B. Written Hazard Communication Program

2.



3.



4.



5.



6.



be careful while in the laboratory. If done, providing this information should be done

carefully to avoid frightening an individual who probably has very little technical

knowledge.

A list of hazardous chemicals in the workplace must be compiled. The provisions in the

standard only require laboratories keep track of incoming chemicals subsequent to the

effective date of the standard for employers, May 25, 1986. Existing inventories were in

a sense “grandfathered.” Eventually, as older stocks are disposed of or used, the list will

come to reflect the actual holdings in a facility. The list should be kept as current as

possible. If one person is assigned the responsibility to maintain the list, and the data

kept in a personal computer data base, it is only necessary to keep track of additions and

deletions in order to maintain a complete, current list. For the purposes of complying with

this portion of the standard, the quantities of each chemical in the laboratory are not

needed, although these data would be important for a sound management program and

would be helpful in planning a safety program. The list of chemicals, in combination with

the list of employees, will serve to help define the training program.

The written program must define how the employees are to be informed of the requirements of the standard. This will include: details of (1) how the employees are to be

informed about the contents of the standard; (2) the contents of the written plan; (3) how

they are to meet the labeling requirements; (4) how they are to learn of the methods

available to them to warn them of exposures; (5) how to obtain and interpret a MSDS for

a given chemical; (6) the hazards associated with the chemicals to which they are exposed;

(7) how they are to be trained in procedures which will eliminate or reduce these chemical

hazards; and (8) how they are to react in an emergency.

Many laboratory uses of chemicals involve repetitive tasks, while others do not.

Employees must be made aware of the risks associated with the latter type of activities

as well as those accompanying the more routine uses of chemicals, and the same basic

type of information provided as in item 3.

Although pipes are not considered containers for the purpose of this standard and need

not be labeled, the plan must include education of employees about the hazards

associated with any unlabeled pipes containing chemicals in their work area and how to

deal with these hazards.

There must be a procedure or statement in the plan as to how transient employees, such

as persons working on contract, are to be informed of the chemical hazards to which they

may be exposed, and for provision of information of protective measures for these

transient employees. It is not specifically spelled out in the standard but there is a need

for the converse as well. Contractors are often called in to do renovations, perform an

asbestos abatement project, or to conduct a pest control program, as examples, and use

hazardous chemicals in the process or expose personnel to airborne hazards. Provision

should be made in the contracts for these groups for them to provide information to the

occupants of the spaces where their work is being done.



1. Personnel Lists

This appears to be relatively straightforward, but in fact can be rather complicated. In a large

academic institution, the actual duties associated with a given job classification often become

blurred over a period of time. For example, a job title of laboratory technician might appear to

logically relate to chemical exposure, but the duties of the individual may have changed so that

the job may never bring the individual into contact with chemicals at all. It is not possible to

simply have the personnel department list all persons in specific job classifications as professional staff or faculty in research areas.

As an initial step to determine which employees need to participate in a formal hazard



©2000 CRC Press LLC



communication program, a questionnaire can be sent to each department or other internal division

asking them (1) to define those areas in which chemicals are used in their department, (2) to list

each employee in those areas with their job title, and (3) for their appraisal of the involvement

of these individuals with chemicals. This should be followed up with a second questionnaire to

the managers of the individual areas, asking for the same information, and then the area should

be visited to confirm the data provided. This sounds unnecessarily involved, but experience has

shown that all three steps are necessary. In many cases, through oversights, individuals are not

identified who should have been included and, occasionally someone is listed who has no

exposure, usually because it was easier to list everyone rather than consider each individual case.

It is important to identify the position (most positions now have internal identification codes)

so that when the position becomes vacant, a mechanism can be established to ensure that the

new person filling the position receives a proper orientation program. Often, such a position

might qualify for participation in a pre-employment medical screening examination as well so the

effort to correlate positions with exposure to chemical hazards might be justified for more than

one purpose.

All persons being considered for positions covered by the hazard communication program

should receive a brief written statement concerning the program so that they may ask appropriate

questions at the time of their job interview. If they are selected, a more extensive document should

be provided so that they will be aware of the explicit requirements of the standard.

