Health Promotion International, Vol. 14, No. 1, 73-82,
March 1999
© Oxford University Press 1999
Could a national skin cancer primary prevention campaign in Australia be worthwhile?: an economic perspective
Health Economics Unit, Monash University, 1 Department of Medicine (Dermatology), The University of Melbourne and 2 Centre for Behavioural Research in Cancer, Anti-Cancer Council of Victoria, Australia
Address for correspondence: Rob Carter, Health Economics Unit, Centre for Health Program Evaluation, PO Box 477, West Heidelberg, Victoria 3081, Australia
SUMMARY
This study aimed to determine the potential cost-effectiveness of initiating a co-ordinated and comprehensive national skin cancer primary prevention campaign in Australia. The study is undertaken from the perspective of the Federal government as potential funder of a national initiative, but other perspectives are analysed. Using the SunSmart Campaign in the State of Victoria, Australia, as a model for the national program, estimates were made of the reduction in the incidence and associated premature mortality for skin cancer that would accompany reductions in sunlight exposure. Cost offsets to government flowing from the reduction in management costs for skin cancer were estimated and deducted from the cost of the health promotion campaign. Costs affecting individuals, such as the cost of sunscreen, were included in the sensitivity analysis. On the basis of the assumptions used (with a ‘do nothing’ comparator), a national campaign involving a 20-year commitment of $AUD 5 million annually (i.e. 28¢ per person) would avoid 4300 premature deaths and cost $AUD 1360 per life-year saved, or $AUD 14 360 per death deferred. If the cost offsets are included, the program should not only prevent premature deaths but also yield a net saving to government of $AUD 103 million (PV). In the sensitivity analysis the results were robust to a variety of cost and outcome variations. Increasing the lag between decreases in UVR exposure and falls in the incidence of melanoma from 5 to 15 years, and inclusion of costs affecting individuals and their families had the largest effects on cost-effectiveness. The cost per life-year saved (excluding cost effects) increased from $AUD 1360 to $AUD 5103 and $AUD 25 134 respectively. A survey of current expenditure by State/Territory government departments and anticancer bodies on skin cancer primary prevention was carried out as part of the analysis, and an annual commitment of $AUD 5 million by the Federal government represents a doubling of current expenditures (of approximately 14¢ per person). The cost-effectiveness (using a ‘current practice’ comparator) yields a cost of $AUD 2715 per life-year saved (ignoring the cost offsets and using a conservative estimate of improved health outcomes attributed to the new national program). The study results, although indicative only, strongly suggest that a comprehensive national skin cancer primary prevention program would be excellent ‘value for money’ from a variety of perspectives, using both ‘do nothing’ and ‘current practice’ comparators.
Key words: cost-effectiveness; primary prevention; skin cancer
BACKGROUND
Australia has the highest incidence and mortality rates for skin cancer in the world. Incidence for non-melanoma skin cancer (NMSC) is about 1000/100 000 people/year and that for melanoma is now in the order of 30/100 000/year (Marks et al., 1993
; Jelfs et al., 1996). This translates into at least 5500 people developing a melanoma and more than 150 000 developing a NMSC annually. More than 1000 people die each year, with 80% of these deaths being due to melanoma (Australian Bureau of Statistics, 1996). Table 1 presents a summary of morbidity, mortality and cost of illness information for melanoma and NMSC in Australia for 1993/94. Australia has rates for these tumours at least three times those in other countries that consider that they have a public health problem with skin cancer (Magnus, 1991; Ko et al., 1994; Miller and Weinstock, 1994).
|
Sunlight is now widely accepted as the major environmental risk factor for all types of skin cancer in people who are genetically predisposed. The exact nature of sunlight exposure necessary to develop the various skin cancers is not entirely clear, but it is likely that both episodic and cumulative exposures are important, particularly if the exposure is sufficient to cause sunburn (Armstrong, 1986; Giles et al., 1988
; Green et al., 1988; Osterland et al., 1988; Marks, 1989; Marks et al., 1990; Marks and Whiteman, 1994).
