Purpose: This study was designed to evaluate the diagnostic accuracy of computed tomography (CT)-guided biopsy for small lung nodules (SLNs) (≤20 mm) and to assess related complication rates.
Methods: We reviewed the Pubmed, Embase, and Cochrane Library databases to identify all relevant studies published as of April 2020. Random effects modeling were then used to evaluate pooled data pertaining to technical success rates, diagnostic accuracy, pneumothorax rates, and rates of hemoptysis. The meta-analysis was conducted using Stata v12.0.
Results: In total, we identified 25 relevant studies for incorporation into this meta-analysis, incorporating 2922 total CT-guided lung biopsy. Pooled technical success rates, diagnostic accuracy, pneumothorax rates, and hemoptysis rates were 94% (95% confidential interval [CI], 0.91-0.98), 90% (95% CI, 0.88-0.93), 19% (95% CI:, 0.15-0.24), and 12% (95% CI, 0.08-0.15), respectively. We observed significant heterogeneity among these studies for all 4 of these parameters (I = 90.0%, 82.7%, 88.6%, and 88.4%, respectively). When we conducted a meta-regression analysis, we did not identify any variables that influenced diagnostic accuracy or technical success, pneumothorax, or hemoptysis rates. Publication bias risk analyses suggested that there was relatively little risk of publication bias pertaining to pneumothorax rates (P = 0.400) or hemoptysis rates (P = 0.377). In contrast, we detected a high risk of publication bias pertaining to reported technical success rates (P = 0.007) and diagnostic accuracy (P = 0.000).
Conclusions: A CT-guided biopsy can be safely and effectively used to diagnose SLNs.