You are here

Article Text

Article menu
Community child health, public health, and epidemiology
Rehospitalisation after birth hospitalisation: patterns among infants of all gestations
  1. G J Escobar1,
  2. J D Greene1,
  3. P Hulac2,
  4. E Kincannon3,
  5. K Bischoff4,
  6. M N Gardner1,
  7. M A Armstrong1,
  8. E K France5
  1. 1Kaiser Permanente Medical Care Program, Division of Research, Perinatal Research Unit, 2000 Broadway, 2nd floor, Oakland, CA 94612, USA
  2. 2University of Colorado Health Sciences Center, Section of Neonatology, 4200 E. 9th Avenue, Denver, CO 80262, USA
  3. 3Department of Neonatology, Colorado Permanente Medical Group, Inc., 1375 East 20th Avenue, Denver, CO 80205, USA
  4. 4Research and Development Department, Kaiser Permanente, 2550 S. Parker Road, Aurora, CO 80114, USA
  5. 5Department of Preventive Medicine, Kaiser Permanente, 10350 East Dakota Avenue, Denver, CO 80231, USA
  1. Correspondence to:
    Dr G J Escobar
    Kaiser Permanente Medical Care Program, Division of Research, Perinatal Research Unit, 2000 Broadway, 2nd floor, Oakland, CA 94612, USA; gabriel.escobarkp.org

Abstract

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 ⩾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, ⩾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.

  • AOR, adjusted odds ratio
  • KPMCP, Kaiser Permanente Medical Care Program
  • LOS, length of stay
  • NICU, neonatal intensive care unit
  • SNAP-II
  • jaundice
  • home phototherapy
  • neonatal intensive care
  • prematurity
  • rehospitalisation

https://doi.org/10.1136/adc.2003.039974

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Recently, rehospitalisation of healthy term newborns has received scientific and media attention.1–,9 Less research has been done on short term rehospitalisation of other newborns.10,11 In 1999 we reported on rehospitalisations among newborns admitted to the neonatal intensive care unit (NICU).12 We found that the highest rehospitalisation rate occurred among newborns 33–36 weeks gestation that had a length of stay (LOS) of less than four days. We also found considerable inter-facility variation in rehospitalisation rates.

    The purpose of this study was to take a more comprehensive look at neonatal rehospitalisation. Unlike our previous study, we included all surviving infants and paid greater attention to inter-facility variation. We sought to identify patterns that could be useful to practitioners making newborn discharge and short term follow up decisions.

    METHODS

    Study setting

    Our study setting consisted of seven Kaiser Permanente Medical Care Program (KPMCP) delivery centres, a managed care organisation resembling the British National Health Service13 in some respects (for example, physicians are salaried, and care is provided on a capitated basis). At two of these sites, alliance facilities, KPMCP providers share the clinical facilities with fee-for-service providers. This study does not include non-KPMCP patients. Board certified neonatologists staff all seven KPMCP newborn services. Infants requiring some forms of specialised care (for example, extracorporeal membrane oxygenation) are transported to university centres but often return to KPMCP facilities before discharge home. High risk newborns discharged from the seven study sites receive follow up services, such as home visits, from the KPMCP and agencies in California and Colorado.

    Within the KPMCP, admission to the NICU is always left to the treating physician’s discretion, except with respect to premature and low birth weight infants. For these infants, mandatory admission to the NICU policies exist but these vary by facility (that is, one facility may mandate automatic admission for birth weight <2500 g while another employs a <2000 g cut-off). Discharge decisions, decisions to initiate specific treatments (for example, for jaundice), and policies for which baby gets a follow up home visit (as opposed to a clinic visit) are generally based on facility specific policies. However, none of these facilities have mandatory chronological age or attained weight criteria for discharge. Since individual KPMCP facilities have fairly broad budgetary autonomy, individual facilities may differ on how they allocate resources (for example, whether or not to employ home phototherapy, whether or not to invest in follow up clinics as opposed to home health visits).

    The KPMCP Institutional Review Board for the Protection of Human Subjects approved this study.

