Diacetyl and Food Flavorings Containing Diacetyl

Stakeholder Meeting
Arlington, Virginia
October 17, 2007

Meeting Summary Report

Draft - Version Date November 27, 2007

Table of Contents

1 Meeting Format

2 Introduction

3 Questions and Comments to OSHA from Stakeholders

4 Questions from OSHA to Stakeholders

5 Closing Remarks

Diacetyl and Food Flavorings Containing Diacetyl

Stakeholder Meeting
Arlington, Virginia
October 17, 2007
Meeting Summary Report

1 Meeting Format

As part of the OSHA process to gather information about the occupational exposure to diacetyl and food flavorings containing diacetyl, OSHA hosted this stakeholder meeting focusing on butter and other food flavorings containing this substance. OSHA specifically sought information regarding uses of diacetyl and food flavorings containing diacetyl, medical screening and surveillance for workers exposed to these substances, and diacetyl exposure assessment and control methods. OSHA representatives and stakeholders participated in an open conversation not only to address specific OSHA information requests, but also to identify stakeholder concerns associated with developing interim and final rules on occupational exposure to diacetyl and food flavorings containing diacetyl.

The following is a summary of the key comments stakeholders made during that meeting. Comments are grouped together by topic, without reference to the identity of the speakers.

2 Introduction

OSHA representatives provided background information on the enforcement and compliance actions taken by OSHA to address the hazards associated with occupational exposure to diacetyl and food flavorings containing diacetyl, OSHA's authority to regulate employee exposure to these substances, and the OSHA stakeholder and standard-setting process.

OSHA enforcement and compliance activities in the past six years have included investigations by regional OSHA offices and development of publications targeting microwave popcorn facilities and employers. Additionally, a National Emphasis Program aimed at minimizing or eliminating employee exposure will result in inspections of microwave popcorn manufacturing facilities that use butter flavorings containing diacetyl. Other efforts include (1) development of a Safety and Health Bulletin (SHIB) that addresses respiratory disease among employees in microwave popcorn processing plants, and (2) the Hazard Communication Guidance for Diacetyl and Food Flavorings Containing Diacetyl, which addresses responsibilities of flavoring manufacturers and employers who must comply with OSHA's Hazard Communication Standard with regard to diacetyl and food flavorings that contain diacetyl.

In response to a petition to issue an emergency temporary standard (ETS), OSHA's Directorate of Standards and Guidance (DSG) carefully reviewed the existing data. This review did not support the issuance of an emergency temporary standard (ETS) under Section 6(c) of the Occupational Safety and Health Act of 1970 (OSH Act) (29 U.S.C. 651 et seq.) to protect employees from a grave danger due to exposure. Hence OSHA will develop an occupational health standard addressing the likely significant risk of material impairment of health due to exposure to diacetyl and food flavorings containing diacetyl under Section 6(b) of the Occupational Safety and Health Act.

Since receiving the petition for an ETS, OSHA's DSG has undertaken several activities to understand the scope of the hazard and the rapidly evolving information about diacetyl and food flavorings containing diacetyl. Through site visits to flavor manufacturing and microwave popcorn facilities, OSHA is gathering information on production processes, worker exposures, and control methods. In addition, OSHA representatives described several collaborative efforts to assist in their rulemaking process including: (1) ongoing dialogue with the California Department of Industrial Relations' Division of Occupational Safety and Health (Cal/OSHA) to better understand its proposed rule to address occupational exposures to food flavorings, (2) reviewing data collected by Cal/OSHA through the voluntary flavoring industry safety and health evaluation program (FISHEP), and by the National Institute of Occupational Health and Safety (NIOSH), and (3) working with National Jewish Medical Research Center to discuss their study of flavor manufacturing facilities. OSHA acknowledged the contributions and support of these organizations to the rulemaking process.

OSHA representatives also described the process for developing standards. This process typically begins with data gathering efforts, such as publishing requests for information, holding stakeholder meetings, and conducting site visits. After gathering information, OSHA drafts a proposed standard and evaluates the potential economic impacts of the proposed rule on employers, particularly small businesses. OSHA then publishes the proposal for public comment. To gather further information and address issues raised by stakeholders in their written comments, OSHA often holds informal public hearings. OSHA prepares the final rule based on all the information and comments in the record. The rulemaking process includes multiple opportunities for formal and informal stakeholder input.

OSHA representatives emphasized that OSHA is in the stage of gathering information on diacetyl and food flavorings containing diacetyl to better understand the issues involved. This stakeholder meeting serves as an informal, conversational forum for relationship building and information sharing between OSHA and stakeholders who are experts in the flavorings and popcorn industries. OSHA considers stakeholder input to be extremely important in helping develop an effective standard that addresses the key elements associated with exposure to diacetyl and food flavorings containing diacetyl, while considering the complex constitution of flavoring mixtures, the proprietary nature of flavorings, and the feasibility concerns of this standard. A team of industrial hygienists and economists will assist OSHA in conducting site visits and developing the feasibility analysis.

3 Questions and Comments to OSHA from Stakeholders

During the open conversation between OSHA representatives and stakeholders, the stakeholders raised questions and concerns about the rulemaking process and the use of other available datasets besides the three (Cal/OSHA, NIOSH, National Jewish Medical Center) mentioned by OSHA.

