Community-acquired pneumonia (CAP) is an increasing problem in the elderly, with literature reporting that about 45–50% of all CAP hospitalizations occur in patients 65 yrs of age or older[12]. Poor clinical outcomes and increased mortality can be associated with the development of parapneumonic effusions[13]. In this study, a retrospective analysis of the clinical data of 132 elderly patients with CAP in our hospital showed that the proportion of elderly patients with CAP combined with PPE was 40.9%, and the in-hospital mortality of pneumonia patients with PPE was significantly higher than that of patients with pneumonia alone. Therefore, the study of clinical characteristics of elderly patients with CAP combined with PPE is conducive to reinforce the importance of early recognition, diagnosis and treatment.
In addition to the inflammatory and virulent features of the lungs and pleura due to direct bacterial invasion that lead to PPE, the physical conditions and underlying diseases of patients also contribute to the pathophysiological development of PPE. In this study, the middle-edged and older adults with CAP mostly have more basic diseases, and 87.1% of the patients combine with at least one disease. The existence of underlying diseases is an important risk factor for elderly pneumonia, which will lead to the reduction of the body's resistance to infection and increase the development of pneumonia and the risk of death[14]. In the previous studies, there were some common complications of PPE, such as diabetes mellitus, malignancy, chronic excessive alcohol consumption, chronic lung disease, immunosuppressive states, aspiration,etc[15–16]. But in this study, the proportion of PPE patients with the nervous system diseases is higher, which is related to the increased risk of aspiration pneumonia in elderly patients with nervous system diseases, aspiration is widely recognized as an important risk factor for pneumonia[17], for the first time after aspiration pneumonia, 1-month mortality was 23.9% in patients with Parkinson's disease, about two-thirds of the patients died within a year after the aspiration pneumonia[18–19]. The PPE group in-hospital mortality is higher in this study, considering these patients had significantly older ages, long-term bedridden, cognitive impairment, or swallowing dysfunction, onset is given priority to with aspiration pneumonia or more lung infiltrates, these tend to make a delayed diagnosis and treatment after onset, or their families have poor willingness in invasive operation such as endotracheal intubation in the treatment of intention. therefore, PPE patients had higher in-hospital mortality rate than NPPE. In addition, the PPE group did not have a high proportion of diabetes, considering the sample size was relatively small, thus we still had a potential selection bias.
In terms of etiology, in this study, 42.6% of patients in the PPE group were detected with pathogenic bacteria in order as follows: G–Bacillus, Candida and Staphylococcus. The commonest infection of G–Bacillus were Enterobacteriaceae, such as Escherichia coli and Klebsiella pneumoniae, 1 case of pleural effusion was cultured as Streptococcus agalactiae. A meta-analysis of Hassan, etc[20] showed that the most common aerobic isolates in pleural effusion culture were Staphylococcus aureus (20.7%), Streptococcus aerophylus group (18.7%), Pseudomonas (17.6%), Enterobacteriaceae (11.9%), Streptococcus pneumoniae (10.8%), Klebsiella (10.7%), Acinetobacter (5%) and coagulase negative Staphylococcus (4.5%). In this study, the distribution of pathogenic bacteria was different from that in the literature. On the one hand, the pathogenic specimens were different, and on the other hand, because of the empirical antibiotic therapy before the specimens were submitted for examination, the positive rate of pathogenic bacteria was not high. With the appearance of whole genome next generation sequencing, we may overcome some of the shortcomings of standard microscopy and culture techniques. In this way, more or less common pathogens will be identified, which will be more helpful for clinical decision making.
