Learning Curve of Ultrasound-Guided Percutaneous Needle Biopsy for Pleural Lesions: A Retrospective Study at Two Tertiary Referral Hospitals

Abstract

Background/Objectives: Ultrasound (US)-guided percutaneous pleural needle biopsy (PCPNB) is increasingly performed as a minimally invasive approach for the diagnosis of pleural lesions. However, no prior studies have investigated the learning curve for this method. The purpose of this study was to assess the learning curve of US-guided PCPNB using the cumulative summation (CUSUM) method and to calculate the number of procedures to achieve proficiency. Methods: A retrospective analysis was performed on 491 US-guided PCPNBs performed by four board-certified radiologists at two tertiary referral hospitals from January 2012 to December 2024. The standard and risk-adjusted CUSUM (RA-CUSUM) techniques were used to evaluate diagnostic success and false-negative rates. The potential impact of previous percutaneous needle biopsy (PCNB) experience on the learning curve was also assessed. Results: The overall diagnostic success and false-negative rate were 89.2% (438/491) and 8.1% (40/491), respectively. Operators achieved acceptable diagnostic success in US-guided PCPNB after a median of 15 (range, 12-45) procedures in standard CUSUM analysis and 22 (range, 10-33) procedures in RA-CUSUM analysis. Acceptable false-negative rates were attained after 18 (range, 14-42) and 24 (range, 12-44) procedures, respectively. Operators with prior PCNB experience required 12 procedures to achieve both acceptable diagnostic success and an acceptable false-negative rate. In contrast, those without experience required 29 and 27 procedures, respectively. Complications occurred in 1.4% (7/491), including one major complication (0.2%). Conclusions: Diagnostic proficiency and procedural safety in performing US-guided PCPNB improved with increasing operator experience. The low complication rate highlights the clinical safety and feasibility of US-guided PCPNB.

Keywords: learning curve; pleura; ultrasound-guided biopsy.

Figure 1

Flowchart of the study. PCPNB = percutaneous pleural needle biopsy.

Figure 2

CUSUM graphs of diagnostic success and false negative rate. Standard (A) and risk-adjusted (B) CUSUM graphs of diagnostic success. Standard (C) and risk-adjusted (D) CUSUM graphs of false negative rate. Diagnostic success results in a downward trend, whereas diagnostic failures lead to an upward trend. CUSUM = cumulative summation, H0 = lower decision limit, H1 = upper decision limit.

Figure 3

CUSUM graphs stratified by operator subgroups based on prior experience in percutaneous needle biopsy during radiology residency. Standard (A) and risk-adjusted (B) CUSUM graphs of diagnostic success. Standard (C) and risk-adjusted (D) CUSUM graphs of false negative rate.

Figure 4

Effect of increasingly stringent failure criteria on CUSUM graphs. Standard (A) and risk-adjusted (B) CUSUM graphs of diagnostic success. Standard (C) and risk-adjusted (D) CUSUM graphs of false negative rate.