Consensus Statement on Electronic Health Predictive Analytics: A Guiding Framework to Address Challenges.

CONTEXT: The recent explosion in available electronic health record (EHR) data is motivating a rapid expansion of electronic health care predictive analytic (e-HPA) applications, defined as the use of electronic algorithms that forecast clinical events in real time with the intent to improve patient outcomes and reduce costs. There is an urgent need for a systematic framework to guide the development and application of e-HPA to ensure that the field develops in a scientifically sound, ethical, and efficient manner. OBJECTIVES: Building upon earlier frameworks of model development and utilization, we identify the emerging opportunities and challenges of e-HPA, propose a framework that enables us to realize these opportunities, address these challenges, and motivate e-HPA stakeholders to both adopt and continuously refine the framework as the applications of e-HPA emerge. METHODS: To achieve these objectives, 17 experts with diverse expertise including methodology, ethics, legal, regulation, and health care delivery systems were assembled to identify emerging opportunities and challenges of e-HPA and to propose a framework to guide the development and application of e-HPA. FINDINGS: The framework proposed by the panel includes three key domains where e-HPA differs qualitatively from earlier generations of models and algorithms (Data Barriers, Transparency, and ETHICS) and areas where current frameworks are insufficient to address the emerging opportunities and challenges of e-HPA (Regulation and Certification; and Education and Training). The following list of recommendations summarizes the key points of the framework: Data Barriers: Establish mechanisms within the scientific community to support data sharing for predictive model development and testing.Transparency: Set standards around e-HPA validation based on principles of scientific transparency and reproducibility. ETHICS: Develop both individual-centered and society-centered risk-benefit approaches to evaluate e-HPA.Regulation and Certification: Construct a self-regulation and certification framework within e-HPA.Education and Training: Make significant changes to medical, nursing, and paraprofessional curricula by including training for understanding, evaluating, and utilizing predictive models.

Risk of preterm delivery in non-diabetic women with polycystic ovarian syndrome.

OBJECTIVE: To examine the risk and etiology of preterm delivery in women with polycystic ovary syndrome (PCOS). STUDY DESIGN: Retrospective cohort study comparing preterm delivery rate among non-diabetic PCOS and non-PCOS women with singleton pregnancy. Multivariable logistic regression was used to identify predictors of preterm delivery among PCOS women. RESULT: Among 908 PCOS women with singleton pregnancy, 12.9% delivered preterm compared with 7.4% among non-PCOS women (P<0.01). Causes of preterm delivery among PCOS women included preterm labor (41%), cervical insufficiency (11%), hypertensive complications (20%), preterm premature rupture of membranes (15%), fetal-placental concerns (9%) and intrauterine fetal demise (5%). Maternal age, race/ethnicity and nulliparity were significant predictors of preterm delivery in PCOS, whereas body mass index and fertility medications were not. CONCLUSION: A higher proportion of PCOS women delivered preterm (12.9%) compared with non-PCOS women, with the majority of cases due to spontaneous preterm birth. Future studies should explore etiologies and strategies to improve pregnancy outcomes in PCOS.

Racial and ethnic differences in the prevalence of placenta previa.

OBJECTIVE: The objective of this study is to determine the prevalence of placenta previa among different racial and ethnic groups. STUDY DESIGN: We conducted a retrospective cohort study to examine the prevalence of placenta previa among five major racial and ethnic groups: African American, Asian, Caucasian, Hispanic and Native American. We included all deliveries ≥ 20 weeks gestation from a large northern Californian Health Maintenance Organization from 1995-2006. A multivariable logistic regression model was used to control for potential confounders. RESULT: Of the 394 083 deliveries in our cohort, 1580 (0.40%) were complicated by placenta previa. The prevalence of placenta previa was: Asian 0.64%, Native American 0.60%, African American 0.44%, Caucasian 0.36%, Hispanic 0.34% and unknown 0.31% (P<0.001). In our multivariable logistic regression model, only Asians (odds ratio (OR) 1.73, 95% confidence intervals (CI) 1.53-1.95) and African Americans (OR 1.43, 95% CI 1.19-1.72) were at increased risk for having placenta previa, compared with Caucasians. CONCLUSION: Asian women have the highest prevalence of placenta previa.

