Journal of the New Zealand Medical Association, 05-November-2004, Vol 117 No 1205
Mortality, morbidity, and asbestosis in New Zealand: the hidden legacy of asbestos exposure
Asbestosis is a diffuse fibrotic lung disease resulting from the inhalation of asbestos fibres.1 High fibre doses (25–100 fibre/ml/yr)2 are generally required to produce clinically significant asbestosis within an individuals’ lifetime3,4 with milder fibrosis at lower dose levels.5 Asbestosis is thus a marker of high asbestos exposure in individuals and its prevalence a potential indicator of high exposure in populations.
Asbestosis was the first disease to be attributed to asbestos exposure in the workplace,6 80 years later it is still a serious public health issue in industrialised countries.7 Not only is asbestosis a grossly debilitating, irreversible, and progressive disease,8 individuals who are afflicted have a greatly increased risk of an number of diseases including lung cancer, cancers of the digestive system, and non-malignant respiratory disease.9–12
National mortality and morbidity statistics underestimate the public health impact of asbestosis as they report the primary cause of death (COD) or primary diagnosis, which may not indicate or identify the underlying disability.13,14 This is particularly concerning as the National Asbestos Registers suggest that asbestosis is more widespread than malignant asbestos disease in New Zealand.15 Predictions of the scale of the asbestos disease epidemic have been modelled on mesothelioma data;16–18 the contribution of non-malignant disease such as asbestosis has been largely ignored. With a latency period inversely proportional to exposure level,2 but generally reported as 15–30 years19,20 the health burden of asbestosis should peak earlier than mesothelioma thus potentially providing useful additional information about the size and extent of the current epidemic of asbestos diseases.
In this study, the recorded health events of individuals diagnosed with asbestosis in New Zealand between 1974–2001 are examined. The purpose of the study is to highlight the public health importance of asbestosis and assess the overall health impact of this disease in New Zealand.
Mortality, cancer, and hospital records of New Zealand men diagnosed with asbestosis are examined to determine:
The potential contribution of morbidity data to existing models of the asbestos epidemic is also discussed.
National mortality records—All individual death, cancer, and hospital-discharge records that were available at the time of study were obtained on CD-ROM from New Zealand Health Information Service (NZHIS). These data comprised 690,198 death records covering a period from 1974–1999; 266,121 cancer records covering a period from 1980–1998; and 327,940 public and private hospital-discharge records covering a period from 1994–2001.
All records were supplied without identifying information as annual records in discrete Microsoft Excel or Borland dBASE IV. Individual death records with asbestosis (ICD-9 code 501) or lung cancer (ICD-9 code 162) anywhere in the death record were extracted from the mortality files.
All hospital discharge records for patients with a diagnosis of asbestosis in any of five diagnosis fields were extracted from the public and private hospital files. Two new databases were constructed from these files using an INFORMIX database; one comprising all relevant mortality records (n=24,590), and another comprising all relevant hospital discharges for male patients with asbestosis (n=706).
Database linkage—Individual death and hospital discharge records in each of the two main databases were linked separately via the unique encrypted healthcare user (ENC_HCU) number to the cancer registration database. All matching ENC_HCU records were crosschecked against sex and date of birth to minimise the chances of case mismatch.
Time series—All data were not available for all periods. Limitations were imposed by the availability of verified database records, ENC_HCU number for cross-referencing and relevant occupation codes. Three times series had the required characteristics:
Where required, the mortality and hospital discharge datasets were linked via their unique ENC_HCU number to each other and to New Zealand Cancer Registry records for 1980–1998. These series determined the overall structure of the study.
Mortality rates—Age standardised mortality rates were calculated using estimated population data obtained from HZHIS publications for the relevant years. For comparative purposes, all rates were standardised to the Segi World population.
