The Influence of Dust Standards on the Prevalence and Severity of Coal Worker's Pneumoconiosis at Autopsy in the United States of America

Eligibility for inclusion in the NCWAS is restricted to coal miners who worked at an underground coal mine for part of their career. Independent pathologists conduct the autopsies and submit cases to NIOSH. Accession and payment to the pathologist for the autopsy requires submission of the following documents and materials: a consent, release, and history form completed by the surviving next of kin; a brief clinical summary of the case; an autopsy report that includes gross and microscopic description (and weights) of the heart and lungs and thicknesses of right and left ventricular walls; and at least 3 representative lung tissue blocks and corresponding slides. The guidelines do not specify sample sites or a requirement for fixation of the lungs in inflation. The submitted tissue blocks and slides are retained by NIOSH and archived for study.

The history form documents dates of birth and death, total years of coal mining tenure at underground and surface mines, titles of principal and last mining jobs, name and location of mine of last employment, and cigarette smoking history. The smoking history includes number of years smoked, number of cigarettes smoked per day, and (if the miner had quit smoking by the time of death) how long before death the miner had quit smoking. The age of retirement is not included on the history form. For the purposes of analyses presented in this report, it was assumed to be 60 years for each miner who lived to at least that age. This age was based on data in the United Mine Workers of America database, courtesy of Linda Fritz, Assistant Director, Eligibility Processing and Records Management, United Mine Workers of America, for the decade 1970–1979, when a majority of cases were accessioned.

Histologic slide sets for the 6103 autopsy cases were randomly divided into 3 equal-sized subgroups, each of which was distributed to 1 of 3 experienced pathologists to be independently reviewed. The pathologists were blinded to all information on the cases. The lesions of pneumoconiosis were diagnosed using criteria developed jointly by the pneumoconiosis committee of the College of American Pathologists and NIOSH, published as a monograph in the Archives of Pathology & Laboratory Medicine in 1979,11 with minor modifications described below. Coal macules were defined as 0.5- to 6-mm-diameter lesions containing coal dust–laden macrophages and located within the walls of respiratory bronchioles with fine reticulin and minimal collagen and associated with focal emphysema. Coal nodules were defined as coal dust–laden fibrotic lesions up to 10 mm in largest dimension with round or irregular borders and with an irregular distribution of collagen fibers. Silicotic nodules were defined as discrete lesions with smooth, sharp borders and concentric laminated collagen fibers. Progressive massive fibrosis (PMF) was defined as a coal dust–pigmented fibrotic lesion with irregular or whorled deposition of collagen fibers with or without areas of necrosis measuring 1 cm or more in diameter. The size criteria were those defined in the Federal Coal Mine Health and Safety Act in 1969.1 This ensured equivalency between radiographic and pathologic size criteria. Coal macules, coal nodules, and silicotic nodules were graded into absent, mild, moderate, and severe categories based on the size and profusion of lesions within the tissue sections, as previously described.4 Progressive massive fibrosis was noted as present or absent and was not further graded.

Data were analyzed using the Statistical Package for Social Sciences (SPSS, Chicago, Illinois), version 15.0. A variety of descriptive measures was used, including frequencies, cumulative percentages, means, and standard deviations. The prevalences of coal macules, coal nodules, silicotic nodules, and PMF were calculated among the autopsied miners and subgroups. When autopsied miners were stratified, prevalence differences were assessed for statistical significance via the χ² test if the class variable was nominal (such as smoking status) or by the Jonckheere-Terpstra test for trend if the class variable was ordinal (such as calendar decade of mining tenure). Differences in continuous variables across greater than 2 categories (eg, demographics by time interval) were assessed via the Kruskal-Wallis test, which is the nonparametric version of the analysis of variance test. Nonparametric tests were used because the continuous variables followed a relatively skewed distribution (and thus did not satisfy normality requirements for parametric tests).

Several approaches were used to evaluate the potential association between the regulatory changes in mandated dust exposure and the prevalence and severity of pneumoconiosis. One analysis examined trends after grouping miners by the decade in which their employment period was estimated to have begun (before 1950, 1950–1959, 1960–1969, 1970 and after). Prevalence of each specific lesion was then calculated separately for each of these groups to examine the trend of disease over time. Dates of beginning employment were estimated from the miner's age at death and total years of mining experience, assuming a retirement age of 60. Because the last group (those beginning in 1970 or later) was necessarily limited to 25 years' experience or less, the other groups in this specific analysis were also limited to workers with 25 years' experience or less.

