented properly, these methods can provide the opportunity to collect quality samples of the hazard that are sucient to be analyzed by traditional statistical tools. If a limited number of samples are available, how- ever, the emerging Bayesian decision models can be used to conduct risk characterization and prioritization (Hager & Johnson, 2015). It is crucial that sampling and analytical methods from agencies such as NIOSH, the Occupational Safety and Health Administra- tion (OSHA), or AIHA are used (Andrews & O’Connor, 2020). This approach is important because the number of samples to be col- lected must meet a threshold minimum of the limit of detection of the analytical laboratory to avoid censored data or non-detect results. In these circumstances, traditional statisti- cal methods might not be the best for analysis, but Bayesian decision analysis tools are find- ing useful applications in these assessments and can provide a good basis for professional judgment about exposures, profiles of expo- sures, and risk characterizations. These tools and approaches rely on the assumption or hypothesis that exposure data are log-nor- mally distributed (Hager & Johnson, 2015). This assumption, though, must be tested before the analysis of the data is carried out. The test is done through graphical methods to obtain goodness of fit as close to 1 before further analysis can be conducted to prioritize and characterize the hazards and risks. Other tools that can be used for assessing hazards include hazard assessment and risk characterization prioritization schemes from the federal government. For example, Hazus- MH (multi-hazard) is a software package of
TABLE 1
Risk Characterization of Multiple Hazards and Exposure Control Category
Exposure Control Category/Exposure Control Banding
Criteria for Characterizing Risks Based on Permissible Exposure Limit (PEL)/ Occupational Exposure Limit (OEL)
Exposure Management Option
0 1 2 3 4
95% ≤ 0.01% × PEL/OEL 95% ≤ 0.1% × PEL/OEL
No action
Highly controlled
0.1 × OEL < 95% ≤ 0.1–0.5 × PEL/OEL 0.5 × OEL < 95% ≤ 0.5–1.0 × PEL/OEL
Well controlled
Controlled
95% > 1.0 × PEL/OEL 95% > 5.0 × PEL/OEL
Poorly controlled
Other categories
Extremely poorly controlled
Adopted from Arnold et al., 2016; Logan et al., 2009.
TABLE 2
Exposure Data Interpretation and Training Questions
Primary Hazard and Risk Characterization
How Often Does Training Occur on the Following Hazards?
1
2
3
4
5
Chemical exposure Ergonomic hazards Material handling Noise exposure Repetitive lifting Sharp injuries
Note. Response numbers are based on a 5-point Likert scale (1 = infrequently and 5 = frequently).
standardized tools to prioritize and estimate risks from multiple hazards and disasters such as hurricanes, floods, tornadoes, and tsunamis (Federal Emergency Management Agency, n.d.), These tools can support pre- and post-natural disaster hazard planning to enhance or increase resilience in the work- place and communities. Similar tools have been adopted and adapted that assess mul- tiple human-made and technological hazards to characterize the risks, some of which use letter coding to indicate probability, severity, and other factors (Lyon & Hollcroft, 2016). The common theme among the tools is the extent of the impacts (i.e., low, medium, high, or significant) and the probability and sever- ity of the impacts (Stickle, 2012). The risk of
negative impacts on property—and subse- quently the bottom line of for-profit compa- nies—has been the driver for this approach. The databases of injuries from the Cen- ters for Disease Control and Prevention and OSHA are tools that can support safety and health professionals in the prioritization of hazards and the associated risk characteriza- tion process. The data sets in these databases are available by ZIP Code or by NAICS or SIC codes for acute and chronic illnesses. The databases are also searchable by the costs of injuries. If the time frame can be established, then the probability of occurrence, location, costs for medical care, and hazard type can be established. A review of the historical injury records and events, and available information
10
Volume 86 • Number 5
Powered by FlippingBook