SEARCH

SEARCH BY CITATION

Keywords:

  • cancer;
  • mortality;
  • risk;
  • veteran;
  • cohort

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting information
  10. Supporting Information

Objective: To quantify the risk of incident cancer and cancer-related mortality in Australian Government Department of Veterans’ Affairs (DVA) clients.

Methods: A population-based record linkage study of 75,482 adult clients residing in New South Wales (NSW) from 2000 to 2007; median age 75 years (interquartile range, 68–79); 57% male. Standardised incidence ratios (SIRs) and mortality ratios (SMRs) for any cancer and by cancer type were calculated, relative to the NSW population.

Results: The risk of any cancer was slightly increased for males (SIR 1.07, 95%CI 1.04–1.10) but not females (SIR 1.00, 95%CI 0.96–1.04). Males exhibited a significantly elevated risk of prostate cancer (SIR 1.08), cutaneous melanoma (SIR 1.19), head and neck cancer (SIR 1.27) and connective tissue cancer (SIR 1.52). Females did not exhibit excess risk for any cancer type. Risk of cancer death was significantly reduced for any cancer (male SMR 0.78, 95%CI 0.75–0.81; female SMR 0.80, 95%CI 0.76–0.85) and for a range of haematopoietic and solid neoplasms including prostate (SMR 0.57), breast (SMR 0.62) and colon cancer (male SMR 0.67; female SMR 0.71).

Conclusion: Cancer incidence rates are largely similar, and mortality rates moderately lower, for DVA clients compared to the NSW general population.

Implications: These risk patterns may reflect service-related history, a healthy-survivor effect, competing risk of death, and/or comprehensive health care entitlements with minimal to no co-payments. Our findings suggest DVA clients are probably accessing cancer screening services. Outcomes after cancer diagnosis are good, most probably due to comprehensive health care entitlements.

Veterans constitute unique populations with respect to medical history, physical and psychological stress and occupational exposures.1,2 The Australian Government Department of Veterans’ Affairs (DVA) subsidises health care for veterans, serving and former defence force members, war widows and widowers and their dependants. DVA clients are a major subgroup of the Australian population, constituting about 6% of those aged 65 years and older and 27% of Australians aged at least 80 years.3 The DVA maintains comprehensive and up-to-date information about their clients and actively promotes research, including the linkage of client health records with other administrative and population-level databases.

Cancer risk and mortality has previously been explored in subsets of the DVA population, specifically Vietnam and Korean War veterans. Self-reported health surveys of male Vietnam and Korean war veterans indicate a higher prevalence of cancer, cardiovascular disease, liver disease, kidney disease, diabetes, atopic disease and mental health conditions compared with the general Australian population.2,4 Cancer registry based incidence studies have found that male Vietnam war veterans have a 15% excess risk of cancer compared to the general population, an increased risk of cancer of the head and neck, lung, prostate, eye and melanoma and Hodgkin disease, and a lower risk of liver and thyroid cancer, multiple myeloma and non-Hodgkin lymphoma.5 Korean war veterans have a 23% increased risk of cancer, driven by smoking-related cancers, cancer of the lung, head and neck, larynx and oesophagus, but also cancer of the prostate, colon, rectum, unknown primary site and melanoma.1 Cancer-related mortality studies are less consistent, with male Vietnam veterans being shown to have reduced cancer-related mortality,5 and male Korean War veterans elevated cancer-related mortality.6 Little research has been conducted on the health outcomes of female veterans. A single self-reported medical survey found excess reports of any cancer and breast cancer in female Vietnam veterans compared with the general population.7

Importantly, there have been no population-based studies examining the cancer risk of the entire DVA client population, irrespective of service history. Data of this kind has implications for our understanding of the health seeking and care needs of the entire DVA client base. Therefore, we quantified the risk of incident cancer and cancer-related mortality in adult male and female DVA clients residing in New South Wales (NSW) during the years 2000 to 2007, using linked health administrative data.

