Reason for Consult: Still on Oxygen, Discharge Pending, Requesting Thoracentesis.

Abstract

Thoracentesis is commonly performed in the management of pleural effusions. In addition to providing valuable diagnostic data, it can be performed to achieve therapeutic benefit. In many patients with pleural effusion, fluid drainage via thoracentesis leads to rapid relief of dyspnea.1 At times, particularly for hospitalized patients, the procedure is requested to manage hypoxemic respiratory failure. Historical literature suggest that pleural effusions may cause hypoxemia due to an increase in shunting through atelectatic lung parenchyma ipsilateral to the effusion.2 Data to support pleural effusion as the sole cause of hypoxemia are inconsistent, however, likely confounded by the fact that pleural effusions are often a manifestation of underlying disease processes which may cause hypoxemia via other mechanisms.3 Original research by Taylor and colleagues addresses an important and common clinical question: will therapeutic thoracentesis help reverse this patient’s hypoxemia? Given the high frequency of this procedure request, the results of this study provide essential information to allow for medical teams to improve patient care in 2 main ways: first, a realization that the procedure being requested may not achieve the desired therapeutic endpoint; and second, to foster a more thorough deliberation of possible alternative etiologies of hypoxemic respiratory failure in a patient that concurrently has a pleural effusion. Importantly, it should initiate the thought process that the pleural effusion itself may not be related to the pathology that is causing hypoxemia. Although historical data may suggest effusion drainage may improve hypoxemia, anchoring this notion could be considered search-satisfaction bias.4 Furthermore, waiting for the drainage to improve the patient’s respiratory failure may stall investigations that test other hypotheses. The authors retrospectively assessed 502 inpatients who underwent therapeutic thoracentesis (defined as drainage of ≥100 mL, with median aspirated volume 1400 mL).3 They found no significant intrasubject before/after changes in SpO2, degree of supplemental oxygen requirement, or SpO2:FiO2 ratio. Oxygen saturation by pulse oximetry in place of PaO2 allowed for this large number of patients to be retrospectively studied, as formal blood gas analysis is uncommon outside the intensive care unit and a requirement for PaO2 would have greatly diminished the number of patients included and skewed the sample toward the critically ill. Since the postprocedure SpO2 could have been influenced by periprocedural up-titration or down-titration of supplemental oxygen delivery, SpO2:FiO2 was also assessed. This latter metric correlates well with PaO2:FiO2 in acute lung injury patients.5 The study’s baseline patient characteristics are a good representation of the patient population that may have a pleural effusion, with common etiologies well-represented, bolstering the generalizability of these findings. Procedures were performed with the use of ultrasound guidance under the supervision of an interventional pulmonologist. Rates of both major and minor complications are equitable with previously reported data.6 The incidence of nonexpandable lung is stated, and although definitions for nonexpandable lung vary, the reported rate is consistent with previously reported literature.7 Overall, these results suggest therapeutic thoracentesis does not routinely lead to an improvement in clinically relevant measures of systemic oxygenation. The authors did find a significant improvement in several postprocedure measures of oxygenation in univariate and multivariate regression analyses, although all were <1% and seem to be safely labeled clinically insignificant. This study had several limitations, primarily related to its retrospective design. Patients were identified over 5 years of data extracted from an electronic health record, and so therefore relied upon accurate clinical data entry, although there appears to be very limited missing oxygenation-related data. In clinical practice, the application and titration of FiO2 is usually in response to observed SpO2 or PaO2 values, with target values often varying between patients (with unique comorbidities), units, hospitals, or institutions, making the SpO2 outcome alone somewhat difficult to interpret. Inclusion of SpO2:FiO2 helps allay related concerns by also accounting for the degree of contemporaneous supplemental oxygen delivery. Oxygenation was followed through 24 hours postprocedure, and the authors point out that it is conceivable a significantly delayed improvement in postthoracentesis oxygenation could have been missed, although no prior data suggests such a delayed response is common. In summary, thoracentesis is a minimally invasive procedure. It is typically low risk but is not risk-free. Risks and benefits should be carefully weighed and discussed with the patient before proceeding. If being requested primarily for hypoxemia treatment, this study suggests that pleural fluid aspiration may not routinely lead to the clinically desired endpoint. It sets an excellent foundation for future prospective studies that can add information to this important topic. Importantly, it may also lead to a reduction in the number of thoracenteses performed for an inappropriate clinical endpoint and may help prevent early closure on an effusion as the main driver of hypoxemic respiratory failure.