Effect of very delayed repair of congenital diaphragmatic hernia on survival and extracorporeal life support use.

BACKGROUND: Since November 1992, operative repair in neonates with congenital diaphragmatic hernia (CDH) at this institution was delayed until respiratory insufficiency had resolved. METHODS: A retrospective analysis was performed (n = 33) comparing delayed repair with our previously reported institutional experience with immediate repair from January 1988 to October 1992 (n = 66). Infants with severe genetic defects or moribund conditions or who were premature were not considered candidates for repair or extracorporeal life support (ECLS), but they were included in the survival analysis. Survival was defined as hospital discharge. Data were compared with an independent t test or Pearson chi-squared test. RESULTS: Mean age at repair was 8.9 +/- 4.5 days (range, 3 to 20 days). Eleven infants in the study group were placed on ECLS (33% versus 68% in the comparison group; p = 0.001). Six of these infants survived (55% versus 58% in the comparison group; p = 0.846). Of these survivors, one patient was repaired while on ECLS, and the remainder underwent repair after decannulation from ECLS. All 20 of the remaining candidates for repair survived without need for ECLS. Overall survival was 79% versus 56% in the comparison group (p = 0.027). CONCLUSIONS: Our current data suggest that very delayed repair of newborns with CDHs is associated with an increase in the overall survival and a decrease in the use of ECLS when compared with previous experience at this institution.

Congenital diaphragmatic hernia survival and use of extracorporeal life support at selected level III nurseries with multimodality support.

BACKGROUND: Congenital diaphragmatic hernia (CDH) has been cited to have a mortality rate of 50%. There have been multiple studies at individual institutions demonstrating potential benefits from various strategies including extracorporeal life support (ECLS), delayed repair, and lower levels of ventilator support. There has been no multicenter survey of institutions offering these modalities to describe the current use of ECLS and survival of these infants. In addition, the relationship between the number of patients with CDH managed at an individual institution and outcome has not been evaluated. METHODS: We queried 16 level III neonatal intensive care centers on the use of ECLS and survival of infants with CDH who were treated during 2 consecutive years (1993 to 1995). Data are presented as mean +/- SEM, median, and range. RESULTS: Data were collected on 411 patients. Of these, 71% +/- 8% were outborn and 8% +/- 3% were considered nonviable. Overall survival of CDH infants was 69% +/- 4% (range, 39% to 95%). The survival rate of infants on ECLS was 55% +/- 4%, whereas survival of infants not requiring ECLS was significantly increased at 81% +/- 5% (p = 0.005). The mean rate of ECLS use was 46% +/- 2%. There was no correlation between the number of cases per year at an individual institution and overall survival, ECLS survival, or ECLS use (r = 0.341, 0.305, and 0.287, respectively). There was also no correlation between case volume at an individual institution and ECLS survival (r = 0.271). CONCLUSIONS: The current survival rate and rate of ECLS use in infants with CDH at level III neonatal intensive care units in the United States are 69% +/- 4% and 46% +/- 2%, respectively. There is no correlation between the yearly individual center experience with managing CDH and rate of ECLS use or outcome.

Protective effect of hypothermia and left heart bypass on spinal ischemia in the dog.

OBJECTIVE: To test the hypothesis that systemic hypothermia (SH) to 30 degrees C in combination with partial left heart bypass (PLHB) at either a high or low distal arterial perfusion pressure (DAPP) following 45 minutes of cross-clamp (XC) occlusion of the thoracic aorta will protect against clinical and histological spinal cord ischemia in the dog. DESIGN: A blinded, prospective, randomized, and controlled experimental trial. SETTING: Tertiary care center animal laboratory. PARTICIPANTS: Seventeen adult mongrel dogs. INTERVENTIONS: The animals were randomized into 5 groups: control group 1: XC plus no protection (n = 3); control group 2; XC plus systemic normothermia plus PLHB, with a DAPP less than 20 mm Hg (n = 3); treatment group 1: XC plus systemic normothermia plus PLHB, with a DAPP greater than 20 mm Hg (n = 3); treatment group 2: XC plus SH plus PLHB, with a DAPP greater than 20 mm Hg (n = 3); treatment group 3: XC plus SH plus PLHB, with a DAPP less than 20 mm Hg (n = 5). MAIN OUTCOME MEASURES: Clinical and histological neurological injury evaluation by separate blinded observers. RESULTS: Control animals were neurologically and histologically ischemic. Treatment animals were neurologically and histologically normal. Partial left heart bypass with a DAPP greater than 20 mm Hg prevented paraplegia, with either systemic normothermia or SH. Systemic hypothermia plus PLHB, even with a DAPP less than 20 mm Hg, protected against spinal cord ischemia during thoracic aortic occlusion. CONCLUSION: Systemic hypothermia to 30 degrees C combined with PLHB at either a high or low DAPP prevented spinal cord ischemia following thoracic aortic XC occlusion in our canine model and merits clinical trial in patients.

