Surgical complications and procedures in neonates on extracorporeal membrane oxygenation

Post-ECMO chest tube placement: A propensity score-matched survival analysis

BACKGROUND

Severe morbidity and mortality has been reported from chest tube (CT) placement during pediatric extracorporeal membrane oxygenation (ECMO).

METHODS

Kids' Inpatient Database (KID) was analyzed for ECMO with CT placed <8days postcannulation (1997-2009).

RESULTS

Overall, 5884 patients were identified (213 CT) (56% male, 49% white), with a median (IQR) age at ECMO cannulation 7 (117)days, length of stay (LOS) 26 (35)days, and total charges (TC) 342,116 (409,573) USD. Diagnoses included congenital diaphragmatic hernia (CDH) 16%, meconium aspiration (MA) 2%, pulmonary hypertension (PH) 13%, respiratory distress syndrome (RDS) 41%, and cardiac (C) 29%. Survival was overall 57%, CDH 47%, MA 88%, PH 75%, RDS 57%, and C 52%. There were no differences in survival between CT and non-CT patients compared overall, or by diagnosis, or by age <30days, or by diagnosis and age <30days. Multivariate analysis and propensity score matching for all ages, or <30days of age by diagnosis showed no difference in survival between CT and non-CT patients.

CONCLUSION

Analysis of KID with correlative propensity score matching demonstrates no increased mortality in pediatric ECMO patients requiring CT placement.

Chest tube placement in children during extracorporeal membrane oxygenation (ECMO)

BACKGROUND

Pleural collections of air and fluid are frequent in infants and children treated with extracorporeal membrane oxygenation (ECMO). In this anticoagulated population, chest tube placement is potentially hazardous, and catastrophic hemorrhage has been reported. We sought to define the risks associated with chest tube placement in a large population of children managed with ECMO.

METHODS

The records of 189 consecutive children managed with ECMO at two children's hospitals were reviewed. Demographics, indications for ECMO, and ECMO courses were reviewed. In particular, the occurrence of pleural collections and the frequency and technique of chest tube placement were evaluated. The incidence of complications and mortality were determined.

RESULTS

The median age of the subjects was 2days. The overall mortality was 26.5%. A pneumothorax was found in 19 (10.1%), a pleural effusion in 26 (13.8%), and a hemothorax in 2 (1.0%). A chest tube was placed in 27 (19 by a needle-guide wire technique and 8 by cut-down). Major bleeding complications occurred in 6 subjects (22%).

CONCLUSIONS

There was a significant incidence of major bleeding complications and death in subjects in whom chest tubes were placed. The placement of a chest tube during ECMO should be done only if it is likely to improve pump flow or promote weaning of support.

Neurological complications of extracorporeal membrane oxygenation in children

OBJECT

Extracorporeal membrane oxygenation (ECMO) is a potentially life-saving treatment for patients in refractory cardiorespiratory failure. Neurological complications that result from ECMO treatment are known to significantly impact patient survival and quality of life. The purpose of this study was to review the incidence of neurological complications of ECMO in the pediatric population and the role of neurosurgery in the treatment of these patients.

METHODS

Data were obtained from the national Extracorporeal Life Support Organization (ELSO) Registry for the years 1990 to 2009. The neurological complications recorded by the registry include CNS hemorrhage, CNS infarction, and seizure. The ECMO Registry at the authors' institution was then searched, and 3 pediatric patients who had undergone craniotomy during ECMO treatment were identified.

RESULTS

Children in the ELSO Registry who were treated with ECMO survived to hospital discharge in 65% of cases. Intracranial hemorrhage occurred in 7.4% of the ECMO-treated patients, with 36% of those surviving to hospital discharge. Hemorrhage was more likely in patients younger than 30 days old and in those requiring ECMO for cardiac indications. Cerebral infarction occurred in 5.7% of all ECMO-treated patients. Clinically diagnosed seizures occurred in 8.4% of all ECMO-treated patients. The ECMO Registry at the authors' institution revealed that 1898 patients were treated there. Intracranial hemorrhage was diagnosed in 81 patients (5.8%), and 3 of these patients were treated with craniotomy. Two of the patients were alive with minimal neurological impairment and normal school performance at 10 and 16 years of follow-up.

CONCLUSIONS

Intracranial hemorrhage is a serious complication of ECMO treatment. While the surgical risk is substantial, there may be a role for surgical evacuation of hemorrhage in well-selected patients.

