[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Figure 1.
One-Way Sensitivity Analysis Showing the Most Influencing Skin Cancer Screening Variables for Men and Women
One-Way Sensitivity Analysis Showing the Most Influencing Skin Cancer Screening Variables for Men and Women

This analysis explored the uncertainty in the key variables individually by varying the mean values of these variables with ±30%. The top bars represent the variables that contribute the most to the variability of the incremental cost-effectiveness ratio. For all variables, the range is 70% to 130%, except for the final variable, which is 84% to 116%. Dark gray bars show the maximum value of the variable, and light gray bars show the minimum value of the variable. BCC indicates basal cell carcinoma. Values in US dollars are $21 344 for €20 000, $42 687 for €40 000, and $64 027 for €60 000.

Figure 2.
Cost-effectiveness Planes Showing the Probabilistic Sensitivity Analysis 5000 Simulations
Cost-effectiveness Planes Showing the Probabilistic Sensitivity Analysis 5000 Simulations

Each point represents the value of 1 simulation (in 5000 simulations) performed from the distribution around each of the key variables in the model. A willingness-to-pay threshold of €35 000 (US $37 541) per quality-adjusted life-year (QALYs) is indicated by the diagonal line. In A and B, values in US dollars are $5.3 million for €5 million, $10.7 million for €10 million, $16.1 million for €15 million, $21.4 million for €20 million, $26.8 million for €25 million, $32.2 million for €30 million, $37.6 million for €35 million, and $42.9 million for €40 million. In C and D, values in US dollars are $2.1 million for €2 million, $4.3 million for €4 million, $6.4 million for €6 million, $8.5 million for €8 million, $10.7 million for €10 million, and $12.9 million for €12 million.

