In the Journals

Study shows population estimates of melanocytic lesions

A diagnosis of melanocytic proliferation was reported in almost 25% of skin biopsies, according to results of a large population-based study.

Researchers conducted a natural language processing-based analysis of electronic medical records for patients at a single health care system in the United States between Jan. 1, 2007 and Dec. 31, 2012. They aimed to determine frequency and distribution information for histologically confirmed lesions among 80,368 biopsies performed on 47,529 patients.

Results showed 23% of skin biopsies were performed in melanocytic lesions. Diagnostic category distribution results showed 83.1% of these were class I lesions, 8.3% were class II, 4.5% were class III, 2.2% were class IV and 1.9% were class IV.

“These data provide the first population-based estimates across the spectrum of melanocytic lesions ranging from benign through dysplastic to malignant,” the researchers concluded. “These results may serve as a foundation for future research seeking to understand the epidemiology of melanocytic proliferations and optimization of skin biopsy utilization.”

Katy J. L. Bell, PhD, of the Sydney School of Public Health, The University of Sydney, NSW, Australia, and colleagues wrote an editorial accompanying the paper.

“Lott et al demonstrated how technological advances in text analytical tools provide opportunities to harvest important epidemiological, clinical and pathologic data on a range of different medical conditions by analyzing pathology reports,” they wrote. “Automated processes, such as the natural language processing tool used by Lott et al, enable a much larger number of cases to be analyzed and assigned with high accuracy. The generated data are important for demonstrating the relative prevalence and incidence of different medical conditions.”

Bell and colleagues added that findings from the current study allowed researchers to re-weight biopsy classifications.

“When the reproducibility study data were considered alone, ‘undercalling’ of melanoma histopathology appeared to be much more of a problem than ‘overcalling,’” they wrote.

They added, “With a better understanding of the epidemiology of melanocytic pathology, we may formulate policy and practice interventions to optimize both the use of skin biopsies and the interpretation of their histopathology. In this way, we may maximize benefits by the early detection and treatment of melanomas that would otherwise advance while minimizing harms from the overdiagnosis of melanomas that would not.” – by Rob Volansky

Disclosures : Lott reports being an employee of Bayer US LLC, which had no involvement in this research. Bell and Scolyer report no relevant financial disclosures.

A diagnosis of melanocytic proliferation was reported in almost 25% of skin biopsies, according to results of a large population-based study.

Researchers conducted a natural language processing-based analysis of electronic medical records for patients at a single health care system in the United States between Jan. 1, 2007 and Dec. 31, 2012. They aimed to determine frequency and distribution information for histologically confirmed lesions among 80,368 biopsies performed on 47,529 patients.

Results showed 23% of skin biopsies were performed in melanocytic lesions. Diagnostic category distribution results showed 83.1% of these were class I lesions, 8.3% were class II, 4.5% were class III, 2.2% were class IV and 1.9% were class IV.

“These data provide the first population-based estimates across the spectrum of melanocytic lesions ranging from benign through dysplastic to malignant,” the researchers concluded. “These results may serve as a foundation for future research seeking to understand the epidemiology of melanocytic proliferations and optimization of skin biopsy utilization.”

Katy J. L. Bell, PhD, of the Sydney School of Public Health, The University of Sydney, NSW, Australia, and colleagues wrote an editorial accompanying the paper.

“Lott et al demonstrated how technological advances in text analytical tools provide opportunities to harvest important epidemiological, clinical and pathologic data on a range of different medical conditions by analyzing pathology reports,” they wrote. “Automated processes, such as the natural language processing tool used by Lott et al, enable a much larger number of cases to be analyzed and assigned with high accuracy. The generated data are important for demonstrating the relative prevalence and incidence of different medical conditions.”

Bell and colleagues added that findings from the current study allowed researchers to re-weight biopsy classifications.

“When the reproducibility study data were considered alone, ‘undercalling’ of melanoma histopathology appeared to be much more of a problem than ‘overcalling,’” they wrote.

They added, “With a better understanding of the epidemiology of melanocytic pathology, we may formulate policy and practice interventions to optimize both the use of skin biopsies and the interpretation of their histopathology. In this way, we may maximize benefits by the early detection and treatment of melanomas that would otherwise advance while minimizing harms from the overdiagnosis of melanomas that would not.” – by Rob Volansky

Disclosures : Lott reports being an employee of Bayer US LLC, which had no involvement in this research. Bell and Scolyer report no relevant financial disclosures.