Controlling Asbestos in Buildings

An Economic Investigation

Donald N. Dewees

• Publisher: Routledge
• Year: 2016
• Topic: Economics
• eBook ISBN:
  9781317335863
• Subtopic: Environmental Economics
• Index: Economics

Asbestos dust is well-known for causing cancer and other life-threatening illnesses yet still contaminates countless schools, factories and office buildings. This raises the issue of the best way to deal with Asbestos; immediate removal, containment or removal at renovation or demolition. Originally published in 1986, this report aims to evaluate these methods of dealing with asbestos in relation to their cost-effectiveness to conclude the most appropriate solution. This title will be of interest to students of Environmental Studies and Economics.

Three

MODELS FOR PREDICTING ASBESTOS DISEASE

While a number of diseases have been associated with exposure to asbestos fibers, three diseases stand out as the principal causes of premature death among those who have worked with asbestos: as-bestosis, lung cancer, and mesothelioma. In the past, asbestosis has been a major cause of death among asbestos-exposed workers. However worker exposure intensities today are a small fraction of those which caused the disease that is now occurring. It is generally believed that there is a safe threshold exposure level below which asbestosis will not occur, while for the cancers no such threshold has been shown to exist. One estimate places this threshold at a cumulative exposure of 25 f/cc-years; that is, twenty-five years of exposure to 1.0 f/cc (RCA, 1984, p. 281). It is likely that at current workplace exposures, the amount of asbestosis that will be caused will be zero or insignificant (U.K. Advisory Committee on Asbestos [UKAC], 1979, p. 58; RCA, 1984, p. 281). Exposures in buildings and in the ambient environment are as little as one one-thousandth of current workplace exposures, so there is no risk that building occupants or the general public outdoors will develop asbestosis. This disease may be ignored in projecting the health risks from exposing new workers to current fiber levels, or from exposing persons to prevailing asbestos concentrations in buildings or outdoors.

This leaves lung cancer and mesothelioma as the primary diseases of interest. Lung cancer has many causes, but it is now accepted that the inhalation of asbestos fibers in sufficient concentrations can increase the risk of this usually fatal disease. Mesothelioma is a rather rare cancer occurring in the surface cells lining the chest or abdominal cavity. Like lung cancer, mesothelioma is usually fatal within a year or two of diagnosis. Most studies now assume that the dose-response function for these cancers is a straight line through the origin, so there is no absolutely safe exposure level; any nonzero exposure causes a nonzero risk.

Gastrointestinal cancers and other cancers have been associated with asbestos exposure, but their incidence rates are considerably lower than that for lung cancer. OSHA (1983b, p. 51129) assumed that gastrointestinal and other cancers would cause mortality rates equal to 10 percent of the lung cancer mortality rate. However, the RCA (1984, p. 102) said that “the association between asbestos exposure and both gastrointestinal cancer and cancer of the larynx is neither as strongly nor as consistently established in the medical literature as the association between asbestos exposure and [lung cancer and mesothelioma].” We may ignore these cancers in estimating future disease rates without substantially underestimating total disease rates.

The problem of predicting the health effects of current exposures to asbestos is therefore reduced to the problem of predicting the incidence of lung cancer and mesothelioma that will result from the exposure of a cohort of specified characteristics to the anticipated concentration of asbestos fibers for a specified period of time. The problem studied here is fundamentally different from the problem of predicting aggregate future asbestos-related disease in North America arising from past asbestos exposure. One example of the latter type of study is that of Walker and coauthors (1983) who use mesothelioma as an indicator of asbestos exposure. They project future mesothelioma incidence and multiply this incidence by fixed ratios of the incidence of asbestosis and lung cancer to the incidence of mesothelioma. Other projections of future disease from past exposure are those of Peto, Henderson, and Pike (1981) and Nicholson and coauthors (1981). These studies also use methods that are appropriate for aggregate analysis, but omit the detail that can be incorporated in a study of a particular cohort exposed to a single asbestos type.

