You are on page 1of 7

ORIGINAL ARTICLE

See related commentary on pg 2745

Regular Sunscreen Use Is a Cost-Effective Approach to Skin Cancer Prevention in Subtropical Settings
Louisa G. Gordon1,3, Paul A. Scuffham2,3, Jolieke C. van der Pols1, Penelope McBride1, Gail M. Williams4 ` le C. Green1,3 and Ade
In many developed countries, total costs to health systems for cutaneous basal cell carcinomas (BCCs) and squamous cell carcinomas (SCCs) are among the highest of all cancers, yet the investment value of preventive measures remains unknown. Using primary data from a randomized controlled trial, we estimated the costeffectiveness of a skin cancer prevention initiative based on regular sunscreen use. Compared with usual practice (discretionary use), the sunscreen intervention cost an additional US$106,449 (2007) to prevent 11 BCCs, 24 SCCs, and 838 actinic keratoses among 812 residents over 5 years. These health outcomes required an annual average investment of US$0.74 per person and saved the Australian government a total of US$88,203 in healthcare costs over the same period. Such community-based interventions promoting regular sunscreen use among Caucasians in subtropical settings can prevent skin cancer and related skin tumors in practical ways and with great cost efficiency.
Journal of Investigative Dermatology (2009) 129, 2766–2771; doi:10.1038/jid.2009.141; published online 18 June 2009

INTRODUCTION In populations with large percentages of people of European descent, basal cell carcinomas (BCCs) and squamous cell carcinomas (SCCs) of the skin impose a substantial burden in human and economic terms. Because of the negligible mortality compared with other cancers, their high incidence rates and attendant costs tend to be overlooked. In the United States alone, Medicare spends US$13 billion each year on skin cancer treatment. Added to this are the costs of associated skin tumors such as actinic keratoses (AKs), which are highly prevalent—around 6–25% in the United Kingdom and the United States (Holmes et al., 2007) and 40–60% in Australia (Darlington et al., 2003; Holmes et al., 2007)—and among the strongest predictors of skin cancer (Darlington et al., 2003). Given that UV radiation is the major environmental cause of skin cancers (International Agency for Cancer Research, 1992), they are preventable by sun-protective behaviors. In combination with other primary sun-avoidance measures, such as shade and sun-protective clothing, the use of broad1

Department of Cancer and Population Health Studies, Queensland Institute of Medical Research, Brisbane, Queensland, Australia; 2School of Medicine, Griffith University, Brisbane, Queensland, Australia; 3Griffith Medical Research College, A Joint Program of Griffith University and the Queensland Institute of Medical Research, QIMR, Herston, Queensland, Australia and 4 Department of Epidemiology and Biostatistics, School of Population Health, University of Queensland, Brisbane, Queensland, Australia Correspondence: Dr Louisa G. Gordon, Cancer and Population Health Studies, Queensland Institute of Medical Research, PO Royal Brisbane, 300 Herston Road, Herston, Brisbane, Queensland 4029, Australia. E-mail: Louisa.Gordon@qimr.edu.au Abbreviations: AK, actinic keratosis; BCC, basal cell carcinoma; SCC, squamous cell carcinoma Received 13 November 2008; revised 5 March 2009; accepted 20 April 2009; published online 18 June 2009

spectrum sunscreens is an effective adjunct measure. The safety and efficacy of sunscreens in protecting human skin from squamous tumors have been established (Thompson et al., 1993; Green et al., 1999). Because of their central role in skin cancer prevention, sunscreens are unanimously recommended by health authorities in the United States (US Preventive Services Task Force, 2001), the United Kingdom (National Radiation Protection Board UK, 2008), and Australia (Cancer Council Australia and Australasian College of Dermatologists, 2007). Using data from a community-based randomized controlled trial (Green et al., 1999), we evaluated actual use of health-care resources, costs, and health outcomes of BCC and SCC prevention in order to address the question of whether an intervention for skin cancer prevention that promotes the daily application of sunscreen among Caucasians in a sunny environment is a sound economic investment. From a societal perspective, a cost-effectiveness analysis was undertaken using direct health-outcomes data obtained when individuals from a Queensland community took part in a single randomized controlled trial (Green et al., 1999) over a 5-year period. RESULTS Of the 1,621 residents enrolled, 1,383 (85%) remained active participants at the end of the trial. Participants who withdrew were younger on average than ongoing participants (47 vs. 49 years, respectively) but otherwise similar (Green et al., 1999). From January 1993 to December 1997 inclusive, 155 (10%) participants developed 277 new histologically confirmed skin cancers. The occurrence of 11 fewer BCCs and 24 fewer SCCs among the participants in the daily-sunscreen arm compared with those in the usual-practice arm (Table 1) was