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting information
  10. Supporting Information

Setting

Australia's publicly funded universal health care system entitles all citizens and permanent residents to a range of health services including: treatment in public hospitals; subsidised treatment in private hospitals; subsidised outpatient services, including consultations with clinicians (Medicare Benefits Scheme, MBS); and subsidised access to medicines prescribed in some hospitals and the community (Pharmaceutical Benefits Scheme, PBS).

This study focuses on clients of the Australian Government DVA. The DVA funds the health care of eligible veterans, war widows and widowers and their dependants. Eligible persons with Gold Repatriation Health Cards (Gold Card Holders) are entitled to treatment for all conditions (i.e. all health services provided to Australian citizens plus additional DVA-approved services and pharmaceutical items not available to the general population). White Card Holders are entitled to treatment for specific conditions approved by the DVA (other conditions will be treated and subsidised according to the general population entitlements). The Orange Repatriation Pharmaceuticals Benefits Card provides eligible British, other Commonwealth or allied veterans subsidised access to approved pharmaceuticals according to clinical need.8 Almost all male DVA clients are veterans of the Australian forces and almost all female clients are dependants.3 Most male DVA clients are Caucasian.5

Study population

Our study population was all 75,803 adult (≥18 years) DVA clients with health care entitlements residing in NSW from 1 May 2000 to 31 December 2007. DVA clients residing in NSW account for 34% of the Australian DVA population and have a similar age and gender profile to clients residing in other Australian states.3 We excluded DVA clients from our analyses if their health care entitlement status was unknown or if there were inconsistencies in date of birth or date of death between the linked administrative datasets (n=321, 0.4%). Our final study population comprised 75,482 clients.

Data sources

The DVA client file records the name, address, sex, date of birth, date of death, and health care entitlement and NSW residential history (from 1 May 2000) for all clients residing in NSW at any time from 1994. The NSW Central Cancer Registry (CCR) is a population-based registry of all incident primary cancers, excluding non-melanoma skin cancer, diagnosed in NSW residents since 1972. The registry is managed according to the International Association of Cancer Registries (IACR) rules and classifies cancers according to the International Classification of Diseases for Oncology, Third edition (ICD-O-3).

We also obtained data from the NSW CCR on the site-specific cancer incidence and cause-specific mortality rates for the entire NSW population by five-year age group, sex, and calendar year from 2000 to 2007.

Record linkage

Incident invasive cancers and cancer-related deaths were ascertained by linking the DVA client (1994–2007) and NSW CCR (1994–2007) records. Linkages were undertaken by a third party, the Centre for Health Record Linkage (CHeReL) using probabilistic matching and best practice privacy-preserving protocols. The DVA provided the CHeReL with unit record level information on client names, sex and dates of birth and death. For each matched record, we obtained the month and year of invasive cancer diagnosis, topography and morphology codes, and the date of cancer-related death and the underlying cancer cause of death. The degree of cancer spread (localised, regional, metastatic, or unknown) was also obtained for each cancer notification. In situ cancers were not included in these analyses as they are not routinely reported to the NSW CCR. Identifying information about DVA clients such as name and address details were not released to the researchers.

Statistical analyses

Person-years (PY) of client follow-up for an incident cancer were accrued from 1 May 2000, the date the client commenced DVA health care entitlements, or the date they moved to NSW – whichever occurred last – until the date of cancer diagnosis, death, the date they no longer resided in NSW, or 31 December 2007 – whichever occurred first. Clients with a history of cancer prior to the start of follow-up (n=3,437), were not excluded, however, they did not contribute person-years at risk for that cancer type. All incident primary cancers were counted, whether they were first or additional cancers, according to international rules for cancer registration. Follow-up for cancer mortality (from prevalent or incident cancer) was from 1 May 2000, the date the client commenced DVA health care entitlements, or residency in NSW, until the date of death, 31 December 2007 or the date they no longer resided in NSW – whichever occurred first. We calculated the expected numbers of incident cancers (or cancer deaths) by multiplying population cancer incidence (or mortality) rates for each five-year age group, sex, and calendar year strata by the corresponding cohort PY at risk. We also calculated standardised incidence ratios (SIR) and standardised mortality ratios (SMR), the ratio of the observed and the expected numbers of cancers (or cancer deaths) with exact 95% confidence intervals (95%CI) using SAS v11.5. SIRs and SMRs were computed for all cancers combined and for each cancer type where there was at least one expected case. For cancers occurring at excess or reduced risk, and for the most frequently occurring cancers, we calculated SIRs for sub-cohorts stratified by card type and gender. We compared patterns of cancer risk for the card type subgroups but could not compare SIRs statistically because of the heterogeneity in subgroup age and sex distributions.9 For breast, colon and prostate cancers, we stratified sub-cohorts by age-group at start of follow-up (50–69, 70–79, 80+ years) and gender to examine any differentials with respect to age groups targeted for cancer-screening.