A second approach examined the entire cohort using logistic regression analysis, with forward selection, to model the prevalence of each CWP lesion as a function of period of employment, years of mining, age at death, and smoking status. Pneumoconiosis prevalence and severity were compared between 2 subgroups of miners with different periods of employment: (1) those who retired or died before (or in the same year as) full implementation of the new regulations in 1973, and (2) those who began employment after (or in the same year as) initial implementation of the transitional dust exposure regulations in 1970. Miners who belonged to both groups were excluded for this analysis. Adjusted odds ratios were presented to quantify the odds of disease among the postimplementation group versus the odds of disease among the preimplementation group.

The agreement between pathologists for the type and severity of pneumoconiosis was assessed using a multi-rater κ statistic12 applying the microfile MKAPPASC in SPSS. Twenty-three cases were reviewed by all 3 pathologists and used in the analysis. The cases were selected to cover the full range of lesions described in this study. A κ statistic of 1 is considered perfect agreement, and 0 is what would be expected by chance.

A sensitivity analysis was performed to determine if there was a minimum of slides sufficient to diagnose the type and severity of pneumoconiosis. For this analysis cases with 4 or fewer slides were compared with cases with 5 or more with regard to severity of macules, nodules, and PMF. Their relationship to mining tenure was assessed with χ² and Kruskal-Wallis tests.

Miners who died in 29 states were represented in the autopsy study. The basic demographic characteristics of the entire population are shown in Table 1. The number of cases accessioned for each year of the program from 1971 to 1996 (the last year for which pathologic analysis was completed) declined from an average of 400 cases at the beginning of the program to approximately 200 cases in 1996. A majority of cases (∼63%) had been embalmed prior to autopsy. Sixty-three percent of cases had a complete autopsy (excluding brain) and 37% of the autopsies were restricted to the thoracic cavity. The median number of slides of lung tissue per case was 4 (range 3–30).

Examples of the pathologic lesions that were graded for the study are shown in the Figure, A through D. The prevalence of each type of CWP lesion by severity is shown in Table 2. The relationships between years of mining and increasing severity of coal macules, coal nodules, and silicotic nodules were highly significant (P < .001), as was the association between years of mining and presence of PMF (P  =  .009). In the sensitivity analysis, no significant differences were found (P > .05) among the different features of pneumoconiosis between 4 or fewer slides compared with 5 or more, suggesting that 3 slides were sufficient to diagnose pneumoconiosis.

The statistics for agreement between pathologists were: macules, κ  =  0.564 (P < .001); coal nodules, κ  =  0.631 (P < .001); silicotic nodules, κ  =  0.788 (P < .001); and PMF, κ  =  0.883 (P < .001). Kappa values 0.40 to 0.59 are considered moderate; 0.60 to 0.79, substantial; and 0.80 to 1, outstanding.

The prevalences of pathologic findings for selected coal mining states where the miner was last employed are shown in Table 3. Percentages of miners with coal macules were similar for these states. Percentages of miners showing coal nodules, silicotic nodules, and PMF were lower in Kentucky and higher in Pennsylvania than in other states.

Pneumoconiosis among miners who received all of their exposure prior to the establishment of current dust standards was substantially higher than among miners who started employment after the regulations came into effect in 1970 (Table 4). However, years of mining, age at death, and smoking status were all significantly different (P < .001) between the preregulation and postregulation groups, and thus the apparent decline in prevalence of pneumoconiosis may have resulted in part from these factors. Therefore, a separate analysis was performed on miners with no more than 25 years of mining tenure. For this analysis, miners were stratified by decade of entering the workforce (Table 5). All types of pneumoconiosis were less prevalent in miners employed entirely after 1970 compared with those employed prior to 1970. The most notable difference was a lower prevalence of PMF (1.2%) among miners who had worked after 1970 (ie, entirely under the new regulations) compared with a prevalence of 9.6% among miners who had worked in the decades prior to 1970.