Sunburn is a measure of biologically active sunlight reaching the skin that is person-specific (i.e. reflects individual differences in skin type and degree of protection) and also correlates with the risk of developing skin cancer. It reflects, therefore, both individual differences in exposure to sunlight, as well as individual susceptibility. Studies have shown that risk of sunburn correlates clearly with volitional behavioural factors that are able to be influenced by public health campaigns (Cockburn et al., 1989; Lowe et al., 1993; Foot et al., 1993; Hill et al., 1993). Thus, there is a considerable public health argument for undertaking programs directed at changing individuals behaviour to reduce sunburn, as a way of reducing the rising incidence of skin cancer (Muir et al., 1987; Russell, 1987; Mackie et al., 1991).
A number of skin cancer prevention programs are being developed and delivered in various countries throughout the world (Cameron and McGuire, 1990; Cody and Lee, 1990; Robinson, 1990; Borland, 1991; Borland et al., 1991; Boldeman et al., 1991). With very few exceptions (e.g. Hill et al., 1993), there have been no behavioural evaluations of comprehensive primary prevention programs published, and certainly no economic evaluations. This economic appraisal should therefore be of international interest.
The SunSmart campaign (on which the proposed national program is modelled) has been running in the State of Victoria, Australia, since 1987 (Hill et al., 1993). It is the most comprehensive population-based primary prevention program for skin cancer reported in any country in the world. It comprises three elements: (i) a comprehensive education strategy including mass media, teaching resources and a sunlight protection policy and practice code; (ii) structural changes including guidelines for workers' sun protection and downward pressure on the price of sunscreens; and (iii) a variety of sponsorships (Hill et al., 1993).
Evaluation of the SunSmart campaign has shown a reduction in the crude proportion of sunburn in the Victorian population from 11% to 7% over the 3 years (1988–91), with the adjusted odds ratio for being sunburnt being as follows: year 2/year 1: 0.75 (CI 0.57–0.99) and year 3/year 1: 0.59 (CI 0.43–0.81). Substantial attitudinal and behavioural shifts have occurred, including increased hat wearing and sunscreen use over the years 1988–91 (Hill et al., 1993). In the last 4 years, the reduction in sunburn and accompanying behavioural variables slowed, but has not reversed direction.
METHODS
The research task was to investigate the potential cost-effectiveness of introducing a national skin cancer campaign modelled on the Victorian SunSmart campaign, from a range of perspectives, involving both the ‘do nothing’ and ‘current practice’ comparators. The base analysis is from the perspective of the Federal government (only) as the potential third party funder of a new national initiative. A ‘do-nothing’ comparator is used as the Federal government has minimal current involvement in skin cancer prevention activities. In sensitivity analysis this perspective is broadened to a health sector public finance/ provider perspective, by adding in State/Territory health departments and State/Territory anticancer bodies. A ‘current practice’ comparator is used to reflect the range of current activities at the State/Territory level in skin cancer prevention. A still broader perspective (health sector: public and private) is also investigated by adding in estimates for skin cancer expenditures by private individuals for sunscreen and hats. By posing the research question in this way a range of issues in assessing the potential worth a national program could be teased out.
SunSmart provides a useful model to appraise a national skin cancer health promotion campaign, not only because of its comprehensive nature, but because it was introduced over a base level campaign that had been running for several years (i.e. the ‘Slip! Slop! Slap!’ campaign), with a more modest budget characteristic of current activities in many parts of Australia and overseas. The effectiveness rates SunSmart achieved, therefore, hold promise of what a co-ordinated and well-run national strategy could achieve.
COSTS
The focus of the cost analysis in the base run was the estimated cost to the Federal government of funding (via the State/Territory anticancer bodies) a comprehensive national health promotion campaign, co-ordinating initiatives in education (mass media and targeted to settings), structural change and sponsorships, less any savings in health care costs that could be anticipated from a reduction in management costs for skin cancer.
The cost of such a national campaign was based on the SunSmart average cost of 28¢ per person in the 1988/89 to 1990/91 period, applied to the Australian population of 18 million people (ACCV, 1997). While simplistic, this per capita costing is based on an actual state-wide program provided to 4.5 million Victorians that embodied the proposed elements of a co-ordinated comprehensive program. The national program is assumed to run over a 20-year period at the same level of real expenditure throughout as the reference year 1995/96. The extent to which federal–state co-ordination would involve additional costs and whether these might be offset by economies of scale is not known. No provision has been made for these off-setting factors.