This study was approved by the New South Wales Population and Health Services Research Ethics Committee (Approval Number 2008/02/060) and the Department of Veterans’ Affairs Human Research Ethics Committee (Approval Number E008/003).

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting information
  10. Supporting Information

Cohort characteristics

The median age of the DVA clients was 75.4 years (interquartile range, IQR, 67.7–78.9) and 43% were female. The cohort accrued a total of 290,608 PY, with a median of 4.7 years (IQR 2.5–5.3) follow-up per client (Table 1). At the start of follow-up, 94% of the cohort held a Gold or White health care entitlement card (76% and 18% respectively). Gold cardholders were 44% male; White and Orange card holders were predominantly male (90% and 98% respectively). Seventy-seven per cent of Gold and more than 99% of Orange card holders were aged ≥70 years, while 64% of White card holders were younger than 70 years.

Table 1. Characteristics of adult DVA clients residing in NSW between 2000 and 2007.
 N = 75,482%Median person-years follow-up (IQR)
  1. * Excluding non-melanoma skin cancer

Gender  Female  Male 32,226  43,256 42.7  57.3 4.8 (2.5–5.6)  4.7 (2.4–5.2)
Age at start of follow-up (years)  18–39  40–44  45–49  50–54  55–59  60–64  65–69  70–74  75–79  80+ 2,274  1,201  1,886  5,715  2,849  2,888  4,768  14,134  25,374  14,393 3.0  1.6  2.5  7.6  3.8  3.8  6.3  18.7  33.6  19.1 3.2 (1.2–4.8)  4.2 (1.8–4.9)  4.2 (1.8–5.0)  4.6 (2.2–5.3)  4.5 (2.3–5.4)  4.7 (2.1–5.5)  4.8 (2.8–5.5)  4.8 (2.9–5.5)  4.8 (2.9–5.4)  4.3 (2.2–5.0)
Health entitlement at start of follow-up  Gold  White  Orange  Orange and White 57,505  13,296  3,985  696 76.2  17.6  5.3  0.9 4.8 (2.6–5.4)  4.1 (1.9–4.8)  5.4 (2.8–5.6)  4.6 (3.0–4.8)
Incident primary invasive cancer*  Yes  No 6,427  69,055 8.5  91.5 4.6 (2.8–5.2)  4.8 (2.4–5.4)
Age at diagnosis of cancer (years)  <60  60–64  65–69  70–74  75–79  80+ 291  203  206  384  1,352  3,991 4.5  3.2  3.2  6.0  21.0  62.1 4.8 (3.6–5.5)  4.7 (3.4–5.5)  4.8 (3.3–5.6)  4.7 (2.7–5.5)  4.6 (2.5–5.2)  4.5 (2.8–5.0)

Cancer incidence

A total of 6,427 (9%) DVA clients matched to at least one incident cancer notification; 6,063 clients had one cancer and 364 had two or more. The most common cancers were prostate (20%), lung (11%), colon (10%), cutaneous melanoma (10%) and breast (5%), see Figures 1a–b. Localised disease accounted for 44% of prostate cancers, 48% of breast cancers and 79% of cutaneous melanomas, while distant metastatic disease contributed 7%, 8% and 5% respectively.

image

Figure 1a. Risk of incident cancer in male DVA clients residing in NSW, 2000 to 2007.

