Factors Affecting Outcome Following Video-Assisted Thoracoscopic Surgery for Empyema in Children: Experience from a Large Tertiary Referring Center

Abstract

Introduction:

We report the results of video-assisted thoracoscopic surgery (VATS) in a large population of children with empyema, focusing on the factors affecting the postoperative length of stay (LOS).

Materials and Methods:

After ethical approval (RES-18-0000-071Q), a retrospective review was performed (2013–2018). Results are reported as number of cases (%) and median (range) and analyzed by Mann–Whitney U and Kruskal–Wallis tests. Correlation analysis was conducted.

Results:

We identified 159 children with empyema; 75 [42 (56%) males] underwent VATS. Median age was 3.6 (0.4–14.5) years. Presentation was: autumn 15 (20%), winter 26 (35%), spring 18 (24%), summer 16 (21%) with no difference in LOS (P = .6). Preoperative symptoms duration was 7 (2–28) days. Postoperatively, chest drain was on suction in 30 (40%) patients, in situ for 3 (2–13) days. Six (8%) children required further procedures. LOS was 8 (3–47) days. Pleural fluid revealed: Streptococcus species. 41 (55%), other species 8 (11%), no bacteria 26 (34%); LOS was longer with positive pleural fluid: 9 (4–47) versus 6.5 (3–16) days (P = .02). There was no correlation between the LOS and preoperative symptoms duration (r = −0.03 [95% CI −0.3 to 0.2]; P = .7), empyema size (r = 0.2 [95% CI −0.07 to 0.5]; P = .1) and chest drain size (r = 0.09 [95% CI −0.14 to 0.3]; P = .4).

Discussion:

In our experience, >90% of children with empyema will be treated with a single VATS with an average LOS of 8 days. Positive microbiology culture significantly affects the LOS.

Introduction

Empyema thoracis (or simply empyema) is a collection of pus in the pleural space that complicates pneumonia in around 0.6% of children.¹

Since the 1990s there has been a steady increase in both the incidence and proportion of complicated cases of pneumonia worldwide.^2–5 In addition, although the introduction of the seven-valent pneumococcal conjugate vaccine (PCV7) has led to a substantial decrease in hospitalizations for childhood pneumonia, there has also been a corresponding increase in the number of hospitalizations for empyema.⁶

Surgical intervention for empyema is indicated when infected fluid or sepsis is not effectively controlled with antibiotics or in the presence of significant respiratory distress. Surgical options include percutaneous aspiration and/or chest drain insertion (with or without intrapleural fibrinolytic agents), video-assisted thoracoscopic surgery (VATS), and thoracotomy. In children requiring surgery, currently there is limited evidence in the literature on which treatment provides the best outcome. A recent systematic review and meta-analysis documented that, although the use of chest drain and fibrinolytic agents is effective in >75% of cases for the surgical management of empyema in children, VATS might be associated with a reduced need for reintervention and shorter postoperative hospital stay without additional morbidity.⁷

However, the article highlighted some heterogeneity among the studies regarding the length of stay (LOS) after intervention. A number of factors, which have not been fully investigated, could potentially affect the LOS, including the type of pathogens, the duration of preoperative symptoms, and the timing of surgical intervention. The aim of our study was to report the results of VATS in a large population of children with empyema, focusing on the factors affecting the LOS.

Materials and Methods

This is a retrospective cohort study of children (<18 years of age) treated with VATS and chest drain for empyema at a single institution. The study was approved by the local ethics committee (RES-18-0000-071Q). Patients were identified using computerized institutional databases searching for “pneumonia,” “pleural effusion,” “empyema,” “video-assisted thoracoscopic surgery,” and VATS associated with cases occurring between January 2013 and June 2018.

