This study investigated the prognostic relevance of pleural and pericardial effusion in patients with acute PE in a German multicenter study. The main finding is that the presence of pleural effusion alone is a prognostic factor in patients with acute PE. The present analysis is one of the largest to date on this topic.

Acute PE is associated with a significant mortality, with reported short-term mortality of up to 20% [1,2,3,4]. The most established prognostic factor is systolic blood pressure [15, 16]. Risk stratification is also mainly based on systolic blood pressure [1]. Other important aspects include cardiac injury with blood parameters, age over 70 years, history of bed rest for more than five days, cancer, chronic obstructive pulmonary disease, renal failure, heart failure, cardiovascular disease, and tachycardia [1, 16].

Risk stratification of acute PE is very important for treatment planning. Patients with low-risk PE can be treated with anticoagulation in most cases, whereas patients with severe PE may require mechanical revascularization [17].

An important aspect of the present analysis is that only the presence of pleural effusion carries prognostic information, whereas quantification by width or density does not carry additional relevant information. At first glance, this seems to contradict clinical routine. Possible confounding factors could be that the volume of the pleural or pericardial effusion is highly variable during the course of the patient’s illness. Moreover, both may be confounded by previous drainage treatment, which could not be accounted for in the present analysis. The exact timing of CT seems to be crucial for correct volumetry and Hounsfield measurement, which may be too heterogeneous in the present cohort.

Nevertheless, this result of the present analysis is of interest because it only reports the presence, which could be easily performed in clinical routine without the need for more complex measurements by the radiologist.

Pleural effusion is very common in critically ill patients with various causes, including viral pleuritis, congestive heart failure, or cancer [9]. Notably, acute PE is the fourth most common cause of pleural effusion after congestive heart failure, cancer, and pneumonia [9].

The prognostic role of the presence of pleural effusion was demonstrated in a recent meta-analysis of 13,430 patients with a reported relative risk of 2.19 (95% CI: 1.53–3.15, p < 0.001) for 30-day mortality [11]. Our present data are very well in line with this previous study. It is important to consider that only the presence of pleural effusion and not the quantification of pleural effusion was used in the meta-analysis. Second, pleural effusion was also measured on other imaging modalities, which could lead to higher heterogeneity of results. CT can be considered as the most standardized imaging modality, although sonography also has a high sensitivity for the diagnosis of pleural effusion [18].

In contrast to pleural effusion, the prognostic role of pericardial effusion in patients with acute PE is less well studied in the literature [12].

A study from Turkey evaluated 570 patients with acute PE. The incidence of pericardial effusion in this study was 7%, which is significantly lower compared to the present results. They showed a statistical signal for the prognostic relevance of pericardial effusion (p = 0.004), but it did not reach statistical significance in multivariate analysis [12].

Previously, the presence of pleural and pericardial effusion was incorporated into a score system, and both contributed equally to the score [19]. This constructed score was more accurate than the clinical score sPESI (AUC of 0.82 vs. 0.75) in the studied cohort of 1698 cases. In this study, the frequency of pericardial effusion was 21.7% in the survivor group compared to 40% in the non-survivor group (p < 0.001), which is slightly higher than in the present cohort [19].

In a meta-analysis of prognostic signs on CT images, right heart dilatation was the only statistically relevant finding with a reported 2.5-fold risk of all-cause mortality [6]. Notably, despite the complexity of the methodology, thrombus burden has been investigated in several studies with inconclusive results [20, 21]. Therefore, it remains a challenge to define prognostically relevant CT findings in acute PE.

An important aspect of the present analysis is that the hazard ratio for the presence of pleural effusion is even higher than that of the most commonly used sPESI score. It is debatable whether the addition of pleural effusion could improve the diagnostic accuracy of this score.

Several limitations of the present study need to be addressed. First, it is a retrospective study from different centers located in Germany. Second, although the measurements performed can be considered reliable, the possibility of some reader bias should be considered. Especially since no central reading of the CT images was performed. Third, there might be some confounding factors induced by previous drainage treatment, which could influence the width of the effusions. Fourth, we could not adjust for other causes of pleural effusion, such as congestive heart failure, which might introduce some confounding bias into the analysis. In addition, there could be some bias introduced by infarct pneumonia, which could also cause pleural effusion. However, only a few patients with pneumonia were identified in the current patient sample and further subgroup analysis was not possible.