2. Chemical List

A list of all hazardous chemicals in the workplace is required as part of the standard. It may

be difficult to convince many managers to take the time to go through their stocks of chemicals

to prepare a list for their facilities. (The need for this information is implied in the laboratory safety

standard, but not required.) This is one of the more burdensome tasks associated with the

standard. However, it is very desirable that an effort be made. Not only does it provide

information on which to base the training program, but it also is needed to prepare an MSDS file

for the facility, although, again, the MSDS file is only required for incoming chemicals. In practice,

however, it is difficult to justify not having an MSDS for a hazardous chemical in use, based on

a technicality. There is no real alternative for the initial survey as a basis for defining the scope

of the program.

It would be desirable if the problem of maintaining the list of incoming chemicals could be

centralized, perhaps as the purchase order is being processed. However, although surprisingly

few chemicals are bought in quantity in most research institutions, even very large ones, there

may be more than 1000 different substances bought during the course of a single year and several

thousand purchased over a number of years. Many of these are bought under a number of

synonyms, or as components of brand-name formulations. Commercial software is available by

which a chemical can be identified by any number of synonyms, standard chemical name, trade

names, or CAS numbers. Local information as to the purchaser and destination (building or

facility) can be provided by the customer. Software is now available which can combine this

information and more, e.g., date received and quantity, and to generate a unique bar code that

can be affixed to each container. These data can be used to maintain a continuing inventory for

an entire organization and to track a chemical from the time of receipt to eventual full consumption

or disposal. It is practical for individual laboratories to use microcomputers, even without these

specialized programs to perform this task for themselves, using commercial database or

s preadsheet programs. At most it would require a week or so, per laboratory, by an employee

to acquire the initial data and enter it, even if existing inventories were included. Maintenance

of the data would involve only adding new containers and removing old ones.

3. Labeling

Current labels on original containers of chemicals as purchased from the distributor or



©2000 CRC Press LLC



manufacturer will almost certainly meet and exceed the requirements of the hazard communication

standard (see Chapter 4, Section V.B). These requirements are:

1. The identity of the hazardous chemical

2. Appropriate hazard warnings

3. The name and address of the chemical manufacturer, importer, distributor, or other

responsible party

Item 2 is the only ambiguous requirement. Most commercially sold chemicals provide this

sort of information on the label in a number of ways, such as:

1.

2.

3.

4.



A risk descriptor, i.e., Danger, Warning, Caution

The NFPA hazard diamond

A descriptive statement of the hazards

By use of stylized symbols, such as a radiation or biohazard symbol



Other useful safety-related information is normally provided as well, such as the flash-point

(if applicable), fire extinguisher type (if applicable), first aid and medical advice, a color code to

aid in avoiding incompatible storage, and standard identifiers, such as a CAS number which can

aid in referring to a MSDS data base, and a UN number which is needed in disposing of the

chemical as a hazardous waste.

Facilities are specifically enjoined by the terms of the standard from removing or defacing

the labels on incoming containers of chemicals. However, it is relatively common to transfer a

portion of the contents from the original container to a secondary container. If this material

remains under the control of the individual responsible for the transfer, and is to be used during

a single work session, then it is not necessary to label the secondary container. If it is not to be

used under these conditions, then the secondary container must be marked with the identity of

the chemical(s) in the container and with “appropriate” hazard warnings for the protection of the

employee. These “appropriate” warnings need not be as comprehensive as the original label, but

must provide adequate safety information.

The most likely occasions when secondary containers are used without proper labeling would

be when chemicals are disbursed from a larger container into a smaller one at a central stockroom,

and when containers are to be taken from the initial workplace into the field. Personnel must be

sure to label the secondary containers in these cases and in any other comparable situation. If

secondary containers are labeled properly, it also will help remedy one of the more troublesome

problems associated with hazardous waste disposal, inadequately identified containers of

chemicals. Once in the laboratory, the tendency is to label the secondary containers less

thoroughly, often with a cryptic label such as “soln. A” or some other non-informative label.

The warning labels must be in English, although they may be provided in other languages

in addition, if appropriate. In many academic institutions, in particular, graduate students who

routinely use a language other than English as a primary language, are becoming numerous, and,

in some cases, consideration may be given to supplementing the commercial label with warnings

in other languages. However, the majority of these graduate students can be expected to

understand written English satisfactorily. Some areas of the U.S. have changed demographically

so that the use of languages other than English may have become predominant. All employees

using chemicals must be instructed in how to interpret the hazard information on the labels.

