Innovations in Risk Assessment and Management


Tuesday, May 24, 2016, 10:30 AM - 12:30 PM


How to Evaluate the Science in a Scientific Study

A. Havics, PH2, LLC, Avon, IN

Situation/Problem: We have achieved a state of polygnosia or too much knowledge. In 1950 the number of journals recorded was about 60,000 and the estimate for year 2000 was about 1,000,000. Having so much data available, most digitally, provides a new problem, how does one value the data available?

Resolution: The question is not an easy one to answer because from a risk standpoint everything is gray. It is a matter of how much gray and then how much gray is too much gray. There are several reasons to evaluate data in this fashion, and perhaps one of the best is the AIHA-ACGIH-ABIH Cannons of Ethical Conduct: “Industrial Hygienists should obtain information regarding potential health risks from reliable sources," and, “Industrial Hygienists should review the pertinent, readily available information to factually inform the affected parties." This leads to the question—how does one evaluate data, whether it is a tox study, a published case study, or another scientist’s unpublished report or notes? First, one must determine the purpose of the evaluation, then one must critically review the a) data presented, b) source of the data, and c) the methodology. From there it all depends upon the findings, but there are a few resources/procedures to use.

Results: Example of these tools include: a) Klimisch approach, b) ToxRTool, c) the Pinto approach, d) Science Policy Council’s Assessment Factors, e) Riegelman’s How to Study a Study and Test a Test, and f) Hill’s Criteria. They each have different applications and approaches. But, the factors considered and the steps involved can be put to good use in evaluating data. Some examples of published and unpublished papers & reports will be used to scratch the surface of how to valuate data: air emission allocation, risk from fracking, fungal spore trap data, determining the weighted consensus from multiple toxicology papers, unpublished asbestos exposure data, etc.

Lessons learned: Even good data can be presented poorly and even good-intentioned scientists can present poor data or poor interpretations. The ability to accept bad data seems to come with a great need consisting first of the absence of data. I too should be very carefully about what I write, lest I become an example of poor value.



Big Data and Assessing Exposures & Risks: When Too Much Seems Like Never Enough

J. Persky, RHP Risk Management Inc., Chicago, IL; B. Heckman, RHP Risk Management Inc., Carlisle, PA; F. Boelter, RHP Risk Management Inc., Boise, ID

Situation/Problem: Avogadro’s Number is a pretty big number, even though numerically it is only a billion times more than the number of cells in the human body. Frankly, big numbers are incomprehensible for most people and the pile of data and variables to consider just keep growing. Whether the challenge is analyzing air samples from a spill response or developing a cumulative risk analysis, when it comes to exposure, stressors and risk, people ask simply What does it mean to me?. We have personal sampling techniques, biological monitoring techniques, modeling techniques, and guidelines and standards for interpretation. Will Big Data help us answer questions or just raise more questions we cannot answer?

Resolution: New technologies and their subsequent innovations don’t directly cause social change; instead they create problems and dilemmas that drive society to seek new solutions from a diverse set of choices. Today’s social ferment from growing global networks is the new breeding ground for radical innovations, for good and for ill. Social experimentation with disruptive technologies will be a prominent feature of the next twenty years and beyond.

Results: We explored several projects where an extraordinary amount of data was available for analysis. Tens of thousands of data points with ranges, distributions, averages, and 99thPCLs were considered. From an industrial hygiene perspective, frequently none of the data shows overexposures and much of the data is censored or doesn’t meet quality objectives. Will the era of big data yield larger amounts of higher quality data? How is industrial hygiene going to evolve to incorporate or deal with genetic diversity, epigenetic variations and arguments about susceptibility to disease?

Lessons learned: Big data holds out the promise to provide the ability to dig into data and answer questions that were not originally envisioned, alter the way we define and evaluate acceptable risk and potentially allow for more personalized exposure assessments and risk characterizations. Our profession needs to develop guidance on how the age of Big Data will influence developing and validating models, define risk acceptability criteria, derivation of site-specific risk-based exposure criteria, and communicate concepts of extremely low probability and de minimis risk.



Expected Quantitative Results Based on Qualitative Evaluation

M. DaSilva, Doulos Ambiental, São Paulo, Brazil

Situation/Problem: In multinational companies with hundreds of employees in several parts of the World. It is very difficult to establish an IH program. There are not enough specialized professionals capable of collecting samples in each site. Therefore, the SHE team has to develop methodologies that may be applied by local people after receiving a basic training in managing occupational risks.