Underimmunization at discharge from the neonatal intensive care unit.

OBJECTIVE: The objectives of this study are to determine immunization rates at discharge from the neonatal intensive care unit (NICU) among infants 2 months of age and above and to evaluate risk factors for underimmunization. STUDY DESIGN: A retrospective cohort study was performed for infants in six NICUs in the Northern California Kaiser Permanente Medical Care Program. Immunization status at discharge was determined for all infants discharged on or after age 60 days. Logistic regression was used to identify risk factors for underimmunization at the time of discharge. RESULT: Of 668 infants discharged on or after age 60 days from the NICU, 51% were up-to-date for routine immunizations. Twenty-seven percent of infants had received no vaccines. Factors associated with higher immunization rates at discharge include history of mechanical ventilation, congenital heart disease and a diagnosis of apnea or bronchopulmonary dysplasia during the NICU stay, whereas surgery was associated with lower immunization rates. CONCLUSION: A significant proportion of infants discharged on or after 2 months of age in the NICU in this health system was unimmunized or underimmunized at discharge. Further efforts should be made to improve immunization rates prior to discharge.

The impact of maternal characteristics on the moderately premature infant: an antenatal maternal transport clinical prediction rule.

OBJECTIVE: Moderately premature infants, defined here as those born between 30°/7 and 346/7 weeks gestation, comprise 3.9% of all births in the United States and 32% of all preterm births. Although long-term outcomes for these infants are better than for less mature infants, morbidity and mortality are still substantially increased in comparison with infants born at term. There is an added survival benefit resulting from birth at a tertiary neonatal care center, and although many of these infants require tertiary level care, delivery at lower level hospitals and subsequent neonatal transfer are still common. Our primary aim was to determine the impact of maternal characteristics and antenatal medical management on the early neonatal course of the moderately premature infant. The secondary aim was to create a clinical prediction rule to determine which infants require intubation and mechanical ventilation in the first 24 h of life. Such a prediction rule could inform the decision to transfer maternal-fetal patients before delivery to a facility with a Level III neonatal intensive care unit (NICU), where optimal care could be provided without the requirement for a neonatal transfer. STUDY DESIGN: Data for this analysis came from the cohort of infants in the Moderately Premature Infant Project (MPIP) database, a multicenter cohort study of 850 infants born at gestational age 30°/7 and 346/7 weeks, with birth weight between 591 to 3540 g. [corrected], who were discharged to home alive. We built a logistic regression model to identify maternal characteristics associated with need for tertiary care, as measured by administration of surfactant. Using statistically significant covariates from this model, we then created a numerical decision rule to predict need for tertiary care. RESULT: In multivariate modeling, four factors were associated with reduction in the need for tertiary care, including non-White race (odds ratio (OR)=0.5, (0.3, 0.7)), older gestational age, female gender (OR=0.6 (0.4, 0.8)) and use of antenatal corticosteroids (OR=0.5, (0.3, 0.8)). The clinical prediction rule to discriminate between infants who received surfactant, versus those who did not, had an area under the curve of 0.77 (0.73, 0.8). CONCLUSION: Four antenatal risk factors are associated with a requirement for Level III NICU care as defined by the need for surfactant administration. Future analyses will examine a broader spectrum of antenatal characteristics and revalidate the prediction rule in an independent cohort.

Risk of fever and sepsis evaluations after routine immunizations in the neonatal intensive care unit.

