Nault A, Zhang C, Kim K, Saha S, Bennett DD, Xu YG. Biopsy Use in Skin Cancer DiagnosisComparing Dermatology Physicians and Advanced Practice Professionals. JAMA Dermatol. 2015;151(8):899-902. doi:10.1001/jamadermatol.2015.0173
Histopathologic evaluation is the criterion standard for diagnosis of skin cancer. Underuse of biopsies may promote misdiagnosis, and overuse will increase cost and morbidity. There is no benchmark with which to quantitatively compare health care professionals’ diagnostic accuracy and biopsy use. Prior studies suggest wide variability in biopsy use among practice settings and health care professionals.1- 5 We conducted a retrospective review on the number of skin biopsies needed per malignant neoplasm in our department. The recent article by Coldiron and Ratnarathorn6 documents that, in 2012, mid-level health care professionals independently billed approximately 2.6 million dermatologic procedures, most of which required clinical distinction between benign and malignant lesions. To our knowledge, our study is the first to compare the number needed to biopsy (NNB) per malignant neoplasm between dermatology physicians and advanced practice professionals (APPs).
We performed a retrospective study of all biopsies submitted to our laboratory by 13 dermatology physicians (5 men and 8 women) and 5 APPs (1 physician assistant and 4 nurse practitioners, all women) between January 1 and February 15, 2010. The study was approved by the University of Wisconsin Institutional Review Board. We reviewed requisition forms and clinical notes, and only biopsy procedures that were performed with the intention to exclude skin cancer were included. Biopsies performed for inflammatory conditions, removal of lesions for cosmetic or functional reasons, and re-excisions, as well as biopsies with insufficient documentation, were excluded. Patient demographics and information on lesion appearance, type of health care professional, and final diagnoses were collected. The NNB for nonmelanoma skin cancer (NMSC) was calculated by dividing the total number of nonpigmented lesions by the number of histologically proven NMSCs. The NNB for melanoma was calculated by dividing the total number of pigmented lesions by the number of histologically proven melanomas. Uncommonly, NMSCs were diagnosed from the pigmented lesions, while melanomas were diagnosed from nonpigmented lesions. The NNB for any skin cancer was calculated by dividing the total number of all lesions by the total number of histologically proven malignant lesions.
A total of 1102 biopsy procedures from 743 patients met the inclusion criteria (Table 1). Nonpigmented lesions made up 55.4% of the total biopsies and pigmented lesions, 44.6% of all biopsies; 26.9% were histologically diagnosed as NMSC and 2.2% as melanoma. The NNB for any skin cancer, NMSC, and melanoma was 3.4, 2.1, and 21.4, respectively. There was a significant difference in NNB between physicians and APPs for any skin cancer (2.9 vs 5.9, P < .001), NMSC (1.9 vs 3.1, P < .001), and melanoma (17.4 vs 32.8, P = .04). The NNB was significantly lower for patients older than 65 years, male patients, and patients with a history of NMSC prior to biopsy. When patients were stratified by age, sex, and history of skin cancer, APPs performed significantly more biopsy procedures than did physicians to diagnose a malignant neoplasm in patients younger than 65 years and in patients without a history of skin cancer (Table 2).
Our NNB differed from those in previous studies.1- 5 We made an effort to recapitulate practice by including only biopsies for which physicians and APPs had intention to exclude malignant neoplasm. Wilson et al5 performed a similar study; their NNB for any cancer, NMSC, and melanoma was 2.2, 1.6, and 15, respectively. Variability between studies may result from differences in patient population, geographic location, and practice settings.
To our knowledge, our study is the first to compare NNB between types of health care professionals. At our institution, APPs see new and established patients, most of whom are not evaluated by a physician; however, a physician is available in the clinic. The mean length of practice for our physicians was 11.9 years (range, 0.5-25.5 years) compared with 6.8 years (range, 0.5-20 years) for APPs. In our study, the NNB of any skin cancer for APPs was double that of physicians, and that difference is most pronounced in younger patients and those without a history of skin cancer. The use of dermoscopy was not consistently recorded in the clinical notes, but we speculate that dermatologists’ training in dermoscopy may play a role in their lower NNB. Future studies may evaluate the use of dermoscopy in relation to NNB and type of health care professional.
At the time of this publication, we are reviewing additional cases submitted by dermatologists, other specialty physicians, and APPs from multiple departments to provide a more robust analysis of NNB by type of health care professional and specialty. Similar studies performed elsewhere may allow comparison of NNB across broader practice settings and may provide benchmarks with which physicians can compare their own practices. Most important, our findings suggest that increased use of biopsies may increase the morbidity and cost of care provided by APPs when compared with that provided by dermatologists.
Accepted for Publication: January 23, 2015.
Corresponding Author: Yaohui G. Xu, MD, PhD, Department of Dermatology, University of Wisconsin School of Medicine and Public Health, 1 S Park St, Seventh Floor, Madison, WI 53715 (email@example.com).
Published Online: March 25, 2015. doi:10.1001/jamadermatol.2015.0173.
Author Contributions: Drs Bennett and Xu had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs Bennett and Xu contributed equally to this article.
Study concept and design: Saha, Bennett, Xu.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Nault, Saha, Bennett, Xu.
Critical revision of the manuscript for important intellectual content: Zhang, Kim, Saha, Bennett, Xu.
Statistical analysis: Zhang, Kim, Saha.
Obtained funding: Xu.
Administrative, technical, or material support: Saha, Bennett, Xu.
Study supervision: Bennett, Xu.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported in part by the Department of Dermatology, University of Wisconsin School of Medicine and Public Health, and Shapiro Research Program at University of Wisconsin School of Medicine and Public Health.
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: Justin Endo, MD, and Joanna McGetrick, MD, Department of Dermatology, University of Wisconsin School of Medicine and Public Health, assisted with the study design; Joshua Tarpley, BA, University of Wisconsin School of Medicine and Public Health, Farinoosh Dadrass, BS, Department of Dermatology, University of Wisconsin School of Medicine and Public Health, Lauren Van Loon, BS, Department of Dermatology, University of Wisconsin School of Medicine and Public Health, and Loren Krueger, BS, University of Wisconsin School of Medicine and Public Health, assisted with data entry; and Diane Bock, BS, Department of Dermatology, University of Wisconsin School of Medicine and Public Health, Tisha Kawahara, MS, CRCC, Department of Biostatistics and Medical Informatics, University of Wisconsin–Madison, and Sandra Olson, MS, Department of Dermatology, University of Wisconsin School of Medicine and Public Health, provided administrative assistance. None of the contributors were compensated for their assistance.