Neonatal intensive care unit: predictive models for length of stay

Factors affecting length of stay according to bronchopulmonary dysplasia severity: a nationwide cohort study in Korea

BACKGROUND

Longer hospitalizations for preterm infants with bronchopulmonary dysplasia (BPD) delay developmental outcomes, increase the risk for hospital-acquired complications, and exert a substantial socioeconomic burden. This study aimed to identify factors associated with an extended length of stay (LOS) at different levels of severity of BPD.

METHODS

A cohort study was conducted using the Korean Neonatal Network registry of very low birth weight infants with BPD between 2013 and 2017 through retrospective analysis.

RESULTS

A total of 4263 infants were diagnosed with BPD. For mild BPD, infants requiring surgical treatment for patent ductus arteriosus needed a longer LOS [eadjusted β coefficients (adj β) 1.041; 95% confidence interval (CI): 0.01-0.08] and hydrocephalus (eadj β 1.094; 95% CI 0.01-0.17). In moderate BPD, infants administered steroids or with intraventricular hemorrhage required a longer LOS (eadj β 1.041; 95% CI 0.00-0.07 and eadj β 1.271; 95% CI 0.11-0.38, respectively). In severe BPD, infants with comorbidities required a longer LOS: pulmonary hypertension (eadj β 1.174; 95% CI 0.09-0.23), administrated steroid for BPD (eadj β 1.116; 95% CI 0.07-0.14), sepsis (eadj β 1.062; 95% CI 0.01-0.11), patent ductus arteriosus requiring surgical ligation (eadj β 1.041; 95% CI 0.00-0.08), and intraventricular hemorrhage (eadj β 1.016; 95% CI 0.05-0.26). Additionally, the higher the clinical risk index score, the longer the LOS needed for infants in all groups.

CONCLUSIONS

The factors affecting LOS differed according to the severity of BPD. Individualized approaches to reducing LOS may be devised using knowledge of the various risk factors affecting LOS by BPD severity.

Development and validation of a novel risk classification tool for predicting long length of stay in NICU blood transfusion infants

Newborns are as the primary recipients of blood transfusions. There is a possibility of an association between blood transfusion and unfavorable outcomes. Such complications not only imperil the lives of newborns but also cause long hospitalization. Our objective is to explore the predictor variables that may lead to extended hospital stays in neonatal intensive care unit (NICU) patients who have undergone blood transfusions and develop a predictive nomogram. A retrospective review of 539 neonates who underwent blood transfusion was conducted using median and interquartile ranges to describe their length of stay (LOS). Neonates with LOS above the 75th percentile (P75) were categorized as having a long LOS. The Least Absolute Shrinkage and Selection Operator (LASSO) regression method was employed to screen variables and construct a risk model for long LOS. A multiple logistic regression prediction model was then constructed using the selected variables from the LASSO regression model. The significance of the prediction model was evaluated by calculating the area under the ROC curve (AUC) and assessing the confidence interval around the AUC. The calibration curve is used to further validate the model's calibration and predictability. The model's clinical effectiveness was assessed through decision curve analysis. To evaluate the generalizability of the model, fivefold cross-validation was employed. Internal validation of the models was performed using bootstrap validation. Among the 539 infants who received blood transfusions, 398 infants (P75) had a length of stay (LOS) within the normal range of 34 days, according to the interquartile range. However, 141 infants (P75) experienced long LOS beyond the normal range. The predictive model included six variables: gestational age (GA) (< 28 weeks), birth weight (BW) (< 1000 g), type of respiratory support, umbilical venous catheter (UVC), sepsis, and resuscitation frequency. The area under the receiver operating characteristic (ROC) curve (AUC) for the training set was 0.851 (95% CI 0.805-0.891), and for the validation set, it was 0.859 (95% CI 0.789-0.920). Fivefold cross-validation indicates that the model has good generalization ability. The calibration curve demonstrated a strong correlation between the predicted risk and the observed actual risk, indicating good consistency. When the intervention threshold was set at 2%, the decision curve analysis indicated that the model had greater clinical utility. The results of our study have led to the development of a novel nomogram that can assist clinicians in predicting the probability of long hospitalization in blood transfused infants with reasonable accuracy. Our findings indicate that GA (< 28 weeks), BW(< 1000 g), type of respiratory support, UVC, sepsis, and resuscitation frequency are associated with a higher likelihood of extended hospital stays among newborns who have received blood transfusions.

Developing a length of stay prediction model for newborns, achieving better accuracy with greater usability

BACKGROUND

One in ten newborn children is born prematurely. The elongated length of stay (LOS) of these children in the Neonatal Intensive Care Unit (NICU) has important implications on hospital occupancy figures, healthcare and management costs, as well as the psychology of parents. In order to allow accurate planning and resource allocation, this study aims to create a generalizable and robust model to predict the NICU LOS of preterm newborns.

METHODS

Data were collected from a large tertiary center NICU between 2011 and 2018 and relates to 5,362 newborns. The selected model was externally validated using a data set of 8,768 newborns from another tertiary center NICU. This report compares several models, such as Random Forest (RF), quantile RF, and other feature selection methods, including LASSO and AIC step-forward selection. In addition, a novel step-forward selection based on False Discovery Rate (FDR) for quantile regression is presented and evaluated.

RESULTS

A high-orderquantile regression model for predicting preterm newborns' LOS that uses only four features available at birth had more attractive properties than other richer ones. The model achieved a Mean Absolute Error (MAE) of 6.26 days on the internal validation set (average LOS 27.04) and an MAE of 6.04 days on the external validation set (average LOS 29.32). The suggested model surpassed the accuracy obtained by models in the literature. It is shown empirically that the FDR-based selection has better properties than the AIC-based step-forward selection approach.