Percutaneous access for venovenous extracorporeal life support in neonates.

BACKGROUND: Although percutaneous cannulation has been previously described in adult and pediatric patients older than 3 years, its use in neonates for venovenous extracorporeal life support (ECLS) has not been previously described. METHODS: Twenty neonates of weight 3.4 +/- 0.6 kg (range, 2.3 to 4.9 kg.) with severe respiratory failure (meconium aspiration syndrome, persistent pulmonary hypertension of the newborn, sepsis) were managed with double-lumen venovenous ECLS. Percutaneous access via the right internal jugular (RIJ) vein with a 12F (n = 13) or 15F (n = 7) double-lumen ECLS cannula was obtained via a modified Seldinger technique. RIJ access was specifically obtained 2 to 3 cm above the clavicle with a 21-gauge needle and a 0.018-in guide wire followed by a technique that allowed insertion of the larger cannula guide wire. Decannulation simply involved removal of the cannula with hemostasis obtained by direct pressure. RESULTS: Percutaneous cannulation was performed without difficulty in 11 infants. Conversion to an open technique for cannula placement was required in four patients early in our experience because percutaneous access to the RIJ with the 0.018-inch guide wire could not be achieved, and in one neonate because the 15F cannula could not be advanced into the RIJ and resulted in laceration of the vein. Inability to achieve RIJ access also led to use of an "exposure-assisted" percutaneous procedure in four patients later in our experience. Complications included a pneumothorax identified on chest radiograph 18 hours after cannulation in one infant and partial thrombotic occlusion of the cannula requiring a change over a guide wire in another. There were no problems associated with decannulation. CONCLUSIONS: This is the first description of percutaneous cannulation in neonates for venovenous ECLS. In the authors' early experience, percutaneous access using a 12F double lumen venovenous cannula may be efficiently performed especially in patients >3.0 kg in weight and has the potential for simplifying and reducing the cost of the ECLS technique.

Use of the placenta as an artificial lung.

Oxygen transfer across a single cotyledon of the human placenta was assessed by using three different perfusates in the maternal circuit: 1) M-199 culture medium, 2) human adult red blood cells (RBCs), and 3) perflubron. These maternal circuit perfusates were oxygenated with a membrane oxygenator. RBCs were perfused on the fetal side of the circuit and samples were taken preplacenta and postplacenta for each maternal perfusate. PO2 and PCO2 were measured and O2 transfer was calculated for each maternal perfusate. O2 transfer per single cotyledon (mean +/- SE) was 0.18 +/- 0.04, 0.20 +/- 0.03, and 0.15 +/- 0.05 ccO2/min when using: 1) M-199, 2) RBCs, and 3) perflubron, respectively. O2 transfer per kilogram of placental tissue was 13.08 +/- 2.78, 14.57 +/- 2.05, and 10.43 +/- 3.79 ccO2/kg per minute when using: 1) M-199, 2) RBCs, and 3) perflubron, respectively. When extrapolated to the individual weights of the entire placenta, the O2 transfer was 9.15 +/- 1.95, 10.20 +/- 1.43, and 7.30 +/- 2.65 when using: 1) M-199, 2) RBCs, and 3) perflubron, respectively. We conclude from these data that O2 transfer can be accomplished during placental perfusion. Larger studies are required to differentiate efficacy among the three maternal circuit perfusates.

Characteristics of an albumin dialysate hemodiafiltration system for the clearance of unconjugated bilirubin.

Extraction of protein bound liver failure toxins, such as unconjugated bilirubin, short chain fatty acids, and aromatic amino acids has been reported using hemodiafiltration with albumin in the dialysate, but the characteristics of such a system have not been described. Therefore, bilirubin clearance using albumin dialysate hemodiafiltration was evaluated in the setting of different dialysate albumin concentrations, varying temperature and pH. An in vitro continuous hemodiafiltration circuit was used with single pass countercurrent dialysis. Unconjugated bilirubin was added to bovine blood and filtered across a polyalkyl sulfone (PAS) hemofilter using matched filtration and dialysate flow rates. The serial bilirubin content was measured and first order clearance kinetics verified. The clearance rate constants were calculated for three dialysate groups of different albumin concentration at constant temperature and pH (group 1: 10 g/dl albumin, n = 5; 2 g/dl albumin, n = 5; normal saline, n = 5), and three groups of different temperature and pH at constant albumin dialysate concentration (group 2: pH = 7.0, temperature = 20 degrees C, n = 5; pH = 7.5, temperature = 20 degrees C, n = 5; pH = 7.0, temperature = 40 degrees C, n = 5). Comparisons were made with ANOVA and Tukey post hoc analysis. When albumin was used in the dialysate, the 2 g/dl group cleared bilirubin 3.1 times faster than saline alone (p = 0.001), and the 10 g/dl group was superior to both (p = 0.001). There were no measurable differences between the 2 g/dl groups at the various temperatures tested (p = 0.08), but the clearance was less at a pH of 7.5 (p = 0.015). The clearance of unconjugated bilirubin is greatly enhanced with the use of albumin containing dialysates when compared to traditional crystalloid hemodiafiltration, is greater at lower pH, and seems to be unaffected by temperature.