    Study population

    The study included all live births between 1 October 1998 and 31 March 2000. The methods employed to link data sources have been published elsewhere.12,14–,19 We used the Kaiser Permanente Neonatal Minimum Data Set (NMDS) database20 to develop a retrospective cohort of NICU admissions. Inclusion criteria consisted of: (1) NICU admission with a date of birth in the study period; (2) birth at one of the seven facilities; and (3) surviving to discharge. We did not include babies transported into these seven NICUs (outborn babies).

    Babies not admitted to the NICU were identified from KPMCP hospitalisation databases. Variables such as gestational age were obtained from the NMDS database20 or from other KPMCP databases. We calculated LOS using time of birth and time of first discharge home (if transport occurred, it includes time spent in multiple facilities). Post-menstrual age at discharge was calculated by adding hospital LOS to gestational age. KPMCP databases tracking provision of home nursing visits and home phototherapy were also scanned. We did not have access to data for follow up services provided by California or Colorado agencies.

    KPMCP information systems use a common medical record number, which permits easy linkage of hospitalisation, rehospitalisation, clinical, and laboratory data for a given member, as well as linkage of a maternal to a neonatal record. Babies born to KPMCP women members are automatically covered for the first month of life, which permits attaining very high follow up rates. We determined which babies completed their automatic membership month during the study time period, then scanned the length of enrolment database to identify which of these babies did not continue KPMCP membership. We also searched KPMCP outpatient, inpatient, and out-of-plan utilisation databases to determine who did not have utilisation during the 21 days following NICU discharge, and would therefore be lost to follow up. These babies were not included in multivariate analyses.

    Newborn groups, diagnoses, and severity of illness

    Based on previous studies,12,15,21–,23 we suspected that some infants were at higher risk of rehospitalisation, and that five diagnostic categories (jaundice only; jaundice with other diagnoses; feeding difficulties; evaluation for bacterial infection (“rule out sepsis”); and confirmed urinary tract infections) accounted for most rehospitalisations. We divided survivors into three groups: very premature (<34 weeks) infants, who are automatically admitted to the NICU; moderately premature infants (34–36.9 weeks); and term infants (⩾37 weeks). Among babies ⩾34 weeks, we examined three admission types: normal newborn care; brief (<24 hours) admission to the NICU; and formal NICU admission (⩾24 hours in the NICU).

    We grouped International Classification of Diseases24 rehospitalisation discharge diagnoses as follows:

    • Jaundice: 773.0, 773.1, 773.2, 773.4, 774.0, 774.1, 774.2, 774.30, 774.31, 774.39, 774.6, 774.7, 782.4 and/or procedure codes 99.83 and/or 99.01

    • Feeding difficulties: diagnoses beginning with 276; 775.5, 779.3, 785.50, and 785.59

    • “Rule out sepsis”: V29.0, V29.1, 079.9, 778.4, and 780.6

    • Confirmed urinary tract infection: diagnoses beginning with 590 and 599.0.

    When more than one category applied, we counted the child as belonging to each category. A rehospitalisation where any of the following occurred was considered “severe/prolonged”: (1) LOS exceeded seven days; (2) assisted ventilation; (3) inter-hospital transport; and/or (4) death.

    All newborns who remained in the NICU for ⩾24 hours were assigned a 12 hour Score for Neonatal Acute Physiology, version II (SNAP-II).25 The SNAP-II score, which is similar in some ways to the CRIB,26 assigns points based on physiological derangements in the first 12 hours in the NICU as measured by the following variables: (1) the lowest mean arterial blood pressure, temperature, ratio of arterial oxygen tension to fraction of inspired oxygen, and pH; (2) whether or not a newborn had seizures; and (3) urine output. Babies’ SNAP scores can range between 0 and 115, although survival is unusual with scores >50. A general rule of thumb is that, above a score of 9, one SNAP-II point is roughly equivalent to a 0.3–0.6% mortality risk. For analytical purposes, newborns not assigned a SNAP-II (healthy newborns or newborns in the NICU for <24 hours) received scores of 0.