Electronic medical records were reviewed, and data collected using an electronic database for statistical analysis (Excel, Microsoft Corporation). Information collected included the patient demographics, clinical presentation, microbiological and laboratory data, hospital course, and follow-up outcome. All patients were initially medically managed by the pediatric respiratory team. They were then referred to the surgical team for possible intervention after clinical deterioration despite optimal medical management. This was defined as a persisting fever despite broad spectrum intravenous antibiotic therapy and/or significant respiratory distress (tachypnea, tracheal tug, and subcostal recessions) requiring oxygen therapy and/or leading to the need for ventilatory support. Routine preoperative evaluation consisted of blood investigations: full blood count, white cell count (WCC), urea and electrolytes, c-reactive protein (CRP), chest X-ray, and chest ultrasound scan. The radiological investigations assessed the volume of the empyema and presence of debris, internal echoes, loculations, or septations. Chest computer tomography was not routinely performed. VATS was performed under general anesthesia according to routine standardized practice. In summary, this included endotracheal intubation with two-lung ventilation and a lateral decubitus position. A 5-mm 30° telescope was inserted through the sixth or seventh intercostal space in the mid-axillary line. One or two additional ports for the operating instruments were inserted under direct visual guidance. Intermittent carbon dioxide insufflation was used when necessary. Loculations were drained and adherent lung was separated from the parietal pleura and fissures were defined. The pleural fibrin deposits were removed with endoscopic grasping devices or by means of extensive irrigation and suctioning. Necrotic lung tissue caused by necrotizing pneumonia was carefully manipulated to prevent any postoperative air leakage. At the completion of the procedure, one or two chest tubes were inserted and put on suction (10 mmHg) or under water seal according to the attending surgeon's preference. Both the pleural fluid and fibrinous tissue were sent for microscopy and culture and polymerase chain reaction assay. After the VATS, patients were transferred in the surgical ward or, if indicated, in the pediatric intensive care unit for postoperative care. Chest drains were removed when there was minimal drainage (<20 mL/day). Postoperative intravenous antibiotic therapy was adjusted by both the respiratory and infectious diseases teams according to the culture results. Intravenous antibiotics were discontinued if patients remained afebrile for 48 hours after the chest drains had been removed. Patients were discharged when both afebrile and experienced no respiratory symptoms. Oral antibiotics were started on discharge and continued for 2–6 weeks, according to the cultures from the pleural fluid, until patients were reviewed in the outpatient clinic.

Results are reported as number of cases (%) and median (range) and were analyzed by chi-square test, Fisher's exact test, Mann–Whitney U test, and the Kruskal–Wallis test. Correlation analysis was performed to identify factors related to prolonged LOS.

Results

We identified 159 children admitted with pneumonia complicated with pleural effusion/empyema during the study period; 48 (30%) were transferred from rural hospitals after a median of 3 (1–8) days. Overall, 75/159 (47%) required VATS and are the subject of this study. Patients' demographics are presented in Table 1. Patients requiring VATS were significantly younger than patients successfully treated with medical therapy alone (P = .004). There was no difference in seasonal presentation between the 2 groups (P = .8) or the median duration of symptoms before hospitalization (P = .8). VATS was completed in all patients with no intraoperative complications. A chest drain was inserted in all patients after VATS. Postoperatively, chest drains were placed on 10 cm H₂O suction for 48 hours in 30 (40%) patients according to the attending surgeon's preference.

Table 1.

Patients' Demographics

	VATS (n = 75)	Non-VATS (n = 84)	P
Median age in years	3.6 (0.4–14.5)	5.9 (0.7–16.6)	.004
Male/female	42/33	44/40	.6
Presentation
Autumn	15 (20)	18 (21)
Winter	26 (35)	24 (29)	.8
Spring	18 (24)	25 (30)
Summer	16 (21)	17 (20)
Symptoms duration in days	7 (2–28)	7 (1–21)	.8
Volume of fluid drained in mL	375 (100–1300)	N/A
Drain on suction	30 (40)	N/A
Duration of chest drain in days	3 (2–13)	N/A
Reoperation or readmission	6	None
LOS in days	10 (3–47)	5 (1–27)	<.0001
Postdischarge antibiotic therapy in days	4 (2–6)	3.5 (2–6)	.8

Data reported as number of cases (%) and median (range).

LOS, length of stay; N/A, not applicable; VATS, video-assisted thoracoscopic surgery.

There were no significant perioperative complications or deaths. Overall LOS in the VATS group was significantly longer compared with the non-VATS group (P < .0001), whereas median duration of oral antibiotic therapy after discharge was similar between the 2 groups (P = .08). There was no difference in LOS in the different seasons (P = .6). The analysis of the microbiology data from the pleural fluid and fibrin deposits revealed as follows: no bacteria in 26 (34.5%) patients, Streptococcus pneumoniae in 29 (38.5%), Streptococcus pyogenes (Group A) in 11 (14.5%), Streptococcus anginosus in 1 (1.5%), Staphylococcus aureus in 5 (6.5%), and other species in 3 (4.5%).

