Digital photography in skin cancer screening by mobile units in remote areas of Brazil

BMC Dermatol. 2014 Dec 24:14:19. doi: 10.1186/s12895-014-0019-1.

Abstract

Background: Non-melanoma skin cancer (NMSC) is one of the most common neoplasms in the world. Despite the low mortality rates, NMSC can still cause severe sequelae when diagnosed at advanced stages. Malignant melanoma, the third most common type of skin cancer, has more aggressive behavior and a worse prognosis. Teledermatology provides a new tool for monitoring skin cancer, especially in countries with a large area and unequal population distribution. This study sought to evaluate the performance of digital photography in skin cancer diagnosis in remote areas of Brazil.

Methods: A physician in a Mobile Prevention Unit (MPU) took four hundred sixteen digital images of suspicious lesions between April 2010 and July 2011. All of the photographs were electronically sent to two oncologists at Barretos Cancer Hospital who blindly evaluated the images and provided a diagnosis (benign or malignant). The absolute agreement rates between the diagnoses made by direct visual inspection (by the MPU physician) and through the use of digital imaging (by the two oncologists) were calculated. The oncologists' accuracy in predicting skin cancer using digital imaging was assessed by means of overall accuracy (correct classification rate), sensitivity, specificity and predictive value (positive and negative). A skin biopsy was considered the gold standard.

Results: Oncologist #1 classified 59 lesions as benign with the digital images, while oncologist #2 classified 27 lesions as benign using the same images. The absolute agreement rates with direct visual inspection were 85.8% for oncologist #1 (95% CI: 77.1-95.2) and 93.5% for oncologist #2 (95% CI: 84.5-100.0). The overall accuracy of the two oncologists did not differ significantly.

Conclusions: Given the high sensitivity and PPV, Teledermatology seems to be a suitable tool for skin cancer screening by MPU in remote areas of Brazil.

MeSH terms

  • Brazil
  • Dermoscopy / methods*
  • Early Detection of Cancer / methods
  • Humans
  • Image Interpretation, Computer-Assisted / methods*
  • Mobile Health Units*
  • Observer Variation
  • Photography*
  • Remote Consultation / methods*
  • Rural Health Services*
  • Skin Neoplasms / diagnosis*
  • Telepathology / methods