Since smart phones have become affordable for the masses, an increased interest in teledermatology as a quick and handy tool for the evaluation of skin diseases has been observed. Patient satisfaction with teledermatology is high, but doubts persist concerning reliability [1, 2]. Although previous studies have demonstrated good agreement between teledermatologic and face-to-face diagnoses , it is unclear whether therapeutic recommendations based on mobile teledermatology are in accordance with conventional dermatology. The aim of this study was to assess concordance between face-to-face and teledermatological consultations and determine the accuracy of teledermatology relative to histopathology, when possible.
A total of 195 patients presenting to the outpatient clinic of our university department of dermatology in 2014 were included. Inclusion criterion was a visible skin finding whereas exclusion criteria were previously diagnosed skin diseases. The study was approved by the local ethics committee and written informed consent was obtained. Patients were clinically examined by five resident dermatologists (“clinicians”), and teledermatology was performed by one experienced dermatologist (a “teledermatologist”). Two photographs (overview and close-up) were taken using an Apple iPhone 4s  and sent to the teledermatologist with standardized anamnesis using the app and online portal Goderma (now KLARA, Berlin, Germany). The same system was used to document the clinicians’ and teledermatologist's assessment: specification of the skin disease (mole, eruption, and other), diagnosis, differential diagnoses, therapeutic recommendations, and an evaluation if a physical consultation was necessary (table 1). In 36 cases, punch biopsies were taken for further histological evaluation.
Mean age of the study population (40.5% female) was 50.6 years (range: 1-89 years). Clinicians’ most common diagnoses were eczema (16.6%), non-melanoma skin cancer including precursors (15.4%), infectious diseases (12.4%), acne (7.1%), and different benign lesions (6.5%). The primary diagnosis and differential diagnoses made by the clinicians and teledermatologist were in accordance in 58.9% and 32.1% cases, respectively. In cases with discordant primary diagnoses, the differential diagnoses were concordant in 25%, leading to a summary of concordance of primary and/or differential diagnoses in 83.9% of all cases. Further diagnostic procedures were requested by the clinicians in 65.7% and by the teledermatologist in 57.4% of all cases. Compared to histology, when available, the clinicians’ and teledermatological diagnoses were correct in 72.2% and 55.6% of cases, respectively. Therapeutic or diagnostic recommendations were given by clinicians in 86.4% and by the teledermatologist in 89.3%, with a total concordance in 60.4%, and a total discordance in 17.2% of all cases. Clinicians performed a necessary standard physical consultation in 57.4%, and the teledermatologist in 44.4% of cases. Among the cases in which physical presentation was recommended, the clinicians and teledermatologist favoured an immediate consultation in 46.4% and 68.0%, respectively.
The concordance rates between the clinicians and teledermatologist observed in this study are consistent with previous studies [5-9]. Compared to histopathology, the teledermatologist's diagnoses were less accurate than the clinicians, which suggests that malignant lesions might be a challenge for teledermatology. The risk of overlooking malignancies has often been raised as a concern in teledermatology  and a policy of referring any suspicion of malignancy in teledermatology to conventional dermatologists might be essential. The high rate of assessment by the clinicians and teledermatologist, that physical presentation is necessary in 57.4% and 44.4% of all cases, respectively, further underlines that the option of referral for teledermatology patients to dermatologists is crucial. Concerning treatment, the findings of our study are in accordance with previous studies . A standardized history using the app does not replace the medical history taken based on a personal conversation between a doctor and a patient, which can be crucial for numerous dermatological diseases. Experienced dermatologists correctly diagnose diseases in up to 77% of cases by visual inspection alone, without any further information, but their knowledge of the patient's medical history substantially improves accuracy , which appears to be an important limitation of teledermatology in general. In conclusion, the results of this study indicate that teledermatological consultations could serve as a screening tool for conventional physical consultation, and some skin findings might be assessed and treated only remotely. With regards to remote areas with limited access to dermatologists in particular, teledermatology has the potential to dramatically improve healthcare, although it does not, and cannot, replace conventional dermatology.
Acknowledgments: The authors thank all patients for participating in the study. Financial support: the study was financially supported by KLARA (Goderma) GmbH. Conflict of interest: Bernadette Eberlein received funding from KLARA (Goderma) GmbH. All other authors have nothing to declare.