OBJECTIVE: Premature infants can experience cardiorespiratory events such as apnea after immunization in the neonatal intensive care unit (NICU). These changes in clinical status may precipitate sepsis evaluations. This study evaluated whether sepsis evaluations are increased after immunizations in the NICU. STUDY DESIGN: We conducted a retrospective cohort study of infants older than 53 days who were vaccinated in the NICU at the KPMCP (Kaiser Permanente Medical Care Program). Chart reviews were carried out before and after all immunizations were administered and for all sepsis evaluations after age 53 days. The clinical characteristics of infants on the day before receiving a sepsis evaluation were compared between children undergoing post-immunization sepsis evaluations and children undergoing sepsis evaluation at other times. The incidence rate of sepsis evaluations in the post-immunization period was compared with the rate in a control time period not following immunization using Poisson regression. RESULT: A total of 490 infants met the inclusion criteria. The rate of fever was increased in the 24 h period after vaccination (2.3%, P<0.05). The incidence rate of sepsis evaluations was 40% lower after immunization than during the control period, although this was not statistically significant (P=0.09). Infants undergoing a sepsis evaluation after immunization were more likely to have an apneic, bradycardic or moderate-to-severe cardiorespiratory event in the day before the evaluation than were infants undergoing sepsis evaluations at other times (P<0.05). CONCLUSION: Despite an increase in fever and cardiorespiratory events after immunization in the NICU, routine vaccination was not associated with increased risk of receiving sepsis evaluations. Providers may be deferring immunizations until infants are clinically stable, or may have a higher threshold for initiating sepsis evaluations after immunization than at other times.

Pediatric care for preterm infants after NICU discharge: high number of office visits and prescription medications.

OBJECTIVE: To determine the frequency and risk factors for clinic and pharmacy use in preterm infants during the first year after neonatal intensive care unit (NICU) discharge. STUDY DESIGN: We analyzed clinic visits and prescriptions in a cohort of 23 to 32 weeks infants. We constructed multivariable regression models to determine risk factors for high use. RESULT: The 892 preterm infants experienced 18 346 pediatric visits (mean 20 visits per infant per year) and filled 2100 prescriptions (mean 5.5 prescriptions per year among infants taking medications). Most visits were non-well child care visits to pediatric primary care providers. Prematurity and bronchopulmonary dysplasia (BPD) are important risk factors: infants at 23 to 26 weeks gestation or infants with BPD had an average 29 visits per year and 9 prescriptions per year among infants taking medication. However, half of the highest using infants were relatively healthy infants at 27 to 32 weeks gestation who escaped BPD, NEC or grade 3 to 4 intraventricular hemorrhage. CONCLUSION: Premature infants had frequent pediatric visits and prescription medications. Extreme prematurity and neonatal morbidities are important risk factors; however, half of the highest using infants are moderately preterm without neonatal morbidities.

Pregnancy plasma glucose levels exceeding the American Diabetes Association thresholds, but below the National Diabetes Data Group thresholds for gestational diabetes mellitus, are related to the risk of neonatal macrosomia, hypoglycaemia and hyperbilirubinaemia.

AIMS/HYPOTHESIS: Gestational diabetes mellitus (GDM) is a risk factor for perinatal complications. In several countries, the criteria for the diagnosis of GDM have been in flux, the American Diabetes Association (ADA) thresholds recommended in 2000 being lower than those of the National Diabetes Data Group (NDDG) that have been in use since 1979. We sought to determine the extent to which infants of women meeting only the ADA criteria for GDM are at increased risk of neonatal complications. MATERIALS AND METHODS: In a multiethnic cohort of 45,245 women who did not meet the NDDG criteria and were not treated for GDM, we conducted nested case-control studies of three complications of GDM that occurred in their infants: macrosomia (birthweight >4,500 g, n = 494); hypoglycaemia (plasma glucose <2.2 mmo/l, n = 488); and hyperbilirubinaemia (serum bilirubin > or =342 micromol/l (20 mg/dl), n = 578). We compared prenatal glucose levels of the mothers of these infants and mothers of 884 control infants. RESULTS: Women with GDM by ADA criteria only (two or more glucose values exceeding the threshold) had an increased risk of having an infant with macrosomia (odds ratio OR = 3.40, 95% CI = 1.55-7.43), hypoglycaemia (OR = 2.61, 95% CI = 0.99-6.92) or hyperbilirubinaemia (OR = 2.22, 95% CI = 0.98-5.04). Glucose levels 1 h after the 100-g glucose challenge that exceeded the ADA threshold were particularly strongly associated with each complication. CONCLUSIONS/INTERPRETATION: These results lend support to the ADA recommendations and highlight the importance of the 1-h glucose measurement in a diagnostic test for GDM.