The health care system cost offsets are estimated using the Macro Economic Evaluation Model (MEEM), the methodology of which has been described elsewhere (Carter, 1994). The percentage reductions in the incidence of melanoma and NMSC (see below) attributed to the national campaign, are multiplied by the MEEM estimates of total recurrent health care expenditure on these diseases (refer Table 1
) to estimate the potential annual offsets. The lag periods before the reduced cancer incidence is realized has been set at 5 years for melanoma and 15 years for NMSC and solar keratoses (with variations tested in sensitivity analysis). The net cost (i.e. cost of national program minus offsets) of the national program over its assumed 20 years of operation is presented in present value (PV) terms using a 5% discount rate.
Other benign skin cancers have not been included in the analysis (with the exception of solar keratoses) as they are not related to sunlight. As MEEM did not include disease costings below the three-digit ICD9-CM level, solar keratoses were costed from health insurance data on the number of treatments and associated expenditures (Medicare item no. 30291), together with provision for general practitioner and dermatology specialist consultation fees.
The ‘incremental’ cost of the new national program was estimated by deducting estimates of current expenditure at the State/Territory level. A survey was undertaken of State/Territory anticancer bodies to ascertain expenditure on skin cancer prevention programs in Australia during 1995/96 and related to the skin cancer incidence within each of these boundaries (Table 2).
|
Costs for individuals and their families in complying with the SunSmart recommendations were also estimated. Costing consisted of the consumption of sunscreen equivalent to one tube of sunscreen (at $AUD10 per tube) per year for every third person (i.e. $AUD3.33 per person per annum for Australia's 18 051 230 population), and the use of hats equivalent to one extra hat (at $AUD10 per hat) every 3 years for every second person. No estimate was included for clothing on the assumption that the use of existing clothing would be sufficient to comply with the SunSmart recommendations. The savings individuals would receive from a reduction in the incidence of skin cancer (the gap between the health insurance rebate given and the fee charged) were deducted from the additional costs.
HEALTH OUTCOMES
The health outcomes are measured in terms of premature deaths deferred and potential years of life saved (PYLS). The PYLS and deaths deferred are estimated by analysis linking predicted changes in sunburn to corresponding reductions in total lifetime ultraviolet radiation (UVR) exposure, and hence to anticipated outcomes in terms of reduced incidence of melanoma and NMSC. Discounting (5%) is applied to benefits (measured as PYLS) in an analogous way to costs.
Data from the evaluation of the SunSmart campaign (Hill et al., 1993), suggest that the incidence of sunburn in Australia could be reduced by at least one-third (the adjusted OR for being sunburnt was 0.59 in 1990/91 compared with 1988/89, i.e. a 41% reduction). We assume that this reduction in levels of sunburn will be maintained over the 20-year life span of the health promotion initiative. The next step is to relate the anticipated change in suntan behaviour to potential reductions in skin cancer morbidity/ mortality through the changes in UVR exposure that the behavioural changes imply. Table 3 summarizes this step, providing the cancer incidence and mortality changes for various decreases in total lifetime UVR exposure. The estimates are produced by relating known incidence rates of melanoma and NMSC by latitude in Australia to the known levels of biologically active UVR. They are based on incidence rates of skin cancer due to UVR exposures (and behaviours) in the population over the last 60–70 years (and are conservative compared with current incidence rates, i.e. they underestimate potential health gains in Australia). Although relatively crude, they do give a reasonable measure of the decrease in skin cancer incidence and mortality to be expected from various reductions in exposure to UVR resulting from prevention programs such as SunSmart. Given the anticipated one-third reduction in the incidence of sunburn, it was assumed that a 20% reduction in total lifetime UVR was possible achieving a reduction of 249 deaths per annum from melanoma and 59 deaths per annum from NMSC (Marks, 1989). Sensitivity analysis was conducted for lower levels of achieved UVR reduction.