CONCLUSION

This paper demonstrates a process to create a predictive model for NICU LOS in preterm newborns, where each step is reasoned. We obtain a simple and robust model for NICU LOS prediction, which achieves far better results than the current model used for financing NICUs. Utilizing this model, we have created an easy-to-use online web application to ease parents' worries and to assist NICU management: https://tzviel.shinyapps.io/calcuLOS.

Predictors of length of hospital stay for preterm infants in Ethiopia: a competing risk analysis

Background

Length of hospital stay (LOS) is one of the essential indicators for evaluating the efficiency and the quality-of-care service delivered. predicting LOS is critical for resource allocation, decision-making, lowering neonatal morbidity and death, enhancing clinical outcomes and parent counseling. In addition, extended hospital stays (long LOS_NICU) place a burden on the healthcare systems decreasing bed turnover rates as well as their financial stand and the mental stress on families. In Ethiopia, there is limited evidence on the determinant factors that influence on LOS.

Objectives

To determine factors affecting neonatal intensive care unit length of stay for all preterm newborns who were discharged alive.

Method

The study used a secondary data source, was collected for the Study of Illness in Preterm (SIP) infants project. The research study was a multicenter, cross-sectional, observational clinical study that took place in five Ethiopia hospitals from July 1, 2016, to May 31, 2018. The predictors of LOS were determined using Fine-Gray's competing risk analysis.

Results

For this study 3,511 preterm infants admitted to the NICU were analyzed. About 28.8% of the preterm infants died during their time in neonatal care while 66.6% were discharged alive. At the end of the study 4.6% babies were still in the NICU. The overall median LOS (death or discharge) was 7 days, with an interquartile range of 8 days. The cumulative incidence of discharge rose with increasing in gestational age and birth weight, on the contrary, the rate of discharge was decreased by 45.7% with the development of RDS (SDH ratio: 0.543), by 75.9% with the development of apnea (SDH ratio: 0.241), by 36.2% with sepsis, and by 43.6% with pneumonia (SDH ratio: 0.564).

Conclusions

Preterm newborns with a low gestational age and birth weight have a greater probability of having a prolonged LOS. Complications of the medical conditions RDS, apnea, sepsis, pneumonia, anemia, asphyxia, and NEC substantially raise LOS considerably.

A case-based reasoning system for neonatal survival and LOS prediction in neonatal intensive care units: a development and validation study

Early prediction of neonates' survival and Length of Stay (LOS) in Neonatal Intensive Care Units (NICU) is effective in decision-making. We developed an intelligent system to predict neonatal survival and LOS using the "Case-Based Reasoning" (CBR) method. We developed a web-based CBR system based on K-Nearest Neighborhood (KNN) on 1682 neonates and 17 variables for mortality and 13 variables for LOS and evaluated the system with 336 retrospectively collected data. We implemented the system in a NICU to externally validate the system and evaluate the system prediction acceptability and usability. Our internal validation on the balanced case base showed high accuracy (97.02%), and F-score (0.984) for survival prediction. The root Mean Square Error (RMSE) for LOS was 4.78 days. External validation on the balanced case base indicated high accuracy (98.91%), and F-score (0.993) to predict survival. RMSE for LOS was 3.27 days. Usability evaluation showed that more than half of the issues identified were related to appearance and rated as a low priority to be fixed. Acceptability assessment showed a high acceptance and confidence in responses. The usability score (80.71) indicated high system usability for neonatologists. This system is available at http://neonatalcdss.ir/ . Positive results of our system in terms of performance, acceptability, and usability indicated this system can be used to improve neonatal care.

When should a premature neonate (24-35 weeks) be discharged? five-year experience from a single-center neonatal intensive care unit in Türkiye

Background/aim

The survival rate among preterm infants has improved, and hospital stays have been prolonged, consistent with positive developments in perinatal and neonatal care. The aim of this study was to provide evidence-based information for healthcare professionals concerning the ideal time for discharge by evaluating the reasons for prolonged hospital stays.

Materials and methods

Six hundred eighty-one premature babies born at 24-35 weeks at the Mersin University Medical Faculty Hospital between January 2016 and May 2020 and admitted to the neonatal intensive care unit were included in the study following a retrospective file examination. Date of birth (gestational age) and discharge week (duration of hospital stay) calculated from the date of final discharge were recorded. Based on the literature, the ideal discharge time was determined to be 40 weeks according to postmenstrual age (week of birth + length of hospital stay). The primary variable was whether the infants were discharged before the ideal discharge week. The secondary variable was the effect of the presence of comorbidity on the length of hospital stay and ideal discharge time.

Results

The mean hospital stay of preterm neonate born at 250-7-260-7, 270-7-280-7 and 290-7-300-7 weeks was significantly shorter in the absence of comorbidity than in the presence of comorbidity (p = 0.001, 0.004, and 0.008, respectively). More than half (53.5%) were discharged before the expected date of birth as gestational weeks increased, despite the prolonged length of stay in the presence of comorbidity.

Conclusion

Health professionals can inform families that, in the absence of comorbidity, discharge is possible at an average of 36 weeks for 250-7-280-7-week gestational ages, and at an average of 34 weeks for 290-7-320-7-week gestational ages.