Unstudied infants: outcomes of moderately premature infants in the neonatal intensive care unit.

BACKGROUND: Newborns of 30-34 weeks gestation comprise 3.9% of all live births in the United States and 32% of all premature infants. They have been studied much less than very low birthweight infants. OBJECTIVE: To measure in-hospital outcomes and readmission within three months of discharge of moderately premature infants. DESIGN: Prospective cohort study including retrospective chart review and telephone interviews after discharge. SETTING: Ten birth hospitals in California and Massachusetts. PATIENTS: Surviving moderately premature infants born between October 2001 and February 2003. MAIN OUTCOME MEASURES: (a) Occurrence of assisted ventilation during the hospital stay after birth; (b) adverse in-hospital outcomes-for example, necrotising enterocolitis; (c) readmission within three months of discharge. RESULTS: With the use of prospective cluster sampling, 850 eligible infants and their families were identified, randomly selected, and enrolled. A total of 677 families completed a telephone interview three months after hospital discharge. During the birth stay, these babies experienced substantial morbidity: 45.7% experienced assisted ventilation, and 3.2% still required supplemental oxygen at 36 weeks. Readmission within three months occurred in 11.2% of the cohort and was higher among male infants and those with chronic lung disease. CONCLUSIONS: Moderately premature infants experience significant morbidity, as evidenced by high rates of assisted ventilation, use of oxygen at 36 weeks, and readmission. Such morbidity deserves more research.

Moderately premature infants at Kaiser Permanente Medical Care Program in California are discharged home earlier than their peers in Massachusetts and the United Kingdom.

OBJECTIVE: To compare gestational age at discharge between infants born at 30-34(+6) weeks gestational age who were admitted to neonatal intensive care units (NICUs) in California, Massachusetts, and the United Kingdom. DESIGN: Prospective observational cohort study. SETTING: Fifty four United Kingdom, five California, and five Massachusetts NICUs. SUBJECTS: A total of 4359 infants who survived to discharge home after admission to an NICU. MAIN OUTCOME MEASURES: Gestational age at discharge home. RESULTS: The mean (SD) postmenstrual age at discharge of the infants in California, Massachusetts, and the United Kingdom were 35.9 (1.3), 36.3 (1.3), and 36.3 (1.9) weeks respectively (p = 0.001). Compared with the United Kingdom, adjusted discharge of infants occurred 3.9 (95% confidence interval (CI) 1.4 to 6.5) days earlier in California, and 0.9 (95% CI -1.2 to 3.0) days earlier in Massachusetts. CONCLUSIONS: Infants of 30-34(+6) weeks gestation at birth admitted and cared for in hospitals in California have a shorter length of stay than those in the United Kingdom. Certain characteristics of the integrated healthcare approach pursued by the health maintenance organisation of the NICUs in California may foster earlier discharge. The California system may provide opportunities for identifying practices for reducing the length of stay of moderately premature infants.

Pre-pregnancy and pregnancy-related factors and the risk of excessive or inadequate gestational weight gain.