Congenital Diaphragmatic Hernia Repair on Extracorporeal Life Support: A Decade of Lessons Learned

Congenital diaphragmatic hernia (CDH) is a vexing anomaly that manifests with variable pulmonary compromise in neonates. More than one-third of neonates with CDH require extracorporeal membrane oxygenation (ECMO) for refractory pulmonary hypertension (PHN). To assess the outcome of neonates having CDH repair on ECMO, we reviewed our experience for babies treated between 1992 and 2003. Of 97 neonates with CDH, 40 required ECMO, and 30 were repaired on bypass. Eighteen were supported by veno-venous bypass (VV) and 12 by veno-arterial bypass (VA). While on ECMO, transfusion requirements increased twofold postoperatively (15 to 33 cc · kg-1 · day-1, P = 0.03) and then significantly decreased after decannulation (1.5 cc · kg-1 · day-1, P < 0.01). Non-intracranial hemorrhage occurred in 7 (23%) infants and intracranial hemorrhage in 3 (10%). Twelve (40%) infants died; one (3%) on ECMO secondary to refractory PHN. The mean length of stay for the 18 (60%) survivors was 48 days. Comparisons between survivors and nonsurvivors showed a significantly increased mortality for infants placed on VA bypass ( P < 0.01). However, no other variable was predictive of survival. We conclude that CDH repair on ECMO is technically feasible, shows similar survival to the Extracorporeal Life Support Organization (ELSO) registry, and is associated with few bleeding complications.

Intraoperative ultrasound reduces ECMO catheter malposition requiring surgical correction

BACKGROUND/PURPOSE

One hundred ninety-three cannulation procedures for extracorporeal membrane oxygenation (ECMO) have been performed at the authors' institution from 1994 to now. Before 1996, their practice had been to position these catheters exclusively by clinical assessment and chest radiograph. Since then, the authors have utilized intraoperative ultrasound guidance during cannulation procedures to confirm proper tip position. This retrospective analysis was undertaken to establish whether this practice has reduced the rate of surgical repositioning of ECMO catheters in these patients.

METHODS

A retrospective chart review was performed for all infants who underwent ECMO cannulation procedures at the authors' institution. Numbers of infants requiring surgery to readjust ECMO catheter position were totaled. Cases were categorized according to the presence or absence of intraoperative ultrasound scan. Statistical significance was determined using X(2) analysis, Student's t test, or analysis of variance where appropriate.

RESULTS

There were 193 ECMO cannulations performed. Of the 101 procedures done without ultrasound scan, 18 necessitated surgical repositioning. In contrast, only 3 of the 92 catheters placed with ultrasound assistance required reoperation. This represents a reduction the rate of repositioning from 17.8% to 3.3% of cannulations (P =.003).

CONCLUSIONS

Based on these findings, the authors advocate the use of intraoperative ultrasound imaging to optimize the position of ECMO catheters. This high rate of initial success helps avoid the potential morbidity of ECMO circuit malfunction, repeat neck dissection, and catheter manipulation in these critically ill, anticoagulated patients.

Risk factors for intracranial hemorrhage in adults on extracorporeal membrane oxygenation

OBJECTIVE

Intracranial hemorrhage is a recognized complication in neonates and infants on extracorporeal membrane oxygenator support and various risk factors associated with this have been defined. The prevalence and risk factors associated with intracranial hemorrhage in adults on extracorporeal membrane oxygenator support are unknown and this study was performed to define these factors.

METHODS

A retrospective study of adults supported with extracorporeal membrane oxygenators at a single institution between January 1992 and December 1996 was performed. Age, gender, weight, body surface area, renal function, anticoagulation, coagulation variables, blood flow, arterial pressure, arterial cannulation sites, duration of support, extracranial bleeding, native cardiac function and presence of intracranial microemboli were analyzed to determine the risk factors for intracranial hemorrhage.

RESULTS

Fourteen out of 74 adults on extracorporeal membrane oxygenator support had intracranial hemorrhage (18.9%). An increased risk of intracranial hemorrhage showed a positive correlation with female gender (P = 0.02, odds ratio 6.5), use of heparin (P = 0.05, odds ratio 8.5), creatinine greater than 2.6 mg/ dl (P = 0.009, odds ratio 6.5), need for dialysis (P = 0.03, odds ratio 4.3) and thrombocytopenia (P = 0.007, odds ratio 18.3). Diminishing renal function and the need for dialysis were associated with increasing duration of support. Multivariable logistic regression showed female gender and thrombocytopenia, especially with platelet counts less than 50000 cells/mm3 to be the most important predictors of intracranial hemorrhage. Intracranial hemorrhage was associated with a mortality of 92.3% compared with a mortality of 61% in those without intracranial hemorrhage (P = 0.027).

