Curt M. Steinhart

Mortality associated with multiple organ system failure and sepsis in pediatric intensive care unit

Seven hundred twenty-six patients from five pediatric intensive care units were studied to determine the association of multiple organ system failure (MOSF) with mortality and to test the hypothesis that MOSF associated with sepsis has a higher mortality rate than MOSF without sepsis. There were 177 (24%) patients with MOSF and 83 (11%) nonsurvivors of MOSF. The mortality rates for two, three, or four or more failed organ systems were 26%, 62%, and 88%, respectively (P less than 0.001). Eighty-four (47%) patients with MOSF had associated sepsis. Sepsis (both bacteremia and clinical sepsis syndrome) did not significantly increase mortality rates in the groups with organ system failure. Mortality rates for patients with sepsis before or within 24 hours of development of MOSF (early sepsis) did not differ from mortality rates for those patients with onset of sepsis more than 24 hours after developing MOSF (late sepsis, 53% vs 33%, P = NS). We conclude that underlying pathophysiologic mechanisms of MOSF other than sepsis are as important as sepsis in critically ill pediatric patients.

Evaluating the frequency rate of hypomagnesemia in critically ill pediatric patients by using multiple regression analysis and a computer-based neural network.

ObjectivesTo determine the frequency rate of hypomagnesemia in patients admitted to the pediatric intensive care unit (ICU), and to identify subsets of patients (grouped by disease) who are at greatest risk of hypomagnesemia. We also compared a neural network model with multiple regression analysis to identify independent variables that would correlate with hypomagnesemia and to predict serum magnesium values in critically ill pediatric patients overall. DesignProspective, multicenter study. SettingTertiary level medical/surgical pediatric ICUs. PatientsData were obtained at admission to the pediatric ICU for 463 patients from newborn to 18 yrs old who were admitted with a variety of surgical and nonsurgical conditions. InterventionsNone. Measurements and Main ResultsTotal serum magnesium values were obtained within the first 24 hrs after admission in 463 pediatric patients admitted to four pediatric ICUs. Hypomagnesemia (defined as total serum magnesium <0.75 mmol/L) was found in 51 (11%) of the 463 patients, with the highest frequency rate (72%) and lowest mean serum magnesium level (0.66 ± 0.17 mmol/L) in patients admitted after surgery with extensive osseous involvement (spinal fusion and craniofacial reconstruction). To determine whether hypomagnesemia could be predicted on the basis of other laboratory and clinical criteria, multiple regression analysis was performed and showed age, weight, and albumin levels weakly associated (r2 = .14, p < .001) with magnesium levels within the different diagnostic groups. These data were used to produce a mathematical model able to predict magnesium levels within 5% of the actual values in 23% of patients. A neural network was also created to compare its predictive capabilities to those of the multiple regression model. Once trained on a random subset (85%) of the patient population, the neural network was able to predict magnesium levels to within 5% of actual values for 88% of the remaining 15% of patients, comparing favorably with the predictions derived from the multiple regression model. ConclusionsHypomagnesemia is not uncommon (11%) in critically ill pediatric patients, but is very common (72%) in patients admitted after surgery for spinal fusion or craniofacial reconstruction. Patients who undergo surgery for correction of scoliosis and craniofacial anomalies should have serum magnesium levels monitored closely after surgery. In other patients, a neural network or multiple regression model could help predict which patients would be at risk of developing hypomagnesemia, thereby focusing testing on patients likely to benefit from such testing.

Accuracy and utility of a continuous intra-arterial blood gas monitoring system in pediatric patients