& 2009 The Society for Investigative Dermatology

2766 Journal of Investigative Dermatology (2009), Volume 129

49) 0.94 (0.103 for the daily-sunscreen group versus $138. P=0.315 5. From a societal perspective. fewer skin cancers.64–0. Combining skin cancer and cost outcomes.87 (0.904 2.059 12.. 1 On the basis of Darlington et al. squamous cell carcinomas. estimated 95% cryotherapy & 5% topical creams (Rosen and Studniberg. thus. Table 2. previous history of skin cancer. 2003) (Table 2).74 annually).03. 3 Difference was statistically significant. cost savings. smoking.72 per person over 5 years (or $0.203 for avoided SCC.438 6. hair color. 5% reduction during 1994–1996 or an equivalent of 0. respectively.50 (0.617 CI.306 for the usual-practice group.149 versus $222. which compares rates of prevalent AKs on sunscreen applied sites in the intervention group relative to the usual practice group. sex.75–1. If 100% www. and hands. 2 Treatment estimates.jidonline. the incremental cost per skin cancer prevented over 5 years was US$3. This equated to an investment of $3. Models were adjusted for age. Number of prevalent actinic keratoses (AKs) prevented.933 37. AKs prevented 50% of AKs from daily sunscreen use1 treated 806 166 403 83 Physician visits ($) 27.041 from a societal viewpoint. From the perspective of the Australian government (which funds the vast majority of health care in such cases). arms. These net costs translated into $405 versus $275 per person for the intervention and usual-practice groups. on the basis of intention-to-treat analyses. from 1993 to 1997.19)1 Participants in daily sunscreen group 806 797 Estimated no.27–0.40–1. Incidence of histologically confirmed BCCs and SCCs on the head. and fewer AKs were attributed to the regular use of sunscreen. BCC. (2003). CI.68 (0. confidence intervals.96)1 1996–1994 0.700 for the usual-practice group (Table 3). number of AKs treated 50% (authors assumption).72–1. and AK (assuming 50% of AKs were treated).208 prevented per person. the net cost over 5 years was $50.92)3 Total 121 156 277 35 BCC.306 474 Total cost saving ($) 61.781 73. These results found a 24% reduction during 2-year period 1992–1994 or an equivalent of 1 AK prevented per person. eye color. the net total cost over 5 years for the daily-sunscreen intervention group was US$329. but it represented a savings to the government (Table 3).34) 0.15) Total 79 76 1552 9 Counts of skin cancers BCC SCC 94 27 105 51 199 78 Difference (cancers prevented) 11 24 0.56–1. Cost-effectiveness ratios were sensitive to the inclusion of conservative cost estimates for AKs and to the proportion of AKs treated (50% in the base analysis). attributable to daily sunscreen use. and estimated cost-savings (US$ 2007) for avoided medical treatment Management2 Rate ratio (95% CI) 1994–1992 0. by randomized sunscreen treatment group Persons affected with skin cancer BCC SCC Daily sunscreen group 59 23 Discretionary sunscreen group 57 34 Total 116 57 Difference (persons with fewer cancers) À2 11 Rate ratio (95% CI)1 1.org 2767 . 2006). basal cell carcinoma. neck. attributed to sunscreen intervention..558 Total 838 486 32. and are not double-counted here.87 (Streeton et al.618 Cryotherapy ($) 31. 2 18 persons affected with cancer had both BCC and SCC. SCC. Medicare items and unit costs are listed in the supplementary file. 1 Rate ratios compared counts of skin cancers in the intervention group with those in the usual practice group.Cost-Effectiveness of Long-Term Sunscreen Use LG Gordon et al.466 Topical cream ($) 2. sunburn propensity. an estimated 838 AKs were prevented in the daily-sunscreen arm from 1992 to 1996 (Darlington et al. confidence interval. 2003).78 (0. average physician visits per AK=2. and occupational exposure. The cost for the intervention group was partially offset by reduced medical care costs of $88. Table 1.04 (0. Similarly.