|
For converting the estimates of premature deaths deferred into potential years of life saved (PYLS), it was assumed that the average age of death from NMSC is 75 (Rosenblatt and Marks, 1996
) and the average age of death from melanoma is 63 (Giles, G., 1997, Victorian Cancer Registry, personal communication). Average life expectancy in Australia, at age 63 (i.e. 83) and age 75 (i.e. 85), respectively (AIHW, 1994), was used to calculate the PYLS per death deferred of 20 years (83–63) for melanoma and 10 years (85–75) for NMSC. In this analysis, benefits accruing after the 20-year life of the program (due to the lag effects) have not been included. While this reduces the estimate of health gain, the effect of discounting on benefits accruing after 20 years makes the bias of little consequence (for example, 1 year of life gained 20 years hence has a current value of 0.38 using a 5% discount rate and 0.15 using a 10% discount rate). A summary of the key cost and benefit assumptions is set out in Table 4. All of these key assumptions are subjected to sensitivity testing.
|
RESULTS
On the basis of the assumptions set out above, a national skin cancer health promotion campaign running over a 20-year period and modelled on the Victorian SunSmart campaign of 1989/90– 1991/92 should prevent approximately 4300 premature deaths compared with the ‘do nothing’ comparator (i.e. 249 less deaths per annum from melanoma in years 5–20; plus 59 less deaths per annum from NMSC in years 15–20; undiscounted). For a financial outlay of approximately $AUD 5 million per annum (or a present value cost of approximately $AUD 62 million for funds expended over the 20-years) the initiative would cost approximately $AUD 1360 per PYLS (ignoring the cost offsets). If the potential cost offsets stemming from the reduction in skin cancer management costs are included, then a national program which proved to be as effective as SunSmart would not only save lives, but would pay for itself twice over (i.e. pay for itself plus yield a net saving in present value terms of approximately $AUD 105 million). The study results, although indicative only, strongly suggest that a comprehensive and well funded national skin cancer health promotion campaign would prove excellent ‘value for money’ from the perspective of the Federal government as third party funder (i.e. funding the program via the anticancer bodies and/or State/Territory health departments and receiving the cost offsets to its hospital, medical and pharmaceutical expenditures, etc.).
Modification of the perspective from Federal government as third party funder to health sector (public) as funder/provider (i.e. adding in the State/Territory governments and anticancer bodies) also yielded encouraging results. This expanded perspective involved changing the comparator from ‘do nothing’ to ‘current practice’ given State/Territory activities in the area. A cost per PYLS of $AUD 2715 (with offsets excluded) was estimated, together with a cost saving of approximately $AUD 11 million (PV). (Refer variation 12 of Table 5.)
|
The analysis incorporating costs to individuals and their families in complying with the SunSmart recommendations involved a quantum leap in costs, but still achieved a quite reasonable result of $AUD 25 134 per PYLS (or $21 531 PYLS including cost offsets). The present value of costs to individuals and their families over the 20-year period ($AUD 1092 million) dwarfed the costs to the third party funder of running the initiative ($AUD 62 million). As the perspective of the study was expanded to include costs impacting on individuals, the costs and cost per PYLS rose substantially. This suggests that governments ought to consider the content of their educational messages carefully, particularly in regard to purchases they encourage the community to make.
Given the uncertainty associated with some of the key assumptions, it was important to explore the effect on cost-effectiveness of plausible variations. Table 5 demonstrates that the cost-effectiveness results proved to be robust to variations in the discount rate, inclusion of a 5-year lag between melanoma incidence and death, and a variety of pessimistic cost and outcome variations. A combination of pessimistic variations (refer variation 13 of Table 5) also yielded quite strong results (i.e. $AUD 8074 per PYLS; $AUD2963 per PYLS with cost offsets included). It should also be noted that optimistic variations are also possible. For example, the inclusion of a 5-year lag for sunshine-related benign tumours (e.g. solar keratoses) rather than 15 years in the cost offset analysis increases the present value of offsets from $AUD 165 million under the standard assumptions to $AUD 302 million, and the savings from $AUD 103 million to $AUD 240 million.