Table 2 summarizes the main findings in patients with positive and negative pleural fluid microbiology. Patients with positive microbiology required drains in situ for a prolonged number of days (P = .0004). The CRP on admission was significantly higher in patients with positive microbiology (P = .001) although the WCC was similar between the 2 groups (P = .08). Patients with positive microbiology had longer duration of symptoms before VATS (P = .01). LOS was significantly longer in the group with a positive pleural fluid microbiology culture (P = .009). Patients with S. pneumoniae had the shortest LOS compared with patients with S. pyogenes (Group A) or other species: 7 (4–20) days versus 12.5 (5–28) days versus 19 (5–47) days; P = .005. There was no difference in the preoperative durations of symptoms among the 3 groups: 6 (3–14) days versus 5 (2–14) days versus 7 (3–21) days; P = .4. Six (8%) children required a second operative intervention (5 redo-VATS, 1 thoracotomy, 1 additional chest drain) at a median of 14 (8–34) days: all of these patients had atypical bacteria on microbiology (Table 3) and had a prolonged LOS of 22 (14–48) days.

Table 2.

Microbiology Results from Pleural Fluid

	VATS with positive microbiology	VATS with negative microbiology	P
Days with drain in situ	4 (1–13)	3 (2–11)	.0004
WCC on admission in 10⁹/L	16.6 (5.2–44.3)	20.2 (2.2–53)	.08
CRP on admission in mg/L	277.5 (68–535)	183.0 (81–505)	.001
Symptoms duration in days	8 (3–28)	6 (2–21)	.01
LOS in days	9 (4–47)	6.5 (3–16)	.009

Data reported as median (range).

CRP, c-reactive protein; LOS, length of stay; VATS, video-assisted thoracoscopic surgery; WCC, white cell count.

Table 3.

Characteristic of Patients Requiring Reintervention

Patient	Microbiology	Reoperation/Findings	CRP on admission, mg/L	WCC on admission, 10 ⁹ /L	Symptoms duration	LOS in days
1	Streptococcus pyogenes	Thoracotomy and decortication/necrotizing lung, persisting air leak	287	16.8	6	21
2	Streptococcus pyogenes	VATS and decortication/persisting air leak	323	23.7	5	24
2	Streptococcus anginosus	VATS and decortication/persisting air leak	180	32.4	10	47
4	Staphylococcus aureus	Thoracotomy and decortication/necrotizing lung, persisting air leak	128	3	8	14
3	Prevotella nanceiensis	VATS and decortication/persisting air leak	162	14.9	14	20
6	Coagulase negative Staphylococcus	VATS and decortication/persisting air leak	333	17	7	48

CRP, c-reactive protein; LOS, length of stay; VATS, video-assisted thoracoscopic surgery; WCC, white cell count.

There was no correlation between the LOS and either the age of the patients (r = 0.08 [95% CI −0.14 to 0.3]), the duration of preoperative symptoms (r = −0.03 [95% CI −0.3 to 0.2]; P = .7), the CRP (r = 0.1 [95% CI −0.1 to 0.3]; P = .3) the WCC (r = 0.02 [95% CI −0.2 to 0.2]; P = .8) on admission, the size of the empyema (r = 0.2 [95% CI −0.07 to 0.5]; P = .1), or the size of the chest drain (r = 0.09 [95% CI −0.14 to 0.3]; P = .4).

Discussion

In this study we have reported on a large cohort of children with pneumonia and empyema managed with VATS at a single institution for a 5-year period. Our findings suggest that the procedure results in discharge at a median of 8 days postoperatively, with positive microbiology negatively affecting the LOS. In our series, 6/75 (8%) required surgical reintervention, which is comparable with previous studies in the literature, with no perioperative morbidity or mortality.^7,8

VATS for the management of empyema in children was originally described in 1993 by Kern and Rodgers.⁹ The procedure has gained popularity over the years and it has been suggested that VATS might lead to a prompt discharge from hospital.⁷ However, the choice of VATS over percutaneous aspiration and/or chest drain insertion (with or without intrapleural fibrinolytic agents) is often subjective and dependent on the local practice; this is largely determined by availability of medical personnel and their individual expertise. At our institution particularly, VATS has been the preferred procedure due to the local expertise with minimally invasive surgery and given the large distance that patients have to travel, many presenting with an advanced disease after several days of illness and antibiotic treatment.