Table 1.  
Results of the Cost-effectiveness Analysis During 20 Years per 1000 Persons
Results of the Cost-effectiveness Analysis During 20 Years per 1000 Persons
Table 2.  
Results of the Scenario Analysis
Results of the Scenario Analysis
Table 3.  
Results of the Budget Effect Analysis During 20 Years
Results of the Budget Effect Analysis During 20 Years
1.
World Health Organization. How common is skin cancer? http://www.who.int/uv/faq/skincancer/en/index1.html. Published 2016. Accessed June 22, 2016.
2.
Gordon  LG, Rowell  D.  Health system costs of skin cancer and cost-effectiveness of skin cancer prevention and screening: a systematic review.  Eur J Cancer Prev. 2015;24(2):141-149.PubMedGoogle ScholarCrossref
3.
Girgis  A, Clarke  P, Burton  RC, Sanson-Fisher  RW.  Screening for melanoma by primary health care physicians: a cost-effectiveness analysis.  J Med Screen. 1996;3(1):47-53.PubMedGoogle ScholarCrossref
4.
Losina  E, Walensky  RP, Geller  A,  et al.  Visual screening for malignant melanoma: a cost-effectiveness analysis.  Arch Dermatol. 2007;143(1):21-28.PubMedGoogle ScholarCrossref
5.
Freedberg  KA, Geller  AC, Miller  DR, Lew  RA, Koh  HK.  Screening for malignant melanoma: a cost-effectiveness analysis.  J Am Acad Dermatol. 1999;41(5, pt 1):738-745.PubMedGoogle ScholarCrossref
6.
Beddingfield  FC. A Decision Analysis to Estimate the Effectiveness and Cost-effectiveness of Screening and an Analysis of the Relevant Epidemiology of the Disease [dissertation]. Santa Monica, CA: Pardee RAND Graduate School; 2002.
7.
Stang  A, Stausberg  J, Boedeker  W, Kerek-Bodden  H, Jöckel  KH.  Nationwide hospitalization costs of skin melanoma and non-melanoma skin cancer in Germany.  J Eur Acad Dermatol Venereol. 2008;22(1):65-72.PubMedGoogle Scholar
8.
Okafor  PN, Stallwood  CG, Nguyen  L,  et al.  Cost-effectiveness of nonmelanoma skin cancer screening in Crohn’s disease patients.  Inflamm Bowel Dis. 2013;19(13):2787-2795.PubMedGoogle ScholarCrossref
9.
Bigby  M.  Why the evidence for skin cancer screening is insufficient: lessons from prostate cancer screening.  Arch Dermatol. 2010;146(3):322-324.PubMedGoogle ScholarCrossref
10.
Hoorens  I, Vossaert  K, Pil  L,  et al.  Total-body examination vs lesion-directed skin cancer screening.  JAMA Dermatol. 2016;152(1):27-34.PubMedGoogle ScholarCrossref
11.
Watts  CG, Cust  AE, Menzies  SW, Coates  E, Mann  GJ, Morton  RL.  Specialized surveillance for individuals at high risk for melanoma: a cost analysis of a high-risk clinic.  JAMA Dermatol. 2015;151(2):178-186.PubMedGoogle ScholarCrossref
12.
Sobin  LH, Gospodarowicz  MK, Wittekind  CH.  TNM Classification of Malignant Tumors. 7th ed. Hoboken, NJ: Wiley-Blackwell; 2009.
13.
Chevolet  I, Hoorens  I, Janssens  A,  et al.  A short dermoscopy training increases diagnostic performance in both inexperienced and experienced dermatologists.  Australas J Dermatol. 2015;56(1):52-55.PubMedGoogle ScholarCrossref
14.
Kirkup  ME, De Berker  DA.  Clinical measurement of dimensions of basal cell carcinoma: effect of waiting for elective surgery.  Br J Dermatol. 1999;141(5):876-879.PubMedGoogle ScholarCrossref
15.
Smoller  BR.  Squamous cell carcinoma: from precursor lesions to high-risk variants.  Mod Pathol. 2006;19(suppl 2):S88-S92.PubMedGoogle ScholarCrossref
16.
Rijksinstituut voor Ziekte en Invaliditeitsverzekering (RIZIV). Nomenclatuur van de geneeskundige verstrekkingen. http://www.riziv.be. Published 2014. Accessed November 20, 2014.
17.
Cleemput  I, Neyt  M, Van de Sande  S, Thiry  N.  Belgian Guidelines for Economic Evaluations and Budget Impact Analyses: Second Edition. Brussels: Belgian Health Care Knowledge Centre (KCE); 2012. KCE report 183C: Health Technology Assessment.
18.
Williams  A.  What Could Be Nicer Than NICE? Office of Health Economics Annual Lecture. London, England: Office of Health Economics; August 2004.
19.
World Health Organization. Threshold values for intervention cost-effectiveness by region. http://www.who.int/choice/costs/CER_levels/en/. Published 2005. Accessed February 15, 2014.
20.
Boniol  M, Autier  P, Gandini  S.  Melanoma mortality following skin cancer screening in Germany.  BMJ Open. 2015;5(9):e008158. doi:10.1136/bmjopen-2015-008158PubMedGoogle ScholarCrossref
21.
Katalinic  A, Waldmann  A, Weinstock  MA,  et al.  Does skin cancer screening save lives? an observational study comparing trends in melanoma mortality in regions with and without screening.  Cancer. 2012;118(21):5395-5402.PubMedGoogle ScholarCrossref
22.
Katalinic  A, Eisemann  N, Waldmann  A.  Skin cancer screening in Germany: documenting melanoma incidence and mortality from 2008 to 2013.  Dtsch Arztebl Int. 2015;112(38):629-634.PubMedGoogle Scholar
23.
Hoorens  I, Vossaert  K, Ongenae  K, Brochez  L.  Is early detection of basal cell carcinoma worthwhile? systematic review based on the WHO criteria for screening.  Br J Dermatol. 2016;174(6):1258-1265.PubMedGoogle ScholarCrossref
24.
Collins  RE, Lopez  LM, Marteau  TM.  Emotional impact of screening: a systematic review and meta-analysis.  BMC Public Health. 2011;11:603.PubMedGoogle ScholarCrossref
Original Investigation
February 2017

Cost-effectiveness and Budget Effect Analysis of a Population-Based Skin Cancer Screening

Author Affiliations
  • 1Department of Public Health, Ghent University Hospital, Ghent, Belgium
  • 2Department of Dermatology, Ghent University Hospital, Ghent, Belgium
  • 3currently in private practice in Maldegem, Belgium
  • 4Department of Medical Oncology, Ghent University Hospital, Ghent, Belgium
  • 5Department of Dermatology, Institut Roi Albert II, Cliniques Universitaires Saint-Luc, Université Catholique de Louvain, Brussels, Belgium
  • 6Institute of Health and Society, Université Catholique de Louvain, Brussels, Belgium
 

Copyright 2016 American Medical Association. All Rights Reserved.

JAMA Dermatol. 2017;153(2):147-153. doi:10.1001/jamadermatol.2016.4518
Key Points

Question  Is population-based screening for skin cancer cost-effective?