and analysis details are available in the supplementary file. Given that this investment is a fraction of what individuals in developed nations alone pay for cosmetic skin-care products. the cost was around US$72 per dose (Department of Health and Ageing PBAC. Although skin cancers treated outside primary care were not included. consumables. visit doctor. governments. the estimated cost of Gardasil (CSL. Boronia.15 per person from $3. 2008). in all sensitivity analyses. based on 5. or $0.382 $50. then up to 19 deaths could potentially be prevented each year with regular sunscreen use in Australia. estimated at a mean US$500 (±$487) per skin cancer episode (Chen et al. and individuals were collectively willing to pay and contribute to preventing skin cancer. The results of the probabilistic sensitivity analyses. The cost-effectiveness ratios changed marginally when base ratios were recalculated using upper and lower 95% confidence interval limits of bootstrapped mean estimates for participant sunscreen application time and purchases as well as the time and expense of physician visits (Table 4). Australia).. sunscreen) (B) Imputed costs for voluntary contributions (unpaid staff (dermatologists. valuation.LG Gordon et al.632 $53.52 per person (rather than the mean of $3. these costs can be up to 10 times higher (Chen et al. there are an estimated 137.72 within a wide 95% confidence interval range (from cost saving to $29.600 new cases and 410 deaths from SCC each year in Australia (Australian Institute of Health and Welfare and Cancer Australia. Our cost-effectiveness ratios compare favorably with those for other primary prevention initiatives currently approved in Australia.394 $138.74 per person annually.. If health providers. unit costs. quantities. apply sunscreen. 1 Detailed information on resource measurement. Furthermore. the incremental annual cost per cancer prevented would have dropped to $1.306 $138. 2001). the human papillomavirus vaccine for preventing cervical cancer. from the government and societal perspectives (US$ 2007) Cost type1 (A) Provider costs (salaries. Australia) against rotavirus gastroenteritis. 2006). rental)) (C) Participant costs (time to attend skin exam. the government’s cost-saving position was preserved. 2006). Costs and skin cancers prevented in the randomized daily and discretionary sunscreen treatment groups. office. out-of-pocket costs for doctors) (D) Government-provided health care costs for treatment of skin cancers and AKs Net government costs (D) Net societal costs (A+B+C+D) Incremental costs—government viewpoint Incremental costs—societal viewpoint Total skin cancers Skin cancers prevented Incremental cost per skin cancer prevented (over 5 years)—government viewpoint Incremental cost per skin cancer prevented (over 5 years)—societal viewpoint Daily sunscreen use (n=812) $145. 2006).72.103 $329. from 1992 to 1996. High proportions of all skin cancers in the United Kingdom and Australia and around 50% in the United States are managed by primary care physicians in office-based settings. this expense seems 2768 Journal of Investigative Dermatology (2009).306 $222. of the AKs prevented by sunscreen use had otherwise been treated. actinic keratoses. when managed in hospital or ambulatory surgical rooms. one in five persons with skin cancer in our sample would need to have been treated as an inpatient to exceed the upper confidence . showed a mean incremental cost-effectiveness ratio of $3. this sunscreen initiative would cost society an outlay of US$3. For example. Sydney.02 per person (Table 4). For governments in a publicly funded health system that incur the cost of skin cancer medicine.000 simulated incremental cost and incremental effect pairings. and participants) for these improved health outcomes. even considering the upper 95% confidence limit of $29. DISCUSSION This economic evaluation of an intervention promoting sunscreen as a means of preventing skin cancer showed that the intervention saved the government money but imposed a small cost on society as a whole (health provider. and for the Rotarix vaccine (GlaxoSmithKline. other staff. governments. Cost-Effectiveness of Long-Term Sunscreen Use Table 3. The wide interval was driven by the large possible variation in medical costs. through the use of sunscreen. if 25% of prevented AKs had been treated.103 $50.. 