A systematic review and meta-analysis of the diagnostic accuracy of fine-needle aspiration cytology for parotid gland lesions.

The clinical usefulness of fine-needle aspiration cytology (FNAC) for the diagnosis of parotid gland lesions is controversial. Many accuracy studies have been published, but the literature has not been adequately summarized. We identified 64 studies on the diagnosis of malignancy (6,169 cases) and 7 studies on the diagnosis of neoplasia (795 cases). The diagnosis of neoplasia (area under the summary receiver operating characteristic [AUSROC] curve, 0.99; 95% confidence interval [CI], 0.97-1.00) had higher accuracy than the diagnosis of malignancy (AUSROC, 0.96; 95% CI, 0.94-0.97). Several sources of bias were identified that could affect study estimates. Studies on the diagnosis of malignancy showed significant heterogeneity (P < .001). The subgroups of American, French, and Turkish studies showed greater homogeneity, but the accuracy of these subgroups was not significantly different from that of the remaining subgroup. It is not possible to provide a general guideline on the clinical usefulness of FNAC for parotid gland lesions owing to the variability in study results. There is a need to improve the quality of reporting and to improve study designs to remove or assess the impact of bias.