injury records, this change in the use of employer records may hinder efforts to

assess the results of the new policies. Third, as mentioned above, a number of

factors besides the effectiveness of OSHA can influence injury rate trends.

Foremost among these is the business cycle.

OTA has compared injury rates with several measures of the business cycle.

Figure 1 shows data for the BLS total recordable injury rate and the

unemployment rate for 1972 to 1982.

The total recordable injury rate declined

from 1973 to 1975 and again from 1979 to 1982 simultaneously with rising

unemployment rates.

In addition, the rising injury rates from 1975 to 1979

coincided with declining unemployment rates.

In Figure 2, these two variables

have been plotted against each other, with the unemployment rate on the horizontal (or x-axis) and the injury rate on the vertical (or y-axis).

Examination of this figure reveals that there appears to be an inverse

relationship between injury rates and unemployment.

That is, as unemployment

rises in a recession, injury rates decline.

Another possible measure of the business cycle is to examine the level of

employment, as opposed to the unemployment rate.

This must be done carefully

because in the last few years, the changes in the level of employment have not

been the same in all industries.

In fact, from 1979 to 1982, employment in the

more hazardous manufacturing and construction industries declined, while

employment in the other major private sector industry groups has stayed the same

or increased.

Figures 3 and 4 show the relationship between employment and

lost-workday injury rates in construction and manufacturing alone.

Again there

appears to be a close relationship.

As employment rises, so do injury rates.

Additional analysis using variables that directly measure the new hire

rate, the number of overtime hours, the rate of production, and capacity

utilization in specific industries may clarify this relationship further.

Examination of the influence of new hires is, however, made more difficult

because BLS no longer publishes statistics on labor turnover, which included the

new hire rate.

But at present, it appears that the effect of the recent

recession, especially in construction and manufacturing, is the most important

factor behind the injury rate declines from 1979 to 1982.

Use of Exposure, Medical, and Death Records

Exposure Records

Information on the extent of worker exposures to health hazards is

especially important because of the lack of an accurate accounting of the extent

of occupationally caused illness.

Exposure information can serve several

purposes: First, it can be used to quantify the extent of the problem, giving

policy makers and citizens a sense of its magnitude. Second, if collected

properly, it can be used to chart progress over time. Third, the development of

accurate and comprehensive exposure data can greatly assist in the conduct of

occupational epidemiology, the development of information for risk assessments

and economic analyses, and the creation of more reliable estimates of the size

of the occupational disease problem.

The NIOSH National Occupational Hazard Survey and National Occupational

Exposure Survey can provide information on the number of employees potentially

exposed to hazardous substances, but they provide no information on the levels

of exposure and only limited information on the durations of exposure.

As part

of its health inspections, OSHA collects and analyzes exposure samples. These

data might be usable for developing employee exposure estimates.

In a contract

report prepared for our assessment of workplace health and safety controls, John

Mendeloff examined this health inspection data.

Based his work, we conclude

that this source of information may be useful for developing estimates of worker

exposures and should be explored further.

Medical and Death Records

Some employers maintain records of medical surveillance of their workers as

well as information on employee deaths and cause of death.

State and local

departments of vital statistics keep death certificates, which contain

information on cause of death and, in many cases, the person's usual occupation.

These records are potentially a very useful source of data for conducting

epidemiologic studies of occupational disease. Epidemiologic investigations to

discover, explore, and confirm associations between particular industries and

occupations and injuries, illnesses, and deaths can be made more efficient and

useful by certain changes in Federal data collection efforts.

Issues Concerning Occupational Injury and Illness Data

In our Assessment report we discuss a number of issues concerning Federal

data collection efforts.

There are several possible changes which could improve

those efforts

improvements which could increase the usefulness of the

collected data, ensure the validity of the data, and to facilitate epidemiology.

Options to address these issues will be discussed in detail in our assessment

report, which will be delivered to the requesting committee, the House Energy

and Commerce Committee, in the near future.

Injury Data

Injury Investigations.

One 188ue concerns the level of activity that OSHA,

NIOSH, and BLS direct towards investigating fatal and nonfatal injuries and in

preparing information from these investigations that would be useful to workers,

employers, and professionals.

Each year, Federal OSHA and the state plan

agencies investigate a large number of accidents involving fatalities and 5 or

more hospitalizations.

Until relatively recently, this effort has been directed

almost solely towards questions of compliance with OSHA standards and little

attention has been paid to using the collected information to prevent future


OSHA has conducted limited analyses of several types of fatalities,

and has recently initiated a limited effort, in some areas, to distribute

summaries of construction accidents to labor unions, trade associations, and

other organizations to provide information for prevention activities.

Complementing the OSHA activities, NIOSH has begun detailed investigation

of a small number of fatal injuries.

In addition, the BLS has obtained

information concerning some types of nonfatal injuries through questionaires

completed by injured workers.

The results of these investigations could provide

useful information for worker, employer, and professional education, serve to

guide OSHA's revisions of its safety standards, and be used in directing OSHA

inspection activities.

BLS Annual Survey.

A second issue concerns whether or not BLS is to

conduct a new "Quality Assurance Program." OTA's examination of various

estimates of the number of occupational fatalities and nonfatal injuries

revealed some differences among the various data sources for the years up to


Since 1981, employer-maintained injury records and the results of the BLS

Annual Survey have been used to grant exemptions from OSHA inspections.


of this the accuracy of these data is especially important.

In the 1970s, BLS

conducted on-site evaluations of a sample of employer responses to the Annual

Survey to verify their accuracy.

This "Quality Assurance Program

has not been

repeated since 1977.

A new effort to verify the accuracy of the data might include only an on

site evaluation of whether employer survey responses corresponded with employer

injury and illness records. Or, more ambitiously, it might attempt to learn if

the employer injury records capture all occupationally related cases.


could involve employee and employer interviews, examination of workers'

compensation, medical insurance, and sick leave records, and examination of

medical records from employers, local doctors, and hospitals.

Illness Data

Mortality Surveys.

Another area concerns creating mortality surveys to

study associations between occupations and industries and causes of death.

These studies are valuable not only for identifying high risks associated with

some types of work but also to indicate occupations and industries that do not

present high risks.

Here there are two issues of importance.

The first

involves efforts to ensure accurate coding of industry and occupation

information on death certificates.

The second is to actually conduct these


One nation-wide mortality survey was done in the United States in the

1950's, and epidemiologists in the States of Washington and Rhode Island have

made similar studies during the last decade.

Currently, NIOSH and the National

Center for Health Statistics (NCHS) are providing technical support to a few

other states for conducting similar mortality surveys. Britain, on the other

hand, has conducted these examinations of death certificates every 10 years

since 1861.

Nationwide mortality analyses could provide important leads for further

study to pin down associations between work and causes of death and valuable

information about hazards in occupations which are scattered across the country,

e.g. carpenters, butchers, dry cleaners.

Currently conducted state wide

« ForrigeFortsett »