Logistic regression was used to further evaluate the significance of associations between disease prevalence, employment period, age at death, mining tenure, and smoking status (Table 6). The effect of mining in the postregulation period on the prevalence of pathologic CWP lesions was then evaluated after adjusting for these demographic variables. Increasing age and mining tenure were both associated with an increased odds ratio for disease. A protective effect of post-1969 mining on all types of pneumoconiosis remained after adjusting for these confounding variables (P < .001). Tobacco smoking was associated with decreased odds ratios for coal macules (P < .001) and PMF (P < .001).

This is the first study to report use of autopsy data to specifically address the efficacy of coal mine dust exposure regulatory standards in the prevention of CWP. Our primary goal was to examine the temporal trends in the classic lesions of CWP before and after implementation of the 1969 Federal Coal Mine Health and Safety Act. The dust standard established by the act established an interim limit of 3 mg/m³ from 1970–1973 and a 2 mg/m³ standard after 1973.

Before the introduction of large-scale mechanization, prior to 1970, dust exposures were often in excess of 6 mg/m³ in the coal mining industry.13 In this study we sought to investigate the effects of the regulatory dust standards promulgated in 1970 in reducing the prevalence and/or severity of pneumoconiosis in underground coal miners. The results show that the dust standards reduced the prevalence and severity of CWP, after adjusting for the potentially confounding effects of age, years of mining, and smoking status. However, we also show a considerable burden of disease in miners who worked entirely under the new standards, indicating that further action to reduce dust exposure is required.

The findings presented here complement those from examinations of living miners in the national Coal Workers' X-Ray Surveillance program10 and the National Study of Coal Workers' Pneumoconiosis.14 After 1969 the overall prevalence of small rounded opacities in living US miners declined from 12.7% (1969–1971) to 11.2% (1972–1975), to 3.0% (1977–1981), before slightly increasing to 3.9% (1985–1988).13,14 However, pneumoconiosis was still occurring in miners who had worked entirely within the new standard, and in 1995, NIOSH recommended reducing the exposure limit for respirable coal mine dust from 2 mg/m³ to 1 mg/m³.15 This recommendation has yet to be enacted.

Recent reports indicate that the prevalences of simple pneumoconiosis and PMF among active underground coal miners have been increasing.16–18 Regional differences have also been identified, with “hot spots” of rapidly progressive CWP documented in certain locations within the United States.19 These hot spots are more prevalent among workers from underground mines with fewer than 50 employees.17 Furthermore, underground coal miners working entirely after implementation of the current dust standards continue to develop advanced simple CWP and PMF.20 

Several causal hypotheses have been proposed to explain this recent reversal of health protection for US coal miners.16 These include poor compliance and enforcement of existing standards21 and new hazards in the mines resulting from changes in technology.16 

In addition to pneumoconiosis, coal mine dust exposure is associated with the development of chronic obstructive pulmonary disease (chronic bronchitis and emphysema).15,22–24 The lung sections submitted to the NCWAS program are not optimal for the evaluation of emphysema, which is best evaluated on whole lung sections.11,24 The assessment for chronic bronchitis is dependent on sections of cartilaginous airways being available for review. The number of cases that included sections of large airways was variable; therefore, we made no attempt to evaluate for chronic bronchitis in this study.

The NCWAS is a voluntary program and the number of participants during the period studied was small relative to the total number of decedents who had CWP listed as a cause of death on death certificates.9 Selection factors for autopsy populations are not well understood and the motivation for requesting an autopsy will vary.25 Changes in society as a whole, such as increased litigiousness, may also influence decisions about consent for autopsy. The consent for autopsy is given by the spouse or other next of kin. The stated intent of the NCWAS program is to provide families of deceased coal miners with evidence in support of black lung benefits.2 It is possible that next of kin were influenced in their decision to consent for an autopsy by the regulatory period during which the miner worked because the availability of screening and diagnosis and accessibility of compensation differed between preregulation and postregulation periods. It is difficult, with the limited data available, to say conclusively how and to what extent these various factors may have influenced our findings.

In this study, the prevalence and severity of CWP were associated with increasing mining tenure. The autopsied miners who had worked only under the more dusty conditions prior to the enactment of the dust standards were older, had worked more years underground, and were less likely to have smoked than those who worked only after implementation of the new dust limits. To adjust for these differences, years of exposure, age at death, and smoking status were fit in a logistic model to predict disease outcome. After adjustments, miners who worked only after the promulgation of dust regulations showed reduced odds ratios for all types of CWP lesions compared with miners who worked only before the implementation of the coal mine dust regulations in 1970.

