Experimental studies demonstrate that hydrostatic and permeability pulmonary edema are followed by pleural fluid accumulation that comes from the lung interstitium [2–4]. Furthermore there is a good correlation between extravascular lung water content and pleural effusion volume. The notion that the major source of hydrostatic effusion is the lung is further supported by clinical studies showing that in patients with chronically elevated hydrostatic pressures, the presence of pleural effusion correlates better with left than right heart filling pressures. These data suggest that the excess fluid exits the lung via the visceral pleura into the pleural space.

Lung collapse associated with pleural effusions may lead to hypoxemia due to ventilation-perfusion mismatch or true shunt. The extent of these abnormalities depends upon the extent of perfusion of the compressed airspace, as determined by local factors such as hypoxic vasoconstriction and vascular compression.

The reduction in lung volume induced by a pleural effusion can be largely attributed to collapse of the dependent portions of the lung most prominent at end expiration. However the change in lung volumes is less than the volume of the pleural effusion and is depended on the compliance of the lungs and chest wall. The more compliant the lung the greater the change in lung FRC; the more compliant the chest wall the greater the thoracic cage adjustment with a smaller impact on lung volume. A number of studies in spontaneously breathing patients with unilateral pleural effusion show a disproportionably small increase in lung volumes after large volume thoracentesis and no or poor relationship between the volume of fluid removed and the increase in lung volume. In mechanically ventilated patients the effect of pleural fluid drainage on lung volumes and gas exchange has been variable, with some studies demonstrating little improvement in PaO₂/FiO₂ while others have demonstrated a significant increase in the PaO₂/FiO₂ ratio [5–7]. The response to fluid drainage may depend on the applied airway pressure. Alveolar pressure generated during the respiratory cycle may not be enough to reopen collapsed lung (see lung recruitment below). Recruitment maneuvers (including Bilevel ventilation) should therefore be considered after drainage of a pleural effusion.

Pleural fluid may be drained by either thoracentesis or placement of a small bore catheter (pig-tail catheter) [5–9]. A pig-tail catheter is recommend for large effusions. A pig-tail catheter may also obviate the need to repeated thoracentesis Pleural fluid drainage should always be performed under ultrasound guidance. Ultrasound allows estimation of the size of the effusion [10, 11]; attempts at thoracentesis should be aborted in patients with small effusions (less than 750–1,000 mL). Furthermore ultrasound allows the procedure to be performed safely, particularly in ventilated patients [8, 9, 12]. Not all patients with a pleural effusion require drainage. This should only be considered in patients with a low PaO₂/FiO₂ and in patients who have failed a spontaneous breathing trial. Goligher et al. reviewed 19 observational studies that evaluated pleural fluid drainage in patients undergoing mechanical ventilation [13]. In this study the mean PaO2:FiO2 ratio improved by 18 %. Reported complication rates were low with the pooled risk of post-thoracentesis pneumothorax being 3.4 %.

Respiratory therapy refers to “treatments” provided by the respiratory therapist to aid in lung expansion and mobilizes retained secretions. This includes techniques to loosen and mobilize secretions including saline instillation, endotracheal suctioning and chest clapping/vibration and recruitment (hyperinflation) maneuvers. Manual hyperinflation delivers a large tidal volume breath over a prolonged inspiratory time, followed by an inspiratory hold and a rapid release of pressure [18]. The goal is to stimulate cough and propel mucous cephalad. There is limited data with respect to the efficacy of manual hyperinflation, however, high airway pressure and large lung volumes may produce adverse homodynamic effects and injure the lung via barotrauma and/or volutrauma. Maa and colleagues performed a randomized controlled trial in which ventilated patients with atelectasis were randomized to manual hyperinflations three times a day or to “standard” care [18]. The manual hyperinflation technique used a rate of 8–13 breaths/min for a period of 20 min each session. The manual hyperinflations were performed by a single investigator using a predefined protocol which limited peak airway pressure to 20 cmH₂O. Spontaneous tidal volumes, oxygenation, sputum volume and the chest radiographic score increased in the treatment group whereas these indices remained largely unchanged in the standard care group. The mechanical ventilator can be used to achieve hyperinflation with similar results [20, 21]. This approach may be safer as the inflation volumes and pressures can be preset. High frequency chest wall vibration/compression/oscillation, rib cage compression (or squeezing) and chest wall “clapping/slapping” have been used to loosen and mobilize secretions. There is however no evidence that any of these interventions have any beneficial effects and they are currently not recommended in mechanically ventilated patients [22, 23].