Commentary on: Downing NR, Scafide KN, Ali Z, Hayat MJ. Visibility of inflicted bruises by alternate light: Results of a randomized controlled trial. J Forensic Sci. 2024;69(3):880–7. https://doi.org/10.1111/1556-4029.15481

Abstract

We read with interest the article by Downing NR, et al. regarding the use of alternate light to view bruises [1]. We agree with the authors that the identification of injuries in patients who have experienced physical violence is important. As colleagues who are forensic medicine experts, we recognize the importance of visualizing bruises in cases of interpersonal violence, especially in individuals with darkly pigmented skin, so we applaud your research on this issue. Practitioners need to be as accurate as possible in their evaluation. The use of tools to assist in the identification of injuries can be very helpful. We also recognize that forensic sciences and the associated analysis or evaluation must remain accurate and valid and not utilize “presumptive tests without confirmation” or tests lacking “reliability, reproducibility, repeatability, and replicability” [2, 3].

When reading this article, we were struck by the remarkable similarity in methods with previous publications by these authors on this topic [4, 5]. All three studies have the same number of 157 participants. This study and the study published in 2020 describe the same methodology of “quota sampling” to “recruit equal numbers of healthy participants.” The study published in 2022, however, stated that “a convenience sample of 157 subjects” was obtained. If these studies are of the same participants, then the statement in the 2022 publication about how the participants were recruited appears to be incorrect and potentially misleading. We are concerned that multiple studies analyzing the same data set create an improper opportunity for bias and post hoc analysis error.

The authors found in a previous analysis of the participants that both the 415 nm ALS with a yellow filter and the 450 nm ALS with a yellow filter to be “optimal for detecting evidence of bruising on individuals with known trauma” [5]. This was also apparent from table 5 of their 2020 study with the same data [4]. With this current publication, the authors state “only 415nm viewed through a yellow filter resulted in a clinically meaningful improvement (0.46 points) in visibility rating when compared to white light.” [1] In figure 1, of this study, we find it difficult to visually discern any difference between the various light frequencies used in very light skin or very dark skin. Additionally, we noted that all modalities of light had reduced visualization of bruises; no modality was clearly better in most or all skin color categories.

Reviewing table 2 in this recent paper shows that the 450 nm ALS with the yellow filter did not have as high a visibility rating improvement, but it was the only other assessment with a visibility rating improvement above 0.2. How should the reader discern what a clinically meaningful improvement in the visibility score would be?

Based upon the methods described by the authors, they constructed a novel bruise visibility score and absorption visibility score by using a visual analog scale ranging from 1 to 5 with defined labels at 1, 3, and 5. The authors provide no discussion in their article about how they determined the reliability, validity, or accuracy of this tool. In their methods, they do not state what amount of scale improvement they believed would constitute a clinically meaningful improvement prior to their analysis. How was the value of 0.50 chosen by the authors? The reader is unable to tell based on their publication. Using this value, the authors appear to have concluded that a 12.5% improvement in the score was a “clinically meaningful improvement.” Their use of a change in the score of 0.50 as clinically meaningful within a scale starting at 1 and ending at 5 represents a 12.5% change as 0.5 is one-eighth of the length of the scale. Based upon the actual value calculated by the authors (0.46), this appears to be an actual 11.5% change, and by the authors own statement, this result only “approaches the clinically meaningful increase” but did not equal or exceed it. Despite this prior statement in their own article, the authors concluded that observation under ALS resulted in a clinically meaningful improvement; we believe this conclusion to be incorrect, and not supported by their own data.

Because no other authors or studies appear to have used a “bruise visibility score” nor an “absorption visibility score”; for comparison, we suggest using an alternate scale by analogy to interpret what a clinically meaningful improvement would be. Visual analog pain scores and scales have a robust research history in the medical literature [6]. For a visual analog pain scale, the minimum clinically important difference published by various authors ranges from around 10–20 mm [7, 8]. These values represent the minimum value that should be considered a clinically significant difference. In this study where a known injury was present in the location being viewed, the authors have barely achieved a value within that range. We are concerned that in the real world, where the patient is not known to have an injury, this observed ideal difference in score improvement will be too low to be clinically useful.

In the conclusions, the authors state that the increased visibility of bruising “could corroborate violence and reduce disparities for persons with dark skin colors.” Table 2, however, appears to not show an improvement in the bruise visibility in dark skin tones compared with very light skin tones. The data show that dark skin tones had a bruise visibility of 0.34, a value both significantly lower than the 0.71 found in very light skin tones and lower than the 0.50 value that the authors indicated was a significant difference in visualization. In figure 1, it is difficult to visually see a difference between the light frequencies used in very light skin to dark skin to discern the presence of bruises, since all modalities of observation had reduced visualization of bruises.

In the implications, the authors encourage clinicians to “follow guidelines” from the AtLAST publication [9]. We note that these guidelines acknowledge the limitations of ALS. “Whether positive or negative for absorption, ALS findings are not diagnostic for the presence or absence of injury.” Following these recommended guidelines leads a reader to question if ALS has any utility in evaluating patients with bruising, particularly if the findings under ALS are not diagnostic of injury.

While not explicitly stated in their paper, we have seen ALS promulgated as a tool to detect bruises not seen under normal white light during examination. The implication of this use is clear in the authors' writing where they say “ALS should be used in conjunction with patients' report of known history of trauma”; if the patient states an injury occurred and the ALS shows absorption, then the clinician can conclude that the absorption is a bruise. The use of ALS to detect unseen or invisible bruises has been declared as a true statement in both court by forensic nurses [10], and in the media by advocates, notably at the research institution of some of the authors [11]. The AtLast guidelines are silent on whether ALS can be used to detect occult bruising (i.e., bruising not seen in white light), and the latest guidelines from the International Association of Forensic Nurses stated that ALS may “help enhance bruises that can be seen under white light” [12]. We firmly believe that ALS is not a tool that can reliably and accurately detect bruises which are not seen in white light. We strongly advocate against the use of ALS as a tool to detect clinically occult bruises.

We suggest the reader use extreme caution when applying these research findings to patients being seen in a clinical forensic practice, especially when interpretation of an injury's presence or absence is of great legal importance. The most optimistic interpretation of this study would be that ALS may minimally enhance the visibility of bruises. Perhaps this small improvement in visualization will be validated in a confirmation study using a control group which does not have blunt injuries or in a confirmation study where blunt injuries can be validated by different diagnostic tool. Until then we believe these authors have provided confusing data that suggest an examiner may be able to see bruises better, in a minimally clinically significant way, under ALS compared with white light in an ideal environment where injury is known to be present in the skin. We do not believe this indicates that ALS should be used routinely for bruise assessment in forensic patients, whether they have light or dark skin, nor that ALS should be used to find bruises that are not visible under normal white light observation.