    Statistical methods

    Bivariate comparisons of potential predictors that were categorical to rehospitalisation were performed using the χ2 test or Fisher’s exact test in the case of small cell sizes. Continuous variables were examined using Student’s t test if normally distributed and the Wilcoxon non-parametric test if not normally distributed.

    Initial selection of variables for multivariate models was based on the results of bivariate comparisons, biological plausibility (for example, we included race because it is associated with differences in bilirubin levels27), and results of previous studies.12,15,18 Centre was included in these models as a fixed effect. Variables were retained in the final model if they reached statistical significance or if the fact that they did not reach statistical significance was clinically significant.

    The small number of centres in our study limited our ability to employ statistical techniques that specifically address the issue of clustering.28 Therefore, we used a simplified approach to explore centre level predictor effects (for example, monthly delivery volume). This consisted of eliminating the facility variable from the multivariate model, substituting the centre level predictor of interest for the facility variable, determining the adjusted odds ratio (AOR) and 95% confidence interval (95% CI) for the predictor, and comparing the c-statistics of the resulting models to assess predictive ability. A c-statistic value near 0.5 indicates no discriminatory power, whereas a value of 1.0 indicates perfect discrimination.29 The value of the c-statistic also equals the area under the receiver operating characteristic (ROC) curve. All statistical analyses employed SAS.30

    RESULTS

    During the study period, 33 374 live births occurred at the study sites and 33 276 (99.7%) survived to discharge. A total of 3811 babies were admitted to the NICU and 3744 survived to discharge. An additional 31 deaths occurred among newborns never admitted to the NICU (newborn comfort care and delivery room deaths). We did not find significant inter-facility differences in in-hospital neonatal mortality rates. Of the 33 276 survivors, 862 (2.6%) were <34 weeks gestation, 2153 (6.5%) were 34–36 weeks, and 30 261 (90.9%) were 37+ weeks. Rehospitalisation rates varied by gestational age range: 26/862 (3.0%) among babies <34 weeks, 94/2153(4.4%) among babies 34–36 weeks, and 618/30 261 (2.0%) among babies 37+ weeks. Table 1⇓ provides data from individual facilities and shows the variation with respect to numbers and characteristics (for example, birth weight) of patients. Additional data on individual hospitals can be found in an appendix (see ADC website).

    View this table:
    Table 1

     Study cohort description

    The most common reason for rehospitalisation was jaundice, with 253/738 (34.3%) of the 738 rehospitalisations falling into the “jaundice only” category and an additional 129/738 (17.5%) having jaundice plus some other diagnosis. Among the infants in the “jaundice only” category, all but one were ⩾34 weeks gestation at birth. Other reasons were less common: feeding difficulties were present in 26.0% of the rehospitalisations, “rule out sepsis” in 11.1%, severe or prolonged illness in 10.0%, and miscellaneous reasons in 24.1%. Many infants had more than one diagnosis. Among babies rehospitalised with “jaundice only”, 84% of the rehospitalisations occurred within 72 hours after discharge while among other rehospitalisations, 46% occurred within 72 hours.

    Considerable variation in overall rehospitalisation rates and in rehospitalisation rates for specific conditions was present across facilities. The degree of variation with respect to phototherapy use was different if one also considered home phototherapy. For example, among babies 34–36.9 weeks gestation, the jaundice rehospitalisation rate ranged from 0.3% to 4.5%, whereas the range of infants treated with phototherapy (occurring at home or in the hospital) was 2.2–5.9%. Among babies ⩾37 weeks gestation, the jaundice rehospitalisation rate ranged from 0.0% to 1.2%, whereas the range of infants treated with phototherapy was 0.7% to 1.3%. Among babies ⩾34 weeks gestation at birth, the overall rehospitalisation rate for “jaundice only” at the five facilities that did not have ready access to home phototherapy was 1.0%, whereas at the other two facilities this rate was 0.3%. This difference was significant (p < 0.0001). The rate of treatment with phototherapy at the five facilities that did not have ready access to home phototherapy was 1.1%, whereas at the two facilities with ready access this rate was 1.4% (p = 0.0052).