OBJECTIVE: Gestational weight gain consistent with the Institute of Medicine's recommendations is associated with better maternal and infant outcomes. The objective was to quantify the effect of pre-pregnancy factors, pregnancy-related health conditions, and modifiable pregnancy factors on the risks of inadequate and excessive gestational weight gain. METHOD: A longitudinal cohort of pregnant women (N=1100) who completed questions about diet and weight gain during pregnancy and delivered a singleton, full-term infant. RESULTS: Gestational weight gain was inadequate for 14% and excessive for 53%. Pre-pregnancy factors contributed 74% to excessive gain, substantially more than pregnancy-related health conditions (15%) and modifiable pregnancy factors (11%). Pre-pregnancy factors, pregnancy-related health conditions, and modifiable pregnancy factors contributed fairly equally to the risk of inadequate gain. CONCLUSION: Interventions to prevent excessive gestational gain may need to start before pregnancy. Women at risk for inadequate gain would also benefit from interventions directed toward modifiable factors during pregnancy.

Rehospitalisation after birth hospitalisation: patterns among infants of all gestations.

AIM: To analyse rehospitalisation of newborns of all gestations. METHODS: A total of 33,276 surviving infants of all gestations born between 1 October 1998 and 31 March 2000 at seven Kaiser Permanente Medical Care Program (KPMCP) delivery services were studied retrospectively. RESULTS: Rehospitalisation rates within two weeks after nursery discharge ranged from 1.0% to 3.7%. The most common reason for rehospitalisation was jaundice. Among babies > or =34 weeks, the most important factor with respect to rehospitalisation was use of home phototherapy. Among babies who were not rehospitalised for jaundice, African-American race (adjusted odds ratio (AOR) = 0.56), and having a scheduled outpatient visit (AOR = 0.73) or a home visit (AOR = 0.59) within 72 hours after discharge were protective. Factors associated with increased risk were: being small for gestational age (AOR = 1.83), gestational age of 34-36 weeks without admission to the neonatal intensive care unit (AOR = 1.65), Score for Neonatal Acute Physiology, version II, > or =10 (AOR = 1.95), male gender (AOR = 1.24), having both a home as well as a clinic visit within 72 hours after discharge (AOR = 1.84), and birth facility (range of AORs = 1.52-2.36). Asian race was associated with rehospitalisation (AOR = 1.49) when all hospitalisations were considered, but this association did not persist if hospitalisations for jaundice were excluded. CONCLUSIONS: In this insured population with access to integrated care, rehospitalisation rates for jaundice were strongly affected by availability of home phototherapy and by follow up. For other causes, moderate prematurity and follow up visits played a large role, but variation between centres persisted even after controlling for multiple factors. Future research should include development of better process measures for evaluation of follow up strategies.

Risk factors and obstetric complications associated with macrosomia.

OBJECTIVE: Macrosomia is associated with adverse maternal outcomes. The objective of this study was to characterize the epidemiology of macrosomia and related maternal complications. METHOD: Live births (146,526) were identified between 1995 and 1999 in the Kaiser Permanente Medical Care Program's Northern California Region (KPMCP NCR) database. Bivariate and multivariate analyses were performed for risk factors and complications associated with macrosomia (birth weight >4500 g). RESULT: Male infant sex, multiparity, maternal age 30-40, white race, diabetes, and gestational age >41 weeks were associated with macrosomia (p<0.001). In bivariate and multivariate analyses, macrosomia was associated with higher rates of cesarean birth, chorioamnionitis, shoulder dystocia, fourth-degree perineal lacerations, postpartum hemorrhage, and prolonged hospital stay (p<0.01). CONCLUSION: Macrosomia was associated with adverse maternal outcomes in this cohort. More research is needed to determine how to prevent complications related to excessive birth weight.

A randomized comparison of home visits and hospital-based group follow-up visits after early postpartum discharge.