1999).6% were cost saving) (Figure 1).449 121 35 Cost saving $3.032 $80.72) from our probabilistic sensitivity analyses.041 Discretionary sunscreen use (n=809) $84.700 156 AK. On the other hand. This explains the relatively low cost of managing individual tumors.52 per person: 99% resulted in better outcomes (cancers prevented) and 8. the ratio would have risen to $5. a quarter of all new cases could be prevented over 5 years (Green et al. and if.149 Cost saving $106. Volume 129 extremely worthwhile.041 per skin cancer prevented over a 5-year period. this intervention yielded considerable cost savings within 5 years under a range of cost scenarios. began at around US$360 per dose (Department of Health and Ageing PBAC.

62 3. 1988. their importance as strong predictors of skin cancer risk is well known. The 95% confidence ellipse is based on 5. Given the growing burden of costs for skin cancer treatment in the Northern Hemisphere.52 3.81 0.. such as photoaging (Stern. although sensitivity testing showed their notable potential to alter cost-effectiveness ratios. it was not possible to precisely measure total AKs because of the extremely high prevalence rates coupled with high rates of spontaneous regression (Frost et al.93 3. 2 Includes the scenario of higher accuracy and less resource use in experienced or specialist doctors. 1999).72 3. our medical costs are likely to be underestimated not only because of excluded inpatient costs but also because of our omission of the costs of treatment morbidity (NHMRC National Health and Medical Research Council.33 3. Even with the exclusion of AK costs. 1999)). They were asked to apply it daily to the head. topical creams. arms.79 3.. the sunscreen intervention remained cost saving to the government. given their similarity to the overall sample (Green et al.200 $900 $600 $300 $0 –$300 –$600 –$900 –$1.500 $1. it seems likely that the fair-skinned populations of the United States and Europe could also benefit from cost savings were sunscreen to be regularly used during summer or during leisure time spent in subtropical climates.80 Medical costs ($) 2 $1.200 – $1. Results of one-way sensitivity analyses1 on key costs: societal perspective (US$ 2007) Incremental costeffectiveness ratio (mean per person) Base analysis Costs involved in skin cancer misdiagnoses (positive predictive value) 0. 1994). Thus. Queensland. We conclude that year-round regular sunscreen use for the prevention of skin cancers produces cost savings for healthcare providers in Australia. Results of probabilistic sensitivity analyses with 95% confidence ellipse. 1994.500 –20 0 20 40 60 80 100 Incremental cancers prevented 3. general practitioner. Our study used primary data from a rigorous intervention study (Green et al.77 3. especially SCC (Marks et al. however.31 5..70 6. Darlington et al. broad-spectrum sunscreen with a sun protection factor of 15 þ . Frost et al. and normal distribution for positive predictor value estimates. 1994). 2003).79 3. Low and high estimates are the lower and upper 95% confidence intervals from bootstrapped mean results with the exception of sunscreen purchases that were tested over a 20% higher and lower cost. and informed written consent was provided by all www.70 3.. 1 limit of medical costs in our sensitivity analysis on the basis of hospital costs for skin procedures. However. 1999. A total of 812 participants randomized to the sunscreen intervention were supplied with a water-resistant. Proportion of AKs treated 0% 25% 100% GP. MATERIALS AND METHODS The Nambour Skin Cancer Prevention Trial investigated the effectiveness of daily application of sunscreen (‘‘the intervention’’) versus the usual discretionary use of sunscreen (‘‘usual practice’’) (Green et al.66 3. On the other hand.15 Low High3 Time to visit a GP ($) Low High Time to apply sunscreen ($) Low High Sunscreen purchases ($) Low High Out-of-pockets for GP visit ($) Low High Figure 1. Although 15% of the participants were not active for the duration. as well as because of additional costs of other common conditions avoidable through sunscreen use.000 Monte Carlo simulations with binomial distributions assigned for skin cancers prevented. Incremental cost ¼ cost of daily application minus cost of discretionary application (mean per person). Full details of methods (Green et al. incremental cancers prevented ¼ cancers in the discretionary group minus those in the daily-sunscreen group. Ethical approval was obtained from the Human Ethics Committee of the Queensland Institute of Medical Research. 