Asbestos registers—The National Asbestos Registers were set up in 1991 upon the recommendation of the Asbestos Advisory Committee21 to register people who had had ‘significant exposure’ to asbestos. The National Asbestos Medical Panel reviewed cases referred via a notifiable occupational disease (NODs) card or a doctor’s letter. The number of cases registered between March 1992–July 1998 was obtained from the 1997–1998 report15 supplemented by figures for individual years obtained directly from OSH. The ACC Scheme Reporting and Forecasting Unit supplied figures for the number of claims for all ‘inhalation diseases’ by sex and year between 1992–1998.
Mortality trends (1974–1999)—316 New Zealand males died with asbestosis between 1974–1999. During the same period, 24,590 male deaths were associated with lung cancer; 90 (0.4%) of these record asbestos exposure or asbestosis in the death record. While male lung cancer declined from 1983, asbestos-related deaths generally increased over the period, rising from 2 deaths in 1977 to 40 deaths in 1999. For lung cancer deaths associated with asbestosis, annual numbers were small and fluctuating but there was a discernible trend in numbers with averages of 0.5, 2.1 and 6.6 cases per year in the 1970s, 80s, and 90s respectively.
The number of males dying with asbestosis increased steeply in the 1990s. This increase was still apparent when pleural cancer (used as a surrogate for mesothelioma) deaths were removed. Age-standardised mortality rates revealed similar trends with a more pronounced downward trend for lung cancer (Figure 1).
Cause of death (COD)—In the period 1988–1999, 264 asbestos associated male deaths were recorded. The majority (86%) were European; the median age for the whole group was 71 years (35–92 years). The contribution of asbestosis to the COD is shown in Table 1.
Table 1. The contribution of asbestosis to the deaths of 264 New Zealand males registered between 1988–1999.
*As coded in the national mortality data sets; †Cause of death (COD) indicated by the death certificate.
Only 44 (17%) of deaths were attributed directly to asbestosis—ie, asbestos was recorded as the ‘underlying’ COD. In the remaining cases (n=220, 83%), asbestos was recorded as a ‘contributing’ COD. The primary COD for these 220 decedents is shown in Table 2.
Table 2. The primary cause of death for 220 New Zealand males dying with asbestosis recorded as a contributing cause of death between 1988–1999
Seventy-three percent of deaths were attributed to cancer with a further 11% attributed to ‘other disease of the respiratory system’ including cases which may have been undiagnosed asbestosis. Circulatory disease accounted for only 12% of all deaths. The primary organ systems involved with malignant and non-malignant disease are shown in Table 3.
Table 3. Primary disease sites for New Zealand males dying with asbestosis 1988–1999
*Includes one cancer of other and ill-defined sites with asbestosis and mesothelioma.
Most (75%) patients died of malignant (64%) or non-malignant (11%) respiratory disease including:
In addition, 14 patients of the males died from circulatory diseases including ischaemic heart disease (63%), cerebrovascular disease (19%), diseases of the arteries (11%), and endocardium disease (7%). Nine patients died from digestive disease: seven (3%) were digestive cancers including neoplasms of the oesophagus (n=1), stomach (n=1), colon (n=2), peritoneum (n=2), and ill-defined site (n=1).
Overall, cancer was the primary cause of death in men dying with asbestosis between 1988-1999. The median age of death for those with cancer was lower than those without (69 and 74 years). Finally, out of a total of 264 deaths 209 (79%) were attributed directly to diseases of the airways with a further five deaths attributed to cancers of ill or undefined sites likely to have been cancers of the respiratory system.
Occupation—The decedent’s last occupation, as recorded on the death certificate, was available for deaths between 1988–1997. Three years were not available due either to known errors in the coding (1991) or coding discontinuity (1998,1999). 189 men dying with asbestosis during the study period were classified to 23 sub-major occupation classes (NZSCO90). One death was not classified.
In each occupational class, the number of deaths that could be attributed primarily to asbestosis, pulmonary fibrosis, pleural, lung or peritoneal cancer was determined (Figure 2). Trades workers (building, metal, and machinery) accounted for 76 (40%) of the deaths; professional and associated occupations accounted for a further 26 (14%) of the deaths.