    Multivariate analyses

    Table 2⇓ shows our final multivariate logistic model for prediction of rehospitalisation for all causes. Asian infants were more likely to be rehospitalised (AOR = 1.49), whereas African-American infants were at lower risk (AOR = 0.61). Male infants were more likely to be rehospitalised than female infants (AOR = 1.28), as were infants who were small for gestational age (AOR = 1.76) and those with a SNAP-II ⩾10 (AOR = 2.00). Infants born at 34–36 weeks gestation who were never in the NICU (AOR = 3.10) were at higher risk. Having both a home visit as well as a scheduled outpatient visit within 72 hours was also associated with an increased risk (AOR = 1.94). Having had a home health visit (AOR = 0.66) or a scheduled outpatient visit (AOR = 0.83) within 72 hours after discharge were associated with a decreased risk. Considerable inter-facility variation persisted in this model.

    View this table:
    Table 2

     Multivariate model for all rehospitalisations*

    Table 3⇓ shows our final multivariate model for prediction of rehospitalisation among all infants when infants in the “jaundice only” category were excluded. Results are similar to those of table 2⇑, with the same predictors being statistically significant except for Asian race.

    View this table:
    Table 3

     Multivariate model for non-jaundice rehospitalisations*

    Table 4⇓ compares the results of the model described in table 3⇑ with those obtained when we substituted centre level predictors (for example, monthly delivery volume in units of 50 births/month) for facility terms. None of these models were superior to our original model (all had c statistics <0.636). Centre characteristics associated with decreased rehospitalisation were availability of home phototherapy, being at an alliance facility, having an increased proportion of non-white births, having a higher monthly delivery volume, and having a higher mean SNAP-II. Centre characteristics associated with increased rehospitalisation rates were having a paediatric residency programme and having a higher NICU admission rate.

    View this table:
    Table 4

     Multiple multivariate models for rehospitalisation with facility terms replaced by centre level predictors

    DISCUSSION

    Moderately premature infants constitute more than 80% of all premature infants in the United States,31 but little is known about their outcomes. A growing body of literature suggests that these babies, who have received much less research attention than very low birth weight infants, deserve scrutiny. Moderately premature infants contribute significantly to overall neonatal mortality rates21 and have been reported to have rates of long term morbidity (for example, cerebral palsy) that are intermediate between those of term and very premature infants.15,22,23 It seems likely that their overall contribution to childhood morbidity is significant. Our finding that moderately premature infants who are never admitted to the NICU experience increased rehospitalisation is important. One possible explanation is that clinicians may be making clinical judgments based on infants’ birth weight rather than gestation. In a separate report, we have reported this concern with respect to neonatal jaundice.32

    Despite the presence of an integrated system, significant variation exists in the use of home phototherapy and follow up visits. Several studies have shown that home phototherapy is safe,33–,35 but our study is the first to show the impact of home phototherapy on a population basis. Not all paediatricians are comfortable with this technology.36 Poland has suggested that this reflects a general scepticism about the need to treat jaundice at all,37 a supposition supported by another of our studies.38 On the other hand, our data suggest that, at facilities employing home phototherapy, clinicians may have a lower threshold for initiating it than for rehospitalisation. If one wanted to assess the benefits of home phototherapy, it would be critical to address the issue of who needs phototherapy in the first place, a point raised by Plastino39 and Newman and Maisels.40 Doing so should include better description of the context in which phototherapy is provided (that is, whether it is part of a broader case management system).41

    Although the KPMCP has compared home visits to clinic visits among term infants,42,43 formal assessments of follow up strategies for premature or high risk infants (for example, term babies who experienced assisted ventilation) are absent from the literature. Our findings also highlight the need for better data about the content of home visits and outpatient follow up visits. In one set of patients—those who had both a home visit as well as a clinic visit—the reason for increased rehospitalisation is probably that the nurse conducting the home visit identified the baby as being at higher risk and referred the baby to the clinic, where the final decision to rehospitalise occurred. However, for the remaining patients, electronic data cannot support any inferences as to whether the effect of a given clinic or home visit was to prevent or detect the need for hospital care. While there are some guidelines for what should occur during the course of these visits,44 there are no guidelines for what sort of information should be captured about these visits.

