Challenges, successes and opportunities for reducing readmissions in a referral-based children's hospital NICU.

AIM: To evaluate readmission data in a level IV neonatal intensive care unit (NICU) to identify patient characteristics and process failures which serve as drivers for readmission. METHODS: Our center is a primary referral center in Central and Southeast Ohio, providing us a unique opportunity to evaluate readmissions. We studied our current discharge process, caregiver perception of discharge readiness, parental comfort and the pre-discharge and post-discharge characteristics of infants. RESULTS: Our readmission rate during the 4 year period has remained stable at 9.8%. 74% of the caregivers rated that their perception of their baby's medical readiness for discharge was good to excellent. Duration of hospitalization and public insurance coverage (Medicaid) were significant risk factors for readmission (p = 0.00). In our setting, the majority of the patients are readmitted through the emergency department and nearly half of all readmissions were for 3 or fewer days. Patients discharged from our Comprehensive Center for BPD had similar readmission rate despite characteristics which should increase their readmission rate. CONCLUSIONS: Readmission rate is a poor indicator of the quality of care provided in the NICU. In addition to patient factors such as longer length of stay and Medicaid, our data suggests that preventing readmission depends on having systems in place to help families cope with transition of care after discharge.

Empowerment programme for parents of extremely premature infants significantly reduced length of stay and readmission rates.

AIM: Extremely premature infants experience long hospitalisation and high readmission rates within 30 days of discharge. This quality control investigation retrospectively explored the impact of the Creating Opportunities for Parent Empowerment (COPE©) programme on these rates in an all referral, Level-IV small baby intensive care unit setting. METHODS: The parents of 303 extremely premature infants participated in the study. Of those, 135 were admitted before the implementation of COPE© and served as baseline historical controls, and the remaining 168 neonates received the intervention. Length of stay and readmission rates were analysed using parametric and nonparametric tests. Additional analyses were used to compare the two groups in terms of gestational age, birthweight and other acuity measures. RESULTS: Neonates who received COPE© required significantly reduced lengths of stay than the control neonates (COPE© 127.1 ± 55.8 days vs. control 139.6 ± 61.9 days, p < 0.05) and significantly lower readmission rates (COPE© 23.9% vs. control 13.2%, p = 0.05). CONCLUSION: The COPE© programme promoted active parental engagement in the unit and significantly reduced hospital stays and readmission rates. Future interventions should identify the specific components of the programme that support the parents of extremely premature infants during the various phases of hospitalisation.

Severity of neonatal retinopathy of prematurity is predictive of neurodevelopmental functional outcome at age 5.5 years. Behalf of the Cryotherapy for Retinopathy of Prematurity Cooperative Group.

OBJECTIVE: The purpose of this study was to assess the relation between neonatal retinopathy of prematurity (ROP) in very low birth weight infants and neurodevelopmental function at age 5.5 years. METHODS: Longitudinal follow-up of children occurred in 2 cohorts of the Multicenter Cryotherapy for Retinopathy of Prematurity Study. The extended natural history cohort followed 1199 survivors of <1251 g birth weight from 5 centers. The threshold randomized cohort (ThRz) followed 255 infants <1251 g from 23 centers who developed threshold ROP and who consented to cryotherapy to not more than 1 eye. At 5.5 years both cohorts had ophthalmic and acuity testing and neurodevelopmental functional status determined with the Functional Independence Measure for Children (WeeFIM). RESULTS: Evaluations were completed on 88.7% of the extended natural history cohort; 87% had globally normal functional skills (WeeFIM: >95). As ROP severity increased, rates of severe disability increased from 3.7% among those with no ROP, to 19.7% of those with threshold ROP. Multiple logistic regression analysis demonstrated that better functional status was associated with favorable visual acuity, favorable 2-year neurological score, absence of threshold ROP, having private health insurance, and black race. Evaluations were completed on 87.4% of the ThRz children. In each functional domain, the 134 children with favorable acuity in their better eye had fewer disabilities than did the 82 children with unfavorable acuity: self-care disability 25.4% versus 76.8%, continency disability 4.5% versus 50.0%, motor disability 5.2% versus 42.7%, and communicative-social cognitive disability 22.4% versus 65.9%, respectively. CONCLUSION: Severity of neonatal ROP seems to be a marker for functional disability at age 5. 5 years among very low birth weight survivors. High rates of functional limitations in multiple domains occur in children who had threshold ROP, particularly if they have unfavorable visual acuity.