DISCUSSION
The purpose of this study was to establish in a general sense whether economic grounds existed to warrant further work to develop a national skin cancer health promotion campaign and based on its potential cost-effectiveness, what priority might be given to that task.
The economic analysis undertaken is preliminary and scene-setting. Adopting the initial perspective of the Federal government as third party funder of a potential national health promotion simplifies the analysis and is adequate for this preliminary assessment, but it means that costs affecting individuals (i.e. costs of sunscreen, hats, etc.) were only included via the sensitivity analysis. Further development of our preliminary estimates in the sensitivity analysis is warranted, given that this variation had the biggest effect on the cost-effectiveness results. The costs to individuals would be greater, for example, if the cost of lycra full bodied swim suits ($AUD 40) or sunglasses ($AUD 20) were included. Sunglasses were not included as eye disease was not part of this analysis.
The effect of sunscreen on the cost to the individual and the cost per PYLS results was, none the less, an important result. Australian primary prevention programs have relegated sunscreen use to second priority now for over 10 years. This has been done for a variety of reasons, including cost, risk of short- and long-term side-effects, and ease and adequacy of application. The substantial increase in costs per year created by sunscreens alone in the sensitivity analysis reinforce this policy. Natural protection in the use of shade, avoidance of sunlight around the middle of the day and clothing remain the first priorities in the program.
Similarly, while it is recognized that State/ Territory governments and State/Territory anticancer bodies do currently spend resources on various skin cancer related programs, it was simpler for this preliminary appraisal to assume a ‘do nothing’ comparator in the basic analysis undertaken from the Federal government perspective. The effect of this assumption should lead to a more conservative cost-effectiveness estimate, as the cost of moving to a comprehensive nationally co-ordinated program is overstated (assuming existing resources would be subsumed within it), while the effectiveness assumptions taken from the SunSmart campaign are net of the base level health promotion activity for skin cancer that had been running in Victoria for several years. Our preliminary estimates of moving from a ‘do nothing’ comparator to a ‘current practice’ comparator (refer variation 12 of sensitivity analysis) also suggest that the incremental cost-effectiveness of upgrading current activities would be rewarding. Even assuming only 25% of the estimated benefits are attributed to the upgrading, the additional expenditure still results in a cost of $AUD 2715 per PYLS.
Sensitivity testing demonstrated that increasing the lag between decreases in UVR exposure and falls in the incidence of melanoma from 5 to 15 years had an important impact on cost-effectiveness. There are no human data available to allow a precise indication of this lag period. We selected 5 years based on reported data on exposure and tumour development, as well as the dose–response equations derived from the Australian incidence for these tumours at different latitudes and radiation levels. (MacKie and Aitchison, 1982; Marks, 1989; Thompson et al., 1993; Autier and Doré, 1998).
The use of sunburn as a surrogate for the proportional reduction in the exposure to sunlight probably underestimates the degree to which the reduction has occurred. For example, individuals can expose themselves to sunlight with biological compensatory mechanisms without producing sunburn. Similarly, at a certain dose of sunlight exposure, sunburn will commence but it is difficult to record an increasing sunburn level in the individual correlated clearly with the dose received by that person. Therefore, a 20% reduction in sunburn in a community could well represent overall a far greater reduction in the total dose of sunlight received by individuals. Because of the difficulty in estimating the total dose received, the conservative estimate of sunburn rates has been used in this analysis.
Note that in terms of the skin cancer burden of suffering presented in Table 1 and used in calculating the cost offsets, ‘indirect costs’ (i.e. earnings forgone due to premature death and absenteeism due to illness) have not been included. Their inclusion would be consistent with a different form of economic appraisal (i.e. cost–benefit analysis) to that adopted in the present study (cost-effectiveness analysis). Benefits in the present study are measured in terms of premature deaths prevented and life-years saved, and the inclusion of a monetary value for mortality benefits in the cost offsets would involve double counting (Gold et al., 1996). Changes in the indirect costs of absenteeism (a morbidity benefit), while relevant to the individual, employers and society, are not relevant to an economic appraisal carried out from the perspective of the health sector third party funder (except to the extent that the health sector is the employer).