    We also found that increased illness severity, as measured by the SNAP-II, was associated with an increased rehospitalisation risk. Although this makes intuitive clinical sense and is consistent with our previous study,12 this finding is important. SNAP-II was designed to predict mortality but seems to have similar properties to SNAP-I, which, in addition to being validated as a predictor for mortality, also predicts other outcomes, such as LOS.45 The SNAP-II is a simple score and can be assigned in less than five minutes. It may become useful for NICUs to employ it for clinical, not just research, purposes (for example, to help assign babies to different follow up programmes).

    We found that African-American infants are less likely, and Asian infants more likely, to be rehospitalised. Raised bilirubin levels are less common in African-American infants and more common in Asian infants in the KPMCP.17,27 Also, in the United States, African-American mothers are less likely to breast feed, which has a strong effect on the second most common reason for rehospitalisation (feeding difficulties), a factor which we18 and others46 have described.

    We cannot explain the increased likelihood of rehospitalisation in male infants. In one of our previous studies,17,27 we reported an increased risk for severe (⩾25 mg/dl) jaundice in male infants. However, in this study, the increased risk in male infants was not limited to rehospitalisation for jaundice.

    We explored whether centre characteristics, such as patient volume, were associated with rehospitalisation rates. Our results in this area are tentative but could guide future cooperative studies. Phibbs et al have reported improved patient outcomes in NICUs with larger volumes,47 so it is conceivable that centres with more volume and/or experience with handling sicker infants (as evidenced by a facility’s mean SNAP-II) might, over time, have developed better approaches to discharge management.

    In conclusion, we have found that variation in short term neonatal jaundice rehospitalisation rates are strongly affected by the availability of home phototherapy and follow up care. Moderate prematurity and follow up visits largely contributed to the non-jaundice rehospitalisation rate, but variation between centres persisted even after controlling for multiple factors. Therefore, variation among newborn rehospitalisation rates cannot be explained completely by either biological or health services factors. Future research should include development of better process measures for evaluation of follow up strategies.

    Acknowledgments

    This project was supported by a grant from the Sidney Garfield Memorial Fund. Funding for the Neonatal Minimum Data Set Database in California was provided by the Kaiser Foundation Health Plan, Inc., and by The Permanente Medical Group, Inc. We wish to thank Ms Emily Breed and Ms Kimberley Harris for assistance with preparation of the manuscript. We also wish to thank Dr Andy Avins for reviewing the manuscript.

