Implications of Early Sonographic Evaluation of Parapneumonic Effusions in Children With Pneumonia
Objective. To devise a clinically relevant grading system for the sonographic evaluation of parapneumonic effusions, and to evaluate length of hospital stay as a function of treatment approach and sonographic grades.
Methods. Chest sonograms of 46 pediatric patients diagnosed with empyemas and admitted to two medical centers in the last 8 years were retrospectively evaluated using a grading system based on the degree of fibrinous organization within the parapneumonic effusions. Hospital charts were reviewed to determine the method of treatment and length of hospital stay. Patients were divided into two treatment groups: nonoperative (n = 26) (antibiotics alone, or combined with thoracentesis, or tube thoracostomy) and operative (n = 20) (open decortication, or video thoracoscopy and pleural debridement). Patients in the nonoperative group were further subdivided into two groups: those who received antibiotics alone (n = 11) and those who received antibiotics plus nonoperative drainage thoracentesis and/or tube thoracostomy (n = 15). Within each treatment group, patients were subdivided into two ultrasound grades: low (no evidence of organization) and high (evidence of organization such as fronds, septations, or loculations). Student'st test was performed to compare the lengths of hospital stay for each of the treatment groups and ultrasound grades.
Results. The length of hospitalization was no different for patients with low-grade ultrasounds in the nonoperative (9.8 days) and operative groups (8.0 days). In contrast, length of hospitalization was significantly shorter for patients with high-grade sonograms in the operative group (8.6 days), when compared with the nonoperative group (16.4 days). Length of hospitalization for patients in the nonoperative group with high-grade sonograms was significantly longer (16.4 days) than for those with low-grade ultrasounds (9.8 days). Furthermore, when the nonoperative patients were divided into an antibiotics alone group and a nonoperative drainage group, the patients with low-grade sonograms had no difference in the length of hospitalization (9.0 days vs. 10.4 days), whereas those patients with high-grade sonograms in the nonoperative drainage group had a significantly longer hospitalization (19.9 days) than the antibiotics alone (high-grade) group (11.4 days).
Conclusions. Patients with a low-grade sonogram had similar length of hospitalization if treated with either nonoperative or operative measures. Patients with high-grade sonograms had significantly shorter length of hospitalization when treated with decortication. Our retrospective study suggests that patients with high-grade ultrasound studies treated nonoperatively do not benefit from pleural drainage procedures or chest tube placement. This study demonstrates the usefulness of early sonographic evaluation of parapneumonic effusions. A prospective study evaluating the usefulness of sonographic assessment of severity of disease in the treatment of children with parapneumonic effusions is warranted on the basis of our retrospective data.
Empyema thoracis is defined as a collection of pus in the thorax. It is usually a result of pneumonia, but may also arise from hematogenous sepsis, thoracentesis, tube thoracostomy, trauma, and subdiaphragmatic infections. Three distinct stages of progression of an empyema have been described. The first stage is exudative and is characterized by free-flowing fluid that can easily be drained. The second stage is fibrinopurulent with the formation of septations and loculations for which tube drainage would be very difficult. The final stage is organization in which surgical debridement is considered by some to be the only treatment option.1 2
Much controversy still exists regarding the management of empyemas, especially in the second stage, with many investigators advocating antibiotics in combination with tube thoracostomy and/or intrapleural thrombolytics,3-9 although others advocate the use of early surgical intervention such as video thoracoscopy and pleural debridement.10-14
The goal of each of these methods of treatment is to evacuate infected debris and re-expand the lung. Decreased hospital stay and morbidity are important objectives.
Although pleural fluid analysis and clinical severity have been used to direct therapy,15 16 we hypothesized that ultrasound would be a useful adjunct in directing therapy because of its ability to visualize the internal composition of the fluid. Thus far, ultrasound has been shown to be useful in the following settings: image-guided catheter drainage, localizing loculations for thoracentesis or tube thoracostomy, or monitoring resolution of empyemas.17 18
To our knowledge, the implications of early sonography of empyemas in directing the course of treatment have not been examined. Initial sonography of the opaque hemithorax offers significant advantages. Sonography: 1) can be performed immediately upon admission, 2) can detect the amount of fluid surrounding the lung and allow distinction between nonareated lung and pleural fluid/debris, which may be impossible to differentiate on plain chest radiograph or computed tomography, and 3) can detect loculation, thus providing valuable information regarding the stage of the empyema and the feasibility of various therapeutic options.19-21
In this retrospective review of children with empyemas treated at two different medical centers, we investigated the association between ultrasound grade and length of hospitalization for patients who underwent nonoperative and operative interventions. We did not analyze the data with respect to pleural fluid chemistry, oxygen requirements, febrile days, radiographic appearance, or age.
