Thoracentesis is a very common procedure, rarely associated with severe complications. One relatively common complication is chest discomfort, which is most of the time felt to be secondary to negative pleural pressures generated during the procedure. While most proceduralists use suction to drain the pleural fluid, some drain effusions by gravity only. The investigators propose to evaluate whether gravity-driven thoracentesis results in less discomfort for patients than suction-drive thoracentesis.

Therapeutic thoracentesis aims to drain fluid from the pleural space to alleviate breathlessness. The amount of and speed with which the fluid can be safely drained in one setting is unclear, and likely depends on the physiology of the pleural effusion. The principle concern when draining a large amount of fluid quickly from the pleural space is that excessively negative pleural pressure may be generated; this occurs if the lung is unable to freely re-expand into the space previously occupied by fluid. Excessively negative pleural pressure and the resulting high transpulmonary pressure gradient are thought to be associated with several complications, including pneumothorax ex vacuo, chest discomfort, and re-expansion pulmonary edema (REPE). Evidence suggests that monitoring pleural pressures during thoracentesis via manometry does not mitigate this problem. In fact, data shows that whether manometry is used or not, most patients do experience clinically significant increase in chest discomfort during thoracentesis. Current methods for draining the pleural fluid include suction- (via vacuum bottles, wall suction or the use of large syringes with a one-way valve tubing system) or gravity-driven thoracentesis. Pressures generated by all suction techniques range from -200 to -500 cmH2O, and far exceed what are considered safe pleural pressures. Accordingly, in case of non-expandable lung, excessively negative pressures may develop quickly, exposing patients to complications. Some clinicians advocate for gravity drainage, which generates less negative pressures in the pleural space (specifically defined as the vertical distance between the catheter and the drainage bag, generally around -50 to -100 cmH20). While this technique is considered standard of care by some as it is potentially more comfortable for patients, it is also likely associated with longer procedures, and is not generally favored by clinicians who in general prefer suction drainage, despite the possible higher risk of complications.The investigators propose to study the impact of gravity- versus suction-driven large volume therapeutic thoracentesis on the development of chest discomfort during the procedure, and consider as secondary endpoints: the duration of the procedure, the amount of pleural fluid drained, the rate of REPE, the rate of pneumothorax ex vacuo.

Inclusion Criteria:

1. Referral to pulmonary services for large-volume thoracentesis

2. Presence of a symptomatic moderate or large free-flowing (non-septated) pleural effusion on the basis of:

   1. Chest radiograph: effusion filling ≥ 1/3 the hemithorax, OR
   2. CT-scan: maximum AP depth of the effusion ≥ 1/3 of the AP dimension on the axial image superior to the hemidiaphragm, including atelectatic lung completely surrounded by effusion, OR Ultrasound: effusion spanning at least three interspaces, with depth of 3 cm or greater in at least one interspace, while the patient sits upright.

3. Age > 18

Exclusion Criteria:

1. Inability to provide informed consent

2. Patient has already been enrolled in this study

3. Study subject has any disease or condition that interferes with safe completion of the study including:

   1. Coagulopathy, with criteria left at the discretion of the operator
   2. Hemodynamic instability with systolic blood pressure <90 mmHg or heart rate > 120 beats/min, unless deemed to be stable with these values by the attending physicians

4. Pleural effusion is smaller than expected on bedside pre-procedure ultrasound

5. Referral is for diagnostic thoracentesis only

6. Presence of more than minimal septations and/or loculations on bedside pre-procedure ultrasound

7. Inability to sit for the procedure

Subjects will be randomly allocated into intervention (gravity-driven) and control (suction-driven) groups by opening an opaque study envelope just prior to starting the procedure containing group assignment. Participants will be blinded to the use of gravity VS. suction drainage to prevent knowledge of their group assignment from biasing their pain assessments.