Risk factors for length of NICU stay of newborns: A systematic review

Background

The improvement in survival of preterm infants is accompanied by an increase in neonatal intensive care unit (NICU) admissions. Prolonged length of stay in the NICU (LOS-NICU) increases the incidence of neonatal complications and even mortality and places a significant economic burden on families and strain on healthcare systems. This review aims to identify risk factors influencing LOS-NICU of newborns and to provide a basis for interventions to shorten LOS-NICU and avoid prolonged LOS-NICU.

Methods

A systematic literature search was conducted in PubMed, Web of Science, Embase, and Cochrane library for studies that were published in English from January 1994 to October 2022. The PRISMA guidelines were followed in all phases of this systematic review. The Quality in Prognostic Studies (QUIPS) tool was used to assess methodological quality.

Results

Twenty-three studies were included, 5 of which were of high quality and 18 of moderate quality, with no low-quality literature. The studies reported 58 possible risk factors in six broad categories (inherent factors; antenatal treatment and maternal factors; diseases and adverse conditions of the newborn; treatment of the newborn; clinical scores and laboratory indicators; organizational factors).

Conclusions

We identified several of the most critical risk factors affecting LOS-NICU, including birth weight, gestational age, sepsis, necrotizing enterocolitis, bronchopulmonary dysplasia, and retinopathy of prematurity. As only a few high-quality studies are available at present, well-designed and more extensive prospective studies investigating the risk factors affecting LOS-NICU are still needed in the future.

Predicting clinical outcomes using artificial intelligence and machine learning in neonatal intensive care units: a systematic review

BACKGROUND

Advances in technology, data availability, and analytics have helped improve quality of care in the neonatal intensive care unit.

OBJECTIVE

To provide an in-depth review of artificial intelligence (AI) and machine learning techniques being utilized to predict neonatal outcomes.

METHODS

The PRISMA protocol was followed that considered articles from established digital repositories. Included articles were categorized based on predictions of: (a) major neonatal morbidities such as sepsis, bronchopulmonary dysplasia, intraventricular hemorrhage, necrotizing enterocolitis, and retinopathy of prematurity; (b) mortality; and (c) length of stay.

RESULTS

A total of 366 studies were considered; 68 studies were eligible for inclusion in the review. The current set of predictor models are primarily built on supervised learning and mostly used regression models built on retrospective data.

CONCLUSION

With the availability of EMR data and data-sharing of NICU outcomes across neonatal research networks, machine learning algorithms have shown breakthrough performance in predicting neonatal disease.

Associação entre peso ao nascer, idade gestacional e diagnósticos secundários na permanência hospitalar de recém-nascidos prematuros

Objetivo: verificar a associação entre peso ao nascer, idade gestacional e diagnósticos médicos secundários no tempo de permanência hospitalar de recém-nascidos prematuros. Métodos: estudo transversal, com 1.329 prontuários de recém-nascidos no período de julho de 2012 a setembro de 2015, em dois hospitais de Belo Horizonte, que utilizam o sistema Diagnosis Related Groups Brasil. Para determinar um ponto de corte para o peso ao nascer e a idade gestacional no nascimento que melhor determinasse o tempo de internação foi utilizada a curva Receive Operator Characteristic. Posteriormente, utilizou-se o teste de Análise de Variância e teste de Duncan para a comparação entre a média de tempo de permanência hospitalar. Resultados: a prematuridade sem problemas maiores (DRG792) foi a categoria mais prevalente (43,12%). O maior tempo médio de internação foi de 34,9 dias, identificado entre os recém-nascidos prematuros ou com Síndrome da angústia respiratória (DRG 790). A combinação de menor peso ao nascer e menor IG ao nascimento apresentaram o maior risco de permanência hospitalar, aumentada quando comparados ao demais perfis formados para esse DRG. Conclusão: os achados poderão direcionar a assistência em relação à mobilização de recursos físicos, humanos e de bens de consumo, além da análise crítica de condições que influenciam os desfechos clínicos. A possibilidade da otimização do uso desses recursos hospitalares aliada à melhoria da qualidade dos atendimentos e da segurança dos pacientes está associada a uma minimização do tempo de permanência hospitalar e da carga de morbidade e mortalidade neonatal.

A High-Fidelity Model to Predict Length-of-Stay in the Neonatal Intensive Care Unit (NICU)

Having an interpretable dynamic length-of-stay (LOS) model can help hospital administrators and clinicians make better decisions and improve the quality of care. The widespread implementation of electronic medical record (EMR) systems has enabled hospitals to collect massive amounts of health data. However, how to integrate this deluge of data into healthcare operations remains unclear. We propose a framework grounded in established clinical knowledge to model patients' lengths-of-stay. In particular, we impose expert knowledge when grouping raw clinical data into medically meaningful variables, which summarize patients' health trajectories. We use dynamic predictive models to output patients' remaining lengths-of-stay (RLOS), future discharges, and census probability distributions based on their health trajectories up to the current stay. Evaluated with large-scale EMR data, the dynamic model significantly improves predictive power over the performance of any model in previous literature and remains medically interpretable.

Length of Stay in the Neonatal ICU is Predictable using Heart Rate: An Opportunity for Optimizing Managed Care

We explore the use of classification and regression models for predicting the length of stay (LoS) of neonatal patients in the intensive care unit (ICU), using heart rate (HR) time-series data of 7,758 patients from the MIMIC-IH database. We find that aggregated features of HR on the first full-day of in-patient stay after admission (i.e. the first day with a full 24-hour record for each patient) can be leveraged to classify LoS in excess of 10 days with 89% sensitivity and 59% specificity. As such, LoS as a continuous variable was also found to be statistically significantly correlated to aggregate HR data corresponding to the first full-day after admission.

Forecasting the length-of-stay of pediatric patients in hospitals: a scoping review

BACKGROUND

Healthcare management faces complex challenges in allocating hospital resources, and predicting patients' length-of-stay (LOS) is critical in effectively managing those resources. This work aims to map approaches used to forecast the LOS of Pediatric Patients in Hospitals (LOS-P) and patients' populations and environments used to develop the models.

METHODS

Using the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) methodology, we performed a scoping review that identified 28 studies and analyzed them. The search was conducted on four databases (Science Direct, Scopus, Web of Science, and Medline). The identification of relevant studies was structured around three axes related to the research questions: (i) forecast models, (ii) hospital length-of-stay, and (iii) pediatric patients. Two authors carried out all stages to ensure the reliability of the review process. Articles that passed the initial screening had their data charted on a spreadsheet. Methods reported in the literature were classified according to the stage in which they are used in the modeling process: (i) pre-processing of data, (ii) variable selection, and (iii) cross-validation.

RESULTS

Forecasting models are most often applied to newborn patients and, consequently, in neonatal intensive care units. Regression analysis is the most widely used modeling approach; techniques associated with Machine Learning are still incipient and primarily used in emergency departments to model patients in specific situations.

CONCLUSIONS

The studies' main benefits include informing family members about the patient's expected discharge date and enabling hospital resources' allocation and planning. Main research gaps are associated with the lack of generalization of forecasting models and limited reported applicability in hospital management. This study also provides a practical guide to LOS-P forecasting methods and a future research agenda.

When is My Baby Going Home? Moderate to Late Preterm Infants are Discharged at 36 Weeks Based on Admission Data

OBJECTIVE

This study evaluates the effect of admission characteristics of uncomplicated moderate to late preterm infants on timing of discharge. One of the first questions that families of infants admitted to the Neonatal Intensive Care Unit (NICU) ask is, "When is my baby going home?" Moderate to late preterm infants are the largest cohort of NICU patients but little data exist about their length of stay (LOS).

STUDY DESIGN

A retrospective electronic chart review was completed on 12,498 infants admitted to our NICU between January 1, 2009 and December 31, 2015. All inborn infants with a gestational age between 32^0/7 and 36^6/7 weeks were studied.

RESULTS

A total of 3,240 infants met our inclusion criteria. The mean postmenstrual age at discharge was 36^3/7 weeks. Infants who were small for gestational age were significantly more likely to have an increased LOS. Infants born between 34 and 36^6/7 weeks had a significantly increased LOS if they had respiratory distress syndrome. Admission diagnoses of neonatal abstinence syndrome, meconium aspiration syndrome, hydrops, hypoxic ischemic encephalopathy, biliary emesis, ABO incompatibly, and a genetic diagnosis all had increased LOS for all late preterm infants.

CONCLUSION

For uncomplicated moderate to late preterm infants, clinicians can counsel families that their infants will likely be discharged at 36 weeks of postmenstrual age. Small for gestational age infants and those with specific diagnoses may stay longer.

Designing a bed-side system for predicting length of stay in a neonatal intensive care unit

Increased length of stay (LOS) in intensive care units is directly associated with the financial burden, anxiety, and increased mortality risks. In the current study, we have incorporated the association of day-to-day nutrition and medication data of the patient during its stay in hospital with its predicted LOS. To demonstrate the same, we developed a model to predict the LOS using risk factors (a) perinatal and antenatal details, (b) deviation of nutrition and medication dosage from guidelines, and (c) clinical diagnoses encountered during NICU stay. Data of 836 patient records (12 months) from two NICU sites were used and validated on 211 patient records (4 months). A bedside user interface integrated with EMR has been designed to display the model performance results on the validation dataset. The study shows that each gestation age group of patients has unique and independent risk factors associated with the LOS. The gestation is a significant risk factor for neonates < 34 weeks, nutrition deviation for < 32 weeks, and clinical diagnosis (sepsis) for ≥ 32 weeks. Patients on medications had considerable extra LOS for ≥ 32 weeks’ gestation. The presented LOS model is tailored for each patient, and deviations from the recommended nutrition and medication guidelines were significantly associated with the predicted LOS.

Developing and validating multivariable prediction models for predicting the risk of 7-day neonatal readmission following vaginal and cesarean birth using administrative databases

BACKGROUND

Approximately 3.5% of deliveries in Canada result in potentially preventable neonatal readmission, often times due to preventable morbidities. With complexities in hospital discharge planning, health care providers may benefit in identifying infants at risk of readmission for additional monitoring.

OBJECTIVES

To develop and validate models for predicting 7-day neonatal readmission following vaginal or cesarean births.

METHODS

All liveborn term singleton infants without congenital anomalies in the province of Alberta who were not admitted to the NICU were identified using perinatal and hospitalization databases. A temporal split-sample was used for model development (2012-2014, vaginal n = 63,378; cesarean n = 21,225) and external validation (2014-2015, vaginal n = 21,583, cesarean n = 7,477). Multivariable logistic regression models using backward stepwise selection were used to identify predictors of 7-day readmission. We evaluated predictors of maternal age, Apgar score, length-of-stay, birthweight, gestational age, parity, residence, and sex. Hosmer-Lemeshow test and c-statistics were used to estimate calibration and discrimination.

RESULTS

The rate of readmission was 3.3% (95% CI 3.1%, 3.4%) and 2.1% (95% CI 1.9%, 2.3%) following vaginal and cesarean births in the development dataset. Prediction model following vaginal birth, excluding predictors of length-of-stay and birthweight, had sub-optimal performance in development (c-statistics 0.69) and validation data (c-statistics 0.68). Prediction model following cesarean birth, excluding predictors of maternal age, birthweight, and residence, had sub-optimal performance in development (c-statistics 0.62) and validation data (c-statistics 0.64). Readmission was observed in 7.9% (95% CI 7.1%, 8.8%) and 4.9% (95% CI 3.9%, 6.1%) of infants of vaginal and cesarean births, respectively, in the top quintile for the risk of 7-day readmission.

CONCLUSION

Using routinely collected administrative data, we developed and validated prediction models for neonatal readmission following vaginal and cesarean births. Presently the model is sub-optimal for use in risk assessment and planning at discharge, however, additional information may improve the predictive performance.

Which Factors Predict Hospital Length-of-Stay for Children Admitted to the Neonatal Intensive Care Unit and Pediatric Ward? A Hospital-Based Prospective Study

Background. The ability to accurately predict hospital length of stay (LOS) or time to discharge could aid in resource planning, stimulate quality improvement activities, and provide evidence for future research and medical practice. This study aimed to determine the predictive factors of time to discharge among patients admitted to the neonatal intensive care unit (NICU) and pediatric ward in Goba referral hospital, Ethiopia. Methods. A facility-based prospective follow up study was conducted for 8 months among 438 patients. Survival analyses were carried out using the Kaplan Meier statistics and Cox regression model. Results. The median length of hospital stay was 7 days (95% confidence interval (CI): 6.45-7.54) and 6 days (95% CI: 5.21-6.78) for patients admitted to NICU and pediatric ward, respectively. In the multivariable Cox regression, the hazard of neonatal patients with less than 37 weeks of gestational age, low birth weight, and those who develop hospital-acquired infection (HAI) after admission had prolonged time to discharge by 54% [adjusted hazard ratio (AHR): 0.46, (95% CI: 0.31-0.66)], 40% [AHR: 0.60, (95% CI: 0.40-0.90)], and 56% [AHR: 0.44, (95% CI: 0.26-0.74)], respectively. The rate of time to discharge among patients who were admitted to the pediatric ward and had HAI delayed discharge time by 49% [AHR: 0.51, (95% CI: 0.30-0.85)] compared to their counterparts. Conclusion. Hospital-acquired infections prolonged hospital stay among neonates and children admitted to the pediatric ward. On a similar note, low gestational age and low birth weight were found to be the independent predictor of longer hospital stay among neonates.

The effect of donor human milk on the length of hospital stay in very low birthweight infants: a systematic review and meta-analysis

BACKGROUND

Donor human milk (DHM) is an alternative to preterm infant formula if the mother's own milk is not available. Since the lactation period and preservation treatment of DHM are different from those of mother's own milk, we aimed to determine the reduction in the length of hospital stay by DHM compared to preterm infant formula.

METHODS

In this systematic review, we searched PubMed/MEDLINE, EMBASE, and the Cochrane Library to retrieve studies on the impact of DHM on the clinical outcomes of preterm infants published before 1 November 2019. The study included very low birthweight (VLBW) infants taking either DHM or infant formula with data on the length of hospital stay. Data were analysed using Review Manager 5.3 software.

RESULTS

The literature search yielded 136 articles, and four randomised controlled trials (RCTs) and eight observational studies met the inclusion criteria. A meta-analysis of the RCTs (N = 725) showed no reduction in the length of hospital stay in both the DHM and infant formula groups (- 0.22 days; 95% CI -6.38, 5.95 days), whereas that of the eight observational studies (N = 2496) showed a significant reduction in the length of hospital stay in the DHM group (- 11.72 days; 95% CI -22.07, - 1.37 days). A subgroup analysis of the RCTs revealed that the incidence of necrotising enterocolitis (NEC) was significantly lower in the DHM group when the analysis included high-quality RCTs (RR = 0.32; 95% CI 0.15, 0.69).

CONCLUSIONS

This systematic review of RCTs showed that DHM neither prolonged nor shortened the length of hospital stay in VLBW infants compared to preterm infant formula; however, it reduced the incidence of NEC, further validating the protective role of DHM in the health and safety of VLBW infants.

The Resource Use Inflection Point for Safe NICU Discharge

OBJECTIVES

(1) To identify a resource use inflection point (RU-IP) beyond which patients in the NICU no longer received NICU-level care, (2) to quantify variability between hospitals in patient-days beyond the RU-IP, and (3) to describe risk factors associated with reaching an RU-IP.

METHODS

We evaluated infants admitted to any of the 43 NICUs over 6 years. We determined the day that each patient's total daily standardized cost was <10% of the mean first-day NICU room cost and remained within this range through discharge (RU-IP). We compared days beyond an RU-IP, the total standardized cost of hospital days beyond the RU-IP, and the percentage of patients by hospital beyond the RU-IP.

RESULTS

Among 80 821 neonates, 80.6% reached an RU-IP. In total, there were 234 478 days after the RU-IP, representing 24.3% of the total NICU days and $483 281 268 in costs. Variability in the proportion of patients reaching an RU-IP was 33.1% to 98.7%. Extremely preterm and very preterm neonates, patients discharged with home health care services, or patients receiving mechanical ventilation, extracorporeal membrane oxygenation, or feeding support exhibited fewer days beyond the RU-IP. Conversely, receiving methadone was associated with increased days beyond the RU-IP.

CONCLUSIONS

Identification of an RU-IP may allow health care systems to identify readiness for discharge from the NICU earlier and thereby save significant NICU days and health care dollars. These data reveal the need to identify best practices in NICUs that consistently discharge infants more efficiently. Once these best practices are known, they can be disseminated to offer guidance in creating quality improvement projects to provide safer and more predictable care across hospitals for patients of all socioeconomic statuses.

Development of the PREMature Infant Index (PREMII™), a clinician-reported outcome measure assessing functional status of extremely preterm infants

Background: Comprehensive measures to evaluate the effectiveness of medical interventions in extremely preterm infants are lacking. Although length of stay is used as an indicator of overall health among preterm infants in clinical studies, it is confounded by nonmedical factors (e.g. parental readiness and availability of home nursing support).Objectives: To develop the PREMature Infant Index (PREMII^™), an electronic content-valid clinician-reported outcome measure for assessing functional status of extremely preterm infants (<28 weeks gestational age) serially over time in the neonatal intensive care unit. We report the development stages of the PREMII, including suggestions for scoring.Methods: We developed the PREMII according to US Food and Drug Administration regulatory standards. Development included five stages: (1) literature review, (2) clinical expert interviews, (3) Delphi panel survey, (4) development of items/levels, and (5) cognitive interviews/usability testing. Scoring approaches were explored via an online clinician survey.Results: Key factors reflective of functional status were identified by physicians and nurses during development of the PREMII, as were levels within each factor to assess functional status. The resulting PREMII evaluates eight infant health factors: respiratory support, oxygen administration, apnea, bradycardia, desaturation, thermoregulation, feeding, and weight gain, each scored with three to six gradations. Factor levels are standardized on a 0-100 scale; resultant scores are 0-100. No usability issues were identified. The online clinician survey identified optimal scoring methods to capture functional status at a given time point.Conclusions: Our findings support the content validity and usability of the PREMII as a multifunction outcome measure to assess functional status over time in extremely preterm infants. Psychometric validation is ongoing.

Modelling the neonatal system: A joint analysis of length of stay and patient pathways

In the United Kingdom, one in seven babies require specialist neonatal care after birth, with a noticeable increase in demand. Coupled with budgeting constraints and lack of investment means that neonatal units are struggling. This will inevitably have an impact on baby's length of stay (LoS) and the performance of the service. Models have previously been developed to capture individual babies' pathways to investigate the longitudinal cycle of care. However, no models have been developed to examine the joint analysis of LoS and babies' pathways. LoS at each stage of care is a critical driver of both the clinical outcomes and economic performance of the neonatal system. Using the generalized linear mixed modelling approach, extended to accommodate multiple outcomes, the association between neonate's pathway to discharge and LoS is examined. Using the data about 1002 neonates, we noticed that there is a high positive association between baby's pathway and total LoS, suggesting that discharge policies needs to be looked at more carefully. A novel statistical approach that examined the association of key outcomes and how it evolved over time is developed. Its applicability can be extended to other types of long-term care or diseases, such as heart failure and stroke.

Estimating neonatal length of stay for babies born very preterm

OBJECTIVE

To predict length of stay in neonatal care for all admissions of very preterm singleton babies.

SETTING

All neonatal units in England.

PATIENTS

Singleton babies born at 24-31 weeks gestational age from 2011 to 2014. Data were extracted from the National Neonatal Research Database.

METHODS

Competing risks methods were used to investigate the competing outcomes of death in neonatal care or discharge from the neonatal unit. The occurrence of one event prevents the other from occurring. This approach can be used to estimate the percentage of babies alive, or who have been discharged, over time.

RESULTS

A total of 20 571 very preterm babies were included. In the competing risks model, gestational age was adjusted for as a time-varying covariate, allowing the difference between weeks of gestational age to vary over time. The predicted percentage of death or discharge from the neonatal unit were estimated and presented graphically by week of gestational age. From these percentages, estimates of length of stay are provided as the number of days following birth and corrected gestational age at discharge.

CONCLUSIONS

These results can be used in the counselling of parents about length of stay and the risk of mortality.

Duration and Time Trends in Hospital Stay for Very Preterm Infants Differ Across European Regions

OBJECTIVES

To compare duration and changes over time in length of hospital stay for very preterm and extremely preterm infants in 10 European regions.

DESIGN

Two area-based cohort studies from the same regions in 2003 and 2011/2012.

SETTING

Ten regions from nine European countries.

PATIENTS

Infants born between 22 + 0 and 31 + 6 weeks of gestational age and surviving to discharge (Models of Organising Access to Intensive Care for Very Preterm Births cohort in 2003, n = 4,011 and Effective Perinatal Intensive Care in Europe cohort in 2011/2012, n = 4,336).

INTERVENTIONS

Observational study, no intervention.

MEASUREMENTS AND MAIN RESULTS

Maternal and infant characteristics were abstracted from medical records using a common protocol and length of stay until discharge was adjusted for case-mix using negative binomial regression. Mean length of stay was 63.6 days in 2003 and varied from 52.4 to 76.5 days across regions. In 2011/2012, mean length of stay was 63.1 days, with a narrower regional range (54.0-70.1). Low gestational age, small for gestational age, low 5-minute Apgar score, surfactant administration, any surgery, and severe neonatal morbidities increased length of stay. Infant characteristics explained some of the differences between regions and over time, but large variations remained after adjustment. In 2011/2012, mean adjusted length of stay ranged from less than 54 days in the Northern region of the United Kingdom and Wielkopolska, Poland to over 67 days in the Ile-de-France region of France and the Eastern region of the Netherlands. No systematic decrease in very preterm length of stay was observed over time after adjustment for patient case-mix.

CONCLUSIONS

A better understanding of the discharge criteria and care practices that contribute to the wide differences in very preterm length of stay across European regions could inform policies to optimize discharge decisions in terms of infant outcomes and health system costs.

Impact of Blood Donor Sex on Transfusion-Related Outcomes in Preterm Infants

OBJECTIVE

Explore the role of red blood cell donor sex on preterm infant neonatal outcomes.

STUDY DESIGN

In a retrospective, exploratory, cohort study, the hospital blood bank database was queried for units of blood released to neonatal intensive care unit patients in 2009-2010. The state blood center provided donor sex, and a department database provided neonatal characteristics and morbidities. Comparisons were made for 2 groups: those who ever received female blood and those who did not.

RESULTS

Among 462 infants <32 weeks of gestation, 190 (41%) received >1 blood transfusion. In univariate analyses, compared with infants who received only male blood, infants who received female donor blood had higher rates of bronchopulmonary dysplasia (38% vs 22%; P = .03), spontaneous intestinal perforation/necrotizing enterocolitis (17% vs 6%; P = .04), and death or any morbidity (60% vs 38%; P < .01), respectively. In adjusted analyses, female blood was associated with any morbidity (P = .0251) and 21 days longer hospitalization (P = .0098). After adding total number transfusions to the model, only an increased number of transfusions was associated with bronchopulmonary dysplasia (P = .0009), any morbidity (P = .0001), and length of stay (P = .0001). In subset regressions comparing exclusively female donor blood with male donor blood, there was a significant interaction of female donor blood and number of transfusions for any morbidity (OR 2.6 95% CI 1.2-5.7, P = .01).

CONCLUSIONS

Preliminary findings suggest that female donor blood was associated with preterm vulnerability to neonatal morbidities.

Estimating discharge dates using routinely collected data: improving the preparedness of parents of preterm infants for discharge home

The length of stay for moderately preterm infants has progressively become shorter in the UK in recent years but staff still commonly inform parents that their baby will go home around their estimated date of delivery (EDD). Parents need as much notice as possible to prepare for the discharge of their baby, and to gain the necessary skills and knowledge to care for their infant safely. We report the use of routinely collected neonatal data to develop and implement a simple centile chart for date of discharge from hospital, which allows staff and parents to predict the likely discharge date more accurately for preterm infants, most of whom now go home more than 3 weeks before their EDD. This information allows better and timelier planning for discharge of such infants, by parents and staff.

Every three-hour versus every six-hour oral feeding in preterm infants: a randomised clinical trial

AIM

This trial compares two oral feeding schedules, every three-hour and every six-hour oral feeding attempts, to determine which schedule allows for more rapid attainment of full oral feeding in preterm infants.

METHODS

Infants born at ≤33-week gestation were randomly assigned to receive oral feeding every three hours or every six hours if feeding cues were present. The primary outcome was time to full oral feeding; secondary outcomes include respiratory and apnoea rates, growth and length of stay.

RESULTS

A total of 55 infants were recruited. There was no difference between the groups in the primary or secondary outcomes.

CONCLUSION

For preterm infants fed when oral feeding cues are present, an every six-hour schedule did not alter the time to full oral feeding and had no effect on rates of tachypnoea, apnoea or length of hospital stay compared to every three-hour feeding schedule. An every six-hour oral feeding schedule led to only small reductions in number of oral feeding attempts per day.

Demonstrating the relationships of length of stay, cost and clinical outcomes in a simulated NICU

OBJECTIVE

Health-care leaders place significant focus on reducing the average length of stay (ALOS). We examined the relationships among ALOS, cost and clinical outcomes using a neonatal intensive care unit (NICU) simulation model.

STUDY DESIGN

A discrete-event NICU simulation model based on the Duke NICU was created. To identify the relationships among ALOS, cost and clinical outcomes, we replaced the standard probability distributions with composite distributions representing the best and worst outcomes published by the National Institutes of Health Neonatal Research Network.

RESULT

Both average cost per patient and average cost per ⩽28 week patient were lower in the best NICU ($16,400 vs $19,700 and $56,800 vs $76,700, respectively), while LOS remained higher (27 vs 24 days).

CONCLUSION

Our model demonstrates that reducing LOS does not uniformly reduce hospital resource utilization. These results suggest that health-care leaders should not simply rely on initiatives to reduce LOS without clear line-of-sight on clinical outcomes as well.

Modelling Neonatal Care Pathways for Babies Born Preterm: An Application of Multistate Modelling

Modelling length of stay in neonatal care is vital to inform service planning and the counselling of parents. Preterm babies, at the highest risk of mortality, can have long stays in neonatal care and require high resource use. Previous work has incorporated babies that die into length of stay estimates, but this still overlooks the levels of care required during their stay. This work incorporates all babies, and the levels of care they require, into length of stay estimates. Data were obtained from the National Neonatal Research Database for singleton babies born at 24–31 weeks gestational age discharged from a neonatal unit in England from 2011 to 2014. A Cox multistate model, adjusted for gestational age, was used to consider a baby’s two competing outcomes: death or discharge from neonatal care, whilst also considering the different levels of care required: intensive care; high dependency care and special care. The probabilities of receiving each of the levels of care, or having died or been discharged from neonatal care are presented graphically overall and adjusted for gestational age. Stacked predicted probabilities produced for each week of gestational age provide a useful tool for clinicians when counselling parents about length of stay and for commissioners when considering allocation of resources. Multistate modelling provides a useful method for describing the entire neonatal care pathway, where rates of in-unit mortality can be high. For a healthcare service focussed on costs, it is important to consider all babies that contribute towards workload, and the levels of care they require.

What factors predict length of stay in a neonatal unit: a systematic review

OBJECTIVE

In the UK, 1 in 10 babies require specialist neonatal care. This care can last from hours to months depending on the need of the baby. The increasing survival of very preterm babies has increased neonatal care resource use. Evidence from multiple studies is crucial to identify factors which may be important for predicting length of stay (LOS). The ability to predict LOS is vital for resource planning, decision-making and parent counselling. The objective of this review was to identify which factors are important to consider when predicting LOS in the neonatal unit.

DESIGN

A systematic review was undertaken which searched MEDLINE, EMBASE and Scopus for papers from 1994 to 2016 (May) for research investigating prediction of neonatal LOS. Strict inclusion and exclusion criteria were applied. Quality of each study was discussed, but not used as a reason for exclusion from the review.

MAIN OUTCOME MEASURE

Prediction of LOS in the neonatal unit.

RESULTS

9 studies were identified which investigated the prediction of neonatal LOS indicating a lack of evidence in the area. Inherent factors, particularly birth weight, sex and gestational age allow for a simple and objective prediction of LOS, which can be calculated on the first day of life. However, other early occurring factors may well also be important and estimates may need revising throughout the baby's stay in hospital.

CONCLUSIONS

Predicting LOS is vital to aid the commissioning of services and to help clinicians in their counselling of parents. The lack of evidence in this area indicates a need for larger studies to investigate methods of accurately predicting LOS.

Does Persistent Inflammatory Catabolic Syndrome Exist in Critically Ill Neonates?

BACKGROUND

Persistent inflammatory catabolic syndrome (PICS) has not been described in the infant population. This study proposes a definition of PICS in critically ill infants.

METHODS

A published adult criterion of PICS was modified using anthropometric and biochemical reference ranges for infants. A prospective chart review of admissions to a tertiary surgical neonatal intensive care unit (NICU) was performed over 65 days. Demographic, anthropometric, biochemical, and other clinical variables such as length of stay and medication use were collected daily throughout admission. Infants were categorized as having or not having PICS.

RESULTS

Twenty percent of admitted infants (n = 15) developed PICS using the proposed criteria. Infants with PICS were more likely to be classified as failure to thrive (53%), meeting only 75% of their anticipated weight gain. Significantly more infants with PICS had undergone surgery (100%; P = .01), received inotropic medication (40%; P = .05), and had longer NICU and total hospital length of stay ( P < .001 and P < .001). Infants with PICS had higher peak glucose levels (11.8 ± 7.3 mmol/L) and elevated urea concentrations (7.9 ± 4.6 mmol/L).

CONCLUSIONS

PICS does exist in a critically ill neonatal population and may be identified using the definition proposed in this study. Infants with PICS displayed metabolic dysregulation, impaired expected growth velocity, and longer length of stay despite no differences in severity scores or diagnosis between the groups. Validation of this work is required, and research into timely identification of infants with PICS is needed to inform whether these infants would benefit from earlier and novel nutrition intervention.

Estimating Adverse Events After Gastrostomy Tube Placement

OBJECTIVE

Gastrostomy feeding tube placement in children is associated with a high frequency of adverse events. This study sought to preoperatively estimate postoperative adverse events in children undergoing gastrostomy feeding tube placement.

METHODS

This was an observational study of children who underwent gastrostomy with or without fundoplication at 1 of 50 participating hospitals, using 2011-2013 data from the American College of Surgeons' National Surgical Quality Improvement Program Pediatric. The outcome was the occurrence of any postoperative complications or mortality at 30 days after gastrostomy tube placement. The preoperative clinical characteristics significantly associated with occurrence of adverse events were included in a multivariate logistic model. The area under the receiver operating characteristic curve was computed to assess model performance and split-set validated.

RESULTS

A total of 2817 children were identified as having undergone gastrostomy tube placement. The unadjusted rate of adverse events within 30 days after gastrostomy tube placement was 11%. Thirteen predictor variables were identified. Notable preoperative variables associated with a greater than 75% increase in adverse event rate were preoperative sepsis/septic shock (odds ratio [OR], 10.76, 95% confidence interval [CI], 3.84-30.17), central nervous system tumor (OR, 3.36; 95% CI, 1.42-7.95), the primary procedure as indicated by the current procedural terminology (CPT) linear risk variable (OR, 1.93; 95% CI, 1.50-2.49), severe cardiac risk factors (OR, 1.88; 95% CI, 1.17-3.03), and preoperative seizure history (OR, 1.90; 95% CI, 1.38-2.62). The area under the receiver operating characteristic curve was 0.71 with the derivation data set and 0.71 upon split-set validation.

CONCLUSIONS

Preoperatively estimating postoperative adverse events in children undergoing gastrostomy tube placement is feasible.

Accounting for variation in length of NICU stay for extremely low birth weight infants

OBJECTIVE

To develop a length of stay (LOS) model for extremely low birth weight (ELBW) infants.

STUDY DESIGN

We included infants from the California Perinatal Quality Care Collaborative with birth weight 401 to 1000 g who were discharged to home. Exclusion criteria were congenital anomalies, surgery and death. LOS was defined as days from admission to discharge. As patients who died or were transferred to lower level of care were excluded, we assessed correlation of hospital mortality rates and transfers to risk-adjusted LOS.

RESULTS

There were 2012 infants with median LOS 79 days (range 23 to 219). Lower birth weight, lack of antenatal steroids and lower Apgar score were associated with longer LOS. There was negligible correlation between risk-adjusted LOS and hospital mortality rates (r=0.0207) and transfer-out rates (r=0.121).

CONCLUSION

Particularly because ELBW infants have extended hospital stays, identification of unbiased and informative risk-adjusted LOS for these infants is an important step in benchmarking best practice and improving efficiency in care.