Download figure to PowerPoint

The overall cancer risk was slightly elevated for males (SIR 1.07, 95%CI 1.04–1.10) but not for females (SIR 1.00, 95%CI 0.96–1.04) (Figures 1a–b). Male DVA clients were at an increased risk of cancer of the prostate, head and neck, connective tissue, and cutaneous melanoma, and at decreased risk of lymphoid leukaemia, cancer of the pancreas and cancer of unknown primary site. Female clients were not at increased risk of any cancer type and were at a decreased risk of pancreatic cancer, cancer of unknown primary site, and other lymphoid haematopoietic neoplasms.

image

Figure 1b. Risk of incident cancer in female DVA clients residing in NSW, 2000 to 2007.

Download figure to PowerPoint

When stratified by health care entitlement, our findings for Gold and White card holders were similar to those observed for all card holders for cancers showing an increased risk. Cancers showing a decreased risk for all card holders were only decreased for Gold card holders (Supplementary Table 1, available online). Male orange card holders aged at least 70 years were at a decreased risk for cancer of the prostate, colon and cutaneous melanoma. Stratification by age-group further showed that only male clients 50–69 years were at increased risk of prostate cancer and female clients in the same age group were at increased risk of breast cancer (data not shown).

Cancer mortality

Of the 14,070 clients who died during follow-up, 3,485 (25%) died of cancer. Lung cancer was the most common (21%), followed by prostate (11%), unknown primary (8%), colon (8%) and rectum (5%). DVA clients were at a decreased risk of dying from any cancer (male SMR 0.78, 95%CI 0.75–0.81; female SMR 0.80, 95%CI 0.76–0.85) and from a number of specific cancer types, including the common cancers, prostate, colon, lung (men only) and breast cancer (Figures 2a–b).

image

Figure 2a. Risk of death from cancer in male DVA clients residing in NSW, 2000 to 2007.

Download figure to PowerPoint

image

Figure 2b. Risk of death from cancer in female DVA clients residing in NSW, 2000 to 2007.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting information
  10. Supporting Information

This population-based cohort study quantified the cancer incidence and mortality profile of all DVA clients residing in NSW. Overall, and for some site-specific cancers including prostate and melanoma, male clients had a slightly elevated risk of cancer. However, both male and female clients had a reduced risk of cancer death. Given the age of the cohort, it is likely a healthy survivor effect has played a role in our cancer incidence and cancer mortality findings. That is, given the increased risk of the Australian male veteran population to death from chronic diseases and external causes,2,6 those who survived beyond 70 years of age may have been healthier than members of the general population in the same age group. It is also possible that a competing risk of death from other causes played a part in the reduced risk of cancer death. Finally, it is also probable that the elevated cancer risk and the reduced risk of death from cancer was associated with clients’ relatively unrestricted access to health care services compared to the general population.

Increased cancer risk was observed for two of the most common cancers in males, prostate cancer and cutaneous melanoma. Males also had increased risk of head and neck cancer and connective tissue cancer. These results were partially consistent with those from previous studies of Australian male veterans, although caution is required with such comparisons because our study population includes DVA clients without a history of war service. We could not establish the service record of our individual cohort members; however, 99.7% of male and 8.1% of female DVA clients are veterans, the remainder are dependants with no service-related history.3 Australian male Vietnam and Korean War veterans (median age <55 years) have been shown to have an increased risk of prostate cancer, cutaneous melanoma and head and neck cancers.1,5 These previous studies also found increased risk of lung cancer, cancers of the head and neck (Vietnam), larynx and oesophagus (Korea) and Hodgkin disease (Vietnam).