3.1 How will OSHA proceed with rulemaking on diacetyl and food flavorings that contain diacetyl so that process is not stretched over many years, as has occurred with beryllium, silica, and chromium?

Some stakeholders expressed concern over structural impediments inherent in OSHA's rulemaking process that would lengthen this process.

Several stakeholders also emphasized the importance of rapidly obtaining the information needed to develop a standard. These stakeholders noted that workers continue to fall sick and suffer the large social and family costs of the disease, while employers and agencies debate rules and processes. The slow response from regulatory authorities and employers is disheartening for sick workers who continue to be exposed or have lived with their disease for many years, up to 9 years in some cases, and are still waiting for decisive action. Stakeholders stated that employers and regulatory agencies should put worker interests at the forefront of the debate.

OSHA representatives emphasized that the diacetyl rulemaking is considered a priority for OSHA and the process is being expedited to the full extent possible. OSHA understands the importance of developing a standard quickly due to the impact exposures to diacetyl and food flavorings containing diacetyl have on workers. OSHA is committed to keep the process moving through all the necessary stages of rulemaking as required under OSHA mandate.

3.2 Is there a timetable for the rulemaking process?

Some stakeholders noted that a timetable for the rulemaking process would greatly reassure stakeholders of OSHA's intent and plan to move expeditiously. A stakeholder requested that the timetable include a schedule listing periods during which OSHA will accept comments and stakeholders may provide OSHA with responses. It was also suggested that stakeholders be allowed to contribute to the agendas for public meetings and the rulemaking process.

OSHA noted that the regulatory agenda, published in the Federal Register, serves as a public timetable of milestones in the rulemaking process and is updated every 6 months.

3.3 The length of traditional OSHA regulatory process is not expected to produce the desired result in the timeframe needed. Hence, given that the State of California is currently developing a standard for occupational exposures to food flavorings, OSHA would greatly benefit from working collaboratively with Cal/OSHA instead of developing a separate process.

A few stakeholders observed that OSHA should examine the entire rulemaking process. In light of the short-term latency issues associated with flavoring-related obstructive airway disease, they urged OSHA to consider creative means to expedite the rulemaking process, such as obtaining data collected by other agencies and organizations with similar aims. A stakeholder noted that Cal/OSHA has similar statutes as OSHA and that a joint effort should be considered.

An OSHA representative pointed out that the agency must work within its requirements for rulemaking and that the analytic burden and feasibility requirements for Cal/OSHA are different compared to those at the federal level. However, OSHA is open to new methods of obtaining data more quickly. OSHA will work collaboratively with Cal/OSHA and NIOSH to take advantage of the progress these organizations have already made with respect to understanding diacetyl and food flavorings containing diacetyl. OSHA has already committed contract resources to help compile the Cal/OSHA data.

Two stakeholders commented that the analytic burdens on OSHA are excessive. They also stated that even though the OSHA and Cal/OSHA regulatory processes are different, the hazard is the same and OSHA should utilize the consensus that has already been built on certain types of information. For example, even though limited information is available from the food manufacturing industry, the consensus developed by the microwave popcorn and flavoring manufacturing industries should be used. Stakeholders also urged OSHA not to consider the rulemaking process merely a legal checklist, but instead to develop a consensus process, even if this process differs from the traditional OSHA rulemaking process.

3.4 Does OSHA have data from the U.S. Environmental Protection Agency (EPA) study?

OSHA representatives responded that they had inquired about the availability of the EPA study and were told that a study report was under industry review to ensure that it did not contain confidential business information. OSHA representatives indicated that the agency would continue to seek the study results. [OSHA finally received the study report in early November]. OSHA representatives also observed that the Cal/OSHA and NIOSH studies were more relevant for the OSHA rulemaking effort, as these studies focused on occupational exposures to food flavorings containing diacetyl, unlike the EPA study which studied emissions from popcorn bags.

One stakeholder emphasized the importance of the EPA study since this study addresses emissions of diacetyl and other butter flavoring compounds from microwave popcorn. The stakeholder encouraged OSHA to obtain the results of the study by formal means, such as via a formal letter of request from the Assistant Secretary, if necessary.

3.5 A strong need exists for obtaining information as quickly as possible. The slow response from regulatory authorities and employers is disheartening for sick workers who continue to be exposed or have lived with their disease for many years.

A few stakeholders observed that many developments have taken place in the past 7 to 8 years and many large companies have already taken steps to institute controls that protect workers. In other cases, some companies may not have made as much progress due to a lack of resources, understanding, or interest. These companies might be looking for guidance from government agencies before acting.

3.6 Why does OSHA's hazard communication guidance not reference the latest Mortality and Morbidity Weekly Reports (MMWR) or the California cases/studies?

OSHA representatives explained that the process for developing and reviewing hazard communication guidance is often stretched over several months. Hence, the information in the guidance may not always reflect the most recent findings. OSHA will have opportunities to update the guidance document in the future and will incorporate new information.

One stakeholder also commended OSHA on developing its hazard communication guidance, which was very useful.

4 Questions from OSHA to Stakeholders

OSHA representatives sought specific information regarding the uses of diacetyl and food flavorings containing diacetyl, effective means of conducting medical screening and surveillance, exposure assessment techniques, and exposure control methods.