OBJECTIVE: Short postpartum stays are common. Current guidelines provide scant guidance on how routine follow-up of newly discharged mother-infant pairs should be performed. We aimed to compare 2 short-term (within 72 hours of discharge) follow-up strategies for low-risk mother-infant pairs with postpartum length of stay (LOS) of <48 hours: home visits by a nurse and hospital-based follow-up anchored in group visits. METHODS: We used a randomized clinical trial design with intention-to-treat analysis in an integrated managed care setting that serves a largely middle class population. Mother-infant pairs that met LOS and risk criteria were randomized to the control arm (hospital-based follow-up) or to the intervention arm (home nurse visit). Clinical utilization and costs were studied using computerized databases and chart review. Breastfeeding continuation, maternal depressive symptoms, and maternal satisfaction were assessed by means of telephone interviews at 2 weeks postpartum. RESULTS: During a 17-month period in 1998 to 1999, we enrolled and randomized 1014 mother-infant pairs (506 to the control group and 508 to the intervention group). There were no significant differences between the study groups with respect to maternal age, race, education, household income, parity, previous breastfeeding experience, early initiation of prenatal care, or postpartum LOS. There were no differences with respect to neonatal LOS or Apgar scores. In the control group, 264 mother-infant pairs had an individual visit only, 157 had a group visit only, 64 had both a group and an individual visit, 4 had a home health and a hospital-based follow-up, 13 had no follow-up within 72 hours, and 4 were lost to follow-up. With respect to outcomes within 2 weeks after discharge, there were no significant differences in newborn or maternal hospitalizations or urgent care visits, breastfeeding discontinuation, maternal depressive symptoms, or a combined clinical outcome measure indicating whether a mother-infant pair had any of the above outcomes. However, mothers in the home visit group were more likely than those in the control group to rate multiple aspects of their care as excellent or very good. These included the preventive advice delivered (76% vs 59%) and the skills and abilities of the provider (84% vs 73%). Mothers in the home visit group also gave higher ratings on overall satisfaction with the newborn's posthospital care (71% vs 59%), as well as with their own posthospital care (63% vs 55%). The estimated cost of a postpartum home visit to the mother and the newborn was $265. In contrast, the cost of the hospital-based group visit was $22 per mother-infant pair; the cost of an individual 15-minute visit with a registered nurse was $52; the cost of a 15-minute individual pediatrician visit was $92; and the cost of a 10-minute visit with an obstetrician was $92. CONCLUSIONS: For low-risk mothers and newborns in an integrated managed care organization, home visits compared with hospital-based follow-up and group visits were more costly but achieved comparable clinical outcomes and were associated with higher maternal satisfaction. Neither strategy is associated with significantly greater success at increasing continuation of breastfeeding. This study had limited power to identify group differences in rehospitalization and may not be generalizable to higher-risk populations without comparable access to integrated hospital and outpatient care.

A critical review of cost reduction in neonatal intensive care. I. The structure of costs.

Neonatal intensive care is expensive. In the current era of intense cost containment in hospital care, neonatologists and hospital administrators are under intense pressure to find strategies for cost reduction for neonatal services. Few neonatal clinicians are trained in economics, management, or accounting, and few hospital administrators are familiar with neonatal intensive care. In this review, we describe the structure and sources of hospital costs and the accounting systems needed to isolate and measure such costs. We discuss where efficiencies might be found and consider specific issues in capitated settings such as health maintenance organizations in the United States, the Canadian health care system and the National Health System in the United Kingdom.

A critical review of cost reduction in neonatal intensive care. II. Strategies for reduction.

Neonatal intensive care is extremely expensive; there is both a financial and an ethical obligation to practice efficiently. In the current era of intense cost containment in hospital care, neonatologists and hospital administrators are under pressure to find strategies for cost reduction for neonatal services. In this review, we address reducing discretionary admissions, the high costs of low-cost testing, minimizing use of selected high-cost technologies (ventilators and parenteral nutrition), shortening length of stay, and optimizing nursing allocation.

SNAP-II and SNAPPE-II: Simplified newborn illness severity and mortality risk scores.