Resolution: A qualitative evaluation is a very helpful tool, if it has a systematic approach and comprehensive criteria. A Brazilian process named Preliminary Risk Analyze - Occupational Hygiene (PRA - OH) allows the estimating of the potential environmental risk in workplaces. It is an Excel Data Sheet having several macros developed to include all data necessary according to AIHA s book “A Strategy for Assessing and Managing Occupational Exposures”. The first part is a basic characterization that presents general information about the company, workplace, position of employee, facility dimensions, etc. The second part sets information of tasks, exposure time, agents, and health effects. A critical part of the PRA - OH is the worker participation (required by Brazilian law). It is an interview with several laborers to register their opinions about concentration or level of environmental agents. Answers are in list boxes or a combo boxes for standardizing the opinions. One matrix with 5 x 5 grades allows classifying the occupational profile (Concentration-Level versus Time Exposure). The Risk Matrix is a cross-correlation of the occupational profile grades and health effects levels. The health effects are the TLV® booklet classified in five levels (1 a 5). The workers received hints about the grading system. Example, noise level 1 is an office environmental and 5 is similar to airport near the airplanes.

Results: The results described here resulted from the PRA - OH applied to maintenance tasks that involved electrical, mechanical, instrumentation and welding services in a petrochemical plant. Around 27 interviews related to a group of 95 employees formed by electrical, mechanical, and boilermaker welder professionals. The workers pointed out around 30 chemical and physical agents including Sulfuric Acid, Vinyl Chloride, Chlorine, Dichloroethane, Sodium Hydroxide, Hydrogen Chloride, Metal Mercury, Noise, Hand and Arm Vibrations, Heat Stress, and others. In total, 200 grades of potential risks were documented. The distribution of the results were 26 irrelevant, 124 attention, 46 moderate, 4 severe and zero grave imminent. Based on the identified qualitative risk, mainly moderate and severe, the monitoring program included 54 noise doses, 10 heat stress (WBGT) evaluations, 6 hand-arm vibration measurements, 63 samples of chemicals (Vinyl Chloride, Welding Fumes, Dichloroethane). The quantitative results showed 61% of noise exposure equal or higher than 80 dBA, 50% of WBGT exceeded the TLV®, 1/3 of hand-arm vibration numbers were in a critical situation, 15 % of Vinyl Chloride concentrations higher than Action Level (TLV® = 0,5 ppm), and all Dichloroethane samples below 5 ppm (TLV® - Action Level)

Lessons learned: The PRA - OH allows the amplification of the vision of environment risk because it includes the opinion of workers who know their tasks very well, chemical and physical agents in workplace. The PRA - OH is a historic document signed by workers and EHS professionals. In general, the worker perception about environmental risk is overestimated. Some occasional exposure is routine. The interviewer needs to be aware of agents that have no subjective indicators, such carbon monoxide. The PRA - OH allows a SEG-HEG based assessment. The PRA - OH provides good information for the occupational physician because it considers environment agents and a monitoring program. A certified industrial hygienist must validate the PRA - OH. The EHS team must have access online to check the PRA - OH developed in Microsoft Excel spreadsheet.



National Aeronautics and Space Administration, NASA, Astronaut Occupational Surveillance Program and Lifetime Surveillance of Astronaut Health, LSAH: Astronaut Exposures and Risk in the Terrestrial and Spaceflight Environment

S. Keprta, W. Tarver, M. Van Baalen, and T. McCoy, NASA Johnson Space Center, Houston, TX

Situation/Problem: Astronauts have a very unique and somewhat understudied occupational exposure profile. In order to understand these risks and properly address them, the National Aeronautics and Space Administration (NASA) originally created the Longitudinal Study of Astronaut Health (LSAH). The first LSAH program was designed to address a variety of needs regarding astronaut health and included a 3 to 1 terrestrial control population in order to compare earth normal disease and aging to that of a microgravity exposed astronaut. Over the years, the program has been modified. One example is the move from short duration Shuttle flights to longer duration International Space Station space flights (and exposures). Also, there was the move to incorporate more of an occupational health and medicine model to the study of astronaut exposure in the space environment. This presentation outlines the baseline exposures and monitoring of the astronaut population, both terrestrial and space.

Resolution: Outline and discuss the exposures and stressors that are part of the profession of a United States NASA astronaut. Understand the purpose and methodology of the programs designed to characterize and evaluate these stressors, the longest running of these programs being the LSAH.