A. Uses of Diacetyl and Food Flavorings Containing Diacetyl

OSHA sought information on products made with diacetyl-containing flavorings, production processes that use these flavorings, variability of flavoring formulations, and worker exposures to diacetyl and food flavorings that contain diacetyl.

4.1 What are the uses of diacetyl and flavorings that contain diacetyl? What is the universe of operations and compounds that use diacetyl? What are the uses of diacetyl in the flavoring and food manufacturing industries besides the microwave popcorn industry? What is the extent and type of diacetyl use in snack foods, bakery goods, and the dairy industries?

Stakeholders provided the following comments and recommendations regarding the uses of diacetyl:

Diacetyl is a naturally occurring substance in many foods, such as butter, strawberries, and other dairy products and fruits. Hence, food manufacturers use diacetyl to provide the same flavor perception of butter in margarines and to mimic the natural flavors of dairy products, fruits, and vegetables in a variety of foods. Diacetyl is generally used in low levels (tens of parts per million [ppm] or less) in foods other than butter-flavored items and popcorn. Diacetyl typically represents less than one percent of the flavor product that leaves the plant and is shipped to food manufacturers. Occupational exposures differ between the flavorists involved in manufacturing the flavoring substance, the food producers who use the flavors in food manufacture, and consumers. Several stakeholders were willing to provide information to OSHA on the wide variety of diacetyl uses in flavor and food products.
Diacetyl appears to be a key contributor to an adverse health effect. However, no clear information exists on the contribution of other components of the flavoring mixtures (such as acetoin and various aldehydes) to adverse effects. Due to the proprietary nature of flavoring formulations and the complex mixtures involved in flavorings, identifying specific constituents is difficult. This issue requires further discussion.

4.2 Are there differences between the use of powdered and liquid diacetyl? Does the difference in the form of diacetyl result in different exposures? Are certain exposure levels associated with handling certain products? Should OSHA be concerned about other chemicals used with diacetyl in flavoring mixtures? Is there data available that will help OSHA determine if a certain level of diacetyl is associated with health effects?

Stakeholders provided the following comments and recommendations regarding the exposure potential of diacetyl:

The opportunity for exposure is greater in the flavor manufacturing industry compared to the food manufacturing industry because workers handle large quantities of relatively pure chemicals. Additionally, flavor production workers handle different chemicals on different days compared to food manufacturing workers who tend to work with the same food item every day. Diacetyl usually represents approximately one percent of the flavoring mixture, but often the levels are lower.
Studies in a test chamber were conducted to look at the amount of diacetyl released from various liquid, dry and wetted forms of diacetyl. The three forms studied included a dry powder encapsulate composed of 1 to 2 percent diacetyl by weight, paste/liquids composed of 4 to 9 percent diacetyl by weight, and dry encapsulated powder moistened to mimic powder in contact with mucous membranes of the respiratory tract. These three forms were individually placed in a chamber, and then exposed to heat (104 degrees Celsius [C]) and humidity that mimic plant conditions. The study showed that all forms emitted diacetyl, with the moistened dry encapsulated form releasing the highest quantity at 54 ppm diacetyl in the static head space of the test chamber. The paste/liquid form also emitted higher amounts of diacetyl than the dry formulation. This data will be provided to OSHA, as requested.
Heat is an important element contributing to the exposure potential of diacetyl. Diacetyl is not typically a volatile material unless heated, such as occurs in flavor manufacturing. The boiling point is 88 degrees C.
One stakeholder claimed to be aware of a clinical case that may be associated with pure diacetyl.
The same stakeholder stated that formulations have occasionally been released in legal proceedings involving worker cases. Hence proprietary formulation information may be available if requested.
Another stakeholder emphasized that worker activities that facilitate exposure are as important as the product being handled by the worker. Frequency of exposure, work processes, and levels of exposure should all be considered, as should worker lung function and health status of workers.

4.3 How can OSHA obtain information on how often workers are exposed, what the work processes are, and what the exposure levels are? Are there potential obstacles to obtaining this type of information? Stakeholders provided the following comments and recommendations regarding the availability of data on worker exposures to diacetyl:

Every major manufacturer is evaluating data related to diacetyl and conducting employee exposure assessments. Thus, OSHA should be able to obtain these data requested. As some stakeholders suggested to OSHA in December 2006, one means of obtaining the information is a formal request by the Assistant Secretary to the companies. The stakeholders' general consensus was that OSHA should not be hesitant about requesting data from industry.
Some stakeholders emphasized the need to focus on the tasks that present opportunities for exposures. The flavor manufacturing industry has a lot of valuable data that have already been provided to Cal/OSHA, the California Department of Health Services (Cal/DHS), and to OSHA consultants. In addition, some other datasets exist that have not yet been made available to OSHA. A few stakeholders showed a willingness to share exposure information with OSHA.
Several stakeholders believe Cal/DHS plans to make all its data publicly available on its website. Cal/OSHA has developed a questionnaire designed to obtain information from the food manufacturing industry.
One stakeholder also suggested that OSHA create a disease reporting registry on the OSHA web site. Currently, there are reported cases distributed across many regions of the country and physicians do not have a consolidated reporting medium. The beryllium registry would be a good model for such an effort. In addition, the stakeholder suggested the disease registry would help in designing much needed clinical trials and therapeutic interventions.
One stakeholder also suggested that occupational histories retained by some lung transplant centers, of which only few exist nationwide, could be tapped as a potential source of information on potential disease cases related to diacetyl and food flavorings containing diacetyl.
A few stakeholders noted that material safety data sheets (MSDSs) frequently do not include detailed information on flavorings because the flavoring constituents are proprietary. Some smaller food manufacturers, therefore, might not be aware of exposures concerns. OSHA representatives noted that the hazard communication guidance was published in order to alert such small companies of the issue.
Some stakeholders emphasized the need to use a broad-based approach in evaluating all the chemicals used in flavorings, not merely focusing on diacetyl, and to study how workers are actually exposed.

