Roadmap to a successful quality improvement project.

Although the benefits of quality improvement initiatives are largely understood by practicing neonatologists and perinatologists, the vast majority have not received any formal training in quality improvement methodology. Even as reporting requirements of quality metrics has increased from a number of outside agencies and public reporting entities, education for physicians regarding how to carry out quality improvement projects has largely remained the individual's responsibility. The first in a series of quality improvement education papers, we focus on the reasons why quality improvement matters and how to develop a team of stakeholders that will be functional and productive in addressing specific quality and safety concerns.

Use of neonatal simulation models to assess competency in bag-mask ventilation.

OBJECTIVE: Providing adequate bag-mask ventilation (BMV) is an essential skill for neonatal resuscitation. Often this skill is learned using simulation manikins. Currently, there is no means of measuring the adequacy of ventilation in simulated scenarios. Thus, it is not possible to ascertain proficiency. The first aim of this study was to measure the pressure generated during BMV as performed by providers with different skill levels and measure the impact of different feedback mechanisms. The second aim was to measure the pressure volume characteristics of two neonatal manikins to see how closely they reflect newborn lung mechanics. STUDY DESIGN: In Phase I to achieve the first aim, we evaluated BMV skills in different level providers including residents (n=5), fellows (n=5), neonatal nurse practitioners (n=5) and neonatologists (n=5). Each provider was required to provide BMV for 2-min epochs on the SimNewB (Laerdal), which had been instrumented to measure pressure-volume characteristics. In sequential 2-min epochs, providers were given different feedback including chest-wall movement alone compared to manometer plus chest-wall movement or chest-wall movement plus manometer plus laptop lung volume depiction. In Phase II of the study we measured pressure-volume characteristics in instrumented versions of the SimNewB (Laerdal) and NeoNatalie (Laerdal). RESULTS: In Phase I, all providers are compared with the neonatologists. All measurements of tidal volume (Vt) are below the desired 5 ml kg(-1). The greatest difference in Vt between the neonatologists and other providers occurs when only chest-wall movement is provided. A linear relationship is noted between Vt and PIP for both SimNewB and NeoNatalie. The compliance curves are not 'S-shaped' and are different between the two models (P<0.001). CONCLUSION: Phase I of this study demonstrates that the SimNewB with the feedback of chest-wall movement alone was the best method of distinguishing experienced from inexperienced providers during simulated BMV. Therefore this is likely to be the best method to ascertain proficiency. Phase II of the study shows that the currently available neonatal simulation manikins do not have pressure-volume characteristics that are reflective of newborn lung mechanics, which can result in suboptimal training.

Cranial sonography in very-low-birth-weight infants: do all infants need to be screened?

The objective of the study was to develop clinical screening criteria to diagnose infants with intraventricular hemorrhage (IVH) and cystic periventricular leukomalacia (PVL). We performed a case-control investigation of two cohorts of very-low-birth-weight infants (n = 505, combined cohorts). Univariate and multivariate analyses were performed from data obtained in cohort 1 to develop screening criteria for IVH and cystic PVL. The screening criteria were then applied to cohort 2. The screening criteria for IVH had a sensitivity of only 51%, a specificity of 62%, a positive predictive value of 31%, and a negative predictive value of 79%. Screening criteria for cystic PVL had a sensitivity of only 22%, a specificity of 58% a positive predictive value of 2%, and a negative predictive value of 95%. These data suggest that using clinical criteria to determine which infants should receive screening cranial sonography for IVH and cystic PVL would miss a substantial number of infants with these conditions.

Infective endocarditis in the premature neonate.

We describe a series of 11 high-risk neonates with infective endocarditis (IE) in this retrospective review. Previously IE has rarely been diagnosed in newborns and is usually fatal. The frequency was 4.3 cases per 100 patients. Five patients survived. Microorganisms included gram positives such as S. aureus and coagulase-negative Staphylococcus, gram negatives such as Klebsiella pneumoniae, Enterobacter cloacae, Enterococcus faecalis, Serratia marcescens, and Acinetobacter calcoaceticus. Echocardiographic location of the lesions showed four left sided, five right sided, and two bilateral. We conclude that IE may be more common than previously described. Prompt diagnosis and treatment led to survival in 45% of our patients. Prospective studies are needed to identify patients at risk and to establish the true incidence of IE in high-risk neonates.

Predicting red blood cell transfusions in very low birth weight infants based on clinical risk factors.