A specific part of the written plan must address how the employees are to be made aware of

the labeling requirements and how they are expected to comply with this standard. It would be

highly desirable to develop a uniform program across an organization, particularly as to labeling

of secondary containers, to avoid unnecessary confusion.



©2000 CRC Press LLC



4. Material Safety Data Sheets

Since the receipt of MSDSs is tied so strongly to the purchase and receipt of chemicals, they

were discussed in some detail in Chapter 4, Section III.C. The exact form of a MSDS is not

mandated by the standard as long as the proper information is provided. Firms use a variety of

formats to provide the required information. ANSI and the CMA have recommended a standard

form which may be adopted.

There are two basic requirements associated with MSDSs in the hazard communication standard. Employees must be trained in how to use the information in them and the MSDSs must be

readily available to the employees.

As discussed in Section III.C., a major problem in a research institution, where the chemical

users may operate virtually independently of each other and are likely to be housed in a number

of different buildings, is to ensure that all users of a given chemical have ready access to a copy

of the most recent version of the MSDS. The distributor is only required to send one copy of

a MSDS to a purchaser and an updated version when a revision is necessary because of new

information. Where several different components of an organization order independently, one

purchaser may receive an update while the others do not, since the vendor technically has

fulfilled its obligation by sending the MSDS to the first unit making the purchase. If a centralized

mechanism for tracking chemical purchases has been established, then all MSDSs could be sent

to a single location from which copies can be forwarded to all groups within the organization that

need them. This is relatively labor intensive and still may not reach all users since chemicals may

be transferred from one laboratory to another with no paper trail. Another alternative, which does

not provide as ready access to all users but does not require the tracking mechanism referred to

above, would be to have all MSDSs received at one location and maintained in a master file, with

copies placed in several secondary master files at locations reasonably convenient to the users.

A third alternative, but still less accessible to users, would be for a single master file to exist, with

copies, of individual MSDSs provided upon request. This might not be considered to meet the

accessibility requirement if the delay in receiving the MSDS is more than 1 or 2 working days.

All of these mechanisms are at best cumbersome and manpower intensive.

Comprehensive generic MSDSs are now commercially available on optical discs which can

be processed by a computer and accessed at any time by the users. These typically are updated

quarterly so that they can satisfy the need for the MSDS file to remain current. They are not

inexpensive, but the cost is much less than for the amount of manpower needed to maintain an

equivalent hard copy file, and they provide a comparable level of access. There also are firms

which maintain computerized MSDS files available to subscribers as a database service. These

are accessed by the users from their terminals using modems, but line and access charges are

incurred.

As noted earlier, MSDSs are widely available on the Internet, either directly from the chemical

manufacturer or distributor (which meets the criteria of directly identifying the chemical supplier

producing the MSDS) or many organizations and universities now maintain and pro-vide generic

MSDSs at their Internet site.

No matter how a facility or organization sets up a MSDS file, a component of the training

program for the employees must include an explanation of how an individual employee can obtain

access to the file. It should be possible for an employee to obtain copies of a MSDS for a given

chemical upon request. In addition to providing access to the MSDS file, part of the program

training must also include instruction in how to interpret a MSDS to obtain appropriate hazard

information.

The information presented in the various categories in a MSDS should pose no real difficulty

to most technically trained persons. Some definitions of terms may need to be provided, such

as LD 5 0 (lethal dose, 50% of the time for the test species), if an individual is not accustomed to



©2000 CRC Press LLC



using such terms, but even these are straightforward. However, some persons will not be as

scientifically sophisticated, and the training program for these individuals will need to be more

thorough. In an instance at the editor's institution, grasping the distinction between a monomer

and a polymer was a major problem for some clerical personnel who felt that they had been

exposed to dangerous levels of a chemical due to some activities in the building where they

worked. The health hazard data and the TLV values given in the MSDS for the chemical stated

that they were for the monomer only and that the effects of exposure of the chemical would be

serious at a few parts per billion in air. The employees exposure was not to the monomer, but to

very small quantities of the stable polymer, for which the health hazards were minimal. Extensive

(and expensive) tests had to be run before the personnel were convinced (some perhaps

continued to have doubts) that they had not been unduly exposed. During training, an effort

needs to be made to ensure that understanding has been achieved.

