Article Text

Download PDFPDF

Original article
Mortality and cancer incidence in a cohort of male paid Australian firefighters
  1. D C Glass1,
  2. S Pircher1,
  3. A Del Monaco1,
  4. S Vander Hoorn2,
  5. M R Sim1
  1. 1Department of Epidemiology & Preventive Medicine, Monash University Australia, Melbourne, Victoria, Australia
  2. 2Statistical Consulting Centre, The University of Melbourne Australia, Melbourne, Victoria, Australia
  1. Correspondence to Dr D C Glass, Department of Epidemiology & Preventive Medicine, Monash Centre for Occupational and Environmental Health (MonCOEH), School of Public Health & Preventive Medicine, The Alfred Centre, 6th Floor, 99 Commercial Road, Melbourne, VIC 3004, Australia; deborah.glass{at}monash.edu

Abstract

Objectives To investigate mortality and cancer incidence of paid male Australian firefighters and of subgroups of firefighters by era of first employment, duration of employment and number and type of incidents attended.

Methods Participating fire agencies supplied records of individual firefighters including their job histories and incidents attended. The cohort was linked to the Australian National Death Index and Australian Cancer Database. SMRs and SIRs were calculated. Firefighters were grouped into tertiles by duration of employment and by number of incidents attended and relative mortality ratios and relative incidence ratios calculated. Analyses were carried out separately for full-time and part-time male firefighters.

Results Compared to the Australian population, there were significant increases in overall risk of cancer, for all paid firefighters SIR 1.09 (95% CI 1.03 to 1.14), in prostate cancer, full-time firefighters 1.23 (95% CI 1.10 to 1.37), part-time 1.51 (1.28 to 1.77), and melanoma full-time 1.45 (95% CI 1.26 to 1.66), part-time firefighters 1.43 (95% CI 1.15 to 1.76). Kidney cancer was associated with longer service in internal analyses for paid firefighters. Prostate cancer was associated with longer service and increased attendance at fires, particularly structural fires for full-time firefighters.The overall risk of mortality was significantly decreased and almost all major causes of death were significantly reduced for paid firefighters.

Conclusions Male paid firefighters have an increased risk of cancer. They have reduced mortality compared with the general population, which is likely to be a result of a strong healthy worker effect and likely lower smoking rates among firefighters compared with the Australian population.

Statistics from Altmetric.com

What this paper adds

  • Studies have shown that firefighters may be at increased risk of cancer, but most studies do not have good exposure metrics.

  • Exposure metrics were developed using number and types of incidents attended.

  • Australian male paid firefighters have an increased risk of overall cancer incidence, of prostate cancer and of melanoma compared with the general population.

  • Prostate and kidney cancers were associated with longer service in internal analyses.

  • Prostate cancer was associated with attendance at fires, particularly structural fires for full-time firefighters.

Background

During the course of their work, firefighters routinely encounter a wide range of hazards. At a fire scene, firefighters are potentially exposed to mixtures of particulates, toxic gases and fumes, oxidation and pyrolysis products, including many carcinogens.1–5

Meta-analyses have shown that firefighting is associated with a significantly increased risk of cancers of the stomach, colon, rectum, skin and brain,6 ,7 although most individual studies show non-significant elevations in risk estimates. Some studies have shown an increased risk for individual cancers such as digestive system cancers,8–10 testicular cancer,11 ,12 prostate cancer,10 ,12–14 non-Hodgkin lymphoma (NHL),7 mesothelioma,8 ,14 malignant melanoma10 ,14 and multiple myeloma.9 ,10 ,14 Some of these studies were based on cancer mortality rather than incidence, which underestimates the occurrence of certain cancers.

The overall mortality of paid firefighters is usually lower than that of the general population, although a recent large study found it to be similar to the general population.8 There is evidence of an association between firefighting and death from cardiovascular disease,15 ,16 and evidence of an increase in cardiac mortality particularly close to an alarm or turn-out.9 ,16 ,17

Exposure assessment has been limited in firefighter studies; the extent of exposure for individual firefighters has not commonly been assessed.7 Duration of employment is the most usual measure of extent of exposure.11 ,13 Some studies have identified the number of runs,9 and others have retrospectively estimated exposure for individuals using exposure matrices developed from fire station records.18 ,19 Little has been published relating quantified attendance at fires or other incidents. This study aims to fill this gap in the scientific literature.

