Search

Our goal was to measure the quality of care provided in the Pediatric Intensive Care Unit (PICU) during Therapeutic Apheresis (TA). We described the care as a step by step process. We designed a flow chart to carefully document each step of the process. We then defined each step with a unique clinical indictor (CI) that represented the exact task we felt provided quality care. These CIs were studied and modified for 1 year. We measured our performance in this process by the number of times we accomplished the CI vs. the total number of CIs that were to be performed. The degree of compliance, with these clinical indicators, was analyzed and used as a metric for quality by calculating how close the process is running exactly as planned or “in control.” The Apheresis Process was in control (compliance) for 47% of the indicators, as measured in the aggregate for the first observational year. We then applied the theory of Total Quality Management (TQM) through our Design, Measure, Analyze, Improve, and Control (DMAIC) model. We were able to improve the process and bring it into control by increasing the compliance to > 99.74%, in the aggregate, for the third and fourth quarter of the second year. We have implemented TQM to increase compliance, thus control, of a highly complex and multidisciplinary Pediatric Intensive Care therapy. We have shown a reproducible and scalable measure of quality for a complex clinical process in the PICU, without additional capital expenditure. J. Clin. Apheresis, 2012. © 2011 Wiley Periodicals, Inc.

The criteria for starting extracorporeal membrane oxygenation (ECMO) therapy in term newborn patients with hypoxemic respiratory failure consist of an oxygenation index (OI) of 25 or higher and alveolar-arterial oxygen (Aao(2)) gradient of more than 600 at sea level. In such conditions, inhaled nitric oxide (iNO) may improve oxygenation and reduce the need for ECMO therapy. We studied early changes in OI and Aao(2) gradients in response to iNO treatment that may indicate a need to continue iNO treatment or the necessity to start an ECMO therapy.In this prospective study, we used 34 outborn neonatal patients that were referred to our pediatric critical care unit in a children's hospital for ECMO therapy with diagnosis of hypoxemic respiratory failure. In all patients, iNO therapy, starting at 80 ppm, was instituted either during transport or on arrival to hospital. Response to iNO was assessed after 1 hour, at which time, iNO concentration was reduced to 40 ppm, provided there was more than 20% improvement in either or both oxygenation indices. Patients who did not respond positively to continuous iNO therapy and met ECMO criteria were given ECMO therapy.Inhaled nitric oxide therapy alone was successful in 10 (29%) of 34 patients. Eighteen patients (53%) required ECMO therapy within the first 10 hours of iNO treatment (early ECMO therapy), whereas 6 other neonates (18%) became eligible for ECMO therapy after prolonged (2-4 days) iNO treatment (late ECMO therapy). No mortality occurred with any treatment. Within 4 hours after iNO therapy, patients who required early ECMO therapy had significantly higher OI and Aao(2) gradients than patients who were treated with iNO therapy alone (P.01, analysis of variance followed by Tukey-Kramer multiple comparison test). Six of 34 patients (18%), categorized as late ECMO therapy, on the average, had initially higher levels of OI and mean airway pressure than neonates in iNO treatment and early ECMO therapy.Persisting levels of OI of more than 20 or Aao(2) gradients of more than 600 after 4 hours of iNO therapy could be indicative of an immediate need for ECMO therapy.