CONCLUSION

Intracranial hemorrhage is a significant complication in adults on extracorporeal membrane oxygenator support. Judicious management of anticoagulation, prevention of renal failure and aggressive correction of thrombocytopenia may help to lower the risk of intracranial hemorrhage in adults on extracorporeal membrane oxygenator support.

Cervical ECMO cannula placement in infants and children: recommendations for assessment of adequate positioning and function

BACKGROUND/PURPOSE

Cervical extracorporeal membrane oxygenation (ECMO) cannula position is often difficult to confirm by chest x-ray alone. Malposition requires a second surgery to rectify the problem. Reoperation places the patient at risk for infection, bleeding, or death. This study analyzes indications for cannula repositioning and suggests an alternative standard for intraoperative evaluation of catheter function as it relates to position.

METHODS

The authors reviewed charts of 73 patients placed on arterio-venous ECMO through cervical vascular access. Reasons for repositioning of either cannula at the initial surgery or postoperatively were recorded.

RESULTS

Of 73 patients, 18 (24.6%) required either arterial cannula or venous cannula repositioning. In 10 (55%) of these patients, cannula malposition was not detected by chest x-ray during the initial cannulation, and they therefore required a second cervical exploration for repositioning.

CONCLUSIONS

Chest x-ray is not sensitive in demonstrating malpositioned cervical ECMO cannulae. Two-dimensional ECHO before wound closure, may be a superior, more cost effective means of assessing cannula placement and function than x-ray alone. Confirmation of cannula position and function, before wound closure, would reduce the risks involved with cervical reexploration.

Open lung biopsy in pediatric patients on extracorporeal membrane oxygenation

Open lung biopsy has proven beneficial in the treatment of life-threatening pulmonary diseases. Its safety and efficacy in infants and children placed on extracorporeal membrane oxygenation (ECMO) for severe respiratory failure is not known. The authors reviewed eight cases (4 neonates, 3 infants, 1 child) who underwent open lung biopsy while on ECMO. The primary diagnoses were pneumonia (4), lymphoma (1), primary pulmonary hypertension (1), and complex congenital heart disease (2). The patients underwent biopsy after they had been on ECMO an average of 9.6 days (range, 1 to 14 days). Biopsy results confirmed the clinical diagnosis in five patients, two of whom had irreversible alveolar destruction resulting in ECMO withdrawal. Three patients had pathological diagnoses, which resulted in major therapy revisions (1 fungal infection and 2 noninfectious lesions that required steroid treatment). The overall average duration of ECMO treatment was 16.3 days (range, 10 to 24 days). Three patients were weaned successfully from ECMO, but only one infant survived to discharge. One nonlethal bleeding complication occurred after biopsy. Open lung biopsy is well tolerated during ECMO. It accurately determines pulmonary pathology and provides valuable prognostic information. Earlier biopsy for patients whose diagnoses are uncertain or who are not responding to ECMO may improve the mortality rate for this high-risk group.

Pulmonary hemorrhage: a novel complication after extracorporeal life support

Pulmonary hemorrhage (PH) occurs infrequently as a complication in neonates with respiratory failure. Major PH has been observed at the authors' institution in several neonates after "successful" completion of extracorporeal life support (ECLS) therapy. The authors sought to determine the incidence of PH and the risk factors associated with this unique and newly described morbidity after ECLS. The hospital records of all patients who had PH after ECLS were reviewed. The control patients were the first three infants who underwent ECLS just before each PH case. PH was defined as the occurrence of bloody tracheal secretions associated with a deterioration in pulmonary status. Demographics, ventilator/ECLS parameters, fluid management, coagulation, and laboratory studies were evaluated in the pre-ECLS, during ECLS, and in the post-ECLS period. From 1985 to 1993, 13 (6%) of 214 neonates suffered major PH, at a mean time of 43.2 +/- 9.2 hours after the ECLS course. The overall mortality rate for children with PH was 38%, compared with 5% among the control patients. In the pre-ECLS phase, patients with PH required more fluid (153.6 +/- 20.2 mL/kg/d v 106.8 +/- 10.2 mL/kg/d) and were acidemic for a longer period (2.3 +/- 1.2 hours v 0.6 +/- 0.2 hours; pH < 7.25). No differences were noted in AaDo2 or oxygenation index criteria. During ECLS, inotropes were required more often (23% v 0%; P < .01) because hypotension was more common (77% v 33%; P < .05). Activated clotting times (ACT) and heparin requirements were equivalent for the two groups. After ECLS the patients with PH required longer ventilatory assistance (184.9 +/- 48.2 hours v 83.4 +/- 16.7 hours) and supplemental oxygen (24.3 +/- 3.0 days v 17.2 +/- 1.9 days). No coagulation abnormalities were identified at the time of PH. Higher SGPT (185.4 +/- 146.4 U/L v 22.6 +/- 3.5 U/L; P < .05) and BUN (69.3 +/- 7.5 mg/dL v 47.2 +/- 5.9 mg/dL; P < .05) also were noted for the patients with PH. PH represents an important and novel morbidity in neonates after ECLS. Prolonged acidosis, a high fluid requirement before ECLS, the need for blood pressure support during ECLS, and evidence of renal and/or hepatic dysfunction serve to identify patients who have a high risk for the development of this complication.