ObjectivesTo determine the accuracy of the Paratrend 7 continuous intra-arterial blood gas monitor (CI-ABGM) in radial and femoral artery catheters placed in children compared with simultaneous measurements of pH, Paco2, and Pao2 performed by intermittent blood gas analysis. To determine sensor longevity in pediatric patients at different arterial sites. To determine the utility of CI-ABGM for tracking unanticipated events related to blood gas deterioration. SettingA pediatric intensive care unit of a university hospital. DesignA prospective clinical investigation. PatientsFifty critically ill pediatric patients, ranging in age from 1 wk to 18 yrs of age, who required either radial or femoral artery catheters for intermittent arterial blood gas monitoring. InterventionsNone. Measurements and Main Results A Paratrend 7 intra-arterial sensor was placed through either an 18- or 20-gauge catheter previously inserted into the radial or femoral artery. At clinically predetermined intervals ranging from every 1 to 8 hrs, the CI-ABGM measurements of pH, Pco2, and Po2 were compared with the values determined by standard intermittent blood gas analysis. The Paratrend 7 system values were individually adjusted to match ABG results when the Paratrend 7 pH differed by greater than ±0.05 units, Pco2 was greater than ±5 torr (0.7 kPa), and Po2 was greater than ±15% of the ABG value. Significant aberrations in gas exchange defined as unanticipated events were categorized as isolated metabolic acidosis (pH <7.20), hypercapnia (Pco2, >70 torr; 9.3 kPa), and hypoxemia (Po2, <50 torr; 6.7 kPa). All unanticipated events were earmarked from consecutive monitoring epochs ranging from 4 to 24 hrs duration from the time of Paratrend 7 sensor insertion to the time of sensor removal. Fifteen sensors were placed into the radial artery, 34 sensors were placed into the femoral artery, and one sensor was initially placed in the radial and moved to a femoral artery location. Mean radial artery insertion duration was 35 hrs. Mean femoral artery duration was 137.2 hrs. A total of 1445 pairs of ABG results were available for comparison. After removal of individual values, which did not meet inclusion criteria, 1411 pH data pairs, 1408 Pco2 data pairs, and 1326 Po2 data pairs were analyzed. The bias and precision for the pH data were 0.00 and 0.04 units, respectively; for the Pco2 data were −0.4 and 4.8 torr (−0.05 and 0.64 kPa), respectively; and for the Po2 data 1.0 and 25 torr (0.1 and 3.3 kPa), respectively. Detection of unanticipated events was evenly spread across the three categories and was most commonly related to iatrogenic causes or cardiac failure. Persistent waveform dampening necessitating sensor removal was more frequently encountered in radial placement compared to femoral placement. ConclusionsThe Paratrend 7 CI-ABGM is accurate within the extremes of physiologic gas exchange typically encountered in the pediatric intensive care setting. The device is capable of tracking extreme fluctuations in gas exchange with a response rate suitable for making real-time therapeutic decisions. The sensor can be recommended for insertion into a femoral artery cannula. There is a high incidence of blood pressure waveform dampening encountered in radial artery use.

Use of brainstem auditory evoked potentials in pediatric brain death.

: Brainstem auditory evoked potentials (BAEPs) were examined prospectively in ten clinically brain-dead and 13 comatose nonbrain-dead children. All clinically brain-dead children failed to demonstrate brainstem impulse transmission, displaying no waves or only wave I. The comatose, nonbrain-dead children maintained evidence of brainstem electrical activity, manifested by at least two waveforms. This suggests that BAEPs are a useful noninvasive method of supporting the clinical diagnosis of brain death in children.

Noninvasive determination of effective (nonshunted) pulmonary blood flow in normal and injured lungs

PEEP is utilized in acute respiratory failure to decrease intrapulmonary shunting and improve oxygenation. Despite these beneficial effects, PEEP may adversely affect cardiac output, thus reducing oxygen delivery. To monitor some of the cardiopulmonary effects of PEEP, we utilized a noninvasive rebreathing technique to measure effective (nonshunted) pulmonary blood flow (&OV0422;epr) and compared the results to those measured by thermodilution (&OV0422;epi) in normal and oleic acid-injured canine lungs.&OV0422;epr was highly correlated with &OV0422;epi (r = .92, r2 = .85, p < .001) despite large variations in PEEP before lung injury (0 to 15 cm H2O) and after lung injury (0 to 20 cm H2O). This close correlation was found even with wide ranges in cardiac output (1.01 to 6.45 L/min) and intrapulmonary shunt fractions (0.03 to 0.67). This technique may prove valuable as a noninvasive method by which to monitor and adjust PEEP therapy in patients with acute lung injury.

Unaltered pulmonary capillary surface area in the presence of changing arterial resistance

We hypothesized that capillary recruitment may not be solely dependent on extracapillary factors. To test this hypothesis, rabbits were anesthetized and placed on total cardiac bypass at a constant, physiological pulmonary blood flow. Vascular occlusion techniques were combined with measurement of the transpulmonary metabolism of an angiotensin-converting enzyme substrate, allowing the concomitant assessment of changes in segmental resistances and dynamically perfused capillary surface area. Intra-arterial serotonin infusion increased upstream pulmonary vascular resistances without affecting dynamically perfused capillary surface area. Intra-arterial isoproterenol infusion diminished serotonin-induced increased upstream resistances, also without affecting capillary surface area. These findings support the hypothesis that pulmonary capillary recruitment may not be solely dependent on extracapillary factors.We hypothesized that capillary recruitment may not be solely dependent on extracapillary factors. To test this hypothesis, rabbits were anesthetized and placed on total cardiac bypass at a constant, physiological pulmonary blood flow. Vascular occlusion techniques were combined with measurement of the transpulmonary metabolism of an angiotensin-converting enzyme substrate, allowing the concomitant assessment of changes in segmental resistances and dynamically perfused capillary surface area. Intra-arterial serotonin infusion increased upstream pulmonary vascular resistances without affecting dynamically perfused capillary surface area. Intra-arterial isoproterenol infusion diminished serotonin-induced increased upstream resistances, also without affecting capillary surface area. These findings support the hypothesis that pulmonary capillary recruitment may not be solely dependent on extracapillary factors.

