Treatment of complicated pneumonia includes antimicrobial therapy and pleural fluid drainage.
Empiric antibiotic therapy should be directed at the most likely pathogens while taking into account the patient's severity of illness and local bacterial resistance patterns. Clindamycin alone or in combination with a third-generation cephalosporin (e.g., cefotaxime, ceftriaxone) is appropriate in most situations. Clindamycin has excellent activity against penicillin-resistant pneumococci32 as well as group A beta-hemolytic streptococci and many MRSA isolates.33–35 Cefotaxime or ceftriaxone is added for enhanced coverage against drug-resistant pneumococci. Patients requiring intensive care management (e.g., noninvasive or invasive ventilation, high concentrations of supplemental oxygen) should receive vancomycin or linezolid instead of clindamycin. Both vancomycin and linezolid provide broader Gram-positive coverage since some community-acquired MRSA isolates are intrinsically or inducibly resistant to clindamycin.
There have not been any trials of complicated pneumonia to determine the optimal length of antibiotic therapy. For patients undergoing pleural drainage, a reasonable approach is to continue antibiotics 1 week after resolution of fever. In general, this will mean about 2 weeks of therapy postintervention. A longer course of antibiotics may be required for patients not undergoing pleural drainage. We typically switch from intravenous to oral therapy once the chest tube has been removed. Reasonable oral antibiotic regimens include clindamycin alone or in combination with either amoxicillin or amoxicillin–clavulanate. Oral linezolid is more appropriate for patients who received empiric vancomycin or linezolid therapy. The results of blood or pleural fluid cultures may permit more narrow spectrum therapy.
The size and character of the effusion should guide the pleural drainage decision. Small effusions do not require any intervention, however close monitoring is warranted as the effusion may increase in size. Moderate, large, and loculated collections generally require early pleural fluid drainage. Early pleural drainage may improve outcomes for several reasons: (1) Infections in the fibrinopurulent phase (stage 2) are associated with a worse outcome than infections in the exudative phase (stage 1)36–39; (2) The fibrinopurulent stage develops early in the course of infection; (3) Noninvasive methods (e.g., radiologic studies) do not reliably differentiate the fibrinopurulent phase from the exudative phase37,40,41; (4) Invasive, nonoperative methods (e.g., needle thoracentesis) using pleural fluid chemistries to classify an effusion as exudative have not been validated in children31,42; and (5) Multiple pleural fluid samples from the same patient do not always reveal concordant results.43,44
The decision on whether to classify an effusion as small, moderate, or large is more difficult. In adults, the American College of Chest Physicians classifies effusions based on the height of the effusion (for those that are free-flowing) on decubitus radiographs as follows: <10 mm, small; ≥10 mm but <1/2 the hemithorax, moderate; and ≥1/2 the hemithorax, large.45 While the adult criteria do not directly apply to young children, they can serve as a useful guide. For example, an 8-mm effusion in a 3-year-old could be considered moderate in size since it occupies a much large proportion of the hemithorax than in a 16-year-old.
Options for pleural fluid drainage include thoracentesis, thoracostomy (i.e., chest tube placement) with or without chemical fibrinolysis, video-assisted thoracoscopic surgery (VATS), and open thoracotomy (Table 34–6). At this time, there is considerable controversy over the ideal management strategy for children with complicated pneumonia and a paucity of studies examining this issue.
Table 34–6. Description of Procedures ||Download (.pdf)
Table 34–6. Description of Procedures
Needle inserted between the ribs on the lateral chest wall into the pleural space, usually with ultrasound or computed tomography guidance.
Local anesthesia, minimal (anxiolysis), or moderate sedation
Large bore, hollow, flexible tube placed between the ribs into pleural space through a 2-cm skin incision on the lateral chest wall. The tube is connected to a canister containing sterile water. Suction is applied to facilitate drainage.
Local anesthesia, moderate or deep sedation
Video-assisted thoracoscopic surgery
Operative technique where a small camera and instruments are inserted into the pleural space through three small (1 cm) incisions of the skin and muscle on the lateral chest wall to mechanically remove purulent material and pleural adhesions. A thoracostomy tube is placed through one of the existing incisions following completion of the procedure.
Operative technique where instruments are inserted into the pleural space through a single 5–8 cm incision of the skin and muscle on the posterolateral chest wall to mechanically remove purulent material and pleural adhesions. A thoracostomy tube is placed through a second smaller 1–2 cm incision following completion of the procedure.
Historically, standard treatment of complicated pneumonia involved operative debridement and conversion of the closed pleural space to an open drainage with rib resection only if there was no definite improvement following tube thoracostomy.46,47 Several authors, noting the rapid resolution of symptoms in children undergoing earlier open thoracotomy, began to advocate the use of thoracotomy as a definitive therapy rather than a procedure of last resort.48–51 The advent of less invasive techniques such as VATS in the late 1990s served as an additional impetus to reconsider the strategy of limiting operative intervention only to cases of tube thoracostomy failure.51–55 The key studies addressing this topic are discussed below.