Hospital charts and thoracic sonograms were reviewed for consecutive pediatric patients admitted to the Children's Hospital of Vanderbilt University Medical Center (between June 1991 and May 1995) and the University of Chicago Children's Hospital (between July 1988 and January 1996) with the diagnosis of either empyema or pleural effusion secondary to pneumonia. Patients were excluded from the study if complete hospital charts and ultrasounds were not available. The final study population consisted of 46 patients.
Ultrasounds were graded using a two-grade system: low-grade (anechoic fluid, no evidence of fibrinous organization) and high-grade (evidence of fibrinous organization such as fronds, septations, loculations, or thickening of the visceral pleural surface) (Fig1A and 1B; Fig2A and 2B).
Charts were reviewed to determine the age of the patient, the length of hospitalization, and the treatment strategies. Pleural fluid analysis was not available for many patients and thus was not included in the study.
Patients were placed into two categories based on treatment groups: nonoperative treatment (antibiotics alone and tube thoracostomy) and operative treatment (video thoracoscopy and pleural debridement or formal thoracotomy and decortication). Patients in the nonoperative treatment group were further subdivided into two groups: antibiotics alone or antibiotics with nonoperative drainage (tube thoracostomy). Length of hospitalization was defined as date of admission to date of discharge. The time from admission to operation in the operative group averaged 3 days.
Statistical analysis was performed using the parametric student'st test to compare lengths of hospitalization for each of the treatment groups and ultrasound grades. A P value < .05 was considered significant.
The total of 46 patients included 25 from Vanderbilt University Medical Center and 21 from the University of Chicago Children's Hospital. The average age was 4.2 years (±4.2 months) with a range of 6 weeks to 17 years. Twenty-seven patients were boys (58.7%).
Operative Treatment Versus Nonoperative Treatment
There were 20 patients (43.5%) in the operative treatment group and 26 (56.5%) in the nonoperative treatment group (Table1). Of the patients in the operative treatment group, 16/20 had received an ultrasound evaluation of high-grade, and 4/20 received an evaluation of low-grade. Of the patients in the nonoperative treatment group, 17/26 had received an evaluation of high-grade and 9/26 had received an evaluation of low-grade. The mean length of hospitalization for the patients with low-grade and high-grade sonograms in the operative treatment group was 8.0 days (±2.0 days) and 8.6 days (±2.4 days), respectively. The mean length of hospitalization for the patients with low-grade and high-grade sonograms in the nonoperative treatment group was 9.8 days (±4.2 days) and 16.4 days (±6.3 days), respectively.
There was no statistically significant difference in the days of hospitalization between the 13 patients (4+9) with low-grade sonograms in the operative and nonoperative groups (P > .05). There was a statistically significant decrease in length of hospitalization in the 16 patients with high-grade sonograms in the operative group compared with the 17 patients with high-grade ultrasounds in the nonoperative group (P < .0001). Of the patients in the nonoperative group, the patients with high-grade ultrasounds had a significantly longer hospitalization than patients with low-grade ultrasounds (P < .005).
Nonoperative Treatment: Antibiotics Alone Versus Nonoperative Drainage
Table 2 demonstrates the difference in lengths of hospitalization for these two nonoperative treatment groups. There were 11 patients in the antibiotics-alone group and 15 patients in the nonoperative drainage group. Of the patients in the antibiotics-alone group, 4/11 received an ultrasound evaluation of low-grade, and 7/11 received an evaluation of high-grade. Of the 15 patients in the nonoperative drainage group, 5/15 patients received a sonographic evaluation of low-grade, and 10/15 received a sonographic evaluation of high-grade. The mean length of hospitalization for the patients with low-grade and high-grade sonograms in the antibiotics-alone group was 9.0 days (±4.3 days) and 11.4 days (±4.4 days), respectively. In the nonoperative drainage group, the mean length of hospitalization for the patients with low-grade and high-grade sonograms was 10.4 days (±4.5 days) and 19.9 days (±5.1 days), respectively.