Effect of rate and inspiratory flow on ventilator-induced lung injury.

BACKGROUND: We examined the effects of decreasing respiratory rate (RR) at variable inspiratory times (It) and reducing inspiratory flow on the development of ventilator-induced lung injury. METHODS: Forty sheep weighing 24.6+/-3.2 kg were ventilated for 6 hours with one of five strategies (FIO2 = 1.0, positive end-expiratory pressure = 5 cm H2O): (1) pressure-controlled ventilation (PCV), RR = 15 breaths/min, peak inspiratory pressure (PIP) = 25 cm H2O, n = 8; (2) PCV, RR = 15 breaths/min, PIP = 50 cm H2O, n = 8; (3) PCV, RR = 5 breaths/min, PIP = 50 cm H2O, It = 6 seconds, n = 8; (4) PCV, RR = 5 breaths/min, PIP = 50 cm H2O, It = 2 seconds, n = 8; and (5) limited inspiratory flow volume-controlled ventilation, RR = 5 breaths/min, pressure-limit = 50 cm H2O, flow = 15 L/min, n = 8. RESULTS: Decreasing RR at conventional flows did not reduce injury. However, limiting inspiratory flow rate (LIFR) maintained compliance and resulted in lower Qs/Qt (HiPIP = 38+/-18%, LIFR = 19+/-6%, p < 0.001), reduced histologic injury (HiPIP = 14+/-0.9, LIFR = 2.2+/-0.9, p < 0.05), decreased intra-alveolar neutrophils (HiPIP = 90+/-49, LIFR = 7.6+/-3.8,p = 0.001), and reduced wet-dry lung weight (HiPIP = 87.3+/-8.5%, LIFR = 40.8+/-17.4%,p < 0.001). CONCLUSIONS: High-pressure ventilation for 6 hours using conventional flow patterns produces severe lung injury, irrespective of RR or It. Reduction of inspiratory flow at similar PIP provides pulmonary protection.

Prolonged partial liquid ventilation in spontaneously breathing awake animals.

OBJECTIVE: To date, studies of partial liquid ventilation (PLV) have examined its effects acutely in anesthetized and mechanically ventilated subjects. We set out to develop a model of prolonged PLV in awake, spontaneously breathing animals. DESIGN: Animal case series SETTING: Cardiopulmonary physiology laboratory. SUBJECTS: Fifteen New Zealand white rabbits (3.2+/-0.39 kg). INTERVENTIONS: Animals were anesthetized and instrumented with a novel technique allowing percutaneously assisted placement of an intratracheal catheter with a subcutaneously tunneled externalized free end. After anesthetic recovery, PLV was performed in spontaneously breathing unsedated animals. MEASUREMENTS AND MAIN RESULTS: Evaporative losses were determined using a single 10 mL/kg perflubron dose (n = 5); hourly radiographs were obtained until residual perflubron was minimal. For prolonged PLV (n = 10), a 10-mL/kg initial perflubron dose was followed every 4 hrs with 5-mL/kg supplements. Radiographs were obtained immediately before and after perflubron administration and were scored (0-5) by a radiologist blinded to dosing regimen and time interval. Data were analyzed with ANOVA and Student's t-test with correction for multiple comparisons. Initial filling was nearly complete (score = 4.8+/-0.42); lungs were maintained approximately half-filled through 4 hrs (score = 2.53+/-0.71). By 6 hrs, the majority of perflubron had evaporated (score = 1.75+/-0.53). Over 24 hrs, radiographs documented continuous perflubron exposure (postffill = 4.53+/-0.64, prefill = 3.40+/-0.71, average = 3.97+/-0.43); scores were comparatively higher after filling (after = 4.53+/-0.64, before = 3.4+/-0.71, p< .001). Left and right lung volumes were equivalent (left = 4.06+/-0.47, right = 3.89+/-0.39, p = .027). All animals survived the 24 hrs of PLV. CONCLUSIONS: Percutaneously assisted intratracheal cannulation with catheter exteriorization permits prolonged PLV in spontaneously breathing, unsedated animals. Continuous perfluorocarbon exposure with this method is reproducible, consistent, and well tolerated for 24 hrs in uninjured animals.