IMPLEMENTATION OF THE THIRD GENERATION NON-INVASIVE REEREATHING CARDIOPULMONARY PARAMETER SOFTWARE IN AN CLEIC ACID LUNG INJURY DOG MODEL

This study was performed to evaluate improved software that non-invasively measures Effective Pulmonary Blood Flew (QEP), Lung Tissue Water (Vt), Lung Diffusion for carbon monoxide (DLoo), and Functional Residual Capacity (FRC).A modified non-invasive method for the measurement of four cardiopulmonary parameters (CPP) which can be measured only with an accepted gas rebreathing method or by invasive means was developed to improve speed of data acquisition and analysis, and ease of use. The rebreathing system includes a Perkin Elmer Respiratory Gas Analyzer MGA-1100, an IBM-XT computer and a valving device. Nine dogs were instrumented to invasively measure thermodilution cardiac output and measure mixed venous and systemic oxygen saturations. Measurements of CPP were made before and after invasive measurements and averaged.Rebreathing QEP correlated well with the invasive QEP (p<0.0001). Vt, DLco and FRC decreased during lung injury (p<0.00001 for all three). FRC increased proportional to increased PEEP from 0 to 15 on H2O (p<0.005).Computation time from the end of a 25 second sampling time to acquisition of results was less than 2 minutes. One program is utilized during the sampling, computation, disk I/O and printing modules. Raw gas concentration vs. time, and log gas concentration vs. time curves are displayed for selection of the window used for calculation of CPP. This study demonstrates a powerful software package for the non-invasive measurement of CPP.

THE NONINVASIVE MEASUREMENT OF EFFECTIVE PULMONARY BLOOD FLOW IN ACUTE LUNG INJURY

Effective (non-shunted) pulmonary blood flow (QEPr) was measured using a modified rebreathing technique and compared to QEPi determined by thermodilution and blood gas analysis at varying PEEP in non-injured and oleic acid injured lungs.Nine mongrel dogs (18-26 Kgs) were anesthetized, intubated with cuffed tubes, and instrumented with intravenous, systemic arterial, and pulmonary arterial catheters. Ventilation was volume controlled at 15 breaths/minute and tidal volume adjusted to maintain PaCO2 between 35-45 torr. Noninvasive measurements of QEPr were done before and after determination of QEPi. PEEP was varied by 5 cm H2O increments from 0 to 15 before lung injury and 0 to 20 after central venous administration of oleic acid (0.10-0.15 ml/kg).QEPr highly correlated with QEPi as seen in the figure (r=.92, R2=.85, n=99, p < .0001) using Pearsons correlation. The noninvasive measurement of QEPr was accurate at varying cardiac outputs (CO, 0.8-6.3 L/min) and intrapulmonary shunt (Qs/Qt, 0.03-0.67). The measurement of QEPr at varying levels of PEEP (0-20 on H2O) and subsequent changes in Qs/Qt demonstrates that QEPr is the pulmonary blood flow that undergoes gas exchange such that QEPr = CO (1-Qs/Qt).The study validates the noninvasive rebreathing technique as a reliable measurement of the effective pulmonary blood flow at varying CO and Qs/Qt in acute respiratory failure indicating clinical utility during PEEP therapy.

Pulmonary Capillary Recruitment and Extracapillary Vascular Tone

Elucidating the mechanisms controlling capillary recruitment could improve our understanding of the mechanisms underlying various pharmacological and pathophysiological processes.

EFFICIENCY OF PEDIATRIC INTENSIVE CARE: A COMPARATIVE ANALYSIS OF 8 PEDIATRIC ICUS

A national standard of ICU efficiency does not exist. This study objectively measures efficiency rates in 8 PICUs to determine if PICUs greatly differ in their efficiencies. Data included demographics, daily mortality risk (Dynamic Risk Index) and therapeutic interventions. Inefficient use of PICU resources was defined using daily mortality risks and a previously developed list of unique PICU resources; (a) low risk, monitored (LRM) pts had daily mortality risks of < 1% and did not use any unique PICU therapies and (b) potential early discharge (PED) pts did use unique resources or have mortality risks >1% during the early ICU stay but their last, consecutive PICU day were equivalent to LRM pts. Total, LRM and PED days of care were used to calculate efficiency rates. RESULTS. Data was collected on 1668 pts and 6962 days of care. LRM pts comprised from 16% to 58% of the PICU pts (p <.0001) and utilized between 5.4% and 34.5% of the days of care (p< .0001). PED pts comprised from 12% to 29% of the PICU pts (p < .0001) and their PED days comprised from 5.1% to 17.2%. Overall, the efficiency ratings ranged from .89 to .55. CONCLUSIONS. (1) Large differences in efficiency of PICU utilization exist. (2) The finding of significant over-utilization by PED patients is unique. (3) Efficiency rates over .80 are reasonable. Supported by MCH grant MCJ-11-527.