Compliance with the training requirements for utilization of a MSDS as a source of hazard

information can be readily achieved for technically trained personnel. For example, a written

handout informing the employees how they can obtain access to a needed MSDS and a short

video tape explaining the contents might be all that is needed. An individual capable of explaining

any confusing points should administer such a program and be available to answer questions.

A statement affirming that the training was received should be signed (and dated) by the

employee after any questions were resolved. This can have significant legal implications. An

employee may maintain that they were never exposed to the information but a signed statement

that they were present at a lecture is hard to refute. For less knowledgeable employees, a formal

training session should be set up and an instructor-student format used. The handout and

videotape mentioned above can still be part of the instruction program, but the instructor should

go over each of the categories in a MSDS and encourage the employee to ask questions. The

employee also should be provided with a written version of the concepts covered, for later

reference.

Some organizations document that an employee has not only been exposed to the information,

but also understands it, by requiring that each employee take a very simple written quiz on the

covered material, in place of signing a simple statement. This is not required to comply with the

standard, but it does provide stronger documentation of an effective training program. Individuals

should not resent imposition of such a requirement, but some professionals feel, rightly or

wrongly, that they have demonstrated sufficient proficiency in their area by fulfilling the required

educational and professional certifications. If a quiz is made part of a program, it should be

expected that a number of individuals will object to taking it. This problem occurs mostly among

highly educated staff. Such a program, though, can be made to work with patience and support

of the organization.

5. Employee Training and Information

Portions of the training program common to every employee: The basic concepts of the

organization's program, how it is administered, the requirements of the standard, and how to read

and understand information labels and MSDSs, have already been covered in the earlier sections.

These can be given by the lead department in the organizations hazard communication program,

usually the Environmental Health and Safety Department. However, a number of other areas will

require a cooperative effort between the administrative department in overall charge of the

program and the individual facilities and departments. In an organization which uses chemicals

in a wide variety of activities, the local managers will need to be the primary parties responsible

for providing much of the required training relating to operations specific to their program and

facility.

The following topics need to be covered routinely in a training program to comply with the



©2000 CRC Press LLC



standard, in addition to those already discussed:

1. The physical and health effects of the specific hazardous chemicals which the employees

may use or to which they may be exposed.

2. Means to detect the presence of toxic materials in the workplace. This should include

means directly available to the employee, such as odor, presence of a respiratory irritant,

and visual means or various symptoms such as dizziness, lassitude, etc. It also should

include types of monitoring that can be done by laboratory personnel, by the

organization's Safety and Health Department, or by outside public and private agencies.

3. Means to reduce or eliminate the exposure of the employee to the risks associated with

the hazardous chemicals in the workplace. This should include work practices that will

reduce the exposures or the use of personal protective equipment.

4. Actions the organization has taken to minimize the exposure of employees to the chemical

hazards. This can include the engineering controls which have been implemented such

as ventilation or monitoring devices. It can also include policy positions which encourage

or require that employees and their supervisors follow good safety practices at all times,

and programs which provide incentives for them to do so or to penalize those who do not.

5. Emergency procedures to follow in the event of an accidental exposure to a hazardous

material.

6. Procedures to warn nonorganizational personnel working in the area of potential

exposures. Generally this will mean persons working under contract to the organization.

It could also include maintenance and other support personnel.

7. Measures to provide information as to the hazards and the protective measures which

both the employer and employee can take to reduce or eliminate the hazards associated

with a nonroutine task involving chemicals.

8. Measures to inform personnel of the hazards associated with unlabeled pipes carrying

chemicals in their work area, and the safety precautions which should be taken.

9. Availability of a medical evaluation should an over exposure have occurred or be

suspected.

The responsibility for training covering these topics should be shared between the local

administrative unit, usually the individual facility or department, and the department assigned

the lead in implementing the organization's program.