Our recent systematic review and meta-analysis documented that, although the use of chest drain and fibrinolytic agents is effective in >75% of cases for the surgical management of empyema in children, VATS might be associated with reduced need for reintervention and shorter postoperative hospital stay without additional morbidity.⁷

Nonetheless, we found some heterogeneity among the studies regarding the LOS after intervention. A number of factors, that have not been fully investigated, could affect the LOS such as the timing of surgery/duration of preoperative symptoms and the causative pathogens. Arguably, the duration of preoperative symptoms and the timing of surgical intervention might play a significant role for the success of the surgical procedure and postoperative recovery. It has been documented that, although the use of chest drain with fibrinolytics might be effective in the early stages of the disease, it is only effective in 60% of cases with advanced disease, and children with advanced disease require an increased number of repeated procedures and have a prolonged hospital stay.^8,10,11 The European Association for Cardio Thoracic Surgery has also suggested that chest drain with fibrinolytic agents should be considered as first line of intervention in patients with a fibro-purulent pleural empyema; however, formal surgical intervention should be preferred in all patients with advanced-stage pleural empyema (Stage 2 or 3 American Thoracic Society classification), if they are able to undergo surgery.¹² To the best of our knowledge, there are no studies in children comparing the LOS after chest drain with fibrinolysis versus VATS at different disease stages. Furthermore, there are very few pediatric studies that have analyzed factors predicting the morbidity in children with empyema. We did not find a correlation between the LOS and the duration of preoperative symptoms, but this could be related to the fact that our population of children undergoing VATS presented with an advanced disease (on average symptomatic for 1 week). The main finding of our study is that patients with a positive pleural fluid microbiology culture had a significant higher CRP on admission, required chest drains in situ for a prolonged number of days, and had a significantly longer LOS. This is despite them having a similar duration of preoperative symptoms when compared with patients with a negative microbiology culture. Furthermore, patients with S. pneumoniae had a shorter LOS compared with patients with S. pyogenes or other species. We suggest that patients with “non-S. pneumoniae” infection might have a more aggressive disease and indeed, the 6 patients who required reintervention had atypical bacteria identified from the pleural fluid, supporting the finding that patients with “non-S. pneumoniae” disease might have a complicated clinical course negatively affecting the LOS. Our hypothesis is further substantiated by epidemiological studies that have documented an increase in number of hospitalization and death rates for “non-S. pneumoniae” infections compared with S. pneumoniae infection.¹³

Results from studies in adults have also documented a negative outcome associated with polymicrobial infections and positive bacterial cultures.^14,15

Furthermore, it has been documented that children with empyema who have positive blood and pleural fluid microbiology cultures have a longer duration of fever after admission to hospital. This includes a longer total number of days of fever regardless of the presence of a pneumococcus infection.¹⁶

We acknowledge that clinician's decision to prolong the intravenous therapy in patients with positive microbiology might have introduced selection bias affecting the LOS. However, the indications for removing the chest drains were the same in all our patients.

It has also been documented that factors that are associated with worse prognosis include a high CRP, low pH, low sugar, and high lactate dehydrogenase levels in pleural fluid.¹⁷

In our series, although the WCC was similar between patients with positive and negative microbiology, the CRP was significantly higher in the group with a positive microbiology, possibly indicating that this group might have had a more severe disease on admission. The role of WCC count and CRP for predicting disease course and clinical outcomes in patients with pneumonia is controversial. However, other authors have confirmed our findings that in children with pneumonia, CRP measured within 24 hours of hospital admission was associated with prolonged LOS and duration of fever, whereas WCC was not independently associated with either outcome.¹⁸ It will be important to investigate these findings in future prospective studies.

Although we could effectively confirm that once patients were admitted to hospital they were commenced on intravenous antibiotics, due to the retrospective nature of the study, we were not able to collect data regarding the duration of preoperative oral antibiotic therapy. This could also affect the severity of the disease, the presence of positive microbiology, and, therefore, the LOS.

Our data show that younger patients are more likely to require surgical intervention. This finding is not surprising considering the worldwide significant increase in hospital admissions for complicated pneumonia and empyema in children <4–5 years after the introduction of the PCV7.⁶ However, we did not identify age a significant factor affecting the LOS.

In conclusion, in our experience, 92% of children with empyema requiring surgical intervention will be treated with a single VATS with an average LOS of 8 days.