3 Includes the scenario of higher costs of skin cancers treated in hospitals.org 2769 . 2003) with a ‘‘usual practice’’ comparator along with high-quality clinical data collected over the study period. and hands. neck. Incremental cost (mean per person) US$ Table 4.. Our AK cost estimates emphasize the high level of health resources that AKs consume when they are managed like skin cancers (for example. a participant flow chart.. The trial was conducted from 1992 to 1996 among a random sample of the residents of the township of Nambour...Cost-Effectiveness of Long-Term Sunscreen Use LG Gordon et al. and application habits were reinforced by study nurses via a quarterly contact. 2004). 1999) have previously been published.77 3.. their health outcomes would not have materially changed the results.60 0.jidonline. gamma distributions for cost variables. with cryotherapy. and key findings (Green et al.02 1. and periodic clinical visits). 2000). The paucity of evidence about AK treatment patterns led to our conservative estimation in costs.

We also thank Terri Jackson for providing critical feedback on an earlier draft of the paper. 2003). at the start (1992). SUPPLEMENTARY MATERIAL Supplementary material is linked to the online version of the paper at http:// www. the participants.72. arm was carried out using Poisson regression. ACKNOWLEDGMENTS We thank all participants and volunteers in the Nambour Trial for their loyalty to this research program. NHMRC National Health and Medical Research Council. Streeton et al. Rosen and Studniberg.LG Gordon et al.nature. Research-driven costs were identified and excluded from the resources for the preventive intervention. Individual-level data relating to sunscreen purchase and application time. with medical record verification. neck. The other authors state no conflict of interest. hospitalisation and mortality.05. 2001). Cost-Effectiveness of Long-Term Sunscreen Use participants.. This analysis was funded through National Health and Medical Research Council (NHMRC) grant 442976. Medical services were valued using Medicare fees. Using diagnosis dates. After the actual patient-level data distributions.. The trial was conducted within the context of a large. Cancers diagnosed during 1992 were excluded a priori. Lee Casey (QIMR) assisted with retrieving key project records. Skin cancer ascertainment All participants received a full skin examination by a dermatologist unaware of treatment allocation.. In the probabilistic sensitivity analyses. In: Cancer Series No. 1999. and the Australian government (through Medicare Australia. given the latent period of skin tumors in humans. Williamstown. and other medical costs were right skewed. midpoint (1994). community supporters. the skin-cancerrelated secondary outcome of the sunscreen intervention (Darlington et al. In addition. adjustments to medical costs were made to account for the percentage of benign lesions that would have been treated provisionally as cancers. binomial distributions for skin cancer counts. and hands (‘‘sunscreen sites’’ under the protocol) were included in our analysis. Resources were identified and measured for the various time periods. College Station. arms. and a comparison of persons affected with cancer in each 2770 Journal of Investigative Dermatology (2009). August 2006–June AG had an unrelated project funded by L’Ore 2009. 2003). Interim skin cancers were ascertained by self-report and/or physicians’ notifications. the probability of skin cancer. The study was conducted according to the Declaration of Helsinki Principles. we also accounted for cost efficiencies relating to physician visits.’’ Final cost and health outcome data were combined into incremental cost–effectiveness ratios. We also accounted for the avoided costs of AKs. Meanwhile. Only histologically confirmed skin cancers on the head. one-way and probabilistic sensitivity analyses were undertaken to test the stability of the base results when individual parameter values were varied (Briggs. as no beneficial effect of sunscreen was expected in year 1. and year-2007 Australian dollars were applied. The analysis did not involve future costs or benefits (because it was retrospective.000 Monte Carlo simulations to assess the variation to the base results. with the costs and outcomes already known). 2007) and were further tested by varying the positive predictive value between 60 and 80% in analyses. valuation methods. we therefore obtained bootstrapped mean costs per participant using 1. usual modes of AK treatment were ascertained by searching all published Australian studies reporting primary care management of actinic skin tumors (Raasch. the assessed costs included those incurred by the program provider. and completion (1996) of the trial. and sources used to quantify resource use are available online in a Supplementary file. Data were analyzed using STATA/SE V9 (StatCorp. Texas) and TreeAge Pro 2008 (TreeAge Software Inc. MA). Volume 129 REFERENCES Australian Institute of Health and Welfare.000 bootstrapped samples and the bias-corrected approach (Briggs. Statistically significant differences were tested using negative binomial regression. gamma distributions were assigned for cost variables.. 2007) across the two groups. cancers diagnosed during 1997 were included to account for the latent effects of the sunscreen intervention immediately after trial cessation. and pathology when a person seemed to have had several cancers treated simultaneously (Richmond-Sinclair et al. publicly funded research program. Margaret Oziemski (dermatologist) gave valuable clinical advice. similarly. Specific details of measurement. we assumed an average treatment rate of AKs of 50% and tested this assumption in sensitivity analyses. using Organization for Economic Cooperation and Development purchasing-power parities). For AKs. Tests were two-sided and results were considered statistically significant when Po0. Costs are presented in year-2007 US dollars (converted at AU$ 1 ¼ US$ 0. To address the uncertainty in our estimations. ratios of successive counts of prevalent AKs indicated the rate of change in AK acquisition in each sunscreen treatment arm (Darlington et al.. using published Nambour Trial outcome data for AKs (Darlington et al. Louisa Gordon is funded through an NHMRC Public Health Fellowship.. 2003. physician-visit time and expense.com/jid Analysis Using intention-to-treat analyses. Wald’s test was used to detect significant group differences with 95% confidence intervals. Estimation of resources and costs Because a wider societal perspective was taken. CONFLICT OF INTEREST ´ al Research. and normal distributions for positive predictor estimates (Youl et al. interpreted as the additional cost for preventing each new skin cancer in the intervention group.. treatment. 2001). 2006). 2008). The measure of benefit for this economic evaluation was ‘‘skin cancers prevented. 2003. Canberra: Australian Institute of Health and Welfare . and Ross Cosmetics and Woolworths supplied the sunscreen (AuScreen).. and therefore discounting was not applied. Cancer Australia (2008) Nonmelanoma skin cancer. all cost variables. and positive predictor values were simultaneously analyzed using 5. 43. general practice consultations. which represent the cost to the Australian government that is reimbursed to physicians through Medicare Australia. 2003). Because it is not feasible to track treatment pathways for individual AKs and the precise proportion of AKs treated (as opposed to monitored) in Australian primary care seems unknown and could vary greatly. the numbers of BCCs and SCCs in the intervention and usual-practice arms were compared and tested for statistically significant group differences using negative binomial regression. These adjustments were made on the basis of a positive predictive value of 70% (Youl et al. which bears the direct costs of treatment).