Objective: To evaluate hemodynamic stability and gas exchange in a neonatal animal model of pumpless arteriovenous extracorporeal membrane oxygenation (AV-ECMO) with extracorporeal shunt flow of up to 15% of cardiac output during variable ventilation and oxygenation. Design: Prospective study. Setting: Research laboratory in a hospital. Subjects: Seven lambs (5.5 ± 0.6 kg, mean ± so). Interventions: The lambs initially were anesthetized by 50 mglkg ketamine intravenously. After tracheostomy, the lambs were mechanically ventilated and paralyzed by using 1 mglkg vecuronium bromide followed by 0.1 mg.kg -1 .hr -1 . One femoral vein was cannulated with a pulmonary artery flotation catheter and used for cardiac output and pulmonary artery pressure measurements. A femoral artery was cannulated for measuring mean arterial blood pressure, measuring heart rate, and blood sampling for gas exchange analyses. Finally, the right internal jugular vein and carotid artery were cannulated and used for the AV-ECMO. Normothermia (38 ± 0.5°C), fluid balance (5 mL.kg -1 .hr -1 normal saline), and anesthesia (5 mg.kg -1 .hr -1 , intravenous ketamine) were maintained. Ventilator settings were adjusted to establish a baseline Paco 2 (25-35 mm Hg) at an Flo 2 of 0.4. The AV-ECMO circuit was established by using a hollow fiber oxygenator, primed with maternal sheep blood (150-200 mL). Measurements and Main Results: The physiologic effects of the AV-ECMO shunt were evaluated at 15, 25, and 40 mL.kg -1 .hr -1 ECMO flow, corresponding roughly to 4%, 8%, and 15% of the cardiac output values. The baseline minute volume was maintained during stepwise increases in arteriovenous shunt. A significant increase in endogenous cardiac output occurred at arteriovenous shunt of 25 and 40 mL.kg -1 .hr -1 (analysis of variance followed by Tukey-Kramer multiple comparisons test), which was attributed to a significant increase of 30% in the heart rate. Effective cardiac output (difference between the thermodilution value and the AV-ECMO flow rate) and mean arterial blood pressure were not significantly changed. CO 2 removal, measured at 15% arteriovenous shunt, was significantly increased with decreasing ventilation to 25% and 50% of the baseline (analysis of variance and Tukey-Kramer test). Oxygenation through the membrane was measured after reducing inspired Flo 2 from 0.4 to 0.21, 0.15, and 0.10 with 15% arteriovenous shunt and baseline minute ventilation. Oxygen delivery by the oxygenator was significantly increased at Flo. 2 of 0.10, providing a maximum of 19.5% of the total oxygen consumption at an arterial hemoglobin-oxygen saturation of 60%. Conclusions: Healthy lambs are capable of maintaining effective cardiac output in the presence of moderate arteriovenous shunts (15%). AV-ECMO may provide efficient ventilatory support in the neonatal population with hypercapnia. The amount of oxygen delivery with AV-ECMO depends on arterial desaturation.

To measure insensible fluid loss from silicone membrane oxygenators during extracorporeal membrane oxygenation (ECMO), an in vitro system was used. A standard neonatal ECMO circuit (Avecor) was connected to a noncompliant reservoir, which was then primed with normal saline. The experiment was conducted by using two silicone oxygenators (Avecor 0.4 and 0.8 m2), three gas flow rates (0.5, 1.0, and 2.0 L/min) (sweep), and two fluid flow rates (200 and 400 ml/min). Two methods were used to measure the water loss. One method was to replace the water to the noncompliant circuit by using a calibrated burette, and the other method was to collect condensed water after cooling the postmembrane sweep gas to 0 degrees C. The influence of the amount of sweep, fluid flow rate, size of membrane, and inlet and outlet sweep gas temperatures on measured water loss was statistically determined. The amount of water loss correlated with sweep (r2 = 0.81; p

Objective: To test if hypothermia, induced by a sustained pentobarbital anesthesia, in rats can reduce ventilatory demands without compromising pulmonary gas-exchange efficiency. Design: Prospective study. Setting: Research laboratory in a hospital. Subjects: One group of 11 female Sprague Dawley rats. Interventions: The rats were anesthetized with 45 mg/kg pentobarbital, tracheostomized and intubated; their femoral veins and arteries were cannulated. After surgery, anesthesia and fluid balance were maintained (10 mg/kg per h pentobarbital, and 5 ml/kg per h saline, i. v.). Rectal temperature, mean arterial blood pressure (MAP), and heart rate (HR) were continuously monitored. The respiratory variables and gas-exchange profiles were determined at 38 °C (normothermia), and during stepwise hypothermia at 37, 35, 33, 31 and 29 °C. The arterial pressure of carbon dioxide (PaCO2), pH and arterial pressure of oxygen (PaO2) during hypothermia were corrected at body temperature. Measurements and results: Graded systemic hypothermia, with maintained anesthesia, produced a strong correlation between reduction in the respiratory frequency and rectal temperature (r2 = 0.55; p < 0.0001; n = 66). The minute volume was significantly reduced, starting at 35 °C, without significant changes in the tidal volume (repeated measures of analyses of variance followed by Dunnett multiple comparisons test). No significant changes occurred in the PaCO2, pH, PaO2, hemoglobin oxygen saturation, the calculated arterial oxygen content and estimated alveolar-arterial oxygen difference during mild hypothermia (37–33 °C). The PaO2, however, was significantly reduced below 31 °C. The MAP remained stable at different levels of hypothermia, whereas HR was significantly reduced below 33 °C. Conclusions: Mild hypothermia in rats, induced by a sustained pentobarbital anesthesia, reduces ventilation without compromising arterial oxygenation or acid-base balance, as measured at body temperature. Theoretically, our observations in spontaneously breathing rats imply that a combination of mild hypothermia with anesthesia could be safely utilized to maintain adequate ventilation, using relatively low minute ventilation. We speculate that such a maneuver, if applied during mechanical ventilation, may prevent secondary pulmonary damage by allowing the use of lower ventilator volume-pressure settings.