We identified positive microbiology affecting the LOS. Patients with S. pneumoniae have a shorter LOS compared with other species. This information should be used to counseling families regarding the postoperative course of children with complicated empyema.

Footnotes

Disclosure Statement

No competing financial interests exist.

References

Jaffe

, Balfour-Lynn

. Management of empyema in children. Pediatr Pulmonol, 2005; 40:148–156.

Byington

, Spencer

, Johnson

, Pavia

, Allen

, Mason

, et al. An epidemiological investigation of a sustained high rate of pediatric parapneumonic empyema: Risk factors and microbiological associations. Clin Infect Dis, 2002; 34:434–440.

Fletcher

, Schmitt

, Syrochkina

, Sylvester

. Pneumococcal empyema and complicated pneumonias: Global trends in incidence, prevalence, and serotype epidemiology. Eur J Clin Microbiol Infect Dis, 2014; 33:879–910.

Gautam

, Wiseman

, Goodman

, Ahmedpour

, Lindsay

, Heyer

, et al. Paediatric thoracic empyema in the tropical North Queensland region of Australia: Epidemiological trends over a decade. J Paediatr Child Health, 2018; 54:735–740.

Grijalva

, Nuorti

, Zhu

, Griffin

. Increasing incidence of empyema complicating childhood community-acquired pneumonia in the United States. Clin Infect Dis, 2010; 50:805–813.

Strachan

, Snelling

, Jaffe

. Increased paediatric hospitalizations for empyema in Australia after introduction of the 7-valent pneumococcal conjugate vaccine. Bull World Health Organ, 2013; 91:167–173.

Pacilli

, Nataraja

. Management of paediatric empyema by video-assisted thoracoscopic surgery (VATS) versus chest drain with fibrinolysis: Systematic review and meta-analysis. Paediatr Respir Rev, 2019; 30:42–48.

Espinosa

, Fallat

, Woods

, Weakley

, Marshall

. An approach to the management of pleural empyema with early video-assisted thoracoscopic surgery and early transition to oral antibiotic therapy. Am Surg, 2016; 82:295–301.

Kern

, Rodgers

. Thoracoscopy in the management of empyema in children. J Pediatr Surg, 1993; 28:1128–1132.

10.

Chiu

, Wong

, Huang

, Lai

, Lin

. Echo-guided management of complicated parapneumonic effusion in children. Pediatr Pulmonol, 2006; 41:1226–1232.

11.

James

, Braswell

, Pezeshkmehr

, Roberson

, Parks

, Moore

. Stratifying fibrinolytic dosing in pediatric parapneumonic effusion based on ultrasound grade correlation. Pediatr Radiol, 2017; 47:89–95.

12.

Scarci

, Abah

, Solli

, Page

, Waller

, van Schil

, et al. EACTS expert consensus statement for surgical management of pleural empyema. Eur J Cardiothorac Surg, 2015; 48:642–653.

13.

Wuerth

, Bonnewell

, Wiemken

, Arnold

. Trends in pneumonia mortality rates and hospitalizations by organism, United States, 2002-2011(1). Emerg Infect Dis, 2016; 22:1624–1627.

14.

Chen

, Chen

, Lin

, Tseng

, Kuo

, Huang

, et al. Acute thoracic empyema: Clinical characteristics and outcome analysis of video-assisted thoracoscopic surgery. J Formos Med Assoc, 2014; 113:210–218.

15.

Lindstrom

, Kolbe

. Community acquired parapneumonic thoracic empyema: Predictors of outcome. Respirology, 1999; 4:173–179.

16.

Picard

, Joseph

, Goldberg

, Mimouni

, Deeb

, Kleid

, et al. Predictive factors of morbidity in childhood parapneumonic effusion-associated pneumonia: A retrospective study. Pediatr Infect Dis J, 2010; 29:840–843.

17.

Soares

, Barreira

, Pissarra

, Nunes

, Azevedo

, Vaz

. Pediatric parapneumonic pleural effusions: Experience in a university central hospital. Rev Port Pneumol, 2009; 15:241–259.

18.

Williams

, Hall

, Auger

, Tieder

, Jerardi

, Queen

, et al. Association of white blood cell count and c-reactive protein with outcomes in children hospitalized with community-acquired pneumonia. Pediatr Infect Dis J, 2015; 34:792–793.