    REFERENCES

    1. ↵
      Braveman P , Kessel W, Egerter S, et al. Early discharge and evidence-based practice. Good science and good judgment. JAMA1997;278:334–6.
      OpenUrlCrossRefPubMedWeb of Science
    2. Congress of the United States of America. Newborns and Mothers Health Protection Act (P.L. 96-104) 1996.
    3. Congress of the United States of America. Departments of Veterans Affairs and Housing and Urban Development, and Independent Agencies Appropriation Act1997.
    4. Gazmararian JA, Koplan JP. Length-of-stay after delivery: managed care versus fee-for-service. Health Aff (Millwood)1996;15 (4) :74–80.
      OpenUrlFREE Full Text
    5. Kessel W , Kiely M, Nora A, et al. Early discharge: in the end, it is judgment. Pediatrics1995;96 (4 pt 1) :739–42.
      OpenUrlAbstract/FREE Full Text
    6. Soskolone EL, Schumacher R, Fyock C, et al. The effect of early discharge and other factors on readmission rates in newborns. Arch Pediatr Adolesc Med1996;150:373–9.
      OpenUrlCrossRefPubMedWeb of Science
    7. Lee KS, Perlman M, Ballantyne M, et al. Association between duration of neonatal hospital stay and readmission rate. J Pediatr1995;127:758–66.
      OpenUrlCrossRefPubMedWeb of Science
    8. Liu L , Clemens C, Shay D, et al. The safety of newborn early discharge. JAMA1997;278:293–8.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      Edmonson MB, Stoddard JJ, Owens LM. Hospital readmission with feeding-related problems after early postpartum discharge of normal newborns. JAMA1997;278:299–303.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      Kotagal UR, Perlstein PH, Gamblian V, et al. Description and evaluation of a program for the early discharge of infants from a neonatal intensive care unit. J Pediatr1995;127:285–90.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      Gray JE, McCormick MC, Richardson DK, et al. Normal birth weight intensive care unit survivors: outcome assessment. Pediatrics1996;97 (6 pt 1) :832–8.
      OpenUrlAbstract/FREE Full Text
    12. ↵
      Escobar GJ, Joffe SJ, Gardner MN, et al. Rehospitalization in the first two weeks after discharge from the neonatal intensive care unit. Pediatrics1999;104:1–9.
      OpenUrlAbstract/FREE Full Text
    13. ↵
      Feachem RG, Sekhri NK, White KL. Getting more for their dollar: a comparison of the NHS with California’s Kaiser Permanente. BMJ2002;324:135–41.
      OpenUrlAbstract/FREE Full Text
    14. ↵
      Selby JV. Linking automated databases for research in managed care settings. Ann Intern Med1997;127 (8 pt 2) :719–24.
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      Cavalier S , Escobar GJ, Fernbach SA, et al. Postdischarge utilization of medical services by high-risk infants: experience in a large managed care organization. Pediatrics1996;97:693–9.
      OpenUrlAbstract/FREE Full Text
    16. Escobar GJ. The neonatal “sepsis work-up”: personal reflections on the development of an evidence-based approach toward newborn infections in a managed care organization. Pediatrics1999;103 (1 suppl E) :360–73.
    17. ↵
      Newman TB, Escobar GJ, Gonzales VM, et al. Frequency of neonatal bilirubin testing and hyperbilirubinemia in a large health maintenance organization. Pediatrics1999;104 (5 pt 2) :1198–203.
      OpenUrlCrossRefPubMedWeb of Science
    18. ↵
      Escobar GJ, Gonzales VM, Armstrong MA, et al. Rehospitalization for neonatal dehydration: a nested case-control study. Arch Pediatr Adolesc Med2002;156:155–61.
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      Meikle SF, Lyons E, Hulac P, et al. Rehospitalizations and outpatient contacts of mothers and neonates after hospital discharge after vaginal delivery. Am J Obstet Gynecol1998;179:166–71.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      Escobar GJ, Fischer A, Kremers R, et al. Rapid retrieval of neonatal outcomes data: the Kaiser Permanente Neonatal Minimum Data Set. Quality Management in Health Care1997;5 (4) :19–33.
      OpenUrlCrossRefPubMed
    21. ↵
      Kramer MS, Demissie K, Yang H, et al. The contribution of mild and moderate preterm birth to infant mortality. Fetal and Infant Health Study Group of the Canadian Perinatal Surveillance System. JAMA2000;284:843–9.
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      Holberg CJ, Wright AL, Martinez FD, et al. Risk factors for respiratory syncytial virus-associated lower respiratory illnesses in the first year of life. Am J Epidemiol1991;133:1135–51.
      OpenUrlPubMedWeb of Science
    23. ↵
      Boyce TG, Mellen BG, Mitchel EF Jr, et al. Rates of hospitalization for respiratory syncytial virus infection among children in Medicaid. J Pediatr2000;137:865–70.
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      NIH Consensus Development Panel on the Effect of Corticosteroids for Fetal Maturation on Perinatal Outcomes. Effect of corticosteroids for fetal maturation on perinatal outcomes. JAMA1995;273:413–18.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      Richardson DK, Corcoran JD, Escobar GJ, et al. SNAP-II and SNAPPE-II: simplified newborn illness severity and mortality risk scores. J Pediatr2001;138:92–100.
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      International Neonatal Network. The CRIB (clinical risk index for babies) score: a tool for assessing initial neonatal risk and comