Investigating the evidence for effective digital skin surveillance methods

Hand dermatitis is a common occupational skin disease and is particularly prevalent among healthcare workers.

To identify which digital skin surveillance methods are effective for detecting hand dermatitis.

A scoping review. Protocol, including search strategy, was developed using the PICO framework and the PRISMA-Scoping Review guide.

Three databases were searched with the following outputs: MEDLINE (n = 150), CINAHL (n = 568), EMBASE (n = 30) along with Cochrane Work (n = 0) and the grey literature. Twenty-eight duplicates were removed. Following screening title, abstract and full-text stages, five articles were included. Across the included studies, different digital skin surveillance tools were developed and tested to screen for hand dermatitis, producing mixed findings. Methods included rapid self-assessment questionnaires (often with photo grading classification of dermatitis to aid assessments) compared with assessments by clinicians using hand photographs. One study demonstrated that a screening method had a high negative predictive value (91%; 95% CI 89-93) but a low positive predictive value (39%; 95% CI 34-45). While able to demonstrate acceptable discriminatory accuracy between people with and without the disease and having a high specificity (86%; 95% CI 84-88), sensitivity was modest at 52% (95% CI 46-59), meaning true cases were not identified. In another study, validity testing results of self-assessment screening compared with screening by a dermatologist using photos showed only fair agreement between the two methods, highlighting variability.

While the findings are promising, there is a compelling case to test such methods among different at-risk workers within a workplace skin surveillance programme.