Purpose: The purpose of this study was to measure nitric oxide (NO) and nitrogen dioxide (NO2) concentrations, at various ventilatory settings and sampling sites, during in vitro inhaled NO and high-frequency oscillatory ventilation therapy [iNO-HFOV]. Materials and Methods: We used a high-frequency oscillatory ventilator (model 3100A, SensorMedics, Yorba Linda, CA), a test lung (model VT-2A Ventilator Tester, Bio-Tek Instruments, Inc., Winooski, VT), nitric oxide delivery and NO/NO2 monitoring (Pulmonox II, Pulmonox, Tofield, Canada), and scavenging systems in this study. The ventilator frequency, amplitude, and inspired oxygen concentration were systematically changed at a fixed flow of NO. The concentrations of NO and NO2, sampled at four sites, were determined by an electrochemical method (Pulmonox II). The NO and NO2 concentrations were measured at the proximal part of the inspiratory limb (site 1), near the Y-piece (site 2), the carina of the test lung (site 3), and the bellows of the test lung (site 4). Results: The concentration of NO decreased significantly (P < .001) from the proximal port (site 1) of the inspiratory circuit (86.16 ± 0.38 ppm) through the lung bellows (site 4) (70.08 ± 0.23 ppm). The concentration of NO2 increased significantly (P < .001) from site 1 (3.25 ± 0.04 ppm) through site 4 (19.4 ± 0.19 ppm). However, the total concentration of NO + NO2 (NOx) remained unchanged at both site 1 and site 4. Increasing the frequency and amplitude of the ventilator significantly altered NO and NO2 concentrations. The NO2 concentration increased significantly (P < .0001) from 5.6 ppm to 18.1 ppm at site 4 when the fraction of inspired oxygen was increased from 0.25 to 0.93. The NO2 concentration also increased significantly (P < .0001) from 0.6 ppm to 18.7 when NO concentrations were independently increased from 12 ppm to 80 ppm. Conclusions: During HFOV, the concentrations of NO and NO2 vary between sampling sites and also are influenced by the frequency, amplitude, and inspired oxygen concentration. NO2 concentrations in the lung were significantly increased above commonly accepted toxic concentrations during ventilation with high concentrations of NO (80 ppm) and high fractional concentrations of oxygen. The excessive increase in NO2 concentration at the “alveolar” level in our test lung model warrants confirmation in an in vivo model.

We studied levels of superoxide dismutase, glutathione, reductase, glutathione peroxidase and lipoperoxides in 12 healthy lambs below 1 year of age (8-19 kg) under therapy with extracorporeal membrane oxygenation (ECMO). Plasma levels of these free oxygen radical scavenging enzymes and lipoperoxides were taken 1 day before the ECMO experiment, at the beginning of ECMO after the first rotations of the roler pump, during, and after ECMO. The pre-ECMO results of days 1 and 2 were compared with the during-ECMO results and those with the post-ECMO results using the t test for paired samples. We found a significant decrease of both superoxide dismutase and glutathione reductase on ECMO, a trend to increased lipoperoxide levels, and unchanged levels of glutathione peroxidase. After discontinuing bypass the levels began to normalize again. We conclude that ECMO reduces some oxygen radical scavenging enzyme levels and exhibits a trend to increased lipoperoxide levels. Near total lung collapse with consecutive reperfusion injury might be harmful considering these results. However, the nonsignificant increase in lipoperoxide levels excludes considerable oxygen toxicity during this short ECMO trial.

The combined clinical and biochemical profile of diabetic ketoacidosis, hyperglycemic hyperosmolar non-ketotic syndrome, complicated by acute pancreatitis, in an 11-year-old with established insulin-dependent diabetes mellitus, is presented. The management requires diligent correction of dehydration and hyperglycemia, while monitoring neurological status and blood chemistry. It is imperative to monitor and avoid potentially fatal complications of the combined entity, namely, cerebral edema, thromboembolism, acute respiratory distress syndrome and rhabdomyolysis. Excluding acute pancreatitis in the face of persistent abdominal pain in this setting is emphasized.