The observed excess risk of prostate cancer and cutaneous melanoma in male DVA clients may reflect service-related exposures. However, a within-cohort analysis of Vietnam servicemen found no significant difference in risk of either of these neoplasms for men who served in Vietnam and those who did not, suggesting non-service related exposures may be responsible for the association.5 Furthermore, in our subgroup analyses, the excess risk of prostate cancer was restricted to those aged 50–69 years, suggesting that DVA clients in this age group may be more likely to access prostate-specific antigen (PSA) tests than the general population. The excess risk of melanoma may be due to the predominance of Caucasian men in the Australian DVA population, as Caucasian ethnicity is an established and strong risk factor for melanoma. It is also possible that DVA clients, with subsidised access to health care, were more pro-active than the general population in seeking medical advice for signs related to cancers. In mixed support of this theory, higher proportions of localised prostate (65% vs 58%; p=0.03), but not melanoma (79% vs 84%; p=0.15), were diagnosed in DVA clients aged 50–69 years compared to the corresponding general NSW population 2000–2007. Furthermore, Orange card holders, who do not have the same level of health care entitlements as Gold and White card holders, had decreased risk of prostate cancer and cutaneous melanoma. However, comprehensive health care access does not explain all the cancer incidence patterns. Subsidised health care with minimal co-payments may facilitate earlier presentation to a health practitioner, and reduce the risk of a diagnosis of cancer of unknown primary site. Thus, it would have been expected that Orange card holders would have an increased risk of cancer of unknown primary site, but the opposite was observed.

We did not observe an excess risk of cancer overall or of any specific cancers in female DVA clients. This finding is to be expected as only 8% of females had a history of war service, thus they would be broadly representative of the general female population. The only prior study of Australian female veterans found an excess risk of self-reported breast cancer in Vietnam veterans.7 We did not find an increased risk of breast cancer in female DVA clients, except for women aged 50–69 years, corresponding to the age range Australian women are eligible for free screening mammograms. This finding may indicate that female DVA clients are more likely to access this health service than the general population.

There was a significantly reduced risk of death from cancer in our study. The “healthy soldier effect” is unlikely to have contributed on account of the age of our cohort and thus the period of time since recruitment into service.10 On the other hand, it is possible that a ‘healthy-survivor’ effect was prevailing in men as they were survivors from a group who had died at a greater rate compared to their peers. The decreased risk of death from cancer in our cohort may also have been related to continued risk of dying from these other causes, and possibly to their unrestricted access to health care services. Prior cancer mortality studies have investigated younger cohorts (mean age <60 years) of male veterans. One study observed a lower risk of death from cancer in national servicemen from the Vietnam War;5 the overall reduction in risk was 16% and the only specific cancer to show a significantly lower risk of death was kidney cancer. In contrast, another study of Australian Vietnam veterans showed significantly elevated all-cancer and lung cancer mortality in those followed up for more than 20 to 30 years.10 Furthermore, veterans of the Korean War have shown an excess risk of cancer death overall, and specifically from oesophageal, colorectal, head and neck, lung, and prostate cancer.6, 10

There are no prior mortality data for female Australian veterans or dependants. Neither a healthy cohort nor healthy survivor effect explains the lower cancer-related mortality we observed in female DVA clients, most of whom were dependants, and not recruited as healthy workers to the armed services. Thus, comprehensive health care entitlements may play a role in the reduced risk of cancer-related death in male and female DVA clients.

This was a large, population-based study of Australian DVA clients, a population that includes veterans and their dependants. The data are comprehensive, encompassing all cancer sites and both incidence and mortality in men and women. However, our analysis could not incorporate individual service history, lifestyle factors such as diet, alcohol or tobacco consumption, or specific environmental and occupational exposures. Furthermore, it included only NSW residents, although NSW-based DVA clients are similar to DVA clients in other Australian jurisdictions in terms of age and sex, except for the Northern Territory and the Australian Capital Territory where DVA clients are younger.3 As linked cancer records were only available until 1994, any individuals diagnosed with cancer prior to this year were not removed from the at-risk population, biasing our cancer risk estimates towards the null. In addition, misclassification of the underlying cause of death cannot be excluded and, if it occurred systematically in favour of non-cancer related conditions, will have contributed to our finding of a reduced risk of cancer-related mortality. Finally, the end of our study period is 2007 so the findings may not necessarily reflect outcomes in the past five years. However, our study's cancer notification and mortality data currently represents the most contemporary data available in Australia.