There was no significant difference between the patients with low-grade sonograms in the two treatment groups (P > .05). There was, however, a much longer hospitalization for patients with high-grade sonograms in the nonoperative drainage group as compared with patients with high-grade ultrasounds in the antibiotics-alone group (P < .005). There was also a significant difference between the patients with low-grade and high-grade sonograms in the nonoperative drainage group (P <. 005).
Rapid recovery of children from serious illness and their expedient return to their home environment is a desirable goal from both the medical and economic standpoints. Parapneumonic effusions in children with pneumonia is an illness for which a multiplicity of treatment options is available to clinicians, and it is clearly important to evaluate objectively which of these treatment options leads to the desired goal.
This retrospective study suggests a potential importance of early sonographic evaluation of empyemas. This retrospective study is based on data acquired at two different institutions with a varied approach to the treatment of children with these conditions. This was necessary to obtain data on a significant number of patients udergoing different types of treatment, including operative and nonoperative management of high-grade effusions. Our data indicate that a patient who presents with a low-grade sonogram will have a similar length of hospitalization if treated by either nonoperative or operative interventions. To avoid high procedural costs and the inherent morbidity of an operation or placement of a chest tube, it would be in the patient's best interest to be treated with antibiotics alone. A diagnostic thoracentesis with Gram stain and pleural fluid culture would be beneficial in selecting the most appropriate antibiotics. In contrast, for our patients with high-grade sonograms, operative intervention was associated with a reduced stay of 47.6% less than nonoperative intervention.
It is interesting to note that patients with high-grade sonograms in the tube thoracostomy drainage group have a 74.6% increase in hospital stay when compared with patients in the antibiotics-alone treatment group. This observation may be due to several factors. Chest wall discomfort secondary to a chest tube does impair pulmonary toilet in children and may retard clearing of the underlying pneumonia. It is also likely that selection of treatment options is biased in this retrospective study, when patients with more toxicity, larger fluid collections, or slower response to antibiotic therapy alone might have been selected for chest tube drainage. In any event, our study would not suggest a therapeutic advantage to drainage plus antibiotics in high-grade empyema.
To date, the role of sonography in the management of empyemas has been limited to image-guided catheter drainage, localization of fluid for thoracentesis and tube thoracostomy, and monitoring clinical improvement. This study clearly demonstrates the benefit of early sonographic evaluation in the classification of empyemas into two clinically useful grades. This information can guide subsequent treatment rationales and optimization of the most appropriate treatment in order to minimize morbidity and convalescence. Our retrospective data suggest that using sonography as an initial adjunct to clinical assessment in the diagnostic evaluation of empyema may effect a reduction in the length of hospitalization. A prospective randomized trial confirming the role of sonography in managing patients with parapneumonic effusions is warranted.
We would like to acknowledge the clinical contributions of our surgical colleagues: Dr John B. Pietsch and Dr Walter Morgan from the Department of Pediatric Surgery, Vanderbilt University Medical Center, and Dr Mindy Statter from the Department of Pediatric Surgery, Chicago Children's Hospital.
- Received March 27, 1997.
- Accepted June 20, 1997.
Reprint requests to (R.M.H.) Department of Radiology, D-1120 Medical Center North, Vanderbilt University Medical Center, Nashville, TN 37232–2675.
- Andrews N,
- Parker E,
- Shaw R,
- Wilson N,
- Webb W
- Bouros D,
- Schiza S,
- Panagou P,
- Drositis J,
- Siafakas N
- Taylor R,
- Rubens M,
- Pearson M,
- Barnes N
- Light R
- Kern J,
- Rodgers B
- ↵Hernanz-Schulman M. Role of sonographic imaging in the pediatric emergency patient. In: Kirks D, ed. Emergency Pediatric Radiology: A Problem-Oriented Approach. Washington, DC: American Roentgen Ray Society; 1995:65–76
- Copyright © 1998 American Academy of Pediatrics