Effects of lactose intake on lactose digestion and colonic fermentation in preterm infants.

The purpose of this study was to determine whether doubling the lactose concentration in formula for preterm infants lowers the fraction of lactose digested and/or increases the fraction of lactose fermented. Six preterm infants, 31 to 36 weeks' postconceptional age, were fed a standard preterm formula (carbohydrate is 50% lactose and 50% glucose polymer)(SC) and/or the same volume of formula modified to contain lactose as the sole carbohydrate (LAC). Relative lactose digestion during the LAC formula feeding compared with SC formula feeding was measured by using a stable isotope approach for quantifying the fractional contribution of formula lactose to plasma glucose enrichment. Relative lactose digestion was 0.98 +/- 0.17 (range, 0.70 to 1.19). Fractional fermentation of lactose was estimated from breath H2 excretion (0.52 +/- 0.34 during LAC feeding and 0.23 +/- 0.22 during SC feeding, P = .11). The rate of breath H2 excretion was much higher with LAC (1.34 +/- 0.98 mL/h) than with SC (0.27 +/- 0.29, P = .029). In conclusion, doubling the lactose concentration had only modest effects on fractional lactose digestion. Increased breath H2 excretion with LAC may relate to fermentation of nonlactose sugar or to ill-defined changes in colonic physiology or motility, which could enhance colonic fermentation of malabsorbed sugar by H2-producing bacteria.

Technical note: pig model for studying nutrient assimilation by the intestine and colon.

We have developed a system for chronically catheterizing 10- to 25-d-old pigs that permits stable isotope tracer studies of intestinal or colonic assimilation of nutrients. This model also can be used to ensure constant enteral feeding or to assess the rate of entry into the terminal ileum of carbohydrates, fats, and amino acids. A plastic cannula with a luminal flange can be surgically placed in the stomach for tracer studies of sugar digestion or for controlled infusion of any formula diet. A similar cannula can be placed in the cecum for infusion of tracer and(or) substrates for studies of fermentation. The cannula has been machined so that a washer and nut can be threaded onto it, allowing the entire apparatus to be fixed to the abdominal wall. The distal end protruding above the skin was tapered to fit standard i.v. extension tubing. A carotid arterial catheter was used to sample substrates for isotopic enrichment measurements.

Estimation of CO2 production in enterally fed preterm infants using an isotope dilution stable tracer technique.

BACKGROUND: Estimates of the rate of CO2 production may be useful in preterm infants, but assessment of the rate of respiratory excretion of CO2 (VCO2) may not always be practical in infants requiring constant care. We hypothesized that the rate of dilution of 13CO2 (RaCO2) would be a valid index of CO2 production in preterm infants. METHODS: Twelve studies of RaCO2 and VCO2 were performed in six enterally fed preterm infants. RaCO2 was measured using a 2-hour, primed, constant, orogastric infusion of NaH13CO3 with formula and an assessment of the plateau 13C enrichment of expired CO2. VCO2 was measured over two 10-minute intervals during the infusion using a flow-through system. Energy expenditure was estimated from these data and the food quotient. RESULTS: Mean (+/- SD) rate of CO2 production using RaCO2 (348 +/- 32 mumol/kg/min) was 114% of that estimated using VCO2 (304 +/- 51 mumol/kg/min). The ratio of VCO2/RaCO2 is equal to the fractional recovery of tracer CO2 in the expired air during the course of the tracer infusion. In studies of short duration, this ratio is generally less than 100% because of isotope exchange. For five pairs of studies performed on consecutive days, each individual value of RaCO2 on day 2 was multiplied by the mean of the individual ratios of VCO2/RaCO2 on day 1 (0.78); corrected RaCO2 was 306 +/- 19 mumol/kg/min compared with 307 +/- 59 mumol/kg/min for VCO2. CONCLUSIONS: Thus, RaCO2, particularly when corrected for isotope recovery, may be a useful index of group mean CO2 production and energy expenditure in preterm infants.

In vivo lactose digestion in preterm infants.