Introduction Clinical application of arteriovenous (AV) extracorporeal membrane oxygenation (ECMO) requires assessment of cardiovascular ability to respond adequately to the presence of an AV shunt in the face of acute lung injury (ALI). This ability may be age dependent and vary with the experimental model. We studied cardiovascular stability in a lamb model of severe ALI, comparing conventional mechanical ventilation (CMV) with AV-ECMO therapy. Methods Seventeen lambs were anesthetized, tracheotomized, paralyzed, and ventilated to maintain normocapnia. Femoral and jugular veins, and femoral and carotid arteries were instrumented for the AV-ECMO circuit, systemic and pulmonary artery blood pressure monitoring, gas exchange, and cardiac output determination (thermodilution technique). A severe ALI (arterial oxygen tension/inspired fractional oxygen 0.05; Fisher's exact test). Surviving and nonsurviving lambs in the AV-ECMO group, unlike the CMV group, required continuous volume expansion and inotropic support (P < 0.001; Fisher's exact test). Lambs in the AV-ECMO group were able to maintain normocapnia with a maximum of 30% reduction in the minute ventilation, as compared with the CMV group (P < 0.05). Conclusion AV-ECMO therapy in lambs subjected to severe ALI requires continuous hemodynamic support to maintain cardiovascular stability and normocapnia, as compared with lambs receiving CMV support.

PURPOSE To evaluate residents' skills in performing basic mathematical calculations used for prescribing medications to pediatric patients. METHOD In 2001, a test of ten questions on basic calculations was given to first-, second-, and third-year residents at Miami Children's Hospital in Florida. Four additional questions were included to obtain the residents' levels of training, specific pediatrics intensive care unit (PICU) experience, and whether or not they routinely double-checked doses and adjusted them for each patient's weight. The test was anonymous and calculators were permitted. The overall score and the score for each resident class were calculated. RESULTS Twenty-one residents participated. The overall average test score and the mean test score of each resident class was less than 70%. Second-year residents had the highest mean test scores, although there was no significant difference between the classes of residents (p =.745) or relationship between the residents' PICU experiences and their exam scores (p =.766). There was no significant difference between residents' levels of training and whether they double-checked their calculations (p =.633) or considered each patient's weight relative to the dose prescribed (p =.869). Seven residents committed tenfold dosing errors, and one resident committed a 1,000-fold dosing error. CONCLUSION Pediatrics residents need to receive additional education in performing the calculations needed to prescribe medications. In addition, residents should be required to demonstrate these necessary mathematical skills before they are allowed to prescribe medications.

We continuously measured proximal and carinal pressures at low and high flow rates and frequencies during conventional mechanical ventilation (CMV) and intratracheal pulmonary ventilation (ITPV), using an artificial lung. The proximal peak inspiratory pressure (PIP), carinal PIP, proximal positive end expiratory pressure (PEEP), and carinal PEEP, or negative end expiratory pressure (NEEP), were measured during simulated CMV and ITPV. Two levels of frequency (30 and 90 per min) and two gas flow rates (3 and 6 L/min) were examined, in both dry and humid states (four combinations of gas flow and frequency at each state). The gas flow and inspiratory time were held constant throughout the CMV and ITPV trials. Humidification of the ventilatory circuit during ITPV prevented the accurate measurement of carinal pressures. This problem was solved by introducing a continuous "bias flow" of 11 ml/min into the pressure monitoring line. A combination of low gas flow and low frequency with CMV showed no significant differences between the proximal and carinal PIP, as well as the proximal and carinal PEEP. The same combination with ITPV, however, resulted in a significantly lower carinal PIP and PEEP, compared to proximal PIP and PEEP. Carinal PIP and PEEP during ITPV were also significantly lower than those observed during CMV with a low flow and low frequency rates. During both CMV and ITPV, using a combination of a high flow rate with a high breathing frequency, carinal PIPs were significantly lower than proximal PIPs. ITPV, however, generated much larger differences between proximal and carinal PIPs than the CMV. A significant NEEP was generated at the carinal level during ITPV with high flow rates, both with high and low frequencies. The NEEP did not occur with a low gas flow, in combination with either a low frequency or a high frequency. The "bias flow" had no significant effect on carinal pressures. In conclusion, ITPV, compared with CMV, generates a significantly lower carinal PIP, but it may also generate carinal NEEP. For safety reasons, therefore, it is essential to monitor carinal pressures continuously in patients treated with ITPV.