The aims of this study were to examine mortality and cancer among Australian firefighters. The full report of the study is on the Monash University website.20 This paper presents findings for full-time and part-time paid firefighters and investigates the cancer and mortality risk of subgroups, based on type of employment, duration of firefighting employment, era of first employment and the number and type of incidents attended.

Methods

Each Australian state and territory has a fire agency that employs full-time paid firefighters, most of whom are based in urban environments. Rural areas are covered by volunteer firefighters. Three fire agencies also employ part-time firefighters. A federal agency covers airport firefighters.

Full-time firefighters are those who have paid full-time operational roles usually based at a metropolitan fire station. Firefighters typically have an 8-day rotation with two 10-hour day shifts, followed by two 14-hour night shifts, followed by 4 days off. Part-time firefighters are on call from home or other jobs, rather than on duty at a fire station. They are typically employed by urban firefighting agencies and attached to stations in semimetropolitan areas. They are on call 24 hours per day to respond to fire or rescue incidents as these arise.

Eight of ten Australian fire agencies supplied records of individual paid firefighters, including name, date of birth, job history and notification of death (if known). The commencement dates of the personnel records varied by agency, ranging from 1976 to 2003, and included all firefighters employed as of or after the commencement date.

The study included 17 394 full-time and 12 663 part-time firefighters. There were 1771 firefighters who had held part-time and full-time jobs and these were included as full-time firefighters in the analyses. Records of incidents attended by each firefighter were supplied by participating agencies. Incident data are collected contemporaneously about events to which individual firefighters are dispatched, including fires, rescues and non-emergency events including false alarms. Incident data are collected by all agencies, and de-identified data are contributed to a national data repository. Data are collected in accordance with a coding and recording protocol described in the national Australian Incident Reporting System (AIRS) Standard Manual. Fire agencies started incident data collection at different times between 1990 and 2003. For this study, incidents were categorised using AIRS codes as all incidents, all fires, and then structural fires, landscape fires and vehicle fires.

The cumulative number of incidents was calculated for each firefighter for each person-year of follow-up. Some paid firefighters may also be volunteer firefighters. All incidents attended were counted for a person regardless of whether they were a paid or volunteer firefighter.

Some firefighters' employment started before incident data collection (n=7480 full time, 3250 part time). In these cases, the mean number of incidents per year for the years where incident data were available was allocated to the years for which data were unavailable.

Firefighters with no recorded incidents (5371 full time and 4952 part time) were excluded from the incident analyses. Some firefighters ceased employment before the commencement of incident data collection (n=2422 full time, 2501 part time). Less than 10% of firefighters employed after incident data collection started had no incidents recorded (n=1466 full time, 2451 part time). Incident data from two small agencies were not included because they were not individually attributed or differed in the type of incidents attended (n=1483 full time).

The study outcomes were cancer incidence and mortality compared to the Australian population. The cohort was linked to the National Death Index (NDI) and the Australian Cancer Database (ACD), both held by the Australian Institute of Health and Welfare (AIHW). At the time of linkage, the NDI was nationally complete from 01 January 1980 to 30 November 2011 for cause of death coding. The ACD was nationally complete from 01 January 1982 to 31 December 2010, except for two states which were only complete until the end of 2009.

The AIHW used a probabilistic program to identify possible matches based on personal identifiers such as surname, first name and birth date. The underlying cause of death is coded to International Classification of Diseases, Ninth Revision (ICD-9)21 or Tenth Revision (ICD-10),22 and cancer incidence records are coded to ICD-10. Each potential match was scored as to the probability of it being a true match. In order to identify likely true matches, possible death matches were independently reviewed by two members of the study team and disagreements adjudicated by a third team member. The AIHW provided the researchers with a de-identified list of cancer cases that were scored as highly certain matches.

