BACKGROUND: Keratinocyte or non-melanoma skin cancer (NMSC) is the commonest malignancy worldwide. Usual treatment is surgical excision. Current guidelines underestimate incomplete excision rates. OBJECTIVES: We aimed to determine the risk of incomplete excision of NMSCs through a systematic review and meta-analysis of primary clinical studies. METHODS: A PRISMA-compliant systematic review and meta-analysis was performed using methodology proposed by Cochrane. A comprehensive search strategy was applied to MEDLINE, Embase, Scopus, CINAHL, EMCare, Cochrane Library and the grey literature (January 2000-27th November 2019). All studies were included except studies on Mohs micrographic surgery, frozen section or biopsies. Abstract screening and data extraction were performed in duplicate. The risk of bias was assessed using a tool for prevalence/incidence studies. The primary outcome was the proportion of incomplete surgical excisions. A random effects model for pooling of binominal data was used. Differences between proportions were assessed by sub-group meta-analysis and meta regression which were presented as risk ratios. PROSPERO CRD42019157936. RESULTS: Searching identified 3477 records, with 110 studies included, comprising 53 796 patients with 106 832 basal cell carcinomas (BCC) and 21 569 squamous cell carcinomas (SCC). The proportion of incomplete excisions for BCC was 11·0% (95% CI 9·7-12·4%) and for SCC 9·4% (95% CI 7·6-11·4%). Incomplete excisions by specialty were: dermatology 6·2% BCCs, 4·7% SCCs; plastic surgery 9·4% BCCs, 8·2% SCCs; general practitioners 20·4% BCCs, 19·9% SCCs. The risk of incomplete excision for general practitioners was four times that of dermatologists for both BCC (RR 3·9 [95% CI 2·0-7·3]) and SCC (RR 4·8 [95% CI 1·0-22·8]). Studies were heterogenous (I2 =98%) and at high risk of bias. CONCLUSIONS: The proportion of incomplete excisions is higher than previously reported. Excisions performed by specialists may lower the risk of incomplete excision.
Br j dermatol