OBJECTIVE: To describe the clinical factors most predictive of red blood cell transfusion in very low birth weight (VLBW) infants. STUDY DESIGN: Retrospective review of VLBW infants cared for at a single level III NICU during a two year period, n = 199. RESULTS: Overall transfusion requirement was 4.6 +/- 6.2 transfusions/infant/hospital course. Length of hospital stay, days of mechanical ventilation, requirement for dopamine support, birth weight, initial hematocrit, periventricular leukomalacia and necrotizing enterocolitis all independently correlated with number of transfusions and donors. Bronchopulmonary dysplasia and patent ductus arteriosus were associated with donor but not transfusion number. CONCLUSIONS: Our data characterize the population of VLBW infants with the greatest blood transfusion and donor requirement. Further investigation is needed to target this population for interventions to reduce blood exposure.

Umbilical venous catheters: evaluation of radiographs to determine position and associated complications of malpositioned umbilical venous catheters.

To evaluate the ability of radiography to detect malpositioned umbilical venous catheters in the left atrium and to determine the frequency of associated complications, we retrospectively compared radiographs and echocardiograms of 31 infants who had malpositioned catheters in the left atrium by echocardiography (cases) and 31 infants who had properly positioned catheters (controls). The case and control infants were of similar gestational age and birthweight (gestational age, 32 +/- 5 weeks; birthweight, 1672 +/- 899 g for cases; gestational age, 31 +/- 5 weeks; birthweight, 1666 +/- 958 g for controls). Malposition was defined as the catheter tip above the seventh thoracic vertebra by radiography. Radiography had sensitivity of 45%, specificity of 87%, positive predictive value of 77%, negative predictive value of 61%, accuracy of 66%, and prevalence of 50%. Thrombus formation in the heart was detected in 8 of 31 (26%) of cases and in 1 of 31 (3%) of controls (p = 0.03). The incidence of complications, such as necrotizing enterocolitis, culture-positive sepsis, total number of sepsis cases, thrombocytopenia, embolism to extremities, and hematuria were similar in both groups (difference not significant). These results suggest that radiography is unreliable in determining incorrect catheter placement. Catheters malpositioned in the left atrium were associated with thrombus formation. There was no significant increase in systemic complications in the infants with a malpositioned catheter.

Severe hemolytic disease of the newborn caused by anti-Gonzales antibody.

We report on a female infant with severe hemolytic disease of the newborn (HDN) secondary to anti-Gonzales antibody (anti-Go(a)) necessitating an exchange transfusion within the first day of life. The infant was born to a mother known to be Gonzales-antigen negative and a father who was Gonzales-antigen positive. The mother had an anti-Go(a) titer of 1:256 at 35 weeks' gestation. The infant was noted to have jaundice shortly after birth, with a bilirubin of 17.8 mg/dl (total) and .05 mg/dl (direct) at 11 hours of life. Coombs' test was positive and cord cells were Gonzales-antigen positive. Eluate on cord cells demonstrated anti-Go(a). Despite aggressive phototherapy, the total bilirubin reached 23.3 mg/dl by 24 hours of age and a double-volume exchange transfusion was performed. Following the exchange transfusion, phototherapy was continued for several days. The hemoglobin, which was 19.8 gm/dl at 11 hours of age, remained stable through the hospitalization, and no further transfusions were required. History revealed that two prior pregnancies resulted in Gonzales-antigen positive infants. The first child experienced mild jaundice requiring no therapy, and the second child required phototherapy for prolonged hyperbilirubinemia. Previous reported cases of anti-Go(a) suggest that this is not a cause of severe HDN. However, in view of the current case, it may be prudent to follow women with anti-Go(a) prenatally with amniotic fluid bilirubin studies, serial antibody titers, and fetal hemoglobin levels.

Soluble interleukin-2 receptor levels in infants with bronchopulmonary dysplasia.

Soluble interleukin-2 receptors (sIL2R) in plasma have been identified as a marker of lymphocyte activation. Lymphocyte activation as a manifestation of inflammation may be important in the pathogenesis of bronchopulmonary dysplasia (BPD). To test the hypothesis that infants with BPD have higher sIL2R levels, 12 infants with or at risk of developing BPD (GA +/- SD, 27 +/- 5 weeks; BW +/- SD 1,053 +/- 733 g) had plasma sIL2R levels determined and were compared to 20 infants being ventilated for respiratory distress syndrome (RDS) (GA +/- SD, 28 +/- 3.5 weeks; BW +/- SD, 1,133 +/- 390 g: P = NS for both GA and BW, t test). Tracheal aspirates in both groups were also analyzed for sIL2R levels. To control for the effects of postnatal age (PNA) and study weight (SW) on the sIL2R levels, another group of 16 nonventilated babies (NVB) had plasma analyzed for sIL2R (PNA +/- SD: 39 +/- 40 days NVB vs. 48 +/- 36 days BPD; P = NS); (SW +/- SD: 1391 +/- 250 g NVB vs. 1212 +/- 700 g BPD; P = NS). The following data were obtained for the plasma sIL2R levels (mean +/- SEM U/mL): RDS controls, 1,231 +/- 80; BPD infants, 1,790 +/- 120; NVB controls, 1,319 +/- 76; P = 0.0005 RDS vs. BPD and P = 0.002 BPD vs. NVB. There was no significant difference in the sIL2R levels for the infants at risk of developing BPD vs. the infants with established BPD. Also, when analyzed separately, infants at risk of BPD and the infants with established BPD had higher sIL2R levels than the RDS and NVB controls. No differences were noted in the tracheal sIL2R levels in the BPD vs. RDS groups. These data indicate that infants with BPD had significantly higher sIL2R levels in plasma than either RDS or NVB controls. Therefore, lymphocyte activation may play a role in the pathogenesis of BPD.