There are a number of generic chemical topics which would be essentially the same, no matter

what type of chemical exposure is involved. Among these are:



! flammable and combustible liquids

! corrosives, acids, and bases

! gases



! explosives



! toxic materials

! irritants



! carcinogens

! allergens

! pathogens



There also are a number of topics on protective measures to minimize exposures that could be

made the topic of standardized presentations, such as:



! safe chemical working practices

! use of personal protective equipment

! fire safety



! safe working practices

! electrical safety

! emergency procedures



Standardized programs can be developed for these topics, as well as others, and videotapes

made which can be used virtually anywhere. If the latter course is taken, much of the hazard

communication standard training requirements could be met by requiring a new employee to view



©2000 CRC Press LLC



selected tapes, supplemented by an opportunity for the employee to ask questions. Standardized

programs are especially useful where there is a continuing turnover, with customized training

being required for single individuals on a sporadic basis.

A large number of commercial companies offer hazard communication training programs, many

of which do provide videotapes as discussed above. They need to be reviewed prior to purchase.

Some, in order to demonstrate compliance with one aspect of a safety program, do not follow

good safety practices in other areas. Cost is not necessarily a guide; some of the better tapes

are among the less costly.

Although much of the training obligation can be met with standardized programs, the primary

responsibility for training must be borne by local managerial personnel or persons delegated by

them. The actual exposures vary from place to place. Generic programs can provide an excellent

foundation for a hazard communication training program, but they must be interpreted and

adapted to local environments and workplace practices. When training has been completed for

a given area, or even for a single chemical and if this is all that is necessary when a new chemical

is brought into the workplace, the employee should be asked to document that the training has

been provided by signing a dated statement to that effect. It is not essential that the employee

agree that all of the information has been understood, although it is certainly hoped that this is

the case, but it is important to have a record documenting that the information has been provided.

The employer may be called upon to provide this documentation during an OSHA inspection

or in the event of litigation.



REFERENCE

1.



Hazard Communication; Department of Labor, Occupational and Health Administration, 29 CFR Parts

1910 (Section 1200), 1915, 1917, 1918, 1926, and 1928, Federal Register 52(163), 31852, August 24,

1987. Also see (for current version): http://frwebgate.access.gpo.gov/cgi-bin/get-cfr.cgi



VIII. HEALTH EFFECTS

In recent years, there has been increasing emphasis placed on the health effects of chemical

exposures. However, as has been frequently noted, health effects are much more difficult to

quantitatively characterize than most physical safety parameters. It is straightforward to define

with reasonable accuracy a number of physical hazards, such as the upper and lower explosive

limits of the vapors of a flammable material. However, the exposure levels (see Figure 4.16, taken

from the Federal Register Vol. 53, No. 109, June 7, 1988, p. 21342) which will cause a given

physiological effect in humans are not nearly as precise, especially if the effect of interest is

delayed or is due to prolonged exposure to low levels of a toxic material.

Even individual reactions to low levels of materials that cause serious immediate or acute

effects at high doses are strongly dependent upon the inherent susceptibilities of individuals.

Some will exhibit a reaction at extremely low levels, while others show no signs of responding

at all to relatively high levels. Part of this is due to the natural range of sensitivity in a population,

but part is also due to contributory effects. There are synergistic effects: for example, a heavy

smoker may have developed emphysema due to the effects of inhaling smoke for an extended

period. Such an individual would be affected by airborne toxic materials which reduce pulmon-ary

function before a person with healthy lungs. The sensitivity of individuals can change with time:

an example is the common oleoresin allergen, poison ivy. The sensitivity of individuals is usually

small to an initial exposure, but with successive exposures the sensitivity increases. Similarly,

the sensitization of individuals to bee stings is well known. The effects of medication also can

modify the sensitivity of individuals. The serious problems associated with simultaneously taking

tranquilizers and drinking alcohol represent a well-known example of this phenomenon.



©2000 CRC Press LLC



The natural differences in individuals due to genetic factors, age, sex, lifestyle, etc., make

evaluation of laboratory tests quite subjective. On examination of the results of a typical blood

panel, one notes that the patient's results are given for each parameter measured, as well as a

range to be expected for a typical healthy person. An individual can have results somewhat

outside the normal range due to hereditary or environmental conditions and still be perfectly

healthy. Fasting and foregoing any medication prior to laboratory tests is an attempt to eliminate

as many variables as possible which would affect the tests. Unless a baseline series of tests are

available from a time prior to an exposure when, presumably the quantities to be measured are

“normal” for the individual, then it is often difficult to determine with certainty whether a given

result is due to an exposure or not. Even then, the results could be distorted by extraneous factors

occurring earlier.

