Variation in initial biopsy technique for primary melanoma diagnosis: A population-based cohort study in New South Wales, Australia.

Background: Factors associated with nonadherence to guideline-recommended complete excision of suspicious cutaneous lesions are unclear.

Objective: The purpose of this study was to analyze patient, melanoma, and clinician factors associated with initial diagnostic biopsy type and determine whether unwarranted variation from guidelines occurred.

Methods: This population-based, cohort study involved the analysis of data from questionnaires completed by clinicians who managed patients with newly diagnosed, histopathologically confirmed primary invasive cutaneous melanomas reported to the New South Wales Cancer Registry between 2006 and 2007.

Results: Of the 2267 biopsies, complete excision was attempted in 69.1% of cases but histologically incomplete in 14.0%. Multivariable regression analyses showed that complete excision was more likely than incision biopsy in patients <70 years ( = .016), shave biopsy in patients <80 years ( = .034), shave biopsy in melanomas of Breslow thickness 0.8-1.0 mm or 2.1-4.0 mm ( = .039) than either punch ( < .001) or shave biopsy ( < .003) in melanomas on trunk or limbs, and punch biopsy when treated by a surgeon ( < .001). Complete excision was less likely than punch biopsies in women ( < .003), with lentigo maligna melanoma or unknown histopathology ( = .004); shave biopsy in patients with lentigo maligna melanoma, or other melanoma subtype ( = .003); punch, shave, or incision biopsy when treated by a dermatologist ( < .001).

Limitations: Generalizability of these findings may be limited to the time of data collection.

Conclusions: Guideline adherence for biopsy type undertaken for clinically suspected melanoma appeared to be suboptimal.