Individuals' at-risk period started from the date of first employment, or from when the agency data were considered complete, whichever was the later. The latter date was based on information from the relevant agency and examination of the yearly death rates by agency. These death rates were examined because of concern that when human resources (HR) systems changed, only those alive and currently employed were transferred. A date was chosen where the yearly death rate became stable.

Any cancers and/or deaths which occurred after each individual's cohort start date were included in the analyses. Follow-up continued until the date of death, or the end of follow-up based on when the national cancer and death data were complete. Only primary cancers were included in the results. However, if a person was diagnosed with more than one primary cancer, then all of them were included in the analyses.

Australian population data were used to calculate the expected numbers of deaths and cancers for each firefighter group based on 5-year age groups and sex-specific rates.23 ,24 The population and the cohort data include all primary incident cancers and multiple primary cancers in the same person.

The SMRs and SIRs were calculated for overall death and cancer, for major cancer and death categories, for all paid and separately for full-time and part-time firefighters.

Duration of employment groups were calculated for firefighters (>3 months <10 years, 10–20 years and 20+ years) and by era first employment (pre-1970, 1970–1994, post-1995). Firefighters employed for <3 months (n=386 full time, 245 part time) were excluded because their firefighting exposure would have been minimal.

Tertiles were calculated for each incident category (separately for full time and part time), based on the cumulative incidents for each person-year for each firefighter. The lowest tertile was then used as the comparison group in internal analyses.

SMRs and SIRs were calculated in comparison to the general population controlling for age in 5-year age groups and for calendar year. Internal comparisons used Poisson regression modelling with an offset term set to the number of person-years. Relative mortality ratio (RMR) and relative incidence ratio (RIR) were calculated. Internal analyses were based on 5-year age groups but collapsed to larger groups where numbers were small. Risks for longer employment duration were compared with the shortest duration (>3 months or <10 years). For the incident analyses, the lowest tertile was used as the comparison group. RMRs and RIRs could not be calculated where deaths and cancer categories had small numbers. All analyses were undertaken using Stata software (StataCorp. Stata Statistical Software: Release 13. College Station, Texas: StataCorp LP. 2013).

Two sensitivity analyses were also undertaken. First, the overall SMR was calculated after including death notifications provided by the agencies, but not found on the NDI. Second, the overall SMR was recalculated after excluding those firefighters employed for <1 year.

More details about the study methodology are provided in the technical report at Monash University website.20

Ethics approval was granted by the human research ethics committees of Monash University, the state and territory cancer registries, the AIHW and the National Coronial Information System.

Results

The cohort description is shown in table 1. The cohort was young; the average age at the end of follow-up (or death date) was 49.9 for full-time firefighters and 44.5 for part-timers and slightly less for firefighters in the incident analyses.

Table 1

Description of the cohort of male firefighters as of 30 November 2011

There were 780 deaths matched on the NDI and 1208 cancers (1107 individual firefighters) matched on the ACD among the 17 394 full-time firefighters and 286 deaths and 485 cancers (455 individuals) matched among the 12 663 part-time firefighters.

Online supplementary figure 1 shows the distribution of the number of incidents attended by individual firefighters. The mean (median) cumulative value for recorded incidents at the end of follow-up was >1230 (820) for full-time firefighters and 291 (112) for part-timers, but a small number of firefighters attended many more incidents; the maximum values were over 10 000. On average, full-time firefighters attended 62 incidents per year of which 47% were fires, compared with 23 incidents of which 79% were fires for part-time firefighters.

Supplementary figure

Distribution of the number of incidents attended by individual full-time and part-time firefighters on a log scale

Firefighters were divided into tertiles based on the cumulative numbers of incidents attended per person-year, for each of the five incident categories. For full-time firefighters, the tertiles for all incidents were: >0–383; >383–1053; and >1053; each tertile included 4500–7000 firefighters contributing 44 000–46 000 person-years. For part-time firefighters, the tertiles were: >0–46; >46–210; and >210; each tertile included 2500–5000 firefighters contributing 19 500–21 500 person-years (see online supplementary tables S1 and S2).

