From the 2023 HVPA National Conference
Stephen Merry MD MPH (Mayo Clinic), Brian McCormick MD (Hampton Family Practice), Kiran Chatha MD, MPH, Vivien Nguyen PharmD, David Leffell MD, Ivana Croghan PhD
The risk for developing skin cancer increases with age, with most cases found in people 65 years of age and older. As older adults are living longer, the need for public health efforts to promote life-long skin health is more critical than ever. Leveraging technology may aid physicians in improving early detection, potentially reducing associated mortality and morbidity. Elastic scattering spectroscopy (ESS) technology utilizes visual light and detects the backscatter of the light by subcellular structures to differentiate between benign and malignant cells. A handheld ESS device has been developed to aid primary care physicians (PCP) in their clinical assessment of concerning skin lesions.
This study aimed to assess the sensitivity and specificity of a handheld ESS device in evaluating skin lesions suspicious for skin cancer while controlling for the confounder of age. To address age, we compared patients eligible for Medicare to those patients in the younger population.
A blinded, prospective, multi-center clinical validation study was performed with 1,579 skin lesions from 1,005 patients, first clinically assessed by PCP investigators and then scanned with the ESS device. PCPs recorded their suspected clinical diagnosis and patient and lesion characteristics. All enrolled lesions were biopsied per standard of care and gold standard dermatopathology results were used as the reference standard for assessing device performance. Primary analysis results were previously reported. In this sub-analysis, patients were stratified and compared by age groups for those eligible for Medicare and those not yet eligible: 378 patients over 65 years (Medicare group [604 lesions]) and 627 patients under 65 years (Non-Medicare group [975 lesions]).
There was an even distribution of males and females in both groups, with the Non-Medicare and Medicare groups having an average age of 49.6 years and 73.3 years, respectively. Similar numbers of skin cancer-related risk factors were noted in both groups. Overall device sensitivity was 95.5% and specificity was 20.7%. For comparison, PCPs had an overall sensitivity of 83.0% for clinically predicting cancer; use of the device could have improved PCP sensitivity from 83.0% to 95.5% (p<0.0001). Device sensitivity was similar in both groups, at 96.9% in the Medicare group and 93.5% in the Non-Medicare group. Meanwhile, PCP sensitivity for the Medicare and Non-Medicare groups were 87.8% and 76.3%, respectively. Device specificity was lower in the Medicare group at 15.2% compared to 23.7% in the Non-Medicare group for PCP-biopsied lesions. The positive predictive value (PPV) of the device was higher for the Medicare group at 24.1% compared to 11.4% for the Non-Medicare group, while negative predictive value (NPV) was similar in the two groups (94.7% for Medicare- and 97.2% for Non-Medicare-eligible patients). Overall diagnostic performance of the device, as measured by area under the curve was similar at 0.774 for the Medicare group and 0.755 in the Non-Medicare group.
The ESS device demonstrated high sensitivity in detecting skin cancer when compared to gold standard dermatopathology. These findings suggest use of the ESS device has the potential to improve PCP skin cancer detection in both the Medicare-eligible and younger population with minimal differences in device effectiveness. The lower specificity in the Medicare group may be due to increased underlying age-related cutaneous changes.