We obtained institutional review board approval for each of the cadaveric experiments and the CT review.
3.1. Cadaver Experiment
Three fresh cadavers (two females and a male; each labeled “Cause of death: old age”), which were donated to our institute for educational and experimental purposes, were available to us. Visual inspection of the cadavers revealed no significant structural abnormalities. One female cadaver was excluded due to the large amount of bilateral pleural fluids detected on ultrasound (US) examination. US of the remaining two cadavers revealed high-echoic, fully expanded right lung and scant amount of pleural fluids.
Both cadavers were placed supine onto a table so that the right hemithorax went over a trapezoid gap (20-cm wide, 20-cm long at the medial margin, 30-cm wide at the lateral margin) created at the top (
Figure 1). An 18-gauge angiocath needle was inserted to the site of US-detected pleural fluid, until a few milliliters of fluid was aspirated. We gently removed the inner needle and then began administering tap water via the catheter, 100 mL at a time. After each session of water administration, DPE was measured using US at the nipple level at each of the middle axillary lines (with transducer angled horizontally) and the posterior axillary line (with transducer angled 45 degrees to the ground). Water administration was stopped when DPE exceeded 1 cm (the minimum generally considered safe for thoracentesis ( 5)) at both lines. Statistical analysis was performed using paired Student t-test. P < 0.05 was considered to indicate statistical significance.
Figure 1. A, A table with gaps at one side of the tabletop was constructed for the cadaver experiment. With the lateral margin of the table carved out, the posterior thorax was exposed for the ultrasound probe and tapping needle to approach it from below. B, Drawing of axial cross section in a typical supine patient with a small-to-moderate effusion. Compared to the conventional method of lateral approach (solid arrow), the effusion can be much more easily visualized and tapped from the posterior and posterolateral aspects of the hemithorax (open arrows) via the carved out (dashed line) portion of the novel supine thoracentesis tabulation.
3.2. CT Review
We reviewed all CT scans (n = 4658) obtained between February 2012 and May 2012. In patients with effusion, concurrent lateral-decubitus plain chest radiographs were reviewed to determine whether that effusion was both nonloculated and at least moderate in amount (the width of effusion exceeding 1.5 cm (
CT scans were obtained on one of two CT scanners (LightSpeed VCT, GE Healthcare, Milwaukee, WI, USA; or SOMATOM definition flash, siemens healthcare, Forchheim, Germany), with or without intravenous administration of contrast medium (100 cc at 2 - 2.5 cc/s). All images were reconstructed into axial images with 5-mm slice thickness at 5-mm intervals and coronal images with 3-mm slice thickness.
Two radiologists independently and blindly measured DPE and CNR and checked the presence or absence of atelectasis. Interpretation of difference in the presence of atelectasis was determined by consensus. Scrolling up and down axial CT scans, we located the longitudinal position of the greatest DPE in each of the middle axillary lines (lateral approach, horizontal direction), posterior axillary line (posterolateral approach, 45 degrees to the ground), and midclavicular line (posterior approach, posteroanterior direction) (
Figure 2A). As the maximal DPE at each approach was measured, the presence or absence of subpleural atelectasis was assessed. At each approach, the CCD was measured by counting the number of axial CT scans showing a DPE greater than 1 cm ( Figure 2B).
Figure 2. A, The depths of pleural effusion (DPEs) in the lateral (mid-axillary, a), posterolateral (posterior axillary, b) and posterior (posteroanterior direction to the mid-clavicular line (dashed line, c) tapping routes. The maximal DPEs along the routes a, b, and c may or may not be seen on the same axial plane. B, Nearby the costophrenic sulcus, craniocaudal distance of pleural effusion (d) has been measured (DPE > 1 cm).
Each of the 32 effusions was placed into a small group according to the fluid volume. For each volume group, mean of maximal lateral, posterolateral, and posterior DPEs and CCDs were calculated. In order to assess the effect of effusion volume on the safety of the three approaches, we compared each group’s mean parameter to our safety standards (CCD of 5 cm and DPE of 1 cm). For the estimation of effusion volume, we summed up the slab volumes [5 mm (scan interval of axial CT) × pleural fluid area (manually measured at each axial section scan)] (
Figure 3. For the estimation of pleural fluid volume, we summed up the slab volumes [pleural fluid area (manually measured at each axial section scan) × 5 mm (scan interval of axial CT)].
For comparisons of DPEs and CCDs, we performed a repeated-measures analysis of variance. When this analysis yielded significant results, it was followed by pairwise comparisons to determine significant differences between the imaginary routes. For comparisons of frequencies of atelectasis, we performed McNemar test. Interobserver variability for measurements of DPEs and CCDs was analyzed by calculating the intraclass correlation coefficient (ICC) with the two-way random effects model. An ICC value lower than 0.4 suggests that the observers are in poor agreement. An ICC value between 0.4 and 0.75 suggests that the level of agreement is fair to good, while an ICC value greater than 0.75 suggests excellent agreement. In all tests, significance was assigned at a P value of less than 0.05.