The overall risk of mortality and of almost all major causes of death was significantly reduced for full-time and part-time firefighters (table 2). Although still reduced compared to the Australian population, the cancer mortality risk for firefighters was comparatively higher than other major causes of death. The SMRs for combined part-time and full-time firefighters are similar to those for the separate groups.

Table 2

SMRs and 95% CIs for firefighter deaths to 30 November 2011 compared with the Australian population

A comparison of mortality outcomes by state shows little difference when compared to national data (data not shown).

Compared with the Australian population, the overall incidence of cancer was significantly elevated for full-time (SIR 1.08; 95% CI 1.02 to 1.14) and part-time firefighters (SIR 1.11; 95% CI 1.01 to 1.21) (table 3). For full-time, but not part-time firefighters, the elevation increased monotonically by duration of employment and there was a significant increase in overall cancers for those employed for 20+ years (SIR 1.09; 95% CI 1.02 to 1.16) (see online supplementary tables S3 and S4). Internal analyses showed no significant trend with duration for overall cancer for full-time or part-time firefighters (table 4). The overall cancer SIR was not increased with era of first employment (see online supplementary tables S5 and S6). There was no trend of increasing risk of overall cancer incidence with increasing tertile of incidents for full-time (table 5) or part-time firefighters (see online supplementary table S7), but there were significant trends with increasing attendance at vehicle fires (full time p=0.04, part time p=0.01). The SIRs for combined part-time and full-time firefighters are similar to those for the separate groups. There was little difference between states in overall cancer risk (data not shown).

Table 3

SIRs and 95% CIs for firefighters to 31 December 2010 compared with the Australian population

Table 4

Relative cancer incidence ratios and 95% CIs for firefighters to 31 December 2010 by duration of employment (adjusted for age and calendar period)

Table 5

Relative cancer incidence ratios and 95% CIs for full-time firefighters to 31 December 2010 by number of incidents and incident types in tertiles (adjusted for age and calendar period)

Melanoma was significantly increased for full-time and part-time firefighters compared with the general population (table 3). When melanoma rates were compared to the relevant state rates, the overall risk did not greatly change (data not shown). Among full-time firefighters, melanoma risk was significantly elevated in New South Wales (NSW), Victoria and Western Australia (WA). For part-time firefighters, the melanoma risk was elevated for NSW firefighters. Firefighters from Queensland did not show an elevated risk. When examined by duration of employment, firefighters showed a significant increase in melanoma for the longest duration categories in external analyses. For full-time firefighters with 10–20 years of employment, the SIR was 1.50 (95% CI 1.12 to 1.98) and with 20+ years, the SIR was 1.46 (95% CI 1.22 to 1.75) (see online supplementary table S3). For part-time firefighters with 20+ years of employment, the SIR was 1.78 (95% CI 1.25 to 2.46) (see online supplementary table S4), but internal analyses did not show an increased risk (table 4). There was a significant increase in melanoma for full-time firefighters in all eras of employment: pre-1970, SIR 1.58 (95% CI 1.24 to 1.98); 1970–1994, SIR 1.35 (95% CI 1.10 to 1.63); and post-1995, SIR 1.58 (95% CI 1.03 to 2.31) (see online supplementary table S5). For part-time firefighters, melanoma was only significantly increased when first employed before 1970 (SIR 2.32; 95% CI 1.38 to 3.67) (see online supplementary table S6). There was no association between melanoma risk and tertile of incidents attended for either full-time or part-time firefighters (table 5 and see online supplementary table S7).

Prostate cancer was significantly increased for full-time and part-time firefighters compared with the general population (table 3). When compared within the cohort, full-time firefighters showed a monotonic increase in risk with a significant trend test for duration of employment (p=0.02) (table 4). Full-time firefighters showed a statistically significantly increased risk in many incident group tertiles and a significant trend of increased risk of prostate cancer with increasing incidents for all incident groups (table 5). For part-time firefighters, some incident categories showed excesses of prostate cancer, but there were no significant trends (see online supplementary table S7).

There was no increase in kidney cancer for full-time or part-time firefighters compared with the general population (table 3). The RIRs were significantly increased for full-time firefighters employed for 20+ years (p=0.05), but this finding was based on only one cancer in the reference group (table 4). The analyses of all paid firefighters also show a significant trend of increasing kidney cancer risk with employment duration. There were no significant trends with incidents (table 5).

