Factors Associated with Complicated Parapneumonic Pleural Effusion/Empyema in Patients with Community-Acquired...

Objectives: To identify factors associated with complicated parapneumonic pleural effusion/empyema (CPPE/empyema) in inpatients with community-acquired pneumonia (CAP) and to build a mathematical model for CPPE/empyema. Methods: This is an observational case-control study nested within a retrospective cohort, based on clinical practice, and including adults hospitalized with CAP from 2009 to 2019. Cases and controls were defined according to diagnosis of CPPE/empyema during admission. For each case, two controls were randomly selected and matched for the period of admission to avoid seasonality bias. Explanatory variables included demographic, analytical, clinical, and radiological data; treatment with corticosteroids on admission; prognostic and CAP severity scales; comorbidity; and the interval between symptoms onset and admission. Results: Of 4372 pneumonias reviewed, 2015 were excluded due to pleural effusion, blunting of the costophrenic angle without thoracentesis, or heart failure. Of the remaining 2357 patients, 106 developed CPPE/empyema (cases), and 212 were selected as controls. Factors associated with CPPE/empyema were pleuritic pain (odds ratio [OR] 7.42, 95% confidence interval [CI] 3.83-14.38), multilobar radiological involvement (OR 4.48, 95% CI 2.26-8.88), and leukocytosis (OR 4.12, 95% CI 1.94-8.76). Corticosteroids showed a protective effect (OR 0.24, 95% CI 0.09-0.61). Age (OR 0.99, 95% CI 0.97-1.02; p = 0.56) and sex (OR 1.91, 95% CI 0.94-3.88; p = 0.074) were adjustment variables. The area under the receiver operating characteristic curve was 0.847 (95% CI 0.772-0.921). Conclusions: Pleuritic pain, multilobar radiological involvement, and leukocytosis are associated with CPPE/empyema in inpatients with CAP. Treatment with corticosteroids upon admission seems to be a protective factor. The discriminative capacity of the resulting multivariable model presents moderate/high accuracy.