4.4 OSHA plans to conduct site visits through contractors to help OSHA better understand the processes involved with use of flavorings containing diacetyl and how workers are exposed in various industries and operations. OSHA encouraged suggestions and support from stakeholders to facilitate these site visits. Stakeholders provided the following comments, questions and suggestions regarding OSHA site visits to establishments that use diacetyl or flavorings containing diacetyl:

Several individual stakeholders thought that site visits may be facilitated in several ways: 1) by employing a voluntary, collaborative model, such as the one used by Cal/OSHA, to involve almost all the flavor manufacturers; 2) by using the OSHA enforcement and consultation tools and NIOSH capabilities as part of an overall strategy to obtain information from food manufacturers; 3) by following the model of an existing industry effort to encourage medical surveillance programs in flavor industry facilities (which has brought medical and health and safety professionals into facilities) and expanding the effort to include flavor manufacturers' food industry customers; 4) by using OSHA's regulatory authority to access employee health and exposure information; and 5) by using other existing OSHA collaborative agreements like VPP to contact members who will be more receptive to OSHA. Some stakeholders also suggested publicizing on OSHA's website the names of companies that refuse to provide OSHA with site visit opportunities.
OSHA representatives emphasized past successes through voluntary cooperation with sites. OSHA hopes to use a similar approach in its site visits rather than using an authority-driven forceful inspection approach. Some stakeholders believed that forceful entry may be a more effective means of visiting a greater number of food plants, since food manufacturing industries use flavorings in the same way every day, which minimizes the need for scheduling visits in advance. Food manufacturing plants are typically harder to enter because they are only 5 to 10 percent unionized. Cooperative site visits are likely to work better with the popcorn and flavoring industries.
Some stakeholders expressed concern that OSHA's data gathering inquiries were unfocused. One stakeholder summarized these concerns, urging better organization and specific identification of which operations would be observed, what flavorings would be targeted, what type of sampling would be performed, and how OSHA intended to use the information. This type of information would be helpful in encouraging members of trade organizations to cooperate more readily. The stakeholder was willing to assist OSHA and its contractor in focusing the site visit protocol.
Some stakeholders asked whether OSHA would be willing to share data with industry prior to development of the standard so that the data can be used by companies to proactively develop protections for workers.
It was also suggested that company medical records (e.g., spirometry) be reviewed as part of site visits. However some obstacles to using such data include (1) most of the lung function (spirometry test) data are not of particularly high quality, although they might have some value in identifying patterns, (2) the disease has a short latency period, and (3) broad-based and well thought out case definitions are needed.

B. Medical Screening and Surveillance

OSHA solicited information on the scope and frequency of medical surveillance, the content of a medical evaluation, and the potential actions that would be triggered by medical screening findings.

4.5 Which employees should be included in the medical surveillance program? Stakeholders provided the following comments and suggestions regarding the types of workers who should be included in a medical surveillance program:

In the flavor manufacturing industry, it was suggested that several groups of employees would be potential candidates for medical surveillance. These include (1) workers who mix and blend liquids and powders who will be exposed to pure flavoring chemicals; (2) workers who help in packaging flavoring products, particularly powdered flavorings; (3) workers who clean vessels; (4) quality assurance technicians and laboratory flavorists who design flavors, although their exposure might be lower than the production workers who blend or package large batches of these products; and (5) office and clerical staff who work in production areas. All these employees should be in the medical surveillance program.
In industries that receive products from the flavoring industry, monitoring workers involved in unpacking the flavoring powders was suggested.
Other groups of workers who may not actively handle flavorings, but nevertheless are exposed because their jobs bring them into contact with diacetyl products, include supervisors and maintenance workers.
It was noted that while some facilities may have as many as 50 to 75 workers; others may only have 2 workers.
A concern was raised about the large number of temporary or contract workers who work in the food manufacturing industry and are often treated differently with respect to respiratory protection and medical evaluation. These workers might be part of the vulnerable population who are potentially exposed but may not receive appropriate protection or surveillance. Some stakeholders thought that these workers should also be included in medical surveillance programs because they are exposed as well and the disease latency is short.

