Abstract
This analysis of survey results examines the accuracy with which medical students, internal medicine residents, and dermatology residents, fellows, and faculty can estimate the diameter of skin lesions in clinical photographs.
Assessing skin lesion diameter is a key component of the ABCDE (Asymmetrical shape, Border, Color, Diameter, and Evolution) criteria to evaluate lesions concerning for melanoma. To our knowledge, the ability of medical trainees and practitioners to estimate lesion size with and without the aid of a ruler has not been previously described. This study seeks to determine the accuracy with which medical students, internal medicine residents, and dermatology residents, fellows, and faculty can estimate the diameter of skin lesions in clinical photographs.
Methods
Study participants included medical students; internal medicine residents; and dermatology residents, fellows, and faculty at Northwestern University Feinberg School of Medicine. Survey response implied informed consent for inclusion in the study, and study approval was obtained from the Northwestern University institutional review board. Survey data was collected anonymously using online Research Electronic Data Capture surveys (Vanderbilt) and paper surveys from August 4 to August 27, 2017. Participants responded to a survey on their use of rulers in the clinic and their confidence in size estimation of skin lesions, and then selected the diameter of 12 lesions in clinical photographs as follows: 4 lesions that included a ruler in the image, 4 lesions that included an anatomic landmark, and 4 lesions that included neither ruler nor landmark (Figure).
Figure. Examples of Survey Images.
Clinical photographs show (A) a skin lesion with a ruler included in the image, (B) a skin lesion with an anatomic landmark included in the image, and (C) a skin lesion with neither a rule nor anatomic landmark included. Images used with permission from VisualDx.
All statistical analyses were performed in R 3.4.1 (R Foundation). The Kruskal-Wallis test was performed to assess group differences between medical students, internal medicine residents, and dermatology residents, fellows, and faculty for 3 sets of images with post hoc pairwise comparisons conducted with the Dunn test for multiple comparisons (Table). Post hoc analysis used the Bonferroni correction and set significance at P = .05.
Table. Characteristics of Medical Students, Internal Medicine Residents, and Dermatology Residents, Fellows, and Faculty Participants.
| Characteristic | Medical Students (n = 39) |
Internal Medicine Residents (n = 27) |
Dermatology Residents, Fellows, and Faculty (n = 12) |
|---|---|---|---|
| Carry ruler, No. (%) | 17 (44) | 6 (22) | 11 (92) |
| Measure diameter with ruler in clinic, % mean | 33.3 | 29.6 | 70.0 |
| Confidence in lesion diameter estimate in clinic if not measured with ruler on a Likert scale of 1 to 5, mean (SD) | 2.7 (1.2) | 3.1 (0.9) | 3.9 (0.7) |
| Assessment of clinical diameter of pigmented lesions, % correct answersa | |||
| Ruler included in image (n = 4) | 67 | 73 | 71 |
| Anatomic landmark in image (n = 4) | 46 | 51 | 48 |
| No ruler, no anatomic landmark (n = 4) | 23b | 19b | 52b |
Summary of diameter assessment is as follows: Kruskal-Wallis test was performed with post-hoc pairwise comparisons conducted with the Dunn test for multiple comparisons.
Indicates a statistically significant comparison.
Results
The dermatology cohort was more likely to carry rulers and measure suspicious lesions in clinic compared with the internal medicine and medical student cohorts (Table). Initial confidence in estimating lesion diameter prior to the test was similar among the 3 groups with mean Likert scale values of 2.7, 3.1, and 3.9 for the medical student, internal medicine resident, and dermatology participants, respectively.
For images that included a ruler, 39 medical students estimated the diameter of 107 of 156 lesions (69%) correctly within 1 mm; 27 internal medicine residents 79 of 108 lesions (73%); and 12 dermatology participants, 34 of 48 lesions (71%). There was no statistically significant difference among groups (P = .73). For images with an anatomic landmark, medical students scored 71 of 156 lesions (46%) correctly within 1 mm; internal medicine, 55 of 108 lesions (51%); and dermatology, 23 of 48 lesions (48%); there was no significant difference among groups (P = .82). For images with no ruler and no anatomic landmark, medical students scored 36 of 156 lesions (23%) correctly; internal medicine 20 of 108 lesions (19%); and dermatology, 25 of 48 lesions (52%), and there was statistical significance between groups (P = .003) (Table). Pairwise post hoc analysis shows a significant difference between the dermatology cohort and internal medicine cohort (P = .003) and between dermatology and student cohorts (P = .02). After receiving the correct responses, participants indicated that they were more likely to measure suspicious lesions with a ruler, with mean (SD) Likert scale score of 3.8 (0.9) (P = .04).
Discussion
All participants in this study were most accurate in assessing diameter in images with a ruler, and most participants were accurate estimating the diameter in images with an anatomic landmark and no ruler. The dermatology cohort was most accurate in assessing diameter in images with no landmark and no ruler but only estimated 52% of lesions correctly within 1 mm. Prior studies have investigated the ability of general practitioners and dermatologists to assess and triage pigmented lesions using photographs that did not include a ruler, but that we know of, this is the first to evaluate estimation accuracy of providers for skin lesions.
The ABCDEs have become a well-established grading system for the evaluation of suspicious skin lesions. The diameter criterion of 6 mm for pigmented lesions is a clinically useful guideline to assess for malignant potential, and assessing diameter is crucial for both dermatologists and nondermatologists to appropriately triage lesions. Furthermore, accurate size estimation also underpins the clinician’s ability to accurately assess lesion evolution. This study shows that an accurate clinical assessment of lesion size requires the use of a ruler for all physicians, including dermatologists, and that by simply increasing physician awareness of this potential pitfall is sufficient to compel change.
References
- 1.Abbasi NR, Yancovitz M, Gutkowicz-Krusin D, et al. Utility of lesion diameter in the clinical diagnosis of cutaneous melanoma. Arch Dermatol. 2008;144(4):469-474. [DOI] [PubMed] [Google Scholar]
- 2.Brochez L, Verhaeghe E, Bleyen L, Naeyaert JM. Diagnostic ability of general practitioners and dermatologists in discriminating pigmented skin lesions. J Am Acad Dermatol. 2001;44(6):979-986. [DOI] [PubMed] [Google Scholar]
- 3.Chen SC, Bravata DM, Weil E, Olkin I. A comparison of dermatologists’ and primary care physicians’ accuracy in diagnosing melanoma: a systematic review. Arch Dermatol. 2001;137(12):1627-1634. [DOI] [PubMed] [Google Scholar]
- 4.Chen SC, Pennie ML, Kolm P, et al. Diagnosing and managing cutaneous pigmented lesions: primary care physicians versus dermatologists. J Gen Intern Med. 2006;21(7):678-682. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Goodson AG, Grossman D. Strategies for early melanoma detection: approaches to the patient with nevi. J Am Acad Dermatol. 2009;60(5):719-735. [DOI] [PMC free article] [PubMed] [Google Scholar]