In vitro studies of intestinal lactase activity and breath-hydrogen studies have suggested that the capacity for lactose digestion in preterm infants is less than the usual intake. To explore this question using an in vivo approach, we determined the fraction of dietary lactose hydrolyzed to glucose (and galactose) in 14 preterm infants with a gestational age of 26-31 wk at the time of birth but a postconceptional age of 31-37 wk at the time of study. The percentage of lactose digested was estimated after 6-h, primed, constant gastric infusions of [1-(13)C]glucose and D-[-1-(13)C]lactose on alternate days. A coefficient of lactose fermentation was derived from the rates of pulmonary excretion of hydrogen and carbon dioxide. Mean (+/- SD) lactose digestion was 79 +/- 26%. There was a significant inverse rank (r = -0.799, P < 0.01) and linear (r = -0.587, P < 0.05) correlation between this variable and postconceptional age. The percentage of lactose fermented averaged 35 +/- 27%.

Pharmacokinetics of dexamethasone in premature neonates.

OBJECTIVE: Dexamethasone is frequently used in premature neonates with bronchopulmonary dysplasia, however little is known about its disposition in this population. METHODS: We evaluated the pharmacokinetics of dexamethasone in 9 premature neonates with a mean gestational age of 27.3 weeks and a postnatal age of 21.8 days. RESULTS: There was a strong relationship between clearance (4.96 ml.min-1.kg-1) and gestational age ( r = 0.884). Pharmacokinetic parameters were grouped based on a gestational age of less than 27 weeks (Group I) and greater than 27 weeks (Group II). Mean clearance in group I and group II was 1.69 and 7,57 ml.min-1.kg-1, respectively. Mean distribution volume in group I and II was 1.26 and 2.19 l.kg-1, respectively. No significant relationships were noted between the disposition of dexamethasone and ventilator requirements or adverse effects. CONCLUSION: The pharmacokinetics of dexamethasone in premature neonates was related to gestational age.

A portable device for estimating the rate of CO2 production in preterm infants.

We developed a simple flow-through system for measuring the net rate of pulmonary excretion of CO2 (VCO2). Its components, connected in series by plastic tubing, are the following: plastic oxygen hood, suction flask used both as a mixing and sampling chamber, Brooks mass flow controller, and wall suction. Gas aliquots withdrawn from the sampling flask are analyzed for CO2 concentration in the laboratory. The system was validated by measuring over a 60-min period the evolution of CO2 from the reaction of sodium carbonate with phosphoric acid. In 9 trials, the mean recovery of CO2 was 99.8%. In 14 studies in 11 preterm infants, we evaluated the reproducibility of duplicate measurements of CO2 production on consecutive days; in 9 of the studies, the coefficient of variation was < 5%, but in other 5 studies, it ranged from 7.7 to 35.8%. We believe that this device might have both research and clinical applicability.

Comparison of methods for estimating fecal carbohydrate excretion in premature infants.

Our previous studies have suggested that there is minimal fecal excretion of carbohydrate-derived energy in premature infants fed lactose or glucose-polymer as the source of carbohydrate. In these studies, carbohydrate energy excretion was equated with the non-nitrogenous, nonfat energy. Although these studies provided consistent results, this methodological approach cannot distinguish actual excretion of dietary carbohydrate from carbohydrate in the form of large molecular compounds derived from infant or bacterial cells (which in fact comprise approximately 90% of the measured value). Therefore, in this study we compared the absorption of carbohydrate energy to the fractional absorption of 13C derived from [D-1-13C]-lactose in 10 premature infants 30-32 weeks gestational age fed either a commercial premature infant formula containing a mixture of carbohydrates (50% lactose: 50% glucose polymer) or the same formula in which lactose was the sole carbohydrate. The two methods provided comparable estimates of carbohydrate absorption (96 and 94%, respectively), although there were significant discrepancies in two infants. These studies provide evidence of external validity for the two methods.

In vivo estimation of lactose hydrolysis in premature infants using a dual stable tracer technique.

To investigate their putative capacity for lactose digestion, primed continuous orogastric infusions of [1-13C]glucose and D-[1-13C]lactose were administered on consecutive days to five premature infants (30-31 wk gestation, 15-32 days of age), who were fed by orogastric infusions of human milk or formula. By monitoring the plateau isotopic enrichment of plasma glucose using isotopomers containing the entire derivatized glucose molecule or C-2 through C-6, we were able to distinguish label appearing in the peripheral circulation deriving from unmetabolized glucose from that arising from recycled or fermented glucose (or lactose). Isotopic enrichment of the C-1 of glucose, corrected for recycling, was then calculated during each tracer infusion, and the fraction of dietary lactose subjected to in vivo hydrolysis was estimated from these values and the respective tracer infusion rates, assuming similar absorptive and metabolic fates of labeled glucose arising from either tracer. This fraction averaged 1.02 +/- 0.16 (SD), suggesting that lactose digestion is efficient by 34-wk postconceptional age.

Efficient assimilation of lactose carbon in premature infants.

Previous studies have shown that there is minimal fecal excretion of carbohydrate energy in premature infants with high breath H2 concentrations, implying that the colon plays an important role in salvaging lactose energy. The present study extends this work by using a stable isotope method for serial assessment of lactose carbon excretion. Thus, in nine premature infants (30-32 weeks of gestation and 11-36 days of age), we conducted 23 longitudinal studies of breath hydrogen concentration and the cumulative fecal excretion of 13C derived from a single dose of [D-1-13C]lactose. The ranges of values (mean +/- SD) for breath hydrogen concentration and lactose absorption were respectively 4-440 ppm/5% CO2 (92 +/- 100) and 82-100% tracer dose (98.4 +/- 3.9). In 21 studies, lactose carbon excretion was less than 2% of the dose. There was a significant correlation between postnatal age and breath hydrogen concentration (R = 0.48; p = 0.02). These studies suggest that colonic fermentation activity is adequate for efficient colonic salvage of lactose even during the second week of life.

Stable isotope model for estimating colonic acetate production in premature infants.

In premature infants, a nutritionally significant proportion of lactose is apparently fermented in the colon to acetate. To estimate the rate of entry of acetate into the peripheral circulation, a model that takes into account extraction of gut-derived acetate by splanchnic and hepatic tissues was developed. Using a [1-13C]acetate orogastric infusion technique, six studies were carried out on five premature infants during constant orogastric feeding. Ranges in gestational age, postnatal age, and breath H2 concentration (corrected for CO2 content) were 28-32 weeks, 16-29 days, and 45-252 microL/L, respectively. The estimated rate of entry of acetic acid (mean +/- SD) was 63.7 +/- 33.8 mumol.kg-1.min-1 (range, 22.9-123.2 mumol.kg-1.min-1), which corresponded to 64.3% +/- 38.6% (24%-136%) of the potential two carbon units from dietary lactose. Thus, a substantial fraction of dietary lactose in premature infants may be converted to acetic acid; this conversion could have a significant effect on protein as well as energy requirements.

Comparison of two neonatal intravenous amino acid formulations in preterm infants: a multicenter study.

Two neonatal intravenous amino acid solutions (Aminosyn-PF and Troph-Amine) were compared in 44 preterm infants. The rate of weight gain, nitrogen balance, and changes in plasma aminograms were determined over 7 days to ascertain whether different outcomes could be identified for the two solutions. At study entry, the infants received a minimum infusion of 2 g amino acid/kg/d with 50 or more nonprotein kcal/kg/d. Group mean amino acid intake over the study period was approximately 2.6 g/kg/d for both groups; nonprotein caloric intake approximated 90 kcal/kg/d. Results showed no significant differences between solutions for the rate of weight gain, nitrogen balance, and nitrogen retention, which approximated intrauterine rates. The rate of weight gain averaged nearly 15 g/kg/d for both solutions. Differences between day 0 and day 7 plasma aminograms showed significant changes between solutions for histidine, lysine, methionine, phenylalanine, threonine, and glutamic acid. However, day 7 plasma aminograms for both solutions compared favorably with those from enterally fed preterm infants reported in the literature. Failure to identify significant differences for the rate of weight gain, nitrogen balance, or nitrogen retention between the two groups suggests that differences in plasma aminograms resulting from use of one solution or the other had no short-term clinical consequences in the premature infants studied.

Necrotizing tracheobronchitis (NTB) following high frequency ventilation: role of an angiotensin converting enzyme inhibitor.

The purpose of this investigation was to study the effect of an angiotensin converting enzyme inhibitor (enalaprilat) on the morphologic manifestations of experimentally induced necrotizing tracheobronchitis (NTB). Twenty piglets were anesthetized before saline lung lavage. High frequency flow interrupter (HFFI) ventilation was used with a strategy known to produce NTB. Animals were randomly assigned to receive IV enalaprilat 0.1 mg/kg (ENP-Hi), enalaprilat 0.01 mg/kg (ENP-Lo), or saline (C). After 8 hours of ventilation, the piglets were sacrificed. Total airway injury scores (mean +/- S.D.) were 1.2 +/- 0.7 for ENP-Hi, 0.2 +/- 0.2 for ENP-Lo, and 21.3 +/- 16 for group C. Enalaprilat minimizes NTB lesions in neonatal piglets exposed to high frequency oscillatory ventilation. Although the origin of NTB is multifactorial, airway mucosa ischemia may play an important role. Enalaprilat may compensate for the reduction of mucosal blood flow by limiting formation of angiotensin II and/or preventing degradation of bradykinin.

Dopamine pharmacokinetics in critically ill newborn infants.

Dopamine is frequently used in critically ill newborn infants for treatment of shock and cardiac failure, but its pharmacokinetics has not been evaluated using a specific analytical method. Steady-state arterial plasma concentrations of dopamine were measured in 11 seriously ill infants receiving dopamine infusion, 5-20 micrograms.kg-1.min-1, for presumed or proven sepsis and hypotensive shock. Steady-state concentrations of dopamine ranged from 0.013-0.3 microgram/ml. Total body clearance averaged 115 ml.kg-1.min-1. The apparent volume of distribution and elimination half life averaged 1.8 l.kg-1 and 6.9 min, respectively. No relationship was observed between dopamine pharmacokinetics and gestational age, postnatal age or birthweight. Substantial interindividual variation was seen in dopamine pharmacokinetics in seriously ill infants, and plasma concentrations could not be predicted accurately from its infusion rate. Marked variation in clearance explains in part, the wide dose requirements of dopamine needed to elicit clinical response in critically ill newborn infants.

28-day survival rates of 6676 neonates with birth weights of 1250 grams or less.

Survival rates specific for birth weight, gestational age, sex, and race are described for 6676 inborn neonates who weighed less than 1251 g at birth and were born during 1986 through 1987. Overall 28-day survival increased with gestational age and birth weight, from 36.5% at 24 weeks' gestation to 89.9% at 29 weeks' gestation, or from 30.0% for neonates of 500 through 599 g birth weight to 91.3% for neonates of 1200 through 1250 g. The expected birth weight-specific survival advantage for female neonates and black neonates diminished when the data were controlled for gestational age, showing that certain previously reported survival advantages are based on lower birth weight for a given gestational age. Multivariate analysis showed that all tested variables were significant predictors for survival, in order of descending significance: gestational age and birth weight, sex, race, single birth, and small-for-gestational-age status. The powerful effect of gestational age on survival highlights the need for an accurate neonatal tool to assess the gestational age of very low birth weight neonates after birth.

The effect of three intravenous fat emulsions containing different concentrations of linoleic and alpha-linolenic acids on the plasma total fatty acid profile of neonates.

We determined the fatty acid profile of total plasma lipids in infants who received one of three intravenous fat emulsions that differed primarily in their linoleic and alpha-linolenic acid content: (I) a safflower oil emulsion, (II) a 50:50 mixture of safflower and soybean oils, or (III) a soybean oil emulsion. After 2 weeks of fat therapy, oleic acid, expressed as a percentage of total plasma lipid fatty acids, decreased in all groups, but less so in group III (p less than 0.01). The linoleic acid percentage increased in all groups, but group I had the greatest increase (p less than 0.05). Group II patients had higher percentages of the linoleic acid metabolites, dihomo-gamma-linolenic acid (II greater than I, p less than 0.05; II greater than III, p less than 0.01) and arachidonic acid (II greater than III, p less than 0.05). Group II patients also had higher levels of alpha-linolenic acid (II greater than I, p less than 0.05) and its metabolite, eicosapentaenoic acid (II greater than I, p less than 0.05). Another alpha-linolenic acid metabolite, docosahexaenoic acid, however, increased in group III, remained stable in group II, and decreased in group I (III and II greater than I, p less than 0.05). We conclude that the content of linoleic acid and alpha-linolenic acid in intravenous fat emulsions results in statistically significant changes in the fatty acid profile of total plasma lipids in infants receiving total parenteral nutrition.

A simple technique to feed newborn piglets.

We describe a simple and relatively inexpensive technique for feeding conventionally reared newborn pigs. The feeding system consists of a nipple assembly and an enteral feeding bag. Piglets are weaned at 4-10 h of age and are initially bottle-fed a simulated sow's milk formula. After adaptation to the synthetic nipple and artificial milk, the piglets are placed on the semiautomatic feeding system. Piglets fed in this manner gain weight at rates comparable to sow-reared animals or animals fed via completely automated systems.