Dependence on observed symptoms to indicate an exposure to a toxic material is also very

subjective. Again, the wide variation in tolerance of individuals to concentrations of toxic agents

causes a corresponding wide variation in the responses to an exposure. Many of the common

symptoms associated with occupational exposures - shortness of breath, headaches, nausea,

dizziness, etc., often can also be the result of other problems, illnesses such as flu, lack of sleep,

psychological problems such as stress due to personal problems or personality conflicts with

a supervisor, overindulgence, etc..An individual needs not only to be aware of the symptoms

which could result from an exposure, but also needs to try to distinguish when these are likely

to be due to an occupational exp osure and when they are likely not to be. If, for example, an

individual normally enjoys good health, has not done anything which might result in any of the

symptoms which the person is experiencing, and there are no “bugs” going around, he might

well suspect that he has suffered an exposure to some environmental hazard. In such a case, he

should mention it to coworkers, report it to his supervisor, and leave the work area. Especially

if others are experiencing similar symptoms, although not always, it is very likely that an exposure

has occurred and appropriate steps should be taken to seek medical aid for the exposed

personnel, limit the exposure of others, and correct the situation. The exception is when hysteria

causes psychosomatic effects among others , although even here it is wisest to treat the situation

as real, until proven otherwise. If only one person is having difficulties, but the immediate work

environment differs for each person, then an exposure may have occurred limited to the

individual, and precautionary steps, such as leaving the area, lying down, and observation by

a colleague, should take place. Often, prompt recognition of a problem is critical in minimizing

the consequences, especially when the possible culprit is a material which provides no other

warning signs. Whenever an exposure has occurred which has resulted in physical effects, an

evaluation by a physician should be obtained promptly. Even if no exposure has occurred, an

individual complaining of an illness should be taken seriously. There could be a medical disorder

requiring care, intervention, or at least documentation. Even if malingering is suspected,

evaluation by a physician can help to confirm that the claim of illness is or is not valid.

Delayed effects due to prolonged exposures to relatively low levels of toxic materials or

radiation rarely are reflected in immediate sensations of malaise sufficient to trigger concern about

possible consequences of the exposures. If a material does not have any warning properties, then

exposures may exist at unsafe levels indefinitely without the occupants of the area being aware

of the exposure. The eventual consequences may be masked by naturally occurring illnesses of

the same type. For example, lung cancer is, unfortunately, common and so the occurrence of lung

cancer might not be recognized as due to an occupational exposure if this occurred. Birth defects

occur in about 3 to 6% of natural births (depending upon how birth defects are defined). What

percentage might be due to an occupational exposure of the mother or father? Similarly, infertility

is a problem for about 15% of married couples. What is the role of occupational exposures for

the unfortunate couple? Some neurotoxins cause det erioration of the central nervous system,

but age and other illnesses may do the same. It is often difficult to establish a correlation between

an occupational exposure and an illness, even statistically for a group, because it is difficult to



©2000 CRC Press LLC



isolate the effect from the influence of other variables or to define an equivalent control group.

Anecdotal evidence citing an apparently unusual rate of a specific illness may be due to a

statistically random occurrence. Often, unless the illness is rare, such as the angiosarcomas

caused by exposure to vinyl chloride, it is impossible to definitely verify a causal relationship

between an occupational exposure and a disease.

Not all delayed effects are due to low levels of exposure. The onset of cancer which may occur

due to exposure to asbestos or radiation is often delayed for periods of 15 or more years, or, and

this is a key point, they may not occur at all. By no means do all individuals exposed to even high

levels of such hazards suffer the consequences.

There are basically three mechanisms by which health hazard data may be acquired: (1)

epidemiological studies of groups of exposed individuals, (2) human experimentation, and (3)

animal studies. There are problems associated with each of these three sources.

The major problem with epidemiological studies is that often one does not have a controlled

experiment; the data is either generated by an ongoing work situation or extracted from past

medical records. In some instances, case reports are sufficiently unusual that they call attention

to themselves, e.g., a reduction in fertility in a group of workers is so large that only a simple

study to determine the cause-effect relationship between a common exposure factor and the

resulting fertility depression is required, the effect being known; it only remains to determine what

experience the workers have in common. Rarely are situations as simple as this, although they

do occur.