Conclusion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting information
  10. Supporting Information

This study provides evidence for the current cancer-related health service needs of DVA clients, and demonstrates their cancer-related survival benefits relative to the general population. Our descriptive analysis shows the cancer incidence and death profile for a major sub-group of the Australian population. There is an increased risk of some common cancers in male, but not female DVA clients. The latter finding is not surprising given that most female clients are dependants of serving veterans and thus broadly representative of the female general population. Risk of cancer death, however, was significantly decreased for all clients and may be related to a healthy-survivor effect in men or DVA clients’ comprehensive health care entitlements that results in relatively few financial barriers to access.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting information
  10. Supporting Information

This study was funded by a Cancer Australia project grant (ID568773). Sallie-Anne Pearson is a Cancer Institute NSW Career Development Fellow (ID09/CDF/2–37). Claire Vajdic is supported by an NHMRC Career Development Fellowship (ID1023159) and a Cancer Institute New South Wales Career Development Fellowship (ID10/CDF/2–42).

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting information
  10. Supporting Information
  • 1
    Australian Institute of Health and Welfare. Cancer Incidence Study 2003: Australian Veterans of the Korean War. AIHW Catalogue No.: PHE 48. Canberra (AUST): AIHW; 2003.
  • 2
    Department of Veterans’ Affairs. Volume 1 – Male. Vietnam Veterans Survey and Community Comparison Outcomes. In: Morbidity of Vietnam Veterans: A Study of the Health of Australia's Vietnam Veteran Community. Canberra (AUST): Commonwealth of Australia; 1998.
  • 3
    Department of Veterans' Affairs. Treatment Population Statistics: Quarterly Report – December 2010. Canberra (AUST): Commonwealth of Australia; 2010.
  • 4
    Sim M, Ikin J, McKenzie D. Health Study 2005: Australian Veterans of the Korean War. Melbourne (AUST): Monash University; 2005.
  • 5
    Wilson EJ, Horsley KW, van der Hoek R. Cancer Incidence in Australian Vietnam Veterans' Study. Canberra (AUST): Commonwealth Department of Veterans’ Affairs; 2005.
  • 6
    Harrex WK, Horsley KW, Jelfs P, van der Hoek R, Wilson EJ. Mortality of Korean War Veterans: The Veteran Cohort Study [report]. Canberra (AUST): Commonweatlh Department of Veterans’ Affairs; 2003.
  • 7
    Department of Veterans’ Affairs. Volume 2 – Female. Vietnam Veterans Survey and Community Comparison Outcomes. In: Morbidity of Vietnam Veterans: A Study of the Health of Australia's Vietnam Veteran Community. Canberra (AUST): Commonwealth of Australia; 1998.
  • 8
    Department of Veterans' Affairs. DVA-funded Health Services: DVA Eligibility Requirements [Internet]. Canberra (AUST): Commonwealth of Australia; 2013 [cited 2013 Jul 8]. Available from: http://www.dva.gov.au/service_providers/services/Pages/health_services.aspx
  • 9
    Rothman KJ, Greenland S, Lash TL. Modern Epidemiology. 3rd ed. Philadelphia (PA): Lippincott, Williams & Wilkins; 2008.
  • 10
    Waller M, McGuire AC. Changes over time in the “healthy soldier effect”. Popul Health Metr. 2011; 9 (7):9.

Supporting information

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting information
  10. Supporting Information

Additional supporting information may be found in the online version of this article:

Supplementary Table 1: Cancer risk (SIR) by health care entitlement card type for male and female DVA clients residing in NSW between 2000 and 2007.

Supporting Information

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting information
  10. Supporting Information
FilenameFormatSizeDescription
azph12168-sup-0001-S1.docx18KSupporting Information

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.