Predicting alternate light absorption in areas of trauma based on degree of skin pigmentation: Not all wavelengths are equal

Abstract

Detection and documentation of bruises on survivors of intimate partner violence (IPV) can provide valuable evidence to support investigation and subsequent prosecution of these crimes. Detection of bruises in persons with darker skin tones is often difficult, contributing to disparities in health and criminal justice responses to IPV. The purpose of this secondary data analysis was to determine whether use of an alternate light source (ALS) increased the predictive probability of successfully detecting bruises on diverse skin tones following a history of physical trauma. In this study, data were analyzed from a convenience sample of 157 subjects inclusive of six skin tones (very light, light, intermediate, tan, brown, dark) with induced bruises. Bruises were assessed under white light and an ALS 21 times over four weeks using 10 different ALS wavelength and goggle color combinations. Data analyzed included 31,841 skin observations obtained over 2897 participant assessments. Multilevel modeling was used to account for the correlation among the repeated measurements for each bruise. Across all categories of skin pigmentation, ALS wavelengths 415 nm and 450 nm viewed through a yellow filter had the most frequent detections of bruises (415 nm: n = 2777, 11.2%; 450 nm: n = 2747, 11.1%) and greater predictive probability of a positive finding (415 nm: 0.90-0.99; 450 nm: 0.85-0.99) than white light (n = 2487; 10%; 0.81-0.90). These two ALS wavelengths were the only combinations that provided greater probability of detection than white light on groups with darker skin (brown or dark), whereas additional ALS wavelengths/filters worked equally well on groups with lighter skin. Findings suggest use of an ALS in clinical assessments of patients of color who report IPV may help reduce health and criminal justice-related disparities.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 13 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.