Plant and machine operators also accounted for 26 (14%) deaths while labourers accounted for 15 (8%), workers with unclassifiable occupations 15 (8%), and corporate managers (including quarry and construction managers) 11 (6%) of all deaths, respectively
The remaining 19 (10%) deaths were distributed between seven occupational groups spanning clerks, technicians, service and sales workers, agricultural and fishery workers, and armed forces. Interestingly, 86% of deaths in physical science and engineering associate professionals were attributed primarily to pleural cancer while 70% of deaths amongst industrial plant operators were attributed to asbestosis, peritoneal or lung cancer—all indicators of high asbestos exposure.
Building trade workers had the highest number of deaths attributed primarily to pleural cancer (17/47), lung cancer (15/47), and asbestosis (5/47).
Morbidity—Credible hospital discharge records were only available from 1994 (NZHIS personal communication). During this period, there were 706 hospital discharges involving 539 patients. Most individuals were only hospitalised once during the period, however a significant number (n=123) of patients averaged three (range 2–11) hospitalisations during the period (1994–2001).
The majority of hospitalisations (n=450, 64%) were coded as acute admissions, 120 (17%) were arranged admissions, 113 (16%) waiting list, and 13 (2%) ACC-related. Over the whole period, not counting 97-day cases, patients were in hospital for an average of 7 days (total 4,319 days, median 5 days). The main reasons for hospitalisation, as indicated by the primary diagnosis, are shown in Table 4.
Table 4. Primary diagnosis for 706 hospital discharges for New Zealand males with asbestosis in the period 1994–2001.
In most cases (n=507; 72%), patients were admitted to hospital for treatment of respiratory disease (41%), circulatory disease (19%), or neoplasms (11%).
Basic hospital costs (excluding drugs) for ninety discharges recorded in 2001 are shown in Table 5.
Table 5. Hospital costs for 69 male New Zealand patients with asbestosis in 2001
*Diagnostic Related Groups
Half of the estimated cost of hospital discharges can be attributed directly to respiratory disease ($172,725)—with a tracheostomy, interstitial lung disease, chronic obstructive disease, and respiratory infections and inflammation accounting for 89% of this cost.
The temporal pattern of these discharges is shown in Figure 3 for 706 discharges involving 539 patients. Discharges rose steeply from 1994, peaked in 1999 and decreased successively in 2000 and 2001.
The National Asbestos Registers and Accident Compensation Corporation (ACC) claims—Between March 1992–July 1998, 136 males (mean age of approximately 66 years) were registered with asbestosis, the majority (80%) were, or had been, smokers. Approximately 25% were likely to have had high exposure to asbestos (ie, asbestos processors, asbestos sprayers, and watersiders handling raw asbestos).
Most of the reminder were employed in secondary industries and included plumbers, fitters, laggers, carpenters, and builders. The mean latency period was reported as 39 years (range 15–71 years). In contrast to all other sources, the National Asbestos Register recorded numbers falling sharply from 1992–1998. Registered ACC claims for the period could only be retrieved for all ‘inhalation diseases – Asbestos/Lead’ (personal communication).
Overall, during the period 1992–1998, 1705 claims were made by New Zealand males. The ACC Workwise Asbestos Claim Database recorded 78 claims in 1997, 42 (54%) claims were successful including 6 (14%) for asbestosis, 33 (79%) for mesothelioma, and 2 (5%) for lung cancer.
It has been estimated that 20%–40% of all adult men are likely to have some past occupation that may have exposed them to asbestos in the workplace.5 In New Zealand, over 8,000 men were directly employed in the asbestos industry (Report of the Asbestos Advisory Committee 1991) with at least a further 1500 men estimated to have been exposed in ‘downstream’(ie, secondary) industries.22–24
Before 1940, virtually all asbestos products were imported into New Zealand. Thereafter, raw asbestos was also imported and manufactured into asbestos-containing products (mainly cladding and pipes). Imports of raw asbestos peaked at around 12,500 tonnes in 1974.18
The import of raw amphibole (blue and brown) asbestos into New Zealand was banned in 1984; chrysotile (white) asbestos was banned in 2002. Workforce regulations to protect employees were not drafted until 1978—in 1983, employers were obliged to inform workers of the particular dangers of smoking in asbestos workers. In New Zealand, asbestos exposure in the working population was probably highest between the 1940s–1980s. The legacy of this exposure is an ageing population of men with asbestos-related disease, which includes some of the most debilitating malignant and non-malignant diseases of the lung.