A limitation of the study was the lack of data on date of entry into the workforce and date of retirement. We also had no data on employment in other industries or of periods of sickness/disability. Several approaches were taken to mitigate these limitations. The first was to estimate an average age of retirement based on data provided by the United Mine Workers of America on miners retiring during the decade 1970–1980. This age of 60 years was then applied to all miners aged 60 years or more at time of death to estimate year of retirement. A second approach was to limit the analyses to groups of miners who, based on their age and year of death, could only have worked either entirely before the full implementation of the new regulations or entirely after their implementation. This restricted analysis confirmed the analyses involving the entire cohort and provided evidence of pneumoconiosis in miners who had entered the workforce after the implementation of the new dust standard.

Miner-specific exposure data were not available for the study subjects. Under the Federal Coal Mine Health and Safety Act of 1969, coal mine owners are required to comply with regulatory limits and monitor respirable dust exposures in their mines. The accuracy of self-reported regulatory data has been questioned and it is likely that there is underreporting of the data for silica26 and coal mine dust.27 

We found that disease severity differed by state in which the coal miner was primarily employed. Cases received from Pennsylvania ranked highest for CWP severity, including PMF and silicotic nodules. Mines in eastern Pennsylvania that extract high-rank anthracite coal have been associated with high disease prevalence, including silicosis.28 Disease prevalence was lowest in Kentucky and intermediate in Western states. Marked regional differences in prevalence of silicosis and PMF have been reported for coal miners from other countries.29 

A negative association between smoking and pneumoconiosis was observed in this study. Nonsmokers showed 3% to 8% greater prevalence of pneumoconiosis compared with ever smokers. Smoking is associated with premature death from many causes, which reduces the time for progression of pneumoconiosis. However, this finding was also evident in a logistic regression model that adjusted for years of mining tenure and age at death. An inverse relationship between pneumoconiosis diagnosed at autopsy and smoking has been reported in studies of miners exposed to silica30 and coal mine dust.31 This observation may have a biological basis. The deposition of particles in cigarette smokers is greater in the central airways,32,33 whereas in nonsmokers alveolar deposition is greater.34,35 Thus decreased peripheral lung deposition of coal mine dust provides a theoretical basis for the lower prevalence of pneumoconiosis in smoking coal miners compared with nonsmoking coal miners with comparable dust exposures.36 

The difference in pneumoconiosis prevalence between smoking and nonsmoking coal miners could also be explained by a participation bias in the NCWAS. Respiratory symptoms are associated with smoking, regardless of the presence of CWP. Considering miners who had not been diagnosed with CWP during life, next of kin might be more likely to request an autopsy on a miner who had been symptomatic than on a miner who had been asymptomatic. Thus, among autopsied cases, smokers without pneumoconiosis would be overrepresented relative to nonsmokers without pneumoconiosis. This could also account for the observation that CWP prevalence was lower among smoking miners than among nonsmoking miners in our autopsy study.

In summary, we conclude that progress has been made in reducing the burden of CWP in US coal miners. Our autopsy findings support the effectiveness of promulgation and enforcement of coal mine dust control regulations. Our findings also indicate that underground coal miners, even those who worked only after implementation of current dust standards, continue to develop advanced simple coal worker's pneumoconiosis and PMF.

The authors thank Robert Castellan, MD, MPH; David Weissman, MD; Robert Cohen, MD; and Tamer El Mays, MD, for their critical review of this manuscript and helpful comments. We thank Michael Andrews for assisting in the statistical analysis and Laura McPhee for manuscript preparation. We also wish to acknowledge the death of the lead author, Dr Val Vallyathan, on Friday, July 23, 2010. Dr Vallyathan was a research pathologist and team leader at NIOSH in the Health Effects Laboratory Division, Morgantown, West Virginia, from 1979 to 2009. Val's influence on occupational health research was widely acknowledged and reached around the world. Val received the Alice Hamilton awards in 1999 and 2001. He served as a book editor, guest editor, and review contributor for many journals and has authored at least 376 full publications in scientific journals. His greatest scientific contributions were to the understanding of the pathogenesis of the pneumoconioses and the role of free radicals in lung disease. Our community has lost a dedicated and committed lung researcher and friend who will be greatly missed.