Findings  A Markov model and budget effect analysis were designed based on a clinical trial comparing 2 screening methods (total-body skin examination [TBSE] and lesion-directed screening [LDS]). Both screening strategies produced a gain in quality-adjusted life-years (QALYs), resulting in incremental cost-effectiveness ratios of €33 072 (US $35 473) per QALY in men and €18 687 (US $20 044) per QALY in women for TBSE and €34 836 (US $37 365) per QALY in men and €19 470 (US $20 884) per QALY in women for LDS.

Meaning  These results suggest that screening using a 1-time TBSE is most cost-effective at a willingness-to-pay threshold of €35 000 (US $37 541) per QALY gained.

Abstract

Importance  Several epidemiological studies show an alarming global increase in incidence of melanoma and nonmelanoma skin cancer.

Objectives  To examine the cost-effectiveness of 2 population-based skin cancer screening methods and to assess their budget effect and the influence on skin cancer epidemiological findings.

Design, Setting, and Participants  A Markov model with a latent period of 20 years and a time horizon of 50 years was used to analyze the cost-effectiveness (societal perspective) and budget effect (public health care payer perspective) of 2 population-based skin cancer screening programs in Belgium compared with the absence of a screening program. A health economic analysis was based on a clinical trial performed in 2014 in Belgium. In the economic model, the total Belgian population 18 years or older was assumed to have been invited for the screening program.

Main Outcomes and Measures  The influence of the screening program on skin cancer epidemiological findings and the cost per quality-adjusted life-year (QALY) gained, as well as the budget effect, expressed as the net costs for the health care payer over 50 years.

Results  All participants (1668 total-body skin examination [TBSE] and 248 lesion-directed screening [LDS]) were screened by a team of 6 dermatologists from March 14 to 18, 2014, for TSBE and April 22 and 25 to 27, 2014, for LDS. Both screening strategies produced a gain in QALYs, resulting in incremental cost-effectiveness ratios of €33 072 (US $35 475) per QALY in men and €18 687 (US $20 044) per QALY in women for TBSE and €34 836 (US $37 365) per QALY in men and €19 470 (US $20 884) per QALY in women for LDS. With a 1-time screening, a 4.0% decrease in the incidence rates of stage III and IV melanoma was predicted at the population level relative to the comparator. The budget effect analysis demonstrated that during 20 years, a 1-time screening would incur a net cost for the health care payer of almost €36 million (US $38.6 million) for TBSE or just over €6 million (US $6.4 million) for LDS (€4.1 [US $4.40] or €0.7 [US $0.80], respectively, per adult).

Conclusions and Relevance  These results can be interpreted as cost-effective at a willingness-to-pay threshold in Belgium of €35 000 (US $37 541) per QALY gained. Based on these findings, a 1-time TBSE in the general adult population 18 years or older is the most cost-effective strategy and is predicted to result in a reduction of skin cancer mortality over 20 years and 50 years. The cost-effectiveness for skin cancer screening is higher in women than in men.

Introduction

The global skin cancer incidence is assessed to be between 2 and 3 million nonmelanoma skin cancers (NMSCs) and 132 000 melanomas each year. It is estimated that 1 in every 3 cancers diagnosed is a skin cancer.1 Despite the health burden and despite the idea that early detection can lead to better cure rates and reduce the costs of disease, few studies have assessed the effectiveness and cost-effectiveness of secondary prevention strategies.2 Screening is a prevention strategy by which early detection is intended to improve prognosis by shifting diagnosis to earlier stages of disease. Most available skin cancer screening studies mainly addressed melanoma,2-6 while NMSC is responsible for an important portion of the direct medical health care costs of skin cancer.7 While early detection campaigns have specifically focused on high-risk groups,8 the majority of skin cancers develop outside of these groups. Most published cost-effectiveness models on skin cancer screening predict mortality reduction; however, no evidence to date indicates that skin cancer screening with total-body skin examination (TBSE) is cost-effective.9 In this study, we compared the cost-effectiveness of 2 population-based screening strategies (a standard TBSE vs lesion-directed screening [LDS]10), organized as a pilot study in Belgium. Although recent efforts were made to assess the financial effect of a skin cancer screening intervention,11 most studies of health system costs do not include such analysis.2