4.6 Which employees should undergo a medical evaluation? Should medical surveillance be tied to a particular action level? Should all employees who work in a particular process be part of the medical surveillance program? What criteria should be used to identify effects of concern? Stakeholders provided the following comments and suggestions on possible triggers for placing an employee in a medical surveillance program:

Many stakeholders agreed that medical surveillance should not be tied to a specific action level but rather to job categories, tasks, potentials for exposure, and disease progression. The frequency of testing should be based on exposure potential and not length of time exposed.
Many stakeholders also thought that the rapid progression and short latency of the disease should be taken into consideration when determining the need for medical surveillance.
One stakeholder cautioned against using time-weighted averages (TWAs) as a trigger for joining a medical surveillance program, citing short, high exposures to flavorings containing diacetyl that are suspected of causing illness.
In California, all employees who work in a particular job process are being placed in the medical surveillance program. Potential for exposure determines placement in the medical screening program.
Some stakeholders pointed to the need for developing a case definition, while others cautioned against using a specific case definition due to the wide range of disease manifestations.
One stakeholder noted that Cal/OSHA is developing guidance on medical surveillance for flavoring-related lung disease to be used by physicians.
In using a case definition, a few stakeholders thought that certain red flags could trigger placement in a medical surveillance program. These red flags include (1) severe airway obstruction (most non-smokers do not get obstructive lung disease), (2) a greater than 10 percent decline in lung function over a 6-month period, excluding individuals with genetic diseases or asthma, and (3) evidence of fixed airway obstruction, especially if flavoring exposure is certain. The type of work and past or present potential for exposure (e.g., a worker who previously worked in the production room who moved to the warehouse would have had a past exposure) should also be considered in the need for medical screening. Some stakeholders thought that all workers should be screened if the employer did not have sound knowledge of the processes and locations of potential exposure.
Another stakeholder suggestion is to employ an epidemiologic concentric circle model to identify the employees that should be placed in the medical surveillance program. Rather than a blanket, all-employee approach, employers should target those with greatest potential for exposure, but err on the side of caution. Under this model, the most exposed employees would be provided medical surveillance and other employees working in the same area would also be candidates for surveillance. Aerosolization of diacetyl should be considered in identifying the target population.

4.7 Is a baseline followed by a 6-month screening frequency appropriate? How should challenges with respect to testing at 6-month intervals be handled for temporary or contract workers?

Several stakeholders agreed that due to the rapid decline and short latency issues associated with bronchiolitis obliterans, screening should be performed at 6-month intervals. In the case of a plant or job operation with a known case of disease, stakeholders suggested that the screening frequency could be increased to less than 6 months. Stakeholders also noted that asymptomatic individuals could also manifest decreased lung function.
Stakeholders acknowledged that temporary workers pose a special challenge to the 6-month screening rule. It was suggested that if temporary workers cannot be screened they should not work in processes with exposure potential.
Another stakeholder suggestion was that the frequency of screening would vary depending on the worker's exposure potential. For workers with greater exposures or exposure potential, screening frequency should be more often than every 6 months. For workers with lowered exposure potential the screening frequency could be less frequent than every 6 months.
Cases of lung disease have occurred in workers of various ages, including young workers, and stakeholders cautioned against using age-based criteria for determining evaluation frequency.

4.8 What types of tests should be included in the medical screening protocol? Should certain tests be required for all workers and additional tests for certain workers? Should a risk-based approach to spirometry testing be used?

One stakeholder suggested that a screening protocol include a physical exam, occupational history, and spirometry. A more complex test, like a CT scan, could be used for follow-up testing, but not for screening. The use of bronchodilators should also be considered in spirometry testing.
The Cal/DHS has already screened 800 people and collected data that could be used in developing definitions and medical protocols. Stakeholders noted that using these data could greatly reduce the number of years required to develop such information as part of the OSHA process.
Stakeholders emphasized the difficulty in limiting the testing protocol to a single complement of tests. The general consensus was that as more information about the exposures and disease becomes available over time, the testing protocol may have to be adjusted. A very highly defined medical screening protocol may not be valid in another 10 years.
Stakeholders noted that strong education and training components to medical testing exist. Considering language issues and ensuring that employees understand the importance of their spirometry testing is critical. Employers should also understand the importance of removing employees from exposures and ordering additional testing.

4.9 What qualified personnel (e.g., a clinician) should an employer have on hand to help in the screening process?

Many stakeholders felt most occupational health clinicians should know what key parameters to monitor and can rely on guidance available from medical sources. Following-up with employers to ensure that they understand the surveillance and screening processes is also important.
Many stakeholders emphasized the need for spirometry to be conducted by well-trained technicians. Test and results can vary by employee effort and technician ability.

4.10 Are uniform questionnaires and spirometry testing important to ascertain quality of the screening program?

Many stakeholders agreed that the spirometry quality is very important and spirometry testing is typically fraught with quality problems. One stakeholder suggestion was to stipulate criteria for spirometry testing in the standard, including the need to follow American Thoracic Society (ATS) criteria. Stakeholders also suggested that technicians receive NIOSH spirometry training, and that employers be required to periodically review the quality of the technician's testing. Saving a copy of the spirometry tracing, not just the numerical results, is also important. One stakeholder recommended it would be more effective for OSHA to suggest technicians receive formal spirometry training but not mandate training in order to achieve greater compliance. Conservative criteria could be used for screening purposes and borderline findings could be followed up with additional evaluation, suggested stakeholders.
A few stakeholders urged OSHA to use validated occupational and medical history questionnaires that exist and are in use, instead of developing a new questionnaire. OSHA can provide sample questions but specific questions regarding work history will vary by industry.
Other stakeholder concerns raised during the discussion included whether or not the standard should address spirometry equipment quality. The majority of equipment comes with American Thoracic Society (ATS) standard software. If the ATS software is routinely updated and the spirometry technician prints out the tracing and result, equipment quality need not be an issue.