The unfolding New Zealand mesothelioma epidemic was described by Kjellstrom and Smartt,18 however the full health impact of occupational asbestos exposure is likely to be much greater. In this study, an increasing trend in non-malignant disease similar to that reported elsewhere13 is documented.
Official mortality statistics compiled from the primary COD are known to underestimate the health impact of asbestosis.10,11,13,14 This is supported in the present study, with only 17% of asbestosis-related deaths being reported in the published mortality tables (Table 2).25
Of the remaining 220 (83%) deaths, cancer (particularly cancer of the respiratory tract) was sited as the primary COD. Indeed, the increased risk of cancer in individuals with asbestosis has been widely reported, with lung/pleural and gastrointestinal cancers predominating.9–12 In the present study, the primary cause of death for a high proportion (29%) of 264 patients dying with asbestosis was pleural cancer (with a further 23% attributed to lung cancer).
It has been estimated that between 6%–23% of lung cancer deaths could be attributable to asbestos exposure in the workplace.26–28 In this study, only a very small proportion (0.4%) of individuals dying with lung cancer had any mention of asbestos exposure or asbestosis in the death record. While this proportion peaked at 0.7% (n=12) in 1997, it is still far short of the minimum expected 6% (n=56) cases predicted by some studies,26,27 or the minimum of 74 cases suggested by Kellstrom and Smartt.18
ACC claims and the National Asbestos Register similarly under-represent lung cancer cases in New Zealand as elsewhere.14,26 Since increases in the number of asbestos-related lung-cancer deaths are predicted to occur at a time when male lung cancer deaths overall are decreasing (Figure 1), the proportion of asbestos-related lung-cancer deaths may be greater than predicted. Thus, more active attempts to diagnose asbestosis and determine likelihood of asbestos exposure in the current male population, particularly those who have been smokers or who have lung cancer, is warranted.
Under-reporting of asbestos-related lung cancer mortality has been attributed to possible lack of awareness of exposure and the possibility that workers may not remember casual exposure 20–40 years earlier. However, the most likely reason that asbestos-related lung cancer deaths are under-reported is the very high prevalence of smoking (80%–85%) in the occupations most likely to be exposed to asbestos in the workplace and an assumption that lung cancer in smokers is most likely to be causally related to tobacco.29
This finding is particularly disturbing, as it has been shown that the likelihood of lung cancer in asbestos-exposed workers who also smoke is 5 times greater than the likelihood of lung cancer in smokers not occupationally exposed to asbestos, and 10 times greater than for non-smoking asbestos workers (seminal work of Selikoff reported in Frank 1979).30
Eighty percent of men registered with asbestosis on the national register were smokers or ex-smokers. Smoking and asbestosis has been less studied than smoking and asbestos-induced lung cancer. However, it is known that smoking inhibits airways clearance of fibres, contributes to the severity of asbestosis, and (in patients with progressing asbestosis) is a significant predictor of lung cancer.31–34 The main point here is that, in smokers, asbestos exposure cannot be ignored because of the hugely magnified risk of lung cancer.
Indeed, in many current and ex-smokers, a clinical diagnosis of asbestosis may be the first indication of this increased risk. Immediate cessation of smoking, prompt treatment for respiratory infections, and regular screening for lung cancer and related malignancies is considered essential in all cases of asbestosis.35
The effect of past exposure to asbestos on workers in the secondary (or ‘downstream’) asbestos industries such as construction, shipbuilding, automobile, and railway repair workers has been widely reported and summarised for chrysotile asbestos by the World Health Organization.36 It has been argued that such exposures have often been overlooked37 particularly if employment was of short duration. However, in terms of health effects, there is no threshold for asbestos exposure, and a strong dose dependency has been demonstrated;36,38 both long-term low-exposure and short-term high-exposure may result in serious lung disease such as asbestosis.2,39,40
In this study, 23 different occupational groups were recorded in the death certificates of men dying with asbestosis in New Zealand (Figure 2). The prevalence of asbestos-related disease (indicated by the last known occupation) was highest in New Zealand trades workers, plant and machine operators, labourers, and casual workers—a picture confirmed in the National Asbestos Registers Reports and studies in Europe and America.14–16,35,41,42 However the data reported here also suggest that some parts of the workforce in all major occupational groups may have had some level of past exposure to asbestos.