The health sector cost offsets are based on ‘direct costs’ only. The estimates are likely to be a conservative estimate of the potential offsets for two reasons. First, because the reference year probably involves lower real costs for skin cancer management than might be experienced in the future when savings from a national health promotion campaign might be realized (reflecting changes in population size and composition, disease incidence, technology, etc.). Secondly, because MEEM provides a conservative estimate of current disease management costs (currently incorporating 90% of total recurrent health care expenditure). Countering these effects, the cost offset estimates may be overstated to the extent that skin cancer management costs based on the prevalence approach currently adopted by MEEM (i.e. total expenditure in 1993/94 on all existing cases in 1993/94) are higher than estimates based on an incidence approach (i.e. PV of total expenditure (1993/94 and beyond) for complete care of all new cases in 1993/94). The prevalence/incidence distinction is relevant because health promotion campaigns typically prevent new cases from occurring, rather than impact on existing cases. To the extent that disease episodes are of a short duration, prevalence-based costings give similar estimates to incidence-based costings. In the case of skin cancer there will be recurring management costs, once treated, for melanoma and NMSC, but solar keratoses are likely to incur a one-time treatment cost (i.e. incidence costs equals prevalence costs).
No attempt has been made to undertake marginal analysis to ascertain how a national program should be designed and implemented, and what level of resourcing was warranted on technical efficiency grounds. The essential question addressed is whether a comprehensive national program along the lines of SunSmart is a potentially cost-effective use of resources. While notions of cost-effectiveness are inherently subjective and country specific, cost-effectiveness results below $AUD 30 000 per PYLS would be regarded favourably in Australia. Cost-effectiveness results of less than $AUD 10 000 per PYLS would be regarded as excellent ‘value for money’ compared with how resources are currently being deployed in Australia's health sector.
In summary, although by necessity the assumptions and analysis is this paper have been simplistic, they do indicate clearly that a comprehensive national skin cancer primary prevention program should prove satisfactory to excellent ‘value for money’ depending on the perspective, comparator and assumptions adopted. As a side-benefit, the cost differential uncovered between public and private costs, provides information which should assist in determining the relative value of various sun protection strategies, particularly sunscreens. More generally, it seems clear that primary prevention campaigns in countries such as Australia with high or increasing incidence and mortality rates for skin cancer should continue. Further work is necessary to explore in more detail the value of the relative components of these campaigns as a way of ensuring that they continue to be cost-effective in the long term.
ACKNOWLEDGEMENTS
The authors acknowledge with thanks the help of Anne Gibbs, of the Anti Cancer Council of Victoria, in collecting and compiling the expenditure data from State/Territory cancer bodies. The authors also wish to thank the Australian Institute of Health and Welfare for access to disease costs data, particularly Ruth Penn, Colin Mathers and Chris Stevenson, who are involved in the DCIS project.
REFERENCES
Anti Cancer Council of Victoria (1997) SunSmart Funding, Internal Information Note.
Armstrong, B. K. (1988) Epidemiology of malignant melanoma: Intermittent or total accumulated exposure to the sun? Journal of Dermatology Surgery and Oncology, 14, 825–849.
Australian Bureau of Statistics. (1996) Causes of Death in Australia. AGPS, Canberra.
Australian Institute of Health and Welfare. (1994) Australia's health 1994: the Fourth Biennial Health Report of the Australian Institute of Health and Welfare. AGPS, Canberra.
Autier, P. and Doré, J. F. (1998) Influence of sun exposures during childhood and during adulthood on melanoma risk. International Journal of Cancer, 77, 533–537.
Boldeman, C., Jansson, B. and Holm, L. E. (1991) Primary prevention of malignant melanoma in Swedish urban pre-school sector. Journal of Cancer Education, 6, 247–253.[Medline]
Borland, R. (1991) Public awareness and reported effects of the 1989–90 SunSmart campaign and associated titles in the Anti-Cancer Council of Victoria, SunSmart Evaluation Studies, Melbourne.
Borland, R., Hocking, B., Godkin, G., Gibbs, A. and Hill, D. (1991) The impact of a skin cancer control education package for outdoor workers. Medical Journal of Australia, 154, 686–688.[Web of Science][Medline]
Cameron, I. H. and McGuire, C. (1990) Are you dying to get a suntan?—the pre- and post campaign survey results. Health Education Journal, 49, 166–170.