4.11 What actions should be taken if a worker has significant findings?

Stakeholders suggested looking at other regulatory models, such as the lead or noise standard, to determine the types of follow-up actions that may be appropriate following identification of significant medical findings. The type of follow-up action should depend on the medical finding. For example, the workers may need to be moved to another task to reduce exposure.

4.12 What should be used for referral criteria?

Many stakeholders emphasized the lack of a case definition. Particular concerns include the large variability in, and limited knowledge currently available on the range of disease manifestations. For example, it is unclear whether health effects due to chronic low level exposures differ from the effects of intermittent peaks of high level exposure.
A few stakeholders noted that relating follow-up actions to exposure and clinical findings is important. One stakeholder suggestion is that the combination of a 10 percent decrease in lung function, fixed airway obstruction, and an exposure pattern that raises a red flag would lead to additional actions.

4.13 OSHA requested that stakeholders submit any available written medical guidelines to OSHA.

One stakeholder agreed to provide written guidelines to OSHA. These include guidelines published in the Journal of Occupational and Environmental Medicine (JOEM) and the ATS criteria on acceptability and reproducibility of spirometry.
The medical screening guidance that is being developed as part of the evaluation of hundreds of workers in California will also be shared with OSHA when ready.

C. Exposure Assessment

OSHA solicited information on the availability of exposure data, any issues in sampling and analytical methodologies, and possible exposure assessment strategies.

4.14 What are the main methods available for use in exposure assessment? Should OSHA and NIOSH cooperate on sampling methods? How can OSHA and NIOSH avoid duplication of efforts? How are humidity effect issues in the NIOSH method being addressed?

Stakeholders and OSHA noted three sampling and analytical methods available for diacetyl:

The NIOSH method (published in 2004), which uses a carbon molecular sieve sorbent tube (150/75 milligrams (mg) sorbent bed size), does not have breakthrough issues, allows a single tube to be used to sample throughout a whole shift, and offers a relatively low limit of detection.

Specific comments and concerns noted with respect to the NIOSH method include:
- The method collection efficiency has been found to be influenced by humidity. Currently, the method is being evaluated in laboratory, field, and chamber studies to characterize the nature and magnitude of diacetyl loss that occurs in humid environments. Additional information is available on the NIOSH website http://www.cdc.gov/niosh/topics/flavorings/. NIOSH no longer considers this the preferred method for diacetyl sampling until the humidity issues have been resolved. The issue of humidity was not previously noticed because humidity effects typically cause breakthrough rather than compound loss on the collection media.
- The chamber evaluation studies on the NIOSH method will confirm the degree to which humidity affects the results and may lead to a correction factor (if necessary). This will be applied to older sampling data that were collected using this method. NIOSH noted that most sampling data collected at popcorn plants were collected during low relative humidity periods. Thus, the correction factor may be minimal or not required.
- One stakeholder noted that relative humidity may also be an issue for dry encapsulated wetted flavorings, which release diacetyl only when moist.
- The OSHA method (PV 2118) uses two silica gel sorbent tubes in series. Each tube is of the150/75 mg sorbent bed size. A sampling air flow rate of 0.05 liters per minute and maximum sample volume of 3 liters of air is recommended for this method. Using this protocol, sorbent tubes must be changed every hour and 16 tubes are required to sample an 8-hour work shift. The OSHA method is not affected by humidity (designed for use at 80 percent relative humidity) and is relatively stable under standard sample storage conditions.
Specific comments and concerns noted with respect to the OSHA method include:
- The OSHA method is currently a partially validated method with a limit of quantification in the range of 0.3 ppm. The OSHA method has a higher detection limit than the NIOSH method, in part because of the smaller sample air volume allowed by the OSHA method. The method is in the process of being fully validated and OSHA anticipates in the eventual fully validated method there might be some improvement in both the limit of detection and the maximum air volume permitted for the fully validated method.
- A modification to the OSHA method involves substituting silica gel tubes of a larger size (400/200 mg sorbent bed), which allows a longer sampling time and larger sample air volume. NIOSH is using this modified method currently.
Specific comments and concerns noted with respect to the OSHA Modified method include:
- A few stakeholders noted one potential problem with the larger sorbent tubes used in the modified OSHA method. In some cases, the motion of a vigorously active flavor worker can cause this longer sampling train to accidentally disconnect. OSHA representatives noted that the tubes can be connected in a series using methods that minimize this problem. A simpler sample train is also being considered.
- NIOSH cautioned against having any media to remove moisture upstream of the sample.
- Two different sampling systems may be required for volatile liquid and particulate samples, regardless of whether the current NIOSH or OSHA methods are used.

In general, stakeholders supported both agencies working together on developing sampling methods and inquired if any results were available from a side-by-side comparison of the OSHA and NIOSH methods. This information will be available once the NIOSH method has been evaluated. Stakeholders urged OSHA and NIOSH to provide information on method detection limits and noted that a lot of task-based sampling is conducted in the field.