Williams G et al. (Update 10/07. (1999) Daily sunscreen application and betacarotene supplementation in prevention of basal-cell and squamous-cell carcinomas of the skin: a randomised controlled trial. Yelverton CB. Neale R. 55. Pandeya N. http://www. Lancet 354:723–9 [erratum The Lancet 1999. McGuire A. pathogenesis. Arch Dermatol 139:451–5 Department of Health and Ageing PBAC (2006) Public Summary Document ‘‘Quadrivalent Human Papillomavirus (Types 6. Leslie D. Williams G. 172–214 Cancer Council Australia. Teuschler HV et al.au/internet/main/publishing. Treatment of photoaging. Rotarix’’.org. Leslie D. Gardasil ‘‘Human Rotavirus Vaccine.jidonline.cancer. Neale R. 16. Vol. Battistutta D. Br J Dermatol 131:455–64 Green A. Jolley D. Selwood TS (1994) Epidemiology of solar keratoses. Neale RE. Smith ED. Williford PM. Marks R.Cost-Effectiveness of Long-Term Sunscreen Use LG Gordon et al. Selwood TS (1988) Malignant transformation of solar keratoses to squamous cell carcinoma. Australas J Dermatol 44:34–9 Stern RS (2004) Clinical practice. Green AC.nsf/Content/ ) Frost C. Am J Prev Med 20:44–6 Youl PH. Dermatol Surg 32:1266–71 Darlington S. Baade PD. Lyon.5 ml. Parsons P et al. Harris AH (2006) How are solar keratoses treated by general practitioners in Australia? Int J Dermatol 45:272–6 Thompson SC. http://www. Polisetty SS. Marks R (1993) Reduction of solar keratoses by regular sunscreen use. Frost C. Dermatol Surg 27:1035–8 Chen JG. (2006) Treatment patterns and cost of nonmelanoma skin cancer management.health. (2008) Incidence of basal cell carcinoma multiplicity and detailed anatomic distribution: longitudinal study of an Australian population. Hart V. injection. N Engl J Med 350:1526–34 Streeton CL. Non-Melanoma Skin Cancer: Guidelines for Treatment and Management in Australia. Green A (2003) A randomized controlled trial to assess sunscreen application and beta carotene supplementation in the prevention of solar keratoses. Canberra: Commonwealth Government Raasch BA (1999) Suspicious skin lesions and their management. In: Economic Evaluation in Health Care: Merging Theory with Practice (Drummond MF. Hart V. van der Pols JC et al. Williams G. Aitken JF (2007) Diagnosing skin cancer in primary care: how do mainstream general practitioners compare with primary care skin cancer clinic doctors? Med J Aust 187:215–20 www. J Invest Dermatol 115:273–7 Frost CA. Williford PM et al.au/policy/positionstatements/ sunsmart/screeningandearlydetectionofskincancer. Chong AH (2007) Solar keratosis: epidemiology. Briggs A (2001) Handling uncertainty in economic evaluation and presenting the results. Green A (2000) High incidence and regression rates of solar keratoses in a Queensland community.org 2771 . Lancet 1:795–7 National Radiation Protection Board UK (2008) Health effects from ultraviolet radiation: Report of an Advisory Group on Non-Ionising Radiation 13/1 NHMRC National Health and Medical Research Council (2003) Clinical Practice Guidelines. Studniberg H (2003) Solar keratoses: analysis in a dermatological practice in Australia. Australasian College of Dermatologists (2007) Position Statement: Screening and early detection of skin cancer. Janda M. (2001) Cost of nonmelanoma skin cancer treatment in the United States. Rennie G. Housman TS. Kancler C. Australas J Dermatol 48:67–74 International Agency for Cancer Research (1992) Solar and ultraviolet radiation. France Marks R. 0. Control Clin Trials 15:512–22 Green A. Gospodarevskaya E. (1994) The Nambour skin cancer and actinic eye disease prevention trial: design and baseline characteristics of participants.gov. 1354:1038] Holmes C. Ware RS. 11. N Engl J Med 329:1147–51 US Preventive Services Task Force (2001) Screening for skin cancer. Goldman ND. eds). Foley P. J Invest Dermatol 31:323–8 Rosen RH. Williams GM. lyophilised powder & solvent for oral administration. Marks G. Fleischer AB Jr. recommendations and rationale. Oxford: Oxford University Press. one ml dose.htm ) Chen JG. Rennie G. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. presentation and treatment. Aust Fam Physician 28:466–71 Richmond-Sinclair NM. (Update 03/06. Freeman M. Williams G. Del Mar CB. Whiteman DC. 18) recombinant vaccine.