Hemorrhagic complications and repair of congenital diaphragmatic hernias: does timing of the repair make a difference? Data from the Extracorporeal Life Support Organization

The authors reviewed the Extracorporeal Life Support Organization (ELSO) data base of all neonates placed on extracorporeal membrane oxygenation for whom CDH was diagnosed between January 1989 and December 1991. For 483 neonates, there were complete data concerning timing of the hernia repair in relation to ECMO. The overall incidence of hemorrhage was 43% (57% among nonsurvivors, 32% among survivors; P < .05). The most common bleeding sites were surgical repair site (24%), head (11.5%), cannulation site (7.5%), and gastrointestinal (5%). Fatal hemorrhage occurred in 4.8% (23 of 483). The most common sites of fatal hemorrhage were head (48%), pulmonary (17%), and abdominal (17%). Bleeding complications were significantly greater for patients repaired on ECMO (58%) versus those repaired before (37%) or after (21%) (P < .05). Surgical-site hemorrhage requiring transfusion occurred in 38% of those repaired on ECMO versus 18% and 6% of those repaired before and after, respectively (P < .05). Gastrointestinal and "other" sites of hemorrhage were significantly more common in those repaired on bypass. The number of patients repaired on ECMO increased from 22% to 48% over the 3 years (P < .05). The incidence of hemorrhagic complications did not differ significantly among the 3 years (P > .05). Repair of the hernia defect while on bypass was associated with significantly greater bleeding complications. These data should be useful in the planning of future prospective trials.

Complications of neonatal extracorporeal membrane oxygenation

From 1973-1985 to 1988 the average patient complications per case were 1.44 per case and significantly increased during 1990 to 1992 to 2.10 per case (Figure 3). During the same periods patient survival significantly decreased from 84% (1973-1985 to 1988, n = 2463) to 80% (1990 to 1992, n = 4005) (Figure 4). The association between total complication rates and survival rate was examined by regression analysis (Table 5). The correlation of patient complication rate and total complication rate with survival is highly significant; however, causality cannot be established. When comparing different entry criteria (Table 2) for incidence of mechanical and patient complications, no significant differences are apparent. This is not surprising since each of the entry criteria were designed to identify the same patient population. When premature neonates (> 35 weeks) were placed on ECMO, 36% of them had intracranial haemorrhage (ICH) with 62% mortality while only 12% of the neonates < 35 weeks had ICH and a 49% mortality. Similar findings were noted with low birthweight neonates (< 2.2 kg), 28% had ICH with 64% mortality while only 12% of the neonates > 2.2 kg had ICH with a 50% mortality. Selection criteria remain problematic for a variety of reasons. They cannot be viewed as absolute because of variability between centres. What represents likely 80% mortality in one centre may not apply to another. Historical controls are misleading because changing respiratory therapy strategies make historical populations difficult to compare. Also, once an ECMO centre becomes established, a more challenging group of patients will be attracted than previously was the case. Further, a single entry criterion cannot be generalized for all entry diagnoses. Criteria for an 80% predicted mortality is probably not the same for MAS, CHN, PPHN, and sepsis. Subsequent patients registered in the Neonatal ECMO Registry of the Extracorporeal Life Support Organization will address these issues more thoroughly, as specific details of the pre-ECMO condition and therapeutic strategies are collected. This collective review should help to identify trends which require reassessment of technique or patient management.

Extracorporeal membrane oxygenation

Extracorporeal membrane oxygenation has now evolved into standard therapy for patients unresponsive to conventional ventilatory and pharmacological support. This article presents a clinical review of extracorporeal life support and its application to neonatal and pediatric patients as well as children requiring circulatory support after open heart surgery.