There was no increase in risk of lympho-haematopoietic (LH) cancer or subcategories compared with the general population. Among full-time firefighters, the RIR for LH cancer was statistically significantly increased for those employed for >10 years, with a significant trend test (p=0.01) (table 4). There was also an increased risk of NHL with increasing employment duration; the risk was significantly raised for those employed for 20+ years, with a significant trend (p=0.01) (table 4). The combined group of all paid firefighters also showed a significant trend for NHL. However, the risk of LH cancer was not statistically significantly raised for part-time firefighters. There were no clear patterns of increased risk with increasing incidents attended (table 5 and see online supplementary table S7).

Online supplementary table S3 shows an increased risk of male breast cancer for full-time firefighters who worked for 20+ years (SIR 3.44; 95% CI 1.12 to 8.04, n=5). The mesothelioma SIR was significantly raised for those who had worked for <10 years (SIR 5.82; 95% CI 1.20 to 17.00, n=3), also based on small numbers.

There were significantly fewer respiratory cancers than expected (table 3). Lung cancer SIR showed a monotonic reduction with employment duration and was significantly reduced for full-time firefighters employed for 20+ years (SIR 0.77; 95% CI 0.60 to 0.98, n=66) (see online supplementary table S3). The RIR for lung cancer showed no relationship with employment duration (table 4).

In sensitivity analyses, all firefighters with <1 year of employment were excluded, but this made little difference to the SMRs and SIRs (data not shown).

When part-time and full-time firefighters were combined, diabetes mortality and liver cancer incidence were significantly reduced.

Discussion

There was a significant increase in overall cancer incidence for full-time and part-time firefighters, while the all malignancy SMR was less reduced than other major causes of death. With earlier diagnoses and better cancer treatments, cancer is not necessarily fatal. This means that cancer incidence is a better measure of cancer rates than cancer mortality. Firefighters are selected, are well paid and are likely to have a higher standard of living than the general Australian population. In addition, they are less likely to live in rural and remote areas which have poorer cancer survival.25

The elevated rates of melanomas are consistent with other studies; in a meta-analysis, the summary risk estimate (SRE) was 1.32 (95% CI 1.10 to 1.57).7 Statistically significant excesses of melanoma have been found among Nordic firefighters aged 30–49 years (SIR 1.62; 95% CI 1.14 to 2.23),14 and among firefighters in California.10 Sun exposure is a significant predictor of melanoma. When state-based comparisons were made, firefighters from Queensland (a sunny state) did not show an excess of melanoma unlike firefighters from NSW, Victoria and WA.

Firefighters were diagnosed with prostate cancer more commonly than the Australian population and the risk increased with increasing years of employment. Elevated rates of prostate cancer have been observed in several other studies,6 and a meta-analysis found an SRE of 1.28 (95% CI 1.15 to 1.43).7 Recent US studies reported prostate cancer SIRs of 1.03 (95% CI 0.98 to 1.09)8 and 1.5 (95% CI 1.3 to 1.7)10 and an SIR for the 30–49 age group 2.59 (95% CI 1.34 to 4.52).14 In recent years, there has been an increase in prostate cancer diagnosis because of the availability of screening tests. However, if the increased rates of prostate cancer (and melanoma) in firefighters were a result of diagnostic bias, firefighters would have to be more likely to be screened than the general population. The fire agencies involved in this study do not offer prostate cancer or melanoma screening.

Duration of employment is a proxy for exposure—an increased risk with increasing duration would suggest that the risk is related to employment. This was seen in external but not internal analyses for prostate cancer and melanoma for full-time firefighters. Trends of increasing risk with increasing duration were seen in internal analyses for all paid and full-time firefighters for kidney cancer and NHL, though some of these findings are based on small numbers. A similar trend was seen for prostate cancer among full-time firefighters.