4.15 What are other methodologies to measure episodic exposures?

One stakeholder suggested that neither the NIOSH nor the OSHA methods were effective for episodic exposures. They suggested that the EPA TEO 17 method, based on thermal desorption, had a lower detection limit than the OSHA method, and was a better alternative.

4.16 What real-time instruments can be used for monitoring?

Stakeholders provided the following suggestions and concerns related to real-time instruments for monitoring:

OSHA identified photo-ionization detectors (PIDs) that retail for $20,000 and use electron capture and gas chromatography. These instruments require an experienced operator. Lower cost PIDs can be used for screening purposes.
The PID may be used to conduct onsite measurements. Stakeholders noted that flavor manufacturing typically contains approximately ten batch operations, composed of 15 or more constituents. Of these, there are generally four to five compounds that are more volatile and present at higher quantities than diacetyl or acetoin. Because of the complex mixtures, limitations to using PIDs in the field may exist.
In response to a stakeholder question about real-time instrument function in a sampling environment that contains complex mixtures like diacetyl-containing flavorings, OSHA representatives noted that the potential for interference exists. An air sample collected on a sorbent tube may need to be sent to the laboratory for additional testing. Although OSHA prefers the specificity offered by gas chromatography, real-time instruments are good screening tools.
A few stakeholders noted that the National Jewish Medical Center and NIOSH are also using Fourier Transfer Infra Red Detection (FTIR) in the popcorn and flavoring industries. However, OSHA does not have experience with this method currently. This computer-based, direct reading monitoring method works by relying on the selectivity of infrared radiation detection to look at a molecule. This method does have potential for interference from other compounds or bands in the spectrum and needs to be used carefully.
One stakeholder thought that more cost-effective screening tools are available in place of FTIR.
It was stated that a private firm is in the process of developing an instrument to sample diacetyl and plans to release the instrument in November 2007.

4.17 What should be the frequency of sampling? How can OSHA describe a task-based sampling strategy in a regulation?

Some stakeholders noted that, in the flavoring industry, sampling ideally occurs when a flavoring that contains diacetyl is being produced. A flavor of interest may only be produced once per month. Hence monitoring requires a great deal of coordination to accommodate this high degree of variability.
In contrast, in food manufacturing where the same product may be produced every day, it was suggested that exposure levels can be fairly constant from day to day for a given task and product type. Monitoring would need to be repeated only with a process change.
Monitoring should be left to the employer with guidance from OSHA and NIOSH.
One stakeholder noted the need to know the levels of concern, such as the permissible exposure limit (PEL), in order to make effective use of air monitoring.
Another stakeholder noted that task-based sampling by NIOSH, the National Jewish Medical Center, and Cal/OSHA shows that higher exposures peaks occur during certain short-term operations. Conversely, a longer full-shift sample reflects the average exposure over the period sampled. The longer sample period is also valuable for comparing sampling methods.
Stakeholders noted the need to think about how best to describe task-based sampling strategies in regulatory language.

4.18 What is the best way to keep stakeholders involved in the exposure assessment process?

OSHA plans to post new methods at the web site www.osha.gov.
A few stakeholders requested that OSHA and NIOSH describe the relative advantages and disadvantages of each method for use in the field.
Some stakeholders also requested that OSHA establish a docket where information relating to the diacetyl rulemaking process can be received from the public and consolidated in a transparent way.

D. Control Methods

OSHA sought information on possible substitutes for diacetyl in diacetyl-containing products, process isolation techniques, use of powders versus liquids, work practice controls, and respiratory protection issues.

4.19 Do substitutions eliminate the problem of diacetyl?

A few stakeholders noted that diacetyl is the single substance most responsible for butter flavor. Substitutes for diacetyl (e.g., diacetyl trimers) exist, but information is lacking about these substitutes and their safety.
A lot of work is being conducted to find a substitute for diacetyl. Replacement compounds are expected to have a similar structure in order to mimic the flavor perception.

4.20 What criteria might OSHA use to guide companies on substitutions?

Stakeholders provided the following comments and recommendations on topics to address regarding substitutions:

Substitute flavoring components are also chemicals and should be carefully evaluated. However, in the absence of concrete information on the identity, health effects, or analytical methods for substitute chemicals, OSHA may not be able to advise users on how best to address them individually. General exposure controls and medical surveillance recommendations for flavorings as a substance class could be more beneficial.
One stakeholder noted that very few food flavoring chemicals are tested for inhalation hazard.
Limited information on MSDSs, due to the proprietary trade secret nature of substitutions, will continue to be an issue.
A few stakeholders noted that limited information exists on the use of diacetyl in food manufacturing. Substitutes will have similar issues. A standard on food flavorings containing diacetyl may help alleviate some of this concern.