Results: We will discuss the typical and non-typical occupational terrestrial and spaceflight exposures to astronauts and the risk assessments used to follow them through their working career and beyond.

Lessons learned: Discuss how standard industrial hygiene risk assessments can, and sometimes cannot, be used to assist in the overall occupational surveillance effort of the NASA Astronaut .



Semi-Quantitative Risk Assessments: It Is About the Controls!

P. Esposito, STAR Consultants, Inc., Arnold, MD

Situation/Problem: Many of today's risk and exposure assessments find numerous references on how to identify hazards, exposures and risks. To that point, classifying these levels of risk are as numerous as are the number of references themselves. NIOSH is leading efforts in the US to identify Banding levels to help define risk factors of severity and likelihood. Classically, we use the hierarchy of controls, qualitatively. However, our assessments typically fail to determine which risk factor (Severity or Likelihood) is impacted by which control? ANSI B11.0, Machine Guarding Risk Assessment, references that only elimination and substitution impacts severity. So, what level or risk reduction do we actually get from engineering controls? ANSI B11.0 says we only get reduction in likelihood, not severity. This greatly impacts what level of protection actually exists when we use engineering, or even less so, administrative or PPE controls.

Resolution: Failure Modes and effects analysis methodology point us toward measuring the reliability of the control to help measure residual risk. The hierarchy of controls: Avoidance, Elimination, Substitution, Administrative, Warnings, and Personal Protective Equipment all come with their own reliability/unreliability factors. In addition, many risk assessments erroneously calculate risk reduction based on the selection of controls, rather than just the risk factor reduction; risk factors being consequences (severity) and our likelihood (probability). Therefore, developing a model where the level of risk reduction is quantitatively tied to the selection of the control or controls can help remove some of the subjectivity in risk assessments, while improving the reproducibility of the process. Also, the model helps you calculate the level of additive risk reductions when there are layers of protection, or defense in depth.

Results: This quantitative aspect of control identification to calculate risk reduction can be a powerful tool in determining when to stop adding controls, or determining the risk apatite of an organization. When combining this approach with some ongoing collection of additional metrics, such as conformance rate of controls, helps verify that the appropriate assumptions were made when calculating the quantitative risk reduction strategies.

Lessons learned: 1. Risk reductions from controls apply to the independent risk factors of severity and likelihood, not directly to the level of risk. 2. The hierarchy of controls, and even the additive effect of layers of protection, can be quantitatively determined. 3. The risk reduction quantification determinations can be independently verified as part of inspection and observations by calculating conformance rates. These conformance rates can then be used to readjust risk reductions to actual work practices, management enforcement, etc.



Risk Management Best Practices to Reduce Injuries and Maximize Economic Benefits in U.S. Mining

S. Griffin, D. Bui, G. Gowrisankaran, E. Lutz, C. He, C. Hu, J. Burgess, The University of Arizona, Tucson, AZ

Objective: Risk management (RM) is a cyclical process of identifying operations or activities that put workers at high risk for injuries, designing controls including engineering changes or standard operating procedures to reduce risks, implementing these controls, and evaluating their effectiveness. While RM is legally required in many countries, U.S. safety and health regulations are typically focused on compliance. The objective of the current study is to determine the effectiveness of RM interventions in reducing injuries and economic costs in the U.S. mining industry.

Methods: Four mining companies with extensive RM expertise, representing the metal, aggregate and coal sectors, participated in the study. Retrospective longitudinal analysis of company internal injury and compensation claims data and Mine Safety and Health Administration (MSHA) injury data was completed to determine the effectiveness of RM programs. MSHA data was also used to compare Injury rates for our partner mines to mines of similar employee size and, for coal mines, total production. Mine employees and managers identified the RM programs most effective at reducing injury. Employees provided costs of program implementation and the resulting changes in injury costs were evaluated, enabling an evaluation of return on investment (ROI). RM best practices were defined as those programs that led to a reduction in injury rates and positive ROI.

Results: Generally, reductions in injuries were observed following implementation of 14 RM programs at our partner mines. Rates of all injuries and lost-time injuries were lower at our partner mines than comparison mines. Implementation costs ranged from $43,000 to $1.2M, with a positive ROI for several programs, including behavioral-based safety interventions and engineering controls.

Conclusions: Several RM programs with reductions in injuries and positive ROI were identified. The results, of the current study, help build a business case for the implementation o​f RM programs, potentially leading to the increased practice of RM in the U.S. mining industry.​