The American Academy of Dermatology’s SPOTme program has provided free skin cancer screenings since 1985.1 Herein, we report the number needed to screen (NNS) for presumptive diagnoses of basal cell carcinoma (pBCC), squamous cell carcinoma (pSCC), and cutaneous melanoma (pCM) among first-time SPOTme screenees from 1992–2010.
From these years, 1,355,945 screening forms were available. After applying previously described age- and missing data-based exclusion criteria1, 1,299,765 forms remained (95.9% of available forms). To prevent repeated sampling, prior attendees of a SPOTme screening were excluded, leaving 945,367 first-time screenees for detailed analysis (69.7% of available forms). Among them, 64,325 pBCC (6.8%), 14,434 pSCC (1.5%), and 11,398 pCM (1.2%) were identified. Presumptive diagnoses are clinically suspicious lesions identified on visual inspection – they are not biopsy-confirmed. The NNS for each presumptive diagnosis was calculated by dividing the number of individuals screened by the number of each lesion-type. This was repeated by age-sex, race, and number of self-reported skin cancer risk factors (personal and family history of skin cancer; recent mole change). Finally, intra-variable NNS comparisons were performed (two-way t-test for age-sex and race; one-way ANOVA for risk factors variable).
Whites (90%), women (63%), those possessing skin cancer risk factors (57%), and persons age 50 years and older (56%) were strongly represented (Table I). The overall NNS was 14.7 for pBCC, 65.5 for pSCC, and 82.9 for pCM. For all diagnoses, the NNS was fewest for men age 50 and older. Whites displayed a four-fold lower NNS for pBCC and pSCC compared to nonwhites, while the NNS for pCM among whites was only 25% lower than for nonwhites. Comparing white men and women age 50 and older, the NNS for pBCC was 23% lower among men, while the NNS for both pSCC and pCM was roughly 37% lower. Having multiple skin cancer risk factors had the greatest impact on the NNS for pSCC (71% reduction in NNS (NNS = 23, 3 risk factors; NNS = 80, 0 risk factors). For each presumptive diagnosis, the p-value was < 0.001 for all intra-variable NNS comparisons.
Table I.
Number needed to screen (NNS) for presumptive screening diagnoses among first-time SPOTme participants, 1992–2010* (n = 945,367)
| Presumptive BCC (n = 64,325; 6.8%**) | Presumptive SCC (n = 14,434; 1.5%**) | Presumptive Cutaneous Melanoma (n = 11,398; 1.2%**) | ||
|---|---|---|---|---|
| Characteristic | n (%) | NNS1 | NNS | NNS |
| Women | ||||
| < 50 years | 274,601 (29.0) | 28.8 | 206.6 | 88.9 |
| ≥ 50 years | 324,834 (34.4) | 12.5 | 57.3 | 101.9 |
| Men | ||||
| < 50 years | 139,964 (14.8) | 20.7 | 109.7 | 74.8 |
| ≥ 50 years | 205,968 (21.8) | 9.4 | 33.4 | 63.4 |
| Race | ||||
| White | 495,637 (89.5) | 14.5 | 54.4 | 100.2 |
| Non-white | 58,069 (10.5) | 57.0 | 209.6 | 124.1 |
| Unavailable2 | 391,661 | |||
| Risk factors3 | ||||
| 0 | 407,133 (43.1) | 18.4 | 80.0 | 108.5 |
| 1 | 391,994 (41.5) | 14.5 | 65.4 | 75.0 |
| 2 | 133,715 (14.1) | 10.3 | 47.6 | 61.4 |
| 3 | 12,525 (1.3) | 5.7 | 23.2 | 52.2 |
| All screenees | 945,367 | 14.7 | 65.5 | 82.9 |
Number needed to screen (NNS) = (total number of screening participants) / (number with the presumptive diagnosis)
No screening forms were available for 1995, 1996, or 2000
Proportion of all participants with that particular presumptive diagnosis
For each presumptive skin cancer, p-value < 0.001 for all intra-variable NNS comparisons (two-way t-test of proportions used for age-sex and race variables; one-way ANOVA used for risk factor variable)
Race not available prior to 2001. No missing data 2002–2010.
Risk factors include: personal history of skin cancer; family history of skin cancer; and a recent change in mole(s) size, color, or shape
A major limitation herein is the lack of histological confirmation for presumptive diagnoses. Consequently, the NNS values are significantly lower than would be anticipated based on incidence data2,3 for these cancers. For comparison, in the SCREEN Project from Schleswig-Holstein Germany (which had pathology confirmation of screening diagnoses), the NNS observed for BCC was 184, melanoma 616, and SCC 920.4
Our study highlights the low NNS for presumptive skin cancers among SPOTme participants. Furthermore, the NNS for presumptive skin cancers was markedly lower among those with self-identified risk factors, suggesting that future outreach strategies could employ these risk factors in their marketing campaigns to increase the yield/cost-effectiveness of screening. Consistent with other epidemiological skin cancer studies,2,5 this analysis of the largest U.S.-based skin cancer screening program reaffirms that mass screening programs would likely achieve their greatest yield among those with multiple skin cancer risk factors and men age 50 years and older – a group that assumes the largest burden of melanoma fatality.
Acknowledgments
Funding sources:
Supported by the American Academy of Dermatology’s SPOTme program, which is made possible through a donation from the Bristol-Myers Squibb Foundation.
Dr. Beaulieu was supported by the National Center for Advancing Translational Sciences, National Institutes of Health (award No. TL1TR001062) and Dr. Tsao was supported by the National Institutes of Health (award No. K24CA149202).
Abbreviations
- NNS
Number Needed to Screen
- pBCC
presumptive Basal Cell Carcinoma
- pSCC
presumptive Squamous Cell Carcinoma
- pCM
presumptive Cutaneous Melanoma
- SCREEN
Skin Cancer Research to Provide Evidence for Effectiveness of Screening in Northern Germany
- ANOVA
Analysis of Variance
Footnotes
Conflicts of interest:
The authors have no conflict of interest to declare
References
- [1].Okhovat JP, Beaulieu D, Tsao H., Halpern AC, Michaud DS, Shaykevich S, Geller AC. The first 30 years of the American Academy of Dermatology skin cancer screening program: 1985–2014. J Am Acad Dermatol. 2018. pii: S0190–9622(18)32047–4. doi: 10.1016/j.jaad.2018.05.1242 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Noone AM, Howlader N, Krapcho M, Miller D, Brest A, Yu M, et al. SEER Cancer Statistics Review, 1975–2015. National Cancer Institute. Bethesda, MD. Available at: https://seer.cancer.gov/csr/1975_2015/. Accessed December 1, 2018. [Google Scholar]
- [3].Rogers HW, Weinstock MA, Feldman SR, Coldiron BM. Incidence estimate of nonmelanoma skin cancer (keratinocyte carcinomas) in the US population, 2012. JAMA Dermatol 2015; 151(10):1081–1086. doi: 10.1001/jamadermatol.2015.1187 [DOI] [PubMed] [Google Scholar]
- [4].Breitbart EW, Waldmann A, Nolte S, Capellaro M, Greinert R, Volkmer B, et al. Systematic skin cancer screening in Northern Germany. J Am Acad Dermatol. 2012; 66:201–211. doi: 10.1016/j.jaad.2010.11.016. [DOI] [PubMed] [Google Scholar]
- [5].Geller AC, Miller DR, Annas GD, Demierre MF, Gilchrest BA, Koh HK. Melanoma incidence and mortality among US whites, 1969–1999. JAMA. 2002; 288:1719–1720. [DOI] [PubMed] [Google Scholar]