Methods
Screening Strategies

The modeled screening strategies were based on a 2014 skin cancer screening trial10 performed in Belgium comparing TBSE with LDS in 2 sociodemographically comparable regions. The TBSE was performed in a community of 9325 adult inhabitants, who received a personal invitation. The LDS was performed in a comparably sized community, whose inhabitants were invited for a free-of-charge skin cancer check of a specific lesion meeting 1 or more of the following criteria: ABCD rule (asymmetry, border irregularity, color variation, and diameter >6 mm), “ugly duckling” sign, new lesion lasting longer than 4 weeks, or red nonhealing lesions. All participants (1668 TBSE and 248 LDS) were screened by a team of 6 dermatologists (K.V., L.B., and other nonauthors) from March 14 to 18, 2014, for TSBE and April 22 and 25 to 27, 2014, for LDS. The participation rate was higher in the TBSE region compared with the LDS region (17.9% [1668 of 9325] vs 3.3% [314 of 9484], P < .01). The skin cancer yield did not differ significantly between the groups (2.3% [39 of 1668] for TBSE vs 3.2% [8 of 248, only including lesions found and detected in patients consulting for a specific LDS lesion] for LDS, P = .40).10 The ethics committee of Ghent University Hospital approved this study, and written patient informed consent was obtained.

Model Structure

A Markov model with a latent period of 20 years and a time horizon of 50 years was developed using a spreadsheet (Excel 2013; Microsoft) complemented with a programming language (Visual Basic, version 10; Microsoft) and incorporated melanoma, basal cell carcinoma (BCC), and squamous cell carcinoma (SCC). The model included undiagnosed skin cancer, diagnosis and treatment, follow-up, and death by stage of disease (eAppendix in the Supplement). All Belgian male and female adults were assumed to have been invited. Modeled clinical outcomes of the screening were pathologically confirmed skin cancer, a false-positive result, or a false-negative result. Spontaneous clinical detection was also assumed in the screening scenario as a comparator situation. The duration of the diagnosis and treatment phase was 6 months (representing 1 cycle) for patients with BCC, stage 0 to II SCC, or stage I or II melanoma and 1 year for patients with stage III or IV SSC or stage III or IV melanoma. To assign a higher probability of death in the first years after diagnosis of stage IV SCC and stage IV melanoma, the follow-up phase was divided into intense and long-term follow-up, which lasted for 4 years, after which the individual moved into long-term follow-up. Patients in follow-up remained in this state until the end of the model’s time horizon or until they died. Melanoma and SCC stages were determined according to the seventh edition of the TNM Classification of Malignant Tumors.12 Stages for BCC were defined as less than 1 cm, 1 to 2 cm, greater than 2 cm, and aggressive histologic findings. Patients with BCC or SCC had a higher risk of developing melanoma. Risk of a recurrent or subsequent similar lesion (for all cancer types) was accounted for in costs. All cohort members started the study in one of the model states based on the baseline prevalence of BCC, SCC, and melanoma (Belgian Cancer Registry, unpublished data, 2013; Integraal Kankercentrum Nederland, unpublished data, 2011). The Markov model served the following 2 aims during 20 years: (1) to evaluate the incremental cost-effectiveness ratio (ICER) (calculated as the net costs divided by the net health effects) from a societal perspective and considering a willingness-to-pay threshold of €35 000 (US $37 541) per quality-adjusted life-year (QALY) gained and (2) to assess the budget effect. The budget effect analysis estimated the net cumulative cost of the screening program (and consequent examinations, treatment, and follow-up) for the public health care payer during 20 years, while allowing new entrance of 18-year-olds each cycle in the lesion-free state, who were representative of the natural progression of skin cancer.

Screening-Related Variables

We did not derive the test characteristics of dermoscopy from the screening trial because only expert dermatologists were involved, which can bias these variables (Table 1). Based on a study by Chevolet et al,13 the test characteristics of dermoscopy used by well-trained and less-trained dermatologists were calculated. The means of these values were used in our model.

Epidemiological and Clinical Data

The prevalence of skin cancer was defined as that of diagnosed (Integraal Kankercentrum Nederland, unpublished data, 2012; Belgian Cancer Registry, unpublished data, 2012) and undiagnosed skin cancer, with the latter calculated as the yield of the screening trial divided by the sensitivity of dermoscopy.13 The prevalence of NMSC was derived from the Dutch cancer registry because these cancers are more accurately registered in the Netherlands. However, a correction factor of 0.51 was used to adapt the NMSC values to Belgium. All-cause mortality risk was applied to all persons in the model (based on Belgian life tables), whereas mortality from skin cancer was applied to only patients with stage IV melanoma or SCC (Belgian Cancer Registry, unpublished data, 2004-2011). A main obstacle is the lack of data on the natural progression of undiagnosed melanomas. Therefore, we applied model calibration based on the yearly number of skin cancer deaths in Belgium. When this natural progression of stage II or III melanoma was set at 0.8% and stage III at 0.7% per 6 months, the output of the model showed approximately 11 100 deaths during 20 years, which was assumed to be acceptable. The natural progression of BCC was derived from a study by Kirkup and De Berker14 showing an evolution of 1 cm per 3.8 years or 1.2 mm per 6 months. The transition risk from stage 0 to II SCC to stage III or IV was estimated as 0.5% per 6 months based on the estimation by Smoller15 (1%-2% per year).

