Portal de Periódicos da CAPES

Mobile teledermatology for melanoma detection: Assessment of the validity in the framework of a population-based skin cancer awareness campaign in northern Italy

2019; Elsevier BV; Volume: 81; Issue: 1 Linguagem: Inglês

10.1016/j.jaad.2019.02.036

1097-6787

Simone Cazzaniga, E Castelli, Anna Di Landro, Marco Di Mercurio, Gianlorenzo Imberti, Gustavo Andrea Locatelli, Francesca Raponi, Pamela Vezzoli, Daniele Gambini, Giovanni Damiani, Alberto Zucchi, Luigi Naldi,

Mobile Health and mHealth Applications

To the Editor: Increasing awareness and promotion of self-examination are possible ways to anticipate melanoma diagnosis and improve survival.1Brunssen A. Waldmann A. Eisemann N. et al.Impact of skin cancer screening and secondary prevention campaigns on skin cancer incidence and mortality: a systematic review.J Am Acad Dermatol. 2017; 76: 129-139Abstract Full Text Full Text PDF PubMed Scopus (80) Google Scholar We designed a study to assess the advantage of using a phone application (ie, an app) in the context of a campaign to promote early melanoma diagnosis.2Cazzaniga S. Castelli E. Di Landro A. et al.Development of a teledermatology system for the melanoma diagnosis. The pilot experience of the project Clicca il neo.Recenti Prog Med. 2016; 107: 440-443Google Scholar Apps for melanoma detection lack validation.3Kassianos A.P. Emery J.D. Murchie P. et al.Smartphone applications for melanoma detection by community, patient and generalist clinician users: a review.Br J Dermatol. 2015; 172: 1507-1518Crossref PubMed Scopus (103) Google Scholar, 4Buechi R. Faes L. Bachmann L.M. et al.Evidence assessing the diagnostic performance of medical smartphone apps: a systematic review and exploratory meta-analysis.BMJ Open. 2017; 7: e018280Crossref PubMed Scopus (33) Google Scholar As a preliminary step, we did a prospective validity study comparing distant assessment of lesions sent by the app with direct clinical evaluation. The study was conducted in the province of Bergamo in Italy during March-July 2017. Persons ≥18 years of age who used the free app advertised in the campaign to send pictures of their suspicious lesions for specialist assessment were invited to undergo a free whole-body examination by dermatologists, different from those providing the online assessment. The study was approved by the Ethics Committee of Papa Giovanni XXIII Hospital (Bergamo, Italy). The teledermatology cycle is shown in Fig 1. A total of 232 patients were included in the study. Those who used the app were significantly younger and more educated than the general population in the area (Table I). They were also more frequently Fitzpatrick phototype I and II, more frequently reported a history of sunburns, and had a larger mean number of nevi on upper limbs (data not shown).Table IDemographics and general characteristics of the study sample compared with the reference adult population of the Bergamo ProvinceCharacteristicOur sample, N = 232, n∗Numbers might not add up to the total due to missing data.%Bergamo, Italy, adult population, N = 910,297,†The 2017 Istat census data were based on resident population aged ≥18 years. Census data on educational attainment were available only for 2011. n%P‡Two-sample t test, assuming unequal variances, and Pearson χ2 test were used for continuous and categorical variables, respectively.Age, years Mean ± standard deviation42.7 ± 13.750.6 ± 18.4<.001 <304318.5134,02414.7 30-4912051.7270,81629.7 ≥506929.7458,80050.4Sex Male10444.8447,17049.1.19 Female12855.2463,12750.9Marital status Unmarried8236.6291,87432.1.07 Married or common-law partner12756.7517,67756.9 Divorced or widow156.7100,74611.1Education Primary school41.8212,20523.1<.001 Lower secondary4218.4333,50936.3 Upper secondary10546.1283,03430.8 University7733.889,6329.8Smoking habits No or former18579.739,56179.5.92 Yes4720.310,22320.5Alcohol consumption Never or sometimes14162.138,80377.0.07 Regular8637.911,58523.0Personal history of skin cancer§Estimates were based on data from the PraKtis study (Naldi L, Colombo P, Placchesi EB, et al. Study design and preliminary results from the pilot phase of the PraKtis study: self-reported diagnoses of selected skin diseases in a representative sample of the Italian population. Dermatology. 