Carter, R. (1994) A macro economic approach to economic appraisal in the health sector. The Australian Economic Review, 106, 105–112.
Cockburn, J., Hennrikus, D., Scott, R. et al. (1989) Adolescent use of sun protection measures. Medical Journal of Australia, 151, 136–140.[Web of Science][Medline]
Cody, R. and Lee, C. (1990) Behaviours, beliefs and intentions in skin cancer prevention. Journal of Behavioural Medicine, 13, 373.
Foot, G., Girgis, A., Boyle, C. A. et al. (1993) Solar protection behaviours: a study of beachgoers. Australian Journal of Public Health, 17, 209–214.[Web of Science][Medline]
Giles, G. G., Marks, R. M. and Foley, P. (1988) Incidence of non-melanocytic skin cancer treated in Australia. British Medical Journal, 296, 13–17.
Gold, M. R., Siegel, J. E., Russell, L. B. and Weinstein, M. C. (1996) Cost-effectiveness in Health and Medicine. Oxford University Press, New York.
Green, A., Beardmore, G., Hart, V. et al. (1988) Skin cancer in a Queensland population. Journal American Academy of Dermatology, 19, 1045–1052.[Web of Science][Medline]
Hill, D., White, V., Marks, R. et al. (1993) Changes in sun-related attitudes and reduced sunburn prevalence in a population at high risk of melanoma. European Journal of Cancer Prevention, 2, 447–456.[Medline]
Jelfs, P., Coates, M., Giles, G. G. et al. (1996) Cancer in Australia 1989–90 (with projections to 1995). Australian Institute of Health and Welfare, Canberra.
Ko, C. B., Walton, S., Keczkes, K., Bury, H. P. R. and Nicholson, C. (1994) The emerging epidemic of skin cancer. British Journal of Dermatology, 130, 269–272.[Web of Science][Medline]
Lowe, J. B., Balanda, K. P., Gillespie, A. M. et al. (1993) Sun-related attitudes and beliefs among Queensland school children: The role of gender and age. Australian Journal of Public Health, 17, 202–208.[Web of Science][Medline]
Mackie, R. M. and Aitchison, T. (1982) Severe sunburn and subsequent risk of primary cutaneous melanoma in Scotland. British Journal of Cancer, 46, 955–960.[Web of Science][Medline]
MacKie, R. M., Österlind, A., Risiter, D. et al. (1991) Report on consensus meeting of the EORTX Melanoma Group on educational needs for primary and secondary prevention of melanoma in Europe. Results of a workshop held under the auspices of the EEC Europe against Cancer programme in Innsbruck. European Journal of Cancer, 27, 13117–13123.
Magnus, K. (1991) The Nordic profile of skin cancer incidence. A comparative epidemiological study on the three main types of skin cancer. International Journal of Cancer, 47, 12–19.
Marks, R. (1989) Possible effects of increased ultraviolet radiation on the incidence of non melanocytic skin cancer. In Health Effects of Ozone Layer Depletion: a Report of the NHMRC Working Party, Melbourne. AGPS, Canberra, pp. 70–81.
Marks, R., Jolley, D., Lecatsas, S. et al. (1990) The role of childhood sunlight exposure in the development of solar keratoses and non-melanocytic skin cancer. Medical Journal of Australia, 152, 62–65.[Web of Science][Medline]
Marks, R., Staples, M. and Giles, G. G. (1993) Trends in non-melanocytic skin cancer treated in Australia: the second national survey. International Journal of Cancer, 53, 585–590.
Marks, R. and Whiteman, D. (1994) Sunburn and melanoma—How strong is the evidence? British Medical Journal, 308, 75–76.
Miller, D. L. and Weinstock, M. A. (1994) Non-melanoma skin cancer in the United States: Incidence. Journal American Academy of Dermatology, 30, 774–775.[Web of Science][Medline]
Muir, C. S., Waterhouse, J., Mack, T. et al. (1987) Cancer incidence in five continents. IARC Scientific Publication No. 88, Vol. V.