In order to treat PPE, the most basic is prompt clean the infection of intrathoracic and using antibiotics, which should be guided by specific pathogen sensitivity[21]. In culture-negative PPE, empirical use of antibiotics should be based on local pathogenic distribution, drug resistance, and antibiotic management policies. The British Thoracic Society[22] (BTS) and American Association for Thoracic Surgery[23] advice that broad-spectrum antibiotics should be used before the pathogens are confirmed, which covers Gram-positive, Gram-negative and anaerobic, such as β-lactamase inhibitors, third-generation cephalosporins and carbapenems, etc. Consistent with recommendations, compared with the NPPE group, the use of antibiotics in the PPE group in this study was mainly β-lactam inhibitors, and the utilization rate of carbapenems or glycopeptides was 10.3% (14/54), which was significantly higher than that in the NPPE group, which was mainly related to the pathophysiological characteristics of the PPE group and the severity of patients' disease.
Clinically, patients often present with pleuritic chest pain due to inflammation of the pleura, but it is uncommon in the PPE group (11.11%), and 6.41% of patients still have chest pain without effusion, so we cannot diagnose pleurisy only by chest pain. Dyspnea is the most common manifestation of pleural effusion, and the severity of dyspnea has little obvious correlation with the amount of effusion, and it may be associated with abnormalities in gas exchange, respiratory mechanics, respiratory muscle function and hemodynamics, which are caused by pleural effusion[24–25]. Although 20.51% of the NPPE patients had chest tightness, the proportion of chest tightness and dyspnea in the PPE group was higher (51.85%), and chest tightness was an independent risk factor for PPE. Therefore, it should be vigilant that whwther the elderly CAP patients with chest tightness complicated with pleural effusion or not. Our study demonstrated that PPE patients were easier to present long-term fever, suggesting that fever, as a clinical marker of inflammation, lasted longer in patients with pleural effusion. Whereas the persistence of fever in patients with pneumonia can complicate the conditions, and suggests that inflammation persists[26].
Because of the persistence of pleural inflammation, inflammatory markers are significantly elevated in PPE and empyema patients[27]. The results of this study showed that there were no significant differences in peripheral white blood cells count, neutrophil count and PCT between the PPE and the NPPE patients, while the CRP level in the PPE group was significantly higher than that in the NPPE group. As a classic inflammatory marker, CRP is widely used in the diagnosis of infectious diseases[28]. In a previous study, pleural fluid CRP levels can be used to distinguish between parapneumonic effusions and other types of exudative effusions[29]. CRP levels < 0.64 mg/dL are likely to indicate a pleural effusion from congestive heart failure, whereas levels ≥ 1.38 mg/dL are suggestive of an infectious etiology. The study of Petrusevska - Marinkovic, etc[16] suggests that the CRP in the complex PPE patients was significantly higher than that in the simple PPE patients [(231.79 ± 112.2) mg/L vs. (163.8 ± 147.9) mg/L, P < 0.01], and both were significantly higher than that in the CAP patients without effusion [(139.48 ± 105.7) mg/L, P < 0.01]. Patients whose CRP does not decline with treatment during the course of the disease are at significantly higher risk for complex PPE or empyema. D-dimer is an objective biomarker for reflecting coagulation and fibrinolysis. The reasons for the formation of PPE include the interaction between inflammation and coagulation as well as the occurrence of intrapleural fibrosis. The microcirculation thrombosis and extracellular fibrin remodeling usually with elevated D-dimer levels[30]. In this study, PPE patients have significant elevated D-dimer levels, indicating the disorder of coagulation and fibrinolysis. Multivariate logistic regression analysis revealed that serum albumin (OR = 0.876, 95%CI: 0.790–0.971, P = 0.012) and serum sodium (OR = 0.896, 95%CI: 0.828–0.969, P = 0.006) is also an independent risk factor for PPE in elderly patients with CAP, which is consistent with previous studies[31].
There were some limitations in this study. Firstly, as a retrospective observational study, the number of cases was small and some data were missing. Second, for some patients with pleural effusion, the B-ultrasound showed less effusion and could not be extracted safely, so the relevant laboratory indicators of pleural effusion were not analyzed. Third, due to the limited sample size, further stratified comparison could not be carried out, and the accuracy and application value of the results still need to be confirmed by further in-depth research.