Costs and Health Effects

Costs included the cost of screening, direct medical costs, and costs because of productivity loss, expressed separately for the health care payer and for the patient. The total cost of the screening per participant was €4.9 (US $5.30) for TBSE and €1.8 (US $1.90) for LDS, mainly because of the difference in duration of the 2 screening methods. The direct costs for treatment and follow-up and the indirect costs because of productivity loss, morbidity, or early mortality were calculated based on a medical consumption questionnaire returned by 287 Belgian patients with skin cancer, multiplied by official Belgian unit costs.16,17

Health effects of the screening were represented by the influence on QALYs, which include the effect on quality of life and life expectancy as a result of the skin cancer stage shift. Consistent with the Belgian guidelines,17 health effects were discounted at 1.5% and costs at 3.0%.

Scenario and Sensitivity Analysis

Several scenarios were tested, including screening from the age of 40 years instead of 18 years (excluding the indirect costs because of productivity loss), extending the time horizon to 50 years instead of 20 years, and screening every 5 or 2 years during 20 years instead of only once. The sensitivity of the results to changes in individual variables was assessed by means of a deterministic 1-way sensitivity analysis and a probabilistic sensitivity analysis. Usefulness and probabilities were varied according to a beta distribution and costs according to a gamma distribution.

Results
Effect on Skin Cancer Epidemiological Findings

During 20 years, the model estimated the 1-time screening to result in a 4.0% decrease in the incidence rates of stage III and IV melanoma at the population level relative to the comparator. Moreover, both screening programs were estimated to have a positive, although modest, effect on mortality from skin cancer, with an absolute reduction of 628 deaths (273 male and 355 female) for TBSE and 118 (57 male and 61 female) for LDS. This effect corresponds to a relative mortality reduction of approximately 5.6% for TBSE and 1.0% for LDS relative to the comparator.

Cost-effectiveness
Base Case

Both screening strategies resulted in a gain in QALYs during 20 years (Table 1). Health effects and costs are in good balance, leading to ICERs of €33 072 (US $35 473) per QALY gained in men and €18 687 (US $20 044) per QALY gained in women for TBSE and €34 836 (US $37 365) per QALY gained in men and €19 470 (US $20 884) per QALY gained in women for LDS, which can be interpreted as a moderate cost-effective result regarding a willingness-to-pay threshold in Belgium of €35 000 (US $37 541) per QALY gained.18,19

Scenario and Sensitivity Analysis

A 1-time TBSE in the general adult population 18 years or older remained the most cost-effective strategy (Table 2). Screening every 2 or 5 years had a lower cost-effectiveness ratio, but since the time horizon was set at 50 years for this scenario because a 20-year time horizon would not capture the effect of screening in, for example, year 18, it should be compared with the scenario of a 1-time screening with a time horizon of 50 years. The 1-way sensitivity analysis showed the most influencing variables for men to be the natural progression of melanoma, the usefulness related to melanoma, the prevalence of undiagnosed melanoma and BCC, the direct cost of BCC long-term follow-up, and the direct cost of stage III and IV melanoma (Figure 1). A higher value for these variables led to a more cost-effective result, except for the prevalence of undiagnosed BCC and the direct cost of BCC follow-up, in which the effect was the opposite. In case of a worse value (bars on the right side of the figure), ICERs exceeded the willingness-to-pay threshold of €35 000 (US $37 541) per QALY gained in men; in women, only the variation in the natural progression of melanoma led to an ICER exceeding the threshold. The probabilistic sensitivity analysis created credibility intervals around the deterministic result (Figure 2). The cost-effectiveness planes show that most simulations are located in the northeast quadrant and are below the willingness-to-pay threshold of €35 000 per QALY gained, although part of the values are situated above the threshold for the simulation in men. The cost-effectiveness acceptability curves in the eAppendix in the Supplement show that, regarding the willingness-to-pay threshold of €35 000 per QALY gained, the probabilities of screening being cost-effective for TBSE and LDS are 79.7% and 59.9%, respectively, in men and 100.0% and 99.9%, respectively, in women.

Budget Effect

The budget effect analysis demonstrated that during 20 years, a 1-time screening would incur a net cost for the health care payer of almost €36 million (US $38.6 million) for TBSE or just over €6 million (US $6.4 million) for LDS (€4.1 [US $4.40] or €0.7 [US $0.80], respectively, per adult). These results are summarized in Table 3.