Neonatal alloimmune thrombocytopenia after maternal immunization with paternal mononuclear cells: successful treatment with intravenous gamma globulin.

We report a case of neonatal alloimmune thrombocytopenia and intracranial hemorrhage in an infant whose mother received immunizations of paternal mononuclear cells. This therapy is designed to prevent unexplained first trimester miscarriages. No previous cases of platelet autoimmunization associated with maternal immunization with paternal mononuclear cells has been reported. Treatment with antenatal maternal infusions of intravenous gamma globulin (IVGG) did not prevent fetal thrombocytopenia, but IVGG may become the treatment of choice for postnatal, antibody-mediated thrombocytopenia of the newborn.

Predicting length of hospitalization of sick neonates from their initial status.

It has been suggested that the estimated date of confinement (EDC) may be used to predict the length of hospital stay for sick newborns. We have found this method unreliable and designed the following study to develop a mathematical model to predict length of stay (LOS). We reviewed the records of 393 neonates. Statistical analysis was performed using multiple linear regression. The factors that reached statistical significance were birth weight (BW), gestational age (GA), and a combined respiratory score (RES). The RES was developed to quantify the degree of initial respiratory illness. Through this model we developed the following formula: log(e) (LOS) = 4.395-0.023 (GA) -0.00054(BW) + 0.0274 (RES). The R value is 0.78. The model predicts an LOS +/- 10 days in 73% of cases. Overall, this model yields a 29% improvement in predictability of LOS compared with a model which used EDC only. This formula may provide useful information for parents and caregivers of hospitalized neonates.

Multicenter controlled trial comparing high-frequency jet ventilation and conventional mechanical ventilation in newborn infants with pulmonary interstitial emphysema.

One hundred forty-four newborn infants with pulmonary interstitial emphysema were stratified by weight and severity of illness, and randomly assigned to receive treatment with high-frequency jet ventilation (HFJV) or rapid-rate conventional mechanical ventilation (CV) with short inspiratory time. If criteria for treatment failure were met, crossover to the alternate ventilatory mode was permitted. Overall, 45 (61%) of 74 infants met treatment success criteria with HFJV compared with 26 (37%) of 70 treated with CV (p less than 0.01). Eighty-four percent of patients who crossed over from CV to HFJV initially responded to the new treatment, and 45% ultimately met success criteria on HFJV. In contrast, only 9% of those who crossed over from HFJV to CV responded well to CV (p less than 0.01), and the same 9% ultimately met success criteria (p less than 0.05). Therapy with HFJV resulted in improved ventilation at lower peak and mean airway pressures, as well as more rapid radiographic improvement of pulmonary interstitial emphysema, in comparison with rapid-rate CV. Survival by original assignment was identical. When survival resulting from rescue by the alternate therapy in crossover patients was excluded, the survival rate was 64.9% for HFJV, compared with 47.1% for CV (p less than 0.05). The incidence of chronic lung disease, intraventricular hemorrhage, patent ductus arteriosus, airway obstruction, and new air leak was similar in both groups. We conclude that HFJV, as used in this study, is safe and is more effective than rapid-rate CV in the treatment of newborn infants with pulmonary interstitial emphysema.

Closure of the ductus arteriosus with indomethacin in ventilated neonates with respiratory distress syndrome. Effects of pulmonary compliance and ventilation.