There were more mesotheliomas than expected for paid firefighters, but this was not a statistically significant increase. Mesothelioma typically has a latent period of 30–40 years26 and few of the firefighters in this study started employment before 1985, so it is unlikely that many employment-related mesotheliomas would have arisen. Firefighting is associated with increased mesothelioma risk in cohorts who started employment in earlier periods, with an SIR 2.29 (95% CI 1.60 to 3.19) in a US cohort,8 and a similar SIR 2.59 (95% CI 1.24 to 4.77) in a Nordic cohort.14

Some cancers in excess in other studies were also increased in some subgroups in this study. These included digestive system cancers,7 ,8 ,10 ,27 ,28 LH cancer,6 ,10 ,28 ,29 kidney cancer,7 ,8 ,10 ,27 ,28 testicular cancer,7 breast cancer30 and thyroid cancer.12

Lung cancer was not increased in firefighters in most previous studies,6 but an increased risk could be difficult to show against background tobacco smoking, particularly if firefighters were less likely to smoke than the comparison population. In this study, the lung cancer SIR reduced with increasing employment duration and was statistically significantly reduced for full-time firefighters employed for 20+ years, which contrasts with excess risk of lung cancer identified in recent studies in the USA8 and Nordic14 countries. This may suggest lower smoking rates in this more recent cohort, but there was no information available on individual smoking rates. Grouped smoking data from one full-time firefighter agency showed about 6% smoked between 2006 and 2011 compared with about 20% of the equivalent general male population over the same period. A reduction in incidence of respiratory cancers, and in mortality from ischaemic heart disease and chronic obstructive pulmonary disease (COPD), may be because firefighters smoke less than the general population.

Deaths from cardiovascular disease were significantly lower than the Australian population in almost all analyses. Evidence suggests that deaths from coronary heart disease are linked to some specific firefighting duties.16 ,31 The activities being carried out just prior to death were not available for this study.

Analyses by era (pre-1970, 1970–1994, post-1995) showed reduced SMRs particularly for the most recent era of first employment, which could be the healthy worker effect or a result of improved work practices and personal protective equipment in recent years. Breathing apparatus had been used by firefighters in most agencies since the early 1980s.

The healthy hire effect is thought to be particularly evident for paid firefighters who are selected partly on the basis of their high level of physical fitness.15 All fire agencies have strict entry requirements for fitness. The healthy hire effect may be compounded by the healthy survivor effect, whereby those who become ill leave the workforce.32 In 2014, there were no Australian national standards in relation to fitness for duty, although one of the agencies has a mandatory 2-yearly fitness assessment.

The study reduced the impact of the healthy worker survivor effect by identifying a date from which each agency considered the HR records were a complete roll of employed firefighters. This means that firefighters who later left because of ill health were included in the cohort.

The contemporarily collected and individually attributable incident data are a major strength of this study, and a noted limitation of previous studies of firefighters. Most previous studies have used duration of employment as a surrogate for exposure;6 the number of runs from the firefighter's fire station has also been used.9 Duration of employment has been shown to be inaccurate as an estimate of exposure.19 Studies have retrospectively estimated exposure for individuals from exposure matrices developed from fire station records and these methods appear more valid than duration of employment.18 ,19 The recent US study showed a strong correlation between fire-runs and time at the fire, and very good correlations between time at fire and fire-runs compared with duration of exposure.19 The incident data in this study identified the type and number of incidents including fires attended for each individual, but did not provide information about attendance duration for individual firefighters. The data had information for each incident on the time of the alarm, the time the incident is under control and the time that duties were completed. Some incidents go on for much longer than others. Landscape fires may go on for days or even weeks and firefighters may have more than one shift attending such fires, which is rare for structural or vehicle fires.

As expected, on average, full-time firefighters attend many more incidents than do part-time firefighters. The number of incidents attended was estimated for some individuals whose employment predated the commencement of incident data collection. This introduces some uncertainty, for example, where more senior firefighters may attend fewer incidents later in their career. In addition, firefighters who had no incidents recorded were excluded from these analyses, which reduced the mean age and the power of the study in these analyses. Thus, the completeness of the incident records is a limitation of the study.

