I read with interest the article by Kılıç et al. evaluating the diagnostic value of handheld thoracic ultrasound (HH-US) for pleural effusion in thoracic surgery consultations [1]. The study highlights HH-US’s high sensitivity (83.3% vs. chest X-ray [CXR], 88.5% vs. computed tomography [CT]) and its utility in bedside evaluations. However, we noted several inconsistencies, discrepancies with existing literature, and methodological concerns that warrant clarification.
The results report pleural effusion in 61 of 91 patients (67%) via CT, with thoracentesis performed in 47 (51.6%) [1]. However, 14 patients had minimal effusion (≤ 600 ml, not requiring intervention), and one refused intervention, suggesting that 46 patients (61 – [14 + 1]) should have undergone thoracentesis. This discrepancy raises questions about whether an additional patient with minimal effusion was included or if there is a reporting error. Additionally, the claim of “no complications” conflicts with three reported post-thoracentesis pneumothorax cases [1]. Pneumothorax is a recognized complication [2]; clarification on its exclusion from the “no complications” statement and details on its management would strengthen the study.
The reported HH-US sensitivity (83.3% vs. CXR, 88.5% vs. CT) is lower than that of prior studies, such as Brooks et al., who reported 100% sensitivity for hemothorax detection [3], and Xirouchaki et al., who found near-perfect sensitivity in critically ill patients [4]. The authors do not discuss potential reasons, such as operator training or the Philips Lumify device’s limitations. Similarly, the low specificity (25.7% vs. CXR, 35.7% vs. CT) contrasts with higher specificities reported elsewhere [5]. Could the authors elaborate on factors such as pleural thickening or loculated effusions contributing to these differences?
The retrospective design introduces selection bias, as acknowledged, but ambiguous inclusion criteria (e.g., undefined “emergencies”) limit generalizability [1]. The 3 cm cutoff (≤ 600 ml) for minimal effusion, based on the Balik formula [6], lacks justification, as literature suggests varying thresholds [2]. Inter-operator variability, critical for ultrasound, is not addressed. Additionally, the statistical analysis omits kappa values for concordance (57.1% vs. CXR, 78.7% vs. CT), hindering assessment of agreement strength [1]. The absence of an explicit hypothesis complicates interpretation, and missing data on fluid volume drained or diagnostic yield (e.g., malignancy confirmation) limit clinical relevance.
I pose the following questions: (1) Can the authors clarify the discrepancy between effusion cases (61) and thoracentesis cases (47)? (2) Why were pneumothorax cases not considered complications? (3) What explains the lower sensitivity/specificity compared to prior studies [3, 4]? (4) How was the 3 cm cutoff determined? (5) Can kappa values for concordance be provided?
I commend the authors’ contribution to HH-US research and encourage further studies to standardize its use in thoracic surgery.