The reported effects of indomethacin on pulmonary compliance are variable depending upon the patient population and on the degree to which indomethacin resulted in successful ductal closure. Eleven fluid-restricted, furosemide-treated premature infants being mechanically ventilated for respiratory distress syndrome (RDS) who also had a significant patent ductus arteriosus (PDA) had pulmonary function testing performed before and after successful closure of the PDA. The diagnosis of a significant PDA was made by clinical and echocardiographic criteria. Indomethacin was administered at a dosage of 0.2 mg/kg/dose every 12 to 18 h for 1 to 3 doses. To control for the 48-h time interval to achieve ductal closure, nine premature infants being ventilated for RDS but who did not have a significant PDA also had pulmonary function evaluations performed before and after the 48 h. Also, to control for the independent effect of fluid restriction and diuretic therapy on pulmonary compliance, eight such premature infants with a PDA had pulmonary function evaluations performed at a 48-h interval. Successful closure of the ductus with indomethacin was associated with an improvement in compliance and ventilation parameters in all infants in the indomethacin-treated infants. In the indomethacin-treated group, the mean percent improvements were noted in the following parameters: CLdyn, 59.2%; CLI, 78.3%; CLE, 63.3%; VT, 63.3%; VE, 54.6%. There were no significant changes in the pulmonary functions in the 48-h RDS or the 48-h PDA fluid-restricted, furosemide-treated control groups. In conclusion, successful closure of the ductus with indomethacin causes a significant improvement in compliance and ventilation parameters in infants being mechanically ventilated for RDS.

Caffeine pharmacokinetics in preterm infants older than 2 weeks.

We studied 17 preterm infants receiving caffeine, and measured their plasma levels of caffeine and the theophylline metabolite by high-pressure liquid chromatography. The half-life was calculated by computer analysis using the least-square method. The mean gestational age of our patients was 29.7 +/- 1.9 weeks (mean +/- SD) and they were studied at 20.7 +/- 6.6 days (mean +/- SD) postnatal age. The caffeine half-life was 52.03 +/- 23.87 h (means +/- SD) and the theophylline half-life was 77.04 +/- 65.01 h (mean +/- SD).

Preductal and postductal transcutaneous oxygen tension measurements in premature newborns with hyaline membrane disease.

Pre- and postductal transcutaneous oxygen tension (tcPO2) was measured in 23 preterm infants with hyaline membrane disease to estimate the degree of right to left ductal shunting. The study was done during the first 24 hours of life and the data were recorded continuously. The studies were 11.9 +/- 4.0- hours long, (mean +/- SD) and the difference between the pre- and postductal tcPO2 (delta TcPO2) was measured every 100 seconds. Of 9,872 determinations, the delta tcPO2 was less than or equal to 15 mm Hg in 90.1%, 16 to 29 mm Hg in 9.3%, and greater than or equal to 30 mm Hg in 0.6%. It was concluded that large differences between pre- and postductal oxygen tension are rare in infants with hyaline membrane disease. The results of this and other published data suggest that it is unlikely that measurement of PaO2 in the descending aorta or tcPO2 below the ductus places infants at an increased risk of having retinopathy of prematurity develop.

Effect of birth order on intraventricular hemorrhage in very low birth weight twins.

We evaluated the effect of birth order on the incidence and severity of intraventricular hemorrhage in 29 sets of very low birth weight twins (1500 g or less). Intraventricular hemorrhage occurred in 55% of first-presenting twins and 62% of second-presenting twins. The incidence of minor intraventricular hemorrhage (grades I and II) was 41% for first twins and 52% for second twins, whereas the incidence of major intraventricular hemorrhage (grades III and IV) was 14 and 10%, respectively. None of these differences reached statistical significance. We conclude that birth order does not appear to have a major effect on the incidence or severity of intraventricular hemorrhage in very low birth weight twins.

Echocardiographic evaluation with hemodynamic correlation of the left ventricle in mitral stenosis.

The value of echocardiographic measurement of left ventricle volume and ejection fraction and of mitral valve area in patients with mitral stenosis has been assessed. All patients referred because of mitral stenosis have been studied by M-mode and two-dimensional echocardiography, by right and left heart catheterization, ventriculography and selective coronary angiography. Patients with other congenital or acquired heart disease or with coronary artery disease were excluded from this study. The selection lead to a series of 39 pts, all with a typical history of Rheumatic Fever: all these pts were divided in three groups according to the degree of mitral stenosis, classified as "mild", if valvular area was greater than 1.8 cm2 (8 pts), "moderate" between 1-1.8 cm2 (15 pts) and "severe" if valvular areas was less than 1 cm2 (16 pts). In all groups echocardiography underestimated left ventricular volumes as well as stroke volumes. Statistical correlation has not been excellent: the best result was again obtained in the calculation of ejection fraction (r = 0.91, P less than 0.001), confirming our previous results in a group of patients with mitral regurgitation. Cross-sectional two-dimensional echocardiography has confirmed, furthermore, as a sensitive and suitable procedure in assessing the mitral valve area (r = 0,87 P less than 0.001); at our experience echocardiography under-estimate mitral area in the cases of severe mitral stenosis.