OBJECTIVES: Illness severity scores for newborns are complex and restricted by birth weight and have dated validations and calibrations. We developed and validated simplified neonatal illness severity and mortality risk scores. The primary outcome was in-hospital mortality. STUDY DESIGN: Thirty neonatal intensive care units in Canada, California, and New England collected data on all admissions during the mid 1990s; patients moribund at birth or discharged to normal newborn care in <24 hours were excluded. Starting with the 34 data elements of the Score for Neonatal Acute Physiology (SNAP), we derived the most parsimonious logistic model for in-hospital mortality using 10,819 randomly selected Canadian cases. SNAP-II includes 6 physiologic items; to this are added points for birth weight, low Apgar score, and small for gestational age to create a 9-item SNAP-Perinatal Extension-II (SNAPPE-II). We validated SNAPPE-II on the remaining 14,610 cases and optimized the calibration. RESULTS: In all birth weights, SNAPPE-II had excellent discrimination and goodness of fit. Area under the receiver operator characteristic curve was .91 +/- 0.01. Goodness of fit (Hosmer-Lemeshow) was 0.90. CONCLUSIONS: SNAP-II and SNAPPE-II are empirically validated illness severity and mortality risk scores for newborn intensive care. They are simple, accurate, and robust across populations.

Early Start: an obstetric clinic-based, perinatal substance abuse intervention program.

Maternal substance abuse is a serious problem with significant adverse effects to mothers, fetuses, and children. The Early Start Program provides pregnant women in a managed care organization with screening and early identification of substance abuse problems, early intervention, ongoing counseling, and case management by a licensed clinical social worker located in the prenatal clinic, where she is an integral part of the prenatal team. We describe the development of the Early Start Program, its administrative history, and how it has interfaced with clinicians and administrators. We also highlight two important program characteristics: the partnership with a perinatal health services research unit and the degree to which the program could be "exported" to other managed care settings.

Prediction and prevention of extreme neonatal hyperbilirubinemia in a mature health maintenance organization.

OBJECTIVE: To investigate biological and health services predictors of extreme neonatal hyperbilirubinemia in a health maintenance organization. DESIGN: Nested case-control study. SETTING: Eleven Northern California Kaiser Permanente hospitals. SUBJECTS: The cohort consisted of 51,387 newborns born at 36 weeks or later weighing 2000 g or more. Cases were newborns with peak total serum bilirubin levels greater than or equal to 428 micromol/L (> or =25 mg/dL) (n = 73). Controls were a random sample of newborns from the cohort with peak bilirubin levels less than 428 micromol/L (<25 mg/dL) (n = 423). MEASUREMENTS: Review of medical records and telephone interviews. RESULTS: Early jaundice was most strongly associated with case status (odds ratio [OR] = 7.3). After excluding subjects with early jaundice, the strongest predictors of hyperbilirubinemia were family history of jaundice in a newborn (OR = 6.0), exclusive breastfeeding (OR = 5.7), bruising (OR = 4.0), Asian race (OR = 3.5), cephalhematoma (OR = 3.3), maternal age of 25 years or older (OR = 3.1), and lower gestational age (OR = 0.6/week). These variables identified 61% of newborns as very low risk (about 1/4200). However, the risk in the remaining 39% was still low (1/370). More cases (79%) than controls (59%) had newborn length-of-stay and follow-up consistent with the American Academy of Pediatrics guidelines, but phototherapy use within 8 hours of the time that the guidelines recommend was uncommon in both cases (26%) and controls (33%). There were no apparent cases of kernicterus. CONCLUSIONS: Prevention of extreme hyperbilirubinemia may require closer follow-up than is currently recommended by the American Academy of Pediatrics and more use of phototherapy than was observed in this study. To prevent extreme hyperbilirubinemia (> or =428 micromol/L [> or =25 mg/dL]) in 1 newborn, many newborns would need to receive these interventions.