Most work highlighting the health impact of asbestos has arisen from the reporting of mortality data; however, morbidity and quality-of-life issues are also an important part of the impact of asbestos on population health. In this study, the burden of hospital care of patients with asbestosis is examined to highlight this issue. Asbestosis sufferers have a large number of acute hospital admissions resulting, for many, in substantial (5–7days) episodes of care. Most are hospitalised with respiratory problems (Table 4) often requiring costly interventions for conditions such as interstitial and chronic obstructive lung disease, respiratory infections, and inflammation.
Interestingly, during a period when lung cancer in males overall has been declining, hospitalisation for patients who have asbestosis has steadily and substantially increased. In these patients, hospitalisations peaked in 1999, then declining progressively through 2000 and 2001 (Figure 3). This decline may be short-lived; however, if it continues, it may contribute significantly to the accumulating information on the scale and timing of the asbestos disease epidemic. Indeed, hospitalisation rates may be used to estimate the prevalence of asbestosis; information that is difficult to obtain from other sources.
Predictions of the scale of the asbestos epidemic have been modelled using mesothelioma data; however since the average latency period for the development of asbestosis is reported to be approximately 10 years shorter than for mesothelioma, trends in health events relating to asbestosis may be expected to provide the first evidence that the asbestos epidemic is peaking.
There are a number of well-known limitations in a study of this type.41 Inaccuracies are known to occur in the information obtained from death certificates. Causes of death can be misclassified because of confounding factors (such as smoking), other occupational exposures, and comorbidities contributing to the under-recognition of asbestos-related mortality. This is often exacerbated further by uncertainty relating to diagnostic criteria for asbestosis and differential diagnoses of idiopathic pulmonary fibrosis or congestive heart failure.8
Occupation recorded in the death record only relates to the last full-time occupation, and may not be an accurate or fair indication of asbestos exposure. However, the occupations highlighted in this study are the same as those highlighted in cohort studies suggesting that the overall picture obtained from these records may have some general validity. Finally, improvements in the diagnosis of asbestosis, raised awareness, and improvements in data collection have undoubtedly contributed to increase trends reported in this study; however, with the exception of asbestos-related lung cancers, the changes observed are too great to be explained in terms of these factors alone.
The bulk of the reported worldwide evidence suggests that occupational diseases (including asbestosis) are under-diagnosed and under-reported22,43,44 and that reported cases are ‘just the tip of the asbestos diseases iceberg’.
Author information: Pamela Smartt, Senior Research Fellow, New Zealand Health Technology Assessment (NZHTA), Department of Public Health and General Practice, Christchurch School of Medicine, Christchurch
Acknowledgements: Most of this work was undertaken while the author was employed as a Senior Research Fellow with the HRC-funded New Zealand Environmental and Occupational Health Research Unit, Division of Community Health, University of Auckland.
I thank Professor Tord Kjellstrom (for his interest in this work and comments on earlier drafts); Chris Lewis, Senior Information Analyst, New Zealand Health Information Service (for his support in supplying anonymous mortality, cancer registration and hospital discharge records); and Roger Marshall, Senior Lecturer in Epidemiology and Biostatistics, School of Population Health, University of Auckland (for critiqueing the paper and provided helpful comments).
Correspondence: Dr Pam Smartt, Senior Research Fellow, New Zealand Health Technology Assessment (NZHTA), Department of Public Health & General Practice, Christchurch School of Medicine, PO Box 4345, Christchurch. Fax: (03) 364 1152; email: firstname.lastname@example.org
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads)
Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals