Focusing on Diagnostic outcomes: The Impact of Image Quality in Global Teledermatology

Summary: Background: Skin disease, the fourth-leading global cause of nonfatal disease burden, is worsening in resource-limited regions.1,2 The Addis Clinic teledermatology (TD) program, collaborating with 12 countries, offers vital care in dermatologist-scarce regions. Although image quality is paramount for TD’s diagnostic efficacy, studies report poor image quality results in 3-8% of TD cases lacking a diagnosis.3–6 This study examines relationships between image quality and TD diagnostic efficacy in resource-limited settings. Study Type: Retrospective cohort Methods: We analyzed image quantity and quality in TD consultations from low-income countries, categorizing images as good or poor quality based on skin visibility, lighting, clarity, and depth perception. Cases were divided by percentage of good-quality images: 0-25%, 25-50%, 50-75%, and 75-100%. Cases were labeled as having a definitive, differential, or no diagnosis. Chi-squared tests were used to evaluate significance of associations between image quality and diagnostic outcomes. Results: We analyzed 1,040 Addis Clinic TD consultations from 91 centers in 58 municipalities across 12 low-resource countries, averaging 4.4 images per consultation. 31.2% of cases had 0-25% good-quality images, 5.4% had 25-50%, 15.6% had 50-75%, and 47.9% had 75-100%. Notably, while 43.5% had all good-quality images, 30.6% had none. The percentage of good-quality images was associated with diagnostic outcomes (p=.00038). Higher percentages were associated with a definitive diagnosis (p=.048), while lower percentages were associated with no diagnosis (p=.000028). Conclusions: These findings highlight the importance of image quality in accurate TD diagnosis of dermatological conditions. Implementing image quality guidelines could significantly improve TD diagnostic accuracy, health outcomes, and healthcare equity in resource-limited settings.