In most instances, epidemiological studies to determine if an exposure to a substance results

in a given effect take the form of cohort studies, in which two separate groups composed of

exposed and unexposed individuals are studied. It is critical that the study be unbiased either

by the way the participants are selected or by the manner in which the outcome is tested. Another

critical factor is whether the two groups are in fact similar in all essential respects, which could

affect the outcome of the study, or that the differences are such that they can be taken into

account either in the design of the experiment or in the analysis of the data. In order to judge the

validity of a study, all of the relevant factors must be completely documented and available for

review.

Most of the epidemiological studies concerning exposure to toxic substances are from the

industrial sector since only in such an environment is it likely that exposures would be limited

to a single chemical or class of chemicals, and where the exposures would be relatively stable

over a prolonged period of time. The majority of the studies that are available tend to come from

Scandinavia, where, for example, Finland maintains a computerized data base of the health records

of all its citizens. Similar records do not exist in the U.S., although some categories of specialized

health data are maintained. Many epidemiological studies of exposures in the U.S. depend upon

records maintained by corporations, or equivalent public agencies such as the national

laboratories, which are managed by industrial firms and have similar medical surveillance

programs to them. The limited range of chemicals for which such work situations provide the

basis for valid epidemiological studies limits the scope of this approach.

Human experimentation is limited by statute and by ethical considerations to studies in which

there is no prospect of permanent harm to the volunteers participating in the study. This

obviously limits the scope of the results obtained by this route, although it can be employed to

determine the onset of early symptoms or to determine threshold levels for detection of odors

or irritation as a potential warning mechanism. Any experiment of this type must be carefully

reviewed by a human subject review committee of the institution or corporate research facility

where the research is being contemplated. Any subject of such experimentation must be fully

informed of any risks or benefits and normally must be given an opportunity to withdraw at any

point. However, even with this restriction, many experiments using volunteers have been

conducted and significant data have been obtained on symptoms initiated by modest levels of



©2000 CRC Press LLC



exp osure. It should also be remembered, though, that “fully informing the volunteer of all known

risks” can in and by itself skew the results. Therefore, published results on subjective symptoms

using a small sample must be viewed as inconclusive or suspect.

There have been data obtained from direct human exposures due to accidental exposures

to high levels of a number of hazardous chemicals. These, of course, are not controlled

experiments and the dose levels must be inferred from the circumstances of the incident, but as

direct evidence of the results of high exposures, they are extremely useful.

Since data from exposures directly to humans are limited, much of the available data on the

toxic effects of chemicals is obtained from animal data. The easiest data to obtain are the median

dose or the median concentration in air which is fatal to an animal under a standardized

experimental protocol although animal rights individuals are taking an increasingly active role

in opposing such tests. The most common animals used for this purpose are strains of rats and

mice because they can be obtained with uniform characteristics relatively inexpensively, and the

cost of housing and feeding them is small compared to most other species. Recently a few

laboratories have succeeded in cloning individual mice to achieve a completely homogenous

population. In addition to rats and mice, many other animals are used, such as primates (monkeys,

chimpanzees), guinea pigs, rabbits, dogs, cats, and chickens, in efforts to obtain a model which

would parallel the effect on humans. In the generic carcinogen standard, relevant animal studies

were intended to specifically involve mammalian species.

The median lethal dose, written LD 50, is given in mg/kg and the species is given. The median

lethal air concentration, LC5 0 , may be given in mg/m3 or ppm for a given species, and the exposure

time interval is usually specified. Lesser amounts of data are given in the literature at other

survival fractions, such as 25% or 75%. These data must be obtained under rigorously controlled

conditions to be useful, and the experimental protocol must be totally documented. Among other

things, enough animals must be used to provide statistical accuracy. Where the effect to be

measured is less well defined than lethality, the number of animals needed to obtain the data may

become quite large. Even after the data using animals has been obtained, the question remains

in many cases of whether the animal model is sufficiently close to that of a human response to

use it to determine the equivalent human response. Much of the controversy of using animal data

to establish human exposure effects revolves around this question.

Another procedure which leads to varying interpretations on the health effects of tested

materials is the practice of using large, nonlethal doses to reduce the number of animals required

when studying other effects such as carcinogenicity. The premise is that a large dose given to

a small number of animals is experimentally equivalent to small doses given to a large number

of animals. A linear extrapolation hypothesis usually is used to estimate the effects at low

exposures. This practice is not uniformly accepted and is often used as an argument to discredit

the results which are obtained, but is the basis of much of the data on these non-acute effects.