2004;208:38-42.) No22899.6906,65799.6.99 Melanoma10.436400.4∗ Numbers might not add up to the total due to missing data.† The 2017 Istat census data were based on resident population aged ≥18 years. Census data on educational attainment were available only for 2011.‡ Two-sample t test, assuming unequal variances, and Pearson χ2 test were used for continuous and categorical variables, respectively.§ Estimates were based on data from the PraKtis study (Naldi L, Colombo P, Placchesi EB, et al. Study design and preliminary results from the pilot phase of the PraKtis study: self-reported diagnoses of selected skin diseases in a representative sample of the Italian population. Dermatology. 2004;208:38-42.) Open table in a new tab From the online assessment, 56 lesions were classified as suspicious. On direct clinical examination, 14 (25%) of those lesions were confirmed as suspicious. Only 1 lesion (0.5%) classified as nonsuspicious online was considered suspicious on direct clinical examination. All the lesions classified as suspicious on direct clinical examination were assessed histologically. A total of 6 (2.6%) lesions, all classified as suspicious online, were confirmed as melanomas (mean Breslow's thickness 0.4 ± 0.2 mm). In total, 2 additional suspicious lesions (0.9%) were classified as pigmented basal cell carcinomas both on direct clinical examination and histologically. The other suspicious lesions were melanocytic nevi. The diagnostic accuracy of the online assessment compared with direct clinical examination was 81.0% (95% confidence interval [CI] 75.4%-85.9%), sensitivity was 92.9% (95% CI 66.1%-99.8%), specificity 80.3% (95% CI 74.4%-85.3), positive predictive value 23.2% (95% CI 13.0%-36.4%), and negative predictive value 99.4% (95% CI 96.9%-100%). The main reasons reported for using the app were the presence of a suspicious lesion (67.2%) and the ease of access to the system (50.5%). The overall satisfaction on the Visual Analogue Scale was 9.5 ± 0.9. Remarkably, about 70% of participants would not have gone to see a dermatologist without the program, indicating that teledermatology might help to promote patient empowerment. Our study had some limitations. We did not collect information on people who did not use the system. Moreover, our sample was probably biased toward those who were more concerned about their health and familiar with technology. Last, self-examination can lead to missing lesions at sites difficult to explore (eg, scalp). Ideally, simple and largely available technologic tools should be adopted to improve access. Our system, which involved the use of remote clinicians to judge lesions, was expensive and time-consuming. In the near future, machine learning systems connected with an app could revolutionise melanoma early diagnosis.5Zakhem G.A. Motosko C.C. Ho R.S. How should artificial intelligence screen for skin cancer and deliver diagnostic predictions to patients?.JAMA Dermatol. 2018; 154: 1383-1384Crossref PubMed Scopus (23) Google Scholar We are indebted to the following for their help: Dr Mara Azzi and Dr Giorgio Barbaglio (Agenzia di Tutela della Salute, Bergamo); Dr Fabio Pezzoli, Dr Paolo Sena, and Dr Alberto Reseghetti (Azienda Socio-Sanitaria Territoriale Papa Giovanni XXIII, Bergamo); Mr Marco D'Agostino (Verona); Mr Ferruccio Arnoldi (Bergamo); Mrs Eugenia Caggese, Dr Cornelia Zinetti, Dr Giovanna Rao, and Dr Laura Marchesi (Centro Studi Italian Group for Epidemiologic Research in Dermatology, Bergamo); and Mrs Lucia De Ponti, Dr Roberto Sacco, and Dr Giuseppe Chiesa (Lega Italiana per la Lotta contro i Tumori, Bergamo). Comment on: "Mobile teledermatology for melanoma detection: Assessment of validity in the framework of a population-based skin cancer awareness campaign in northern Italy"Journal of the American Academy of DermatologyVol. 81Issue 6PreviewTo the Editor: We read with interest the recent article by Cazzaniga et al1 on mobile teledermatology for detecting malignant melanoma in northern Italy. We believe that this work is essential because it emphasizes the importance of mobile devices to detect melanoma in an early stage. Full-Text PDF