4.21 What control mechanisms are used to reduce diacetyl exposure?

Stakeholders provided the following comments on and concerns with controls used to protect employees from diacetyl exposure:

Standard exhaust ventilation designs are the basis for engineering controls in the blending operations of flavoring powder manufacture. However this type of control still requires refinement, such as determining the level to which the air flow rate (Q) should be set.
PPE also continues to be a standard control in the popcorn and flavoring industries.
Examples were provided of other control methods, such as using cold water to clean vessels and minimize volatilization, safe handling methods, and safe pouring techniques. In general, common sense solutions greatly reduce exposure.
Trends are less well defined among the controls used in the food manufacturing industry, which is diverse.
The type of control methods that appear to be effective for managing exposure to diacetyl and flavorings containing diacetyl need not be sophisticated, but employee education, awareness, and training are critical to their effective implementation.
Flavoring manufacturing facilities vary in terms of existing controls. Some facilities are new and have controls in place, while other facilities are much older and may not have fully installed controls. Only about 100 to150 facilities exist nationwide.
Small companies are looking for expert guidance in using standard controls efficiently.
The effectiveness of exposure controls for flavor handling is still being determined.

4.22 What types of respirators are used? Are there any concerns associated with respirator use?

Stakeholders provided the following comments and concerns related to respirator use:

Full-faced respirators have been used since 2000 and are often configured as air supplying respirators.
There is a need to develop engineering controls that would eliminate reliance on respirators. In the meantime, respirators can protect workers, especially if the company's management is serious about enforcing use.
Employee training and willingness to properly use respirators are essential.
A full-face respirator can restrict essential verbal communication between employees.
Respirator programs need to be effectively managed, with special attention to the area of employee medical clearance to wear a respirator.

4.23 Can companies provide information on the cost of implementing control methods to OSHA?

A few stakeholders expressed a willingness to provide OSHA with cost information. Case studies can be assembled to demonstrate costs of controls in the past 5 years.
Two stakeholders pointed to the need for guidance on the exposure level to which diacetyl should be controlled. This level influences the design of controls and the associated installation and usage costs. A stakeholder added that new ventilation systems are being designed to achieve the lowest technically feasible exposure levels and this strategy seems to be effective.

4.24 Is there value in a standard without a Permissible Exposure Limit (PEL), but which includes medical surveillance requirements and engineering controls? What are some advantages and disadvantages of setting a PEL? How would companies know they had done enough without a PEL?

Stakeholders provided the following comments and concerns related to development of a PEL for diacetyl:

Stakeholders emphasized the need to develop a standard based on the current knowledge in order to move forward expeditiously.
It was generally agreed that a PEL based on a safe level of exposure would require further research, but the need exists to quickly develop a standard that relies on currently available information. An interim standard could be based on engineering controls, while additional research for a PEL is completed. It was noted that some companies lacked the most basic controls, so an interim standard in the absence of a PEL would have great benefit.
Other stakeholders suggested that, in addition to engineering controls, the first step could include a standard focused also on exposure assessment, medical surveillance, employee education, and respiratory protection., all geared toward the microwave popcorn and flavorings manufacturing industries. OSHA's influence on these industries would likely stimulate action in those areas by other industries, thus benefiting a broader workforce.
However, some stakeholders cautioned OSHA against taking a long time (e.g., 5 years) to develop a standard for microwave popcorn and flavor manufacturing and then more time (e.g., another 5 years) to develop a food manufacturing industry standard. The two-step process was proposed as a means to expedite the rulemaking process and to protect employees.
Some stakeholders expressed a strong preference for having a PEL in the standard in order to provide target criteria for engineers to work towards in implementing controls, like local exhaust ventilation. Without a PEL to guide what constitutes an acceptable level, employer will need to continue enforcing use of respirators.
Some stakeholders cautioned against assuming that a PEL afforded the best protection for employees. If a PEL is based on the lowest limit detectable, that may not necessarily protect against health effects.
It was noted that deriving a PEL for a complex mixture like diacetyl-containing flavoring substances is a difficult task.
Issues with monitoring methods could further complicate PEL development.
There should be a high priority effort to encourage the necessary research that will allow determination of a credible PEL.
Stakeholders noted that while a standard focused solely on diacetyl would be useful, a performance standard could address other chemicals.
OSHA representatives noted that OSHA must show material impairment of health and significant risk (1/1000 risk of fatality or serious illness) to develop a health-based standard. Without such a rigorous approach, the standard runs the risk of being challenged in court.

4.25 Would a standard focused exclusively on diacetyl be effective?

Some stakeholders thought that controlling for diacetyl would probably also limit most potentially hazardous exposures to other flavoring components, but this depends on the products manufactured.
It was suggested that total volatile organic chemicals (VOCs) should be considered in addition to diacetyl.
Some stakeholders expressed concern about targeting one compound (diacetyl) out of the 2,000 potential compounds in the volatile flavorings industry.
Other stakeholders cautioned that a standard for just diacetyl could miss other potentially hazardous chemical exposures from flavor ingredients. While one stakeholder reported being unaware of any flavorings-related lung disease in the absence of diacetyl, another stakeholder suggested other flavoring components may have a role in disease progression; hence, other chemicals besides diacetyl should be considered in a standard.
Members of the food manufacturing industry often find it difficult to determine the presence and, if present, the amount of diacetyl in the food flavorings they use. A standard that encourages control of the wider group of flavoring ingredients would be more beneficial under these circumstances.

5 Closing Remarks

OSHA representatives thanked the stakeholders for their participation. To further assist OSHA in developing the standard, OSHA representatives asked stakeholders to invite OSHA to visit their facilities, if possible, as part of the information gathering process.

* * * *