The power of this study to determine cancer and mortality was limited by the small numbers of some events in some categories, the short period of follow-up and the relatively young age of the cohort. Continued follow-up of the cohort is recommended to improve power and to increase the proportion of the time for which incident data will be available.

For some outcomes, there are a number of potentially confounding risk factors, including individual genetic or lifestyle factors such as ethnicity, smoking, alcohol consumption, diet or non-firefighting job exposures, for example, in previous jobs and for other jobs held by firefighters. Genetic factors play a part in the risk of some diseases, and differences in cancer risk have been found between Hispanic, black and white firefighters.10 For example, prostate cancer is more common among African-Americans than Caucasian-Americans.33 Neither race nor ethnic origin was identified in agency records, nor are they captured in the AIHW cancer or mortality data and therefore could not be investigated. (Aboriginal and Torres Strait Islanders make up ∼2% of the Australian population.) Individuals with paler skin34 are more susceptible to melanoma. Sun exposure is an important risk factor for melanoma, but no information was available on this for the cohort members on an individual basis.

A major strength of this study is that cancer diagnosis and death registration are mandatory in all Australian states and territories, and registration is virtually complete for the time period of this study.35 Cancers diagnosed in states other than the resident state are identified. Cancers diagnosed overseas and not treated in Australia and deaths occurring overseas are not included, but this is unlikely to occur in numbers that would affect these findings. Previous validation studies of the NDI have found good sensitivity and specificity of the matching process, with 88% sensitivity36 and 98% specificity.36 Matching the cohort to the NDI and ACD is a probabilistic process. A thorough clerical review of possible death matches was carried out by the investigators, and a clerical review of possible cancer matches was carried out by the AIHW, using well-established protocols, so this factor is not likely to have an impact on the validity of the findings. All primary cancers are collected by the ACD and about 8% are cancers that arose after the primary cancer. It is possible that some of these may be treatment related, but these cannot be identified in the national reference data (nor therefore in the cohort); therefore, it was not possible to carry out a sensitivity analysis to assess their impact.

There is a likely latent period between first exposure and diagnosis of cancer, perhaps 10 years for leukaemia,37 around 10–15 years for solid tumours38 and 30–40 years for mesotheliomas.26 It was unlikely, therefore, that many cancers in this cohort would have arisen before 1982 when the ACD was complete. However, if employment only started in 1995, it is probably too early for many solid tumours to have been diagnosed that are related to workplace exposure.

In conclusion, despite some limitations, overall cancer incidence was significantly raised for male paid firefighters compared with the Australian population, and this was most strongly associated with increases in prostate cancer and melanoma. These cancers were significantly elevated in several analyses and were most evident for those employed for 20+ years in external analyses. Prostate cancer showed trends with duration and incidents in internal analyses, while melanoma showed a trend with duration but not incidents. The pattern of mortality and cancer incidence findings were similar for full-time and part-time firefighters, which is suggestive that these are employment-related effects.

Acknowledgments

The study was funded by the fire agencies through the national council and industry peak body, the Australasian Fire and Emergency Service Authorities Council (AFAC). Monash University would like to acknowledge the assistance of the participating agencies in compiling the data for the cohort. The authors thank the members of the Advisory Committee and the Technical Reference Group for their assistance. They also acknowledge the assistance provided by the Australian Institute of Health and Welfare for the timely linkage of the cohort and provision of reference data.

References

Footnotes

  • Collaborators AIHW.

  • Contributors DCG and MRS received funding from the Australasian Fire and Emergency Council for a cancer and mortality study of Australian firefighters which has been completed. They obtained the funding, designed and oversaw the study. SP, ADM and DCG carried out the data cleaning and matching. SVH carried out the statistical analyses. DCG wrote the draft manuscript, which has been seen, edited and approved by the other authors.

  • Funding The study was funded by a grant from the Australasian Fire and Emergency Service Authorities Council.

  • Competing interests None declared.

  • Ethics approval Ethics approval was granted by the human research ethics committees of Monash University, the state and territory cancer registries, the AIHW and the National Coronial Information System.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement Data sharing will be considered on a case-by-case basis, but will need to be discussed with the funding body. Contact the corresponding author initially.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.