This chapter reviews the health effects models, used in several recent publications, that present quantitative risk assessments of the lung cancer and mesothelioma incidence likely to result from asbestos exposure. The sponsors of the studies are the Ontario Royal Commission on Asbestos (RCA), the Chronic Hazard Advisory Panel (CHAP) of the Consumer Product Safety Commission, the U.S. Occupational Safety and Health Administration (OSHA), and the U.S. National Academy of Sciences (NAS). The general principles behind the models are presented, followed by a detailed description of the specific model used by the Ontario Royal Commission on Asbestos. The remaining models are compared with the RCA model. The coefficients estimated for these models are presented and differences among them analyzed. The health risks predicted by each model for the exposure of a typical building occupant and for a typical workplace exposure are presented. The reader not interested in technical details of the disease models may skip the mathematical portions of this chapter without loss of continuity.

Lung Cancer Models

General Considerations

The lung cancer models assume that there is a background lung cancer rate existing without asbestos exposure that varies with the age and gender of the individual and his smoking habit. Exposure to asbestos multiplies the background lung cancer rate by an amount proportional to the extent of the asbestos exposure (Hammond, Selikoff, and Seidman, 1979). Since smokers have a much higher lung cancer rate than nonsmokers, an asbestos exposure which, for example, doubles the lung cancer rate will cause a much greater absolute increase in the number of lung cancer cases among smokers than among a similar population of nonsmokers. The background lung cancer rate is extremely low until almost age forty and rises thereafter. The maximum risk is for individuals in their late fifties and older.

The exposure of the individual to asbestos may be described by the average intensity of that exposure, measured in fibers per cubic centimeter (f/cc), and by the duration of the exposure. The models assume that the cumulative exposure, defined as the product of intensity times duration, is a satisfactory measure of the dose and that the response, the probability of disease, is linearly related to this dose. While this assumption is reasonable for the intensities considered here, it may not be reasonable for very high intensities. In fact, the Ontario RCA concluded that short (several months), intense (tens of f/cc) bursts of exposure might overwhelm the body’s defense mechanisms and cause a disproportionate risk of disease (RCA, 1984, chap. 5, p. 306).

The models assume that the excess risk of lung cancer is proportional to the cumulative exposure to asbestos as just defined. Epidemiological data are not now, and probably never will be, sufficient to prove that a linear relationship fits better than a nonlinear one. However the linear relationship is plausible biologically (RCA, 1984, chap. 5, E 3b), is easy to work with, and is conservative in that at low doses it predicts more disease than the competing models. All recent risk assessments have used the linear model.

Lung cancer is often referred to as a disease in which there is a latent period between exposure and disease. It might be argued that the observed delay is an artifact of the worker’s first exposure usually occurring years before the age at which lung cancer can occur, as demonstrated by the age at which the background lung cancer risk becomes significant. Existing data are consistent with both the age relationship and with a latent period. Most of the models assume a ten-year delay between first exposure to asbestos and any elevation of the disease rate, although this assumption is not implemented in the same way in all models that use it.

In these models, the elevation of risk will continue until death, from whatever cause. This elevation of risk even in old age is particularly uncertain because few cohorts have been followed for a sufficiently long time after exposure ceased to test whether the risk continues or declines. One cohort that was studied did show an apparent trail-off of risk after thirty years after the first exposure (Nicholson, 1981). However, in the absence of robust data defining this phenomenon, the models generally assume no trail-off.

The Ontario Royal Commission on Asbestos Model

The RCA model calculates lung cancer mortality rates for a coho...

Table of contents

• Cover
• Title Page
• Copyright Page
• Original Copyright Page
• Table of Contents
• ACKNOWLEDGMENTS
• one INTRODUCTION
• two ASBESTOS EXPOSURES IN BUILDINGS
• three MODELS FOR PREDICTING ASBESTOS DISEASE
• four CONTROL OF ASBESTOS IN BUILDINGS
• five ECONOMIC FRAMEWORK
• six ECONOMIC ANALYSIS OF ASBESTOS CONTROL OPTIONS
• seven CONCLUSIONS
• APPENDIX A Model of Removal Costs and Building Life
• APPENDIX B Model of the Firm
• APPENDIX C The Cost of Saving Lives