Objective: To compare fluctuations in systemic arterial pressure (SAP) resulting from changes in systemic vascular resistance (SVR) during closed- and open-bridge extracorporeal life support (ECLS). Design: In vitro laboratory study. Setting: Physiology laboratory of a tertiary care pediatric hospital. Methods: A standard neonatal ECLS circuit with simulated SAP was established using normal saline as circulating fluid. Our reference setting included an extracorporeal flow rate of 300 mL/min, a simulated SAP of 60 mm Hg, and a postoxygenator pressure of 150 mm Hg. The simulated SVR was modified by changing the degree of occlusion of the arterial catheter distal to the bridge. For this purpose, we used a graduated clamping device. Subsequently, the pressure changes were measured at four ports in the circuit. They were located as follows: a) on the venous tubing of the circuit between the bridge and the reservoir; b) on the arterial tubing of the circuit between the heat exchanger and the bridge; c) between the first and the second resistance clamps on the arterial tubing of the circuit for monitoring the simulated systemic arterial pressure; and d) at the reservoir. The experiment was repeated with various extracorporeal flow rates to the reservoir (100-300 mL/min) and through the bridge (100-300 mL/min using a custom-made clamp). Variations in the simulated SAP created by varying degrees of occlusion and flow rates were compared with repeated measures analysis of variance followed by the Tukey-Kramer test. Measurements and Main Results: The open-bridge ECLS significantly reduced the variations in the simulated SAP by 15% to 45% (p

Objective: To compare normative ventilatory and gas-exchange data and anesthetic requirements in male and female rats subjected to critical care conditions. Design: Prospective study. Setting: Critical care research laboratory in a hospital. Subjects: Twenty-two age-matched young male and female rats (Sprague-Dawley, Long Evans strain). Interventions: Anesthesia was induced with 65 and 45 mg/kg pentobarbital in male and female rats, respectively. The rats were then tracheostomized and cannulated in one femoral vein and artery. Anesthesia was maintained using 8-15 mg/kg/hr pentobarbital (iv) and controlled by continuous hemodynamic monitoring. Measurements and Main Results: Normoxic baselines for breathing frequency, tidal volume, minute volume, inspiratory-to-expiratory ratio, inspiratory drive (tidal volume/inspiratory time), respiratory system compliance, peak airway pressure, and gas-exchange profiles were established. Ventilatory and gas-exchange responses to oxygen and CO 2 were then determined by exposure to 10 mins of hyperoxia (100% oxygen), two levels of mild and severe hypercapnic hyperoxia (inspired Pco 2 of 30 and 60 torr; 4 and 8 kPa), and two levels of mild and severe normocapnic hypoxia (inspired Po 2 of 81 and 48 torr; 10.7 and 6.3 kPa). The average anesthetic requirement (during a 5- to 6-hr experiment) was 30% less In the female rats than in the male rats (p

Purpose : The purpose of this article was to compare the safety and patient charges of two postextubation treatment regimens. Materials and Methods : Twenty-two pediatric patients, between the ages of 7 months and 13 years, who were mechanically ventilated for less than 5 days were studied in a prospective randomized nonblinded study at a multidisciplinary pediatric intensive care unit. Immediately after extubation all patients received supplemental oxygen, administered via mask or nasal cannulae, at a flow rate or concentration sufficient to maintain the pulse oximetric arterial oxygen saturations >95%; arterial blood gas analyses were performed at 30 minutes after extubation. The subjects were randomly assigned to one of two protocols. Protocol A (our standard management) consisted of (1) three nebulized albuterol treatments administered 1 hour apart, and (2) a chest radiograph obtained within 60 minutes of extubation. Protocol B included one nebulized albuterol treatment administered immediately after extubation. We measured the heart rate, respiratory rate, and arterial blood pressure immediately after and at 60, 120, and 180 minutes following extubation. The following data were also recorded: arterial blood gas analysis results and continuous pulse oximetric arterial oxygen saturation levels. Any significant complications, such as stridor, respiratory distress, or requirement for reintubation, were noted if they occurred within 24 hours of extubation. Patient charge costs were calculated after obtaining the prevailing hospital and physician charges at the time of the study. Results : Eleven patients completed each arm of the study (total = 22). There were no statistically significant differences between the two groups with respect to arterial pH, serum bicarbonate, pulse oximetric arterial oxygen saturation, arterial blood pressure, respiratory rate, or heart rate (P > .05). Patients treated with Protocol A had a statistically, but not clinically, significant higher mean Pa O 2 and Pa CO 2 (P = .02 and P = .05, respectively) than those in Protocol B. Associated charges per patient for Protocol A were $863.50 versus $476.00 for Protocol B. This is a savings of $387.50 per patient. Our pediatric intensive care unit provides care to over 600 intubated patients per year, which would equate to a charge savings of $232,500.00 per year. Conclusion : A modified postextubation management protocol, consisting of fewer interventions, resulted in significant patient charge savings with no increased risk to the patient.