Avansino et al. systematically reviewed the literature to perform a pooled analysis of observational studies published between 1981 and 2004.56 Primary operative therapy (VATS or thoracotomy) reduced the length of stay (LOS) by 45% (199 patients from four studies) and the requirement for repeat procedures by 90% (492 patients from nine studies) compared with primary nonoperative therapy (primary thoracostomy or thoracentesis). This review had several limitations. Firstly, the study design did not allow the authors to adjust for important confounding variables such as the timing of the intervention or the choice of empiric antibiotic therapy. Secondly, the included studies had heterogeneous study designs. Finally, the results were biased toward favoring primary operative therapy since patients undergoing primary thoracostomy were grouped together with patients undergoing primary thoracentesis.
Li et al. conducted a retrospective study of 1173 patients using the Kids’ Inpatient Database to compare primary operative management (decortication within 2 days of admission) to primary nonoperative management (all other children with complicated pneumonia, including those with initial decortication 3 or more days after admission). The primary endpoint was LOS.57 Primary operative management was associated with a 4.3-day shorter LOS [95% confidence interval (CI): −6.4 to −2.3 days) compared with primary nonoperative management. However, if the analysis was limited to the subset of patients with empyema as their primary diagnosis, the reduction in LOS was more modest (1.7-day reduction; 95% CI: –0.4 to –3.0 days). The authors also found a significant difference in therapeutic failure between the groups (5.5% for primary operative management vs. 39.3% for nonoperative management).
Shah et al. conducted a study of 961 patients using administrative data from 27 free-standing children's hospitals.2 In contrast to the study by Li et al., only patients undergoing pleural drainage within 48 hours of hospitalization were included in this study. The primary outcomes were LOS and the requirement for repeat pleural fluid drainage procedures. This study found a 2.7-day reduction (approximately 24% shorter) in LOS for patients undergoing VATS compared with primary tube thoracostomy. In addition, patients in the VATS group had an 84% reduction in need for additional procedures compared with patients initially treated with a thoracostomy (adjusted odds ratio, 0.16; 95% CI: 0.06–0.42).
Kurt et al. conducted a single center randomized trial in the United States comparing VATS with conventional thoracostomy drainage with or without fibrinolysis.58 The primary endpoints were LOS and days with thoracostomy tube.58 Secondary endpoints included duration of fever, duration of supplemental oxygen, narcotic use, and number of drainage procedures. Children undergoing VATS (n = 10) had a significantly shorter LOS (mean LOS, 5.8 days) compared with children undergoing primary thoracostomy (n = 8) (mean LOS, 13.3 days; P < 0.004). In addition, the duration of chest tube drainage, narcotic use, number of radiographs, and number of procedures were significantly lower in children undergoing VATS compared with children undergoing primary chest tube placement.
Sonnappa et al. conducted a single-center randomized trial in the United Kingdom comparing children undergoing VATS with children undergoing primary thoracostomy with intrapleural urokinase with a primary outcome of LOS after the intervention.59 The secondary endpoints included total duration of hospitalization, duration of thoracostomy tube, and failure rate. There were no differences in either the primary or secondary outcomes between the two groups. The lack of differences in outcomes stands in stark contrast to studies conducted in the United States where primary VATS has consistently been associated with shorter hospitalizations and fewer repeat pleural drainage procedures than primary thoracostomy.2,57,58 Differences in causative organisms, timing of presentation for pleural drainage, frequency of chemical fibrinolysis, operative technique, and systems of care could potentially account for such differences in outcomes of children undergoing VATS in the United Kingdom compared with the United States.
The currently available data suggest that early primary VATS is associated with a substantial reduction in the requirement for repeat procedures but a more modest reduction in LOS compared with early primary chest tube placement. While it is tempting to accept early surgical intervention as the new practice standard, we feel that additional data are required. Since some children do remarkably well with chest tube drainage alone, studies should focus on ways to accurately identify this low-risk subgroup of patients. Furthermore, VATS requires specialized surgical training and most community hospitals do not have surgeons with the technical training and expertise to perform this procedure. Therefore, any benefit of primary VATS to patients with complicated pneumonia initially evaluated at community hospitals should be balanced against the delays in pleural drainage that could result from the transfer process. We feel that current guidelines should emphasize the importance of early pleural drainage (within 24–36 hours of hospitalization), regardless of drainage procedure.
Chemical fibrinolysis (e.g., streptokinase, urokinase, tissue plasminogen activator) does not appear to offer significant benefit though pediatric trials have not had sufficient statistical power to accurately evaluate safety or long-term outcomes. In a randomized trial comparing intrapleural urokinase (n = 29) with intrapleural saline (n = 29), there was no difference between the two groups in the proportion of patients requiring subsequent surgical intervention (9% overall).60 However, the urokinase group had a 28% shorter LOS (7.5 days vs. 9.5 days; P = 0.027).60 In the randomized trial by Singh et al.,61 more patients in the saline group (5/21) than in streptokinase group (0/19) required subsequent surgical intervention though this difference did not reach statistical significance; LOS was not assessed. A multicenter randomized trial of 427 adults found that there was no benefit of streptokinase compared with placebo in terms of mortality, requirement for subsequent surgery, radiographic outcomes, or LOS.62 Adverse events including chest pain, fever, and allergic reaction were more common with streptokinase (7%) than with placebo (3%).62 We do not currently recommend the use of chemical fibrinolysis in patients with complicated pneumonia.