Carpenters
|
MORTALITY STUDY RESULTS OF
CARPENTER UNION MEMBERS Cynthia Robinson |
DR. ROBINSON: The study I'm going to talk about is the mortality study where we looked at mortality patterns 1987 through 1990 of carpenters in the construction or wood products industry. This kind of study takes awhile to conduct. It's labor intensive and I'd like to acknowledge all the assistance I received from the Carpenters Union, my colleagues and co-workers at NIOSH and the Center to Protect Workers' Rights. Just a little background. The Bureau of the Census in 1992 estimated that there was about 1,057,000 carpenters in the U.S. And the Union represents about 200,000 of them or about 20 percent of all U.S. carpenters. And probably everyone is aware that this study is part of the larger Health Initiative for Construction Workers conducted by NIOSH and the Center to Protect Workers' Rights. Most previous studies of carpenters have been limited to the wood industry. Very few have looked at construction site carpenters. The International Brotherhood of Carpenters asked NIOSH to do the study. And the goal of the study was to evaluate the mortality patterns of the union members and to identify preventable causes of death that were associated with occupational exposures in the construction or wood products industry. In 1972 Sam Milham conducted a mortality study of the Carpenters Union and this study was to follow up on the results of that study. Sam's study found excess deaths from stomach cancer, bladder cancer, lung cancer, Hodgkin's disease, lymphatic cancer, and fatalities, and nasal cancer. And nasal cancer has long been associated with furniture making in several of the other studies, mostly British studies, but some U.S. So we wanted to particularly look for those causes of death. Carpenters may be exposed to noise, solvents, glues, lacquers, wood preservatives, wood dust, asbestos, silica paints, and electric and magnetic fields from tools. So under the Constitution of the Carpenters, a deceased carpenter's family is entitled to death benefits if he died while he was active or retired. We were able to use the Union death benefit records to identify the eligible members of the study. And, however, we did obtain some death certificates from the state in the case where it was missing, about 2100 of those. We compared the Union mortality with the U.S. proportional mortality rate and we computed age-adjusted PMRs and PCMRs. A PMR is the proportionate mortality ratio and all you need to know is that if it's over 100 it is elevated. The local unions have skill classifications because carpenters tend to join locals based on their skills. And so we used those classifications as risk subgroups to do a subanalysis. We obtained the death certificates. They were coded by a trained nosologist. And we used the NIOSH life table program to calculate age-adjusted PMRs. We also did a more conservative analysis called proportionate cancer mortality analysis. This was done to adjust for potential bias that can result from PMR analysis. And the next five or six slides are the results of the study. Death certificate status showed that we obtained death certificates for about 27,362 carpenters for that four year time period or 99 percent of the deaths. We excluded deaths that occurred in foreign countries because they would not be in our denominator, which was the U.S. mortality rate. And we couldn't locate nine certificates. Next slide. This shows the distribution by race and gender. Most of the study was white males, although I did analysis of the other race/sex groups. I'm not going to focus on them today. Just the white males. I do have some reprints here that have the analysis of the other race/sex groups in the whole study, though, which I guess we'll pass out in a bit. This slide just shows the distribution by age at death. And as you would expect, you see a greater number of deaths for the age 20 group where fatalities tend to occur, and then in the over-75 group. Next slide. This shows the distribution by years in the Union. The blue bar on the right represents all the benefit-eligible carpenters. And there, over 50 percent had greater than 40 years in the Union. The analysis of the study showed many elevated causes of death. I'm just showing some of selected causes. For the entire Union, 11 sites of cancer in addition to the other causes of death were significantly elevated. When we ran the proportionate cancer mortality analysis, which is the more conservative analysis, three sites of cancer remained significantly elevated. They were stomach cancer for both white males and non-white males. The PMR is the black number on the left of the red bar. On the bottom of the slide, the numbers represent PMRs and the 100 line represents background. So elevated is anything greater than 100. Stomach cancer. Lung cancer remained significantly elevated. And mesothelioma was the highest PMR in the study. It was a three-fold excess of deaths. We observed for heart disease a decreased PMR but we did observe an increase in a kind of heart disease called cardiomyopathy. AUDIENCE MEMBER: How many deaths were there in all from mesothelioma? DR. ROBINSON: For the total study -- Well, it's kind of tricky because these are only the mesotheliomas that occurred in the lung. For the entire study there were about 120 deaths due to mesothelioma, which is a considerable excess, since they only occur from asbestos exposure. Next slide. Okay. For the subanalysis, when we looked at the skill classification codes of the local unions, the two largest subgroups were for the construction industry with 22,758 deceased members and the wood products industry. Next slide. So the carpenters in the construction industry had one pattern of death and the PCMRs for lung cancer, mesothelioma and bone cancer were significantly elevated for this group. Asbestosis was elevated, as were falls. Next slide. However, when we looked at the carpenters and the wood products industry, a little different pattern occurred. Stomach cancer remained elevated. Male breast cancer had a significantly elevated PMR, but it was based on only four deaths. And injuries due to transportation such as railway or either traffic or non-traffic motor vehicle deaths was also elevated. Just a few points of discussion. The highest PMRs in the study were for diseases caused by asbestos exposure on construction sites. Exposure to asbestos has been documented even in recent years on construction sites by Nicholson, NIOSH, and others. Our PMR of 301 for mesothelioma was consistent with other studies of construction workers, of carpenters on construction sites. TITA reported in, I believe it's Connecticut, a PMR of 2.2. Our study observed an excess of stomach cancer. This has been previously associated with Canadian workers with long exposure to sawdust by Semiatecki in 1986. He observed that excess risk for stomach cancer was greatest in Canadian workers with long exposure to sawdust, like carpenters. We observed elevated bone and breast cancer PMRs. This has not been reported in previous studies except for a study by McGlaughin who looked at Swedish carpenters who worked in the construction industry and reported excess male breast cancer there also. Heart disease, cardiomyopathy and the healthy worker. The healthy worker effect is defined as selection into employment of healthier individuals and is usually reflected as decreased PMRs for coronary heart disease. We observed elevated risk for a different heart disease risk, cardiomyopathy, which has been associated with occupational exposures and might be looked at further. Just a word about limitations and strengths of this type of study. For limitations, there is potential bias because of death certificate underlying causes of death. We have no work histories, no exposure measurements, no smoking data, and no information on members of the Union who quit and there were no records for that. But the strengths of study is that sometimes Union records are the only way to study occupational groups where there are few large employers who maintain records. And it is based on a financial incentive, which has been shown by Bomont and others to yield accurate estimates of the death rate because of that financial incentive to report the death. It can identify areas for intervention at the work place and can be used to prioritize analytic studies. It encourages further investigation and improvements in data design and quality, hopefully. And in conclusion, despite our limitations, our study observed significantly elevated mortality for the diseases caused by asbestos: lung cancer, and malignant mesothelioma, and asbestosis. We observed moderately elevated mortality for fatal injuries. And we observed some unexpected findings of stomach, bone and male breast cancer that might require further evaluation of possible occupational risk factors. These data suggested that carpentry is a very hazardous trade and that further preventive action against asbestos and other hazards is needed. |