Osterland, A., Tucker, M. A., Stone, B. J. et al. (1988) The Danish case-control study of cutaneous malignant melanoma. II. Importance of UV light exposure. International Journal of Cancer, 42, 319–324.
Robinson, J. R. (1990) Behaviour modification obtained by sun protection education coupled with removal of a skin cancer. Archives of Dermatology, 126, 477–481.
Rosenblatt, L. and Marks, R. (1996) Deaths due to squamous cell carcinoma in Australia: Is there a case for a public health intervention? Australasian Journal of Dermatology, 137, 26–29.
Russell, J. R. (1987) Ozone depletion and cancer risk. Lancet, ii, 443–446.
Thompson, S. C., Jolley, D. and Marks, R. (1993) Reduction of solar keratosis by regular sunscreen use. New England Journal of Medicine, 329, 1147–1151.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D. J. Hill, S. J. Dobbinson, and J. K. Makin Interventions to Lower Ultraviolet Radiation Exposure: Education, Legislation, and Public Policy ASCO Educational Book, January 1, 2009; 2009(1): 526 - 531. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Dixon, S. Dobbinson, M. Wakefield, K. Jamsen, and K. McLeod Portrayal of tanning, clothing fashion and shade use in Australian women's magazines, 1987-2005 Health Educ. Res., October 1, 2008; 23(5): 791 - 802. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. W. Kyle, J. K. Hammitt, H. W. Lim, A. C. Geller, L. H. Hall-Jordan, E. W. Maibach, E. C. De Fabo, and M. C. Wagner Economic Evaluation of the US Environmental Protection Agency's SunWise Program: Sun Protection Education for Young Children Pediatrics, May 1, 2008; 121(5): e1074 - e1084. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. G. Dixon, M. Lagerlund, M. J. Spittal, D. J. Hill, S. J. Dobbinson, and M. A. Wakefield Use of Sun-Protective Clothing at Outdoor Leisure Settings from 1992 to 2002: Serial Cross-sectional Observation Survey Cancer Epidemiol. Biomarkers Prev., February 1, 2008; 17(2): 428 - 434. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. G. Dixon, D. J. Hill, D. J. Karoly, D. J. Jolley, and S. M. Aden Solar UV Forecasts: A Randomized Trial Assessing Their Impact on Adults' Sun-Protection Behavior Health Educ Behav, June 1, 2007; 34(3): 486 - 502. [Abstract] [PDF]
|
|
|
|
 |
|
 W. R. Stanton, M. Janda, P. D. Baade, and P. Anderson Primary prevention of skin cancer: a review of sun protection in Australia and internationally Health Promot. Int., September 1, 2004; 19(3): 369 - 378. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Neale, G. Williams, and A. Green Application Patterns Among Participants Randomized to Daily Sunscreen Use in a Skin Cancer Prevention Trial Arch Dermatol, October 1, 2002; 138(10): 1319 - 1325. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. B. Lowe, L. J. McDermott, W. R. Stanton, A. Clavarino, K. P. Balanda, and B. McWhirter Behavior of caregivers to protect their infants from exposure to the sun in Queensland, Australia Health Educ. Res., August 1, 2002; 17(4): 405 - 414. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. J. Smith, C. Ferguson, J. McKenzie, A. Bauman, and P. Vita Impacts from repeated mass media campaigns to promote sun protection in Australia Health Promot. Int., March 1, 2002; 17(1): 51 - 60. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. M. Livingston, V. M. White, A. M. Ugoni, and R. Borland Knowledge, attitudes and self-care practices related to sun protection among secondary students in Australia Health Educ. Res., June 1, 2001; 16(3): 269 - 278. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Montague, R. Borland, and C. Sinclair Slip! Slop! Slap! and SunSmart, 1980-2000: Skin Cancer Control and 20 Years of Population-Based Campaigning Health Educ Behav, June 1, 2001; 28(3): 290 - 305. [Abstract] [PDF]
|
|
|
|
 |
|
 M. Wise The role of advocacy in promoting health Global Health Promotion, January 1, 2001; 8(2): 69 - 74. [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||