An alcohol-naive 16-year-old male is presented with alcohol-induced atrial fibrillation. Past medical history, review of systems, and presentation were all otherwise benign. Atrial fibrillation occurred early in the intoxication at an alcohol level slightly higher than the legal limit for intoxication (153 mg/dL). His complete cardiac evaluation was otherwise normal. The atrial fibrillation was not treated aggressively and resolved as the alcohol level quickly fell to zero, consistent with his "nonalcoholic" metabolism. Complete follow-up found the adolescent with no evidence of cardiac or other disease.

At birth this baby girl was found to have left pneumohydrothorax and marked pulmonary hypoplasia that was treated with mechanical ventilation. After continued deterioration, she received high-frequency oscillatory ventilation. Her intraarterial pressure wave form dampened, and surgical exploration revealed air, but no blood, in the carotid artery.

Background The plasma colloid osmotic pressure (COP) plays a major role in transcapillary fluid balance. There is no information on plasma COP of healthy infants beyond the first post-natal week. The normal COP in healthy adult subjects (25 mmHg) is currently also applied as a reference value for healthy infants. This study was designed to test whether plasma COP values in healthy infants are the same as those in normal adults. Methods Plasma COP was measured in 37 male and female healthy infants from 1 to 11 months old. For this purpose, 1 ml blood was collected during the patient’s regularly scheduled visit if the patient required any type of blood test for routine laboratory analyses. Results Plasma COP levels correlated slightly with increasing age from 1 to 9 months old (linear regression analysis; r2 = 0.1, P < 0.049). We found no correlation between plasma COP and body weight at the same age (r 2 = 0.05, P = 0.155). The mean and standard deviation of COP in all infants was 25.1 ± 2.6 mmHg, which is almost identical to an average COP of 25 mmHg in healthy adult subjects. Arbitrary division of the infants into three different age groups (1–3 months [n = 11], 5–8 months [n = 13] and 9–11 months [n = 13]) showed an average increase of approximately 2 mmHg in COP of 9-month-old to 11-month-old infants, compared with 1-month-old to 3-month-old infants (one-way analysis of variance; P = 0.26). There was no gender difference in the COP level (unpaired t-test), with an average of 25.1 ± 2.4 mmHg in 19 male infants compared with 25.2 ± 2.9 in 18 female infants. The 95% confidence interval for COP in both male and female infants (n = 37) was between 24.3 to 26.0 mmHg, ranging from 19.5 to 30.3 mmHg, with a median value of 25.2 mmHg. Conclusions The data accept the null hypothesis that the COP range in infants younger than 1 year old is similar to those observed in adult subjects. Our observations, compared with previously reported neonatal COP values, suggest that there is a sharp increase in COP within the first months after birth.

Introduction: Accurate measurement of airway pressures (Paw) are crucial during mechanical ventilation. Intratracheal pulmonary ventilation(ITPV) utilizes a reverse thruster device to deliver a continuous flow of gas at the level of the carina, thereby generating a different carinal Paw than conventional ventilation (CV). This study compares the proximal and carinal Paw during CV and ITPV, in both an artificial lung an in a rabbit model.

Sir .—We read with interest the article by Shulman et al 1 in the August issue of AJDC and herein report our experience with 45 patients with cystic fibrosis in whom a total of 66 catheters were placed. Patients with cystic fibrosis are similar to patients in the study by Shulman and colleagues in that they also frequently require long-term intravenous therapy. 2-7 Our use of a peripherally inserted central catheter for administering antibiotics and supplemental alimentation intravenously has certain advantages. Patients and Methods .—The patients' ages ranged from 7 to 45 years and their weights ranged from 22 to 65 kg. Our experience began with 23 patients in whom 38 catheters were placed. 2 The polyvinylchloride catheters were placed in the basilic or cephalic vein for a mean duration of 15.4 days (range, 5 to 49 days). An 86% success rate was obtained. No major complications such as sepsis, catheter

Two cases of nonbacterial endocardial thrombi (NBET) on the atrial surface of the tricuspid valve in the full-term infant are presented. The pathophysiology and the events peculiar to the full-term infant that suggest susceptibility to this rare pathological phenomenon are reviewed. The difficulties in diagnosis and treatment are discussed.