Other possibilities to estimate the effects of low doses would be to assume that the response will

approach zero more or less rapidly than the dose. A linear extrapolation generally is considered

conservative. Data from experiments such as this are frequently used in arriving at health

standards, by regulatory agencies, pharmaceutical manufacturing companies and the media.

The use of animals has provided the greatest amount of health hazard data, but the practice

has come under increasing attack by animal rights activists. Much of the public support for this

movement originated from widely publicized instances in which animals were not well treated and

undoubtedly suffered more than was necessary. As a result of public pressure and a concern

on the part of many scientists, many new safeguards have been instituted to minimize the amount

of pain and suffering experienced by laboratory animals. Animal care committees are now required

to review experimental protocols and must approve the procedures in order to qualify the research

for federal support. The number of animals involved in the research is limited to the number

required to achieve meaningful results and the pain experienced by the animals must be no more



©2000 CRC Press LLC



than absolutely necessary. These committees must include persons not affiliated in any other

way, directly or indirectly, with the institution, and who might be expected to be caring about

the well-being of the animals.

Although conceding that improvements have been made in the care of the experimental

animals, the animal rights activists goal is to prevent the use of animals in any research which

would adversely affect the animals. There have been instances in which animals have been

“liberated” from facilities and instances where these “liberated” animals have been released into

the environment. There are two practical problems with such actions, regardless of the ethics:

(1) the animals usually are not accustomed to surviving in the wild and most often do not, and

(2) there generally have been no efforts to ensure that the animals are healthy and, hence, a

disease could be introduced into the environment.

The argument that there is no alternative to using animals to gain knowledge to prevent

disease or to cure human illnesses, since experimentation on humans cannot be done, is rebutted

in two ways by the animal activists, the first being a purely moral stance of “why do we assume

that it is morally right to cause pain to animals to help humans?” This is an issue that each

individual must answer for himself, unless a legal restriction is imposed. The second argument

is that animal experimentation is no longer necessary. It is claimed that computer modeling can

provide equivalent information. Relatively few scientists accept this latter argument as a

generalization, although it is agreed that computer modeling can be used in some cases and as

an indication of productive research.

There is merit on both sides, although the extremists of both groups are undoubtedly too

extreme, and some middle position will eventually become acceptable practice. However, animal

health hazard data may be less available in the future.

A newer but effective modality which can be statistically relevant is to use cell cultures. Direct

effects can be seen and measured as to benefits or toxicity. It is more complicated and has seen

limited use as yet. There is also the question as to whether effects seen in individual cells can

be extrapolated to a complex organism.

Recent mapping of the human genome have greatly accelerated an understanding of how

diseases are caused and are opening many more options in treating diseases. It may be possible

to forego animal experimentation entirely in the future. A current program taking place in Iceland

where a remarkably homogenous population exists along with extensive genealogical data may

prove especially helpful in determining what genes are involve with specific diseases, leading

to better approaches to treating these diseases.

A recent report by the EPA indicates that research now appears to indicate that some toxic

substances are less dangerous than formally supposed, based on animal studies, specifically in

regard to carcinogenicity. This finding, according to EPA spokespersons, is based on a better

knowledge of how the metabolism of chemicals differs in various species, a better knowledge of

how much of a chemical that has been taken into the body actually reaches an organ where it may

do harm, and a better understanding of how the chemical influences the mechanisms that cause

cancer. This is a controversial position since, in general, it leads to higher acceptable exposure

levels of the chemicals under discussion, such as dioxin and arsenic. There are scientists that

feel that the level of scientific knowledge does not as yet justify moving away from a very

conservative approach. Both sides should avoid treating the issue as one that can be settled by

politically biased discussion, and the data on which the findings are based should be evaluated,

as should any other hypothesis, on the basis of their scientific merit.

As stated in the section on the hazard communication standard, Section VII.B of this chapter,

OSHA defines health effects, for the purposes of the standard, in Appendix A to 29 CFR

1910.1200. The definitions given below are from that appendix. The laboratory safety standard

also specifically suggests using these same definitions for guidance in defining hazardous



©2000 CRC Press LLC