Discussion

During 20 years, a 1-time TBSE was predicted to gain 2380 healthy life-years in the total population (8.8 million), and LDS gained 397 (Table 1). In addition, TBSE was projected to reduce skin cancer mortality by 5.6% during 20 years. While the transient decrease in melanoma mortality in the Schleswig-Holstein region of Germany, followed by return to prescreening levels,20 could reflect a temporal modification in reporting causes of death, it could also be the result of insufficient elapsed time from skin cancer screening to realize a change in mortality.21 No decrease in melanoma mortality has been documented since nationwide skin cancer screening was introduced in 2013 in Germany.22

Because of the screening cost and the extra costs for treatment and follow-up, implementing a 1-time screening has a fiscal effect for the health care payer. Nevertheless, the balance between costs and health effects is beneficial both for TBSE and LDS (with ratios below the willingness-to-pay threshold in Belgium of €35 000 [US $37 541] per QALY gained). In men, both screening strategies approach this threshold. However, the probability of the screening’s cost-effectiveness being below the threshold was 79.7% for TBSE and 59.9% for LDS in men. The ICER for TBSE exceeded that for LDS. Lesion-directed screening did not seem to have a substantial effect on increasing healthy life-years or reducing deaths owing to the low participation rates in the LDS screening arm.10

Because the skin cancer detection rates were comparable in both screening arms and because LDS screening was less time-consuming, it may be worthwhile to investigate how participation in this type of screening could be increased. If the participation rates for TBSE could be reached in LDS, then LDS would be more cost-effective than TBSE. Screening from the age of 40 years instead of 18 years only slightly reduced the cost-effectiveness, probably because screening could gain more health benefits in younger persons and because older persons have a greater risk of mortality from causes other than skin cancer. If the time horizon of the model is extended to 50 years, then the cost-effectiveness ratio would be better than a 20-year time horizon, meaning that the effect of a 1-time screening is estimated to still continue for 50 years. While successive screening could be cost-effective, the 1-time screening would still be the most cost-effective strategy.

Sensitivity analysis showed that the natural progression of skin cancer has a major influence on the cost-effectiveness outcome. Further research on the evolution of the disease would be beneficial. Other important variables were the direct medical costs of stage III and IV melanoma and the sensitivity of dermoscopy for melanoma. It is possible that the cost for treating stage III and IV melanoma will keep rising because of new treatments, which would result in screening becoming more cost-effective. In addition, the prevalence of undiagnosed BCC and melanoma was influential; therefore, a higher prevalence of undiagnosed melanoma would lead to greater health benefits that compensate for the extra costs, which would make the screening more cost-effective. With a higher prevalence of undiagnosed BCC, screening would be less cost-effective because currently available QALY data for BCC do not indicate that detecting and treating BCC leads to benefits in quality of life. Although treating small BCCs is less expensive than treating more advanced BCC lesions,23 the extra costs for treating small BCCs in the population are prohibitive. Furthermore, because greater sensitivity of dermoscopy leads to better cost-effectiveness, training initiatives for this procedure are recommended.

Other cost-effectiveness studies of skin cancer screening in Australia and the United States have been limited to melanoma. Most of these studies expressed the cost-effectiveness of melanoma screening in cost per life-year saved. In Australia, screening men during 50 years biennially by general practitioners resulted in a ratio of A$12.14 (US $9.08) per life-year saved.3 In the United States, a 1-time screening by dermatologists in a self-selected population resulted in US $51 481 per life-year saved6 and in a high-risk population resulted in US $39 600 per life-year saved.5 One study4 calculated the cost per QALY gained of a visual 1-time screening beginning at age 50 years to be US $10 100 (approximately €9256). When 1-time screening was implemented biennially, the ratio rose to US $80 700 per QALY gained (approximately €73 882 per QALY gained) and if annually to US $586 800 per QALY gained (approximately €537 220 per QALY gained). Our study supports the latter result of better cost-effectiveness with a 1-time screening. However, it is difficult to compare studies because of different methods of screening (visual screening vs dermoscopy screening, as well as the composition of the screening team), various model designs, and variable epidemiological makeup of the populations (eg, the incidence of melanoma is higher in the United States and Australia than in Belgium).

A major strength of this study is that the effectiveness measures are based on a large population-based screening trial, in addition to the detailed analysis of the costs and benefits of a skin cancer screening program. Not only were the benefits of screening captured in the model, but the effect of a false-positive screening result on quality of life in terms of psychological harms was included as well. In our model, the screening examination itself did not have an effect on quality of life. Collins et al24 showed that screening (in general) does not appear to have an adverse emotional effect in the long term and noted that few studies have assessed the short-term emotional influence of screening. Hoorens et al10 examined the anxiety of study participants just before and after skin cancer screening and demonstrated a general decrease immediately after screening. However, baseline levels before screening were not recorded, so no conclusions on the psychological effect (induction of anxiety) of skin cancer screening could be obtained from their study.

Limitations

Some limitations of our study should be addressed. First, there is no accurate registration of NMSC in Belgium. Therefore, we relied on epidemiological results of the Dutch cancer registry, which has a systematic registration of NMSC. Second, accurate information on the natural progression of skin cancer is not available to date. Sensitivity analysis showed this item to be an important determinant of cost-effectiveness. In our model, the natural progression was estimated according to calibration, which is generally a more reliable approach than making assumptions on variables based on limited studies. Third, diversity in skin cancer epidemiological findings worldwide and screening variables (eg, participation rate, diagnostic performance of the screening team, and unit costs of detection, treatment, and follow-up) are context specific, limiting the direct transferability of our results across different countries. However, we believe that not only the mean results but also the findings from the scenario and sensitivity analysis can inform other nations.

Conclusions

In terms of policy implications, skin cancer screening proved to be cost-effective at a willingness-to-pay threshold of €35 000 (US $37 541) per QALY gained. Based on this finding, a 1-time TBSE in the general adult population 18 years or older with screening for both melanoma and NMSC is the most cost-effective strategy and was projected to result in a substantial reduction of 5.6% in skin cancer mortality during 20 years at a cost of €4.1 (US $4.40) per adult. The cost-effectiveness for skin cancer screening is higher in women than in men.

Back to top
Article Information

Accepted for Publication: September 30, 2016.

Corresponding Author: Lieve Brochez, MD, PhD, Department of Dermatology, Ghent University Hospital, De Pintelaan 185, 9000 Ghent, Belgium (lieve.brochez@ugent.be).

Published Online: December 14, 2016. doi:10.1001/jamadermatol.2016.4518

Author Contributions: Drs Hoorens and Brochez had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Ms Pil and Dr Hoorens contributed equally to this study. Drs Annemans and Brochez contributed equally to this study.

Study concept and design: Pil, Hoorens, Vossaert, Annemans, Brochez.

Acquisition, analysis, or interpretation of data: Pil, Hoorens, Vossaert, Kruse, Tromme, Speybroeck, Brochez.

Drafting of the manuscript: Pil, Hoorens, Brochez.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Pil, Hoorens, Speybroeck.

Administrative, technical, or material support: Hoorens, Brochez.

Conflict of Interest Disclosures: Dr Hoorens reported being funded by a PhD fellowship grant from the Klinisch Onderzoeksfonds of Ghent University Hospital. No other disclosures were reported.

Funding/Support: This study was supported in part by a research grant from The LEO Foundation and by a grant from the Belgian Federation Against Cancer (Dr Annemans).

Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions: Barbara Boone, MD, PhD, Sofie De Schepper, MD, PhD, and Katia Ongenae, MD, PhD (all with the Department of Dermatology, Ghent University Hospital) assisted with the screening.

References
1.
World Health Organization. How common is skin cancer? http://www.who.int/uv/faq/skincancer/en/index1.html. Published 2016. Accessed June 22, 2016.
2.
Gordon  LG, Rowell  D.  Health system costs of skin cancer and cost-effectiveness of skin cancer prevention and screening: a systematic review.  Eur J Cancer Prev. 2015;24(2):141-149.PubMedGoogle ScholarCrossref
3.
Girgis  A, Clarke  P, Burton  RC, Sanson-Fisher  RW.  Screening for melanoma by primary health care physicians: a cost-effectiveness analysis.  J Med Screen. 1996;3(1):47-53.PubMedGoogle ScholarCrossref
4.
Losina  E, Walensky  RP, Geller  A,  et al.  Visual screening for malignant melanoma: a cost-effectiveness analysis.  Arch Dermatol. 2007;143(1):21-28.PubMedGoogle ScholarCrossref
5.
Freedberg  KA, Geller  AC, Miller  DR, Lew  RA, Koh  HK.  Screening for malignant melanoma: a cost-effectiveness analysis.  J Am Acad Dermatol. 1999;41(5, pt 1):738-745.PubMedGoogle ScholarCrossref
6.
Beddingfield  FC. A Decision Analysis to Estimate the Effectiveness and Cost-effectiveness of Screening and an Analysis of the Relevant Epidemiology of the Disease [dissertation]. Santa Monica, CA: Pardee RAND Graduate School; 2002.
7.
Stang  A, Stausberg  J, Boedeker  W, Kerek-Bodden  H, Jöckel  KH.  Nationwide hospitalization costs of skin melanoma and non-melanoma skin cancer in Germany.  J Eur Acad Dermatol Venereol. 2008;22(1):65-72.PubMedGoogle Scholar
8.
Okafor  PN, Stallwood  CG, Nguyen  L,  et al.  Cost-effectiveness of nonmelanoma skin cancer screening in Crohn’s disease patients.  Inflamm Bowel Dis. 2013;19(13):2787-2795.PubMedGoogle ScholarCrossref
9.
Bigby  M.  Why the evidence for skin cancer screening is insufficient: lessons from prostate cancer screening.  Arch Dermatol. 2010;146(3):322-324.PubMedGoogle ScholarCrossref
10.
Hoorens  I, Vossaert  K, Pil  L,  et al.  Total-body examination vs lesion-directed skin cancer screening.  JAMA Dermatol. 2016;152(1):27-34.PubMedGoogle ScholarCrossref
11.
Watts  CG, Cust  AE, Menzies  SW, Coates  E, Mann  GJ, Morton  RL.  Specialized surveillance for individuals at high risk for melanoma: a cost analysis of a high-risk clinic.  JAMA Dermatol. 2015;151(2):178-186.PubMedGoogle ScholarCrossref
12.
Sobin  LH, Gospodarowicz  MK, Wittekind  CH.  TNM Classification of Malignant Tumors. 7th ed. Hoboken, NJ: Wiley-Blackwell; 2009.
13.
Chevolet  I, Hoorens  I, Janssens  A,  et al.  A short dermoscopy training increases diagnostic performance in both inexperienced and experienced dermatologists.  Australas J Dermatol. 2015;56(1):52-55.PubMedGoogle ScholarCrossref
14.
Kirkup  ME, De Berker  DA.  Clinical measurement of dimensions of basal cell carcinoma: effect of waiting for elective surgery.  Br J Dermatol. 1999;141(5):876-879.PubMedGoogle ScholarCrossref
15.
Smoller  BR.  Squamous cell carcinoma: from precursor lesions to high-risk variants.  Mod Pathol. 2006;19(suppl 2):S88-S92.PubMedGoogle ScholarCrossref
16.
Rijksinstituut voor Ziekte en Invaliditeitsverzekering (RIZIV). Nomenclatuur van de geneeskundige verstrekkingen. http://www.riziv.be. Published 2014. Accessed November 20, 2014.
17.
Cleemput  I, Neyt  M, Van de Sande  S, Thiry  N.  Belgian Guidelines for Economic Evaluations and Budget Impact Analyses: Second Edition. Brussels: Belgian Health Care Knowledge Centre (KCE); 2012. KCE report 183C: Health Technology Assessment.
18.
Williams  A.  What Could Be Nicer Than NICE? Office of Health Economics Annual Lecture. London, England: Office of Health Economics; August 2004.
19.
World Health Organization. Threshold values for intervention cost-effectiveness by region. http://www.who.int/choice/costs/CER_levels/en/. Published 2005. Accessed February 15, 2014.
20.
Boniol  M, Autier  P, Gandini  S.  Melanoma mortality following skin cancer screening in Germany.  BMJ Open. 2015;5(9):e008158. doi:10.1136/bmjopen-2015-008158PubMedGoogle ScholarCrossref
21.
Katalinic  A, Waldmann  A, Weinstock  MA,  et al.  Does skin cancer screening save lives? an observational study comparing trends in melanoma mortality in regions with and without screening.  Cancer. 2012;118(21):5395-5402.PubMedGoogle ScholarCrossref
22.
Katalinic  A, Eisemann  N, Waldmann  A.  Skin cancer screening in Germany: documenting melanoma incidence and mortality from 2008 to 2013.  Dtsch Arztebl Int. 2015;112(38):629-634.PubMedGoogle Scholar
23.
Hoorens  I, Vossaert  K, Ongenae  K, Brochez  L.  Is early detection of basal cell carcinoma worthwhile? systematic review based on the WHO criteria for screening.  Br J Dermatol. 2016;174(6):1258-1265.PubMedGoogle ScholarCrossref
24.
Collins  RE, Lopez  LM, Marteau  TM.  Emotional impact of screening: a systematic review and meta-analysis.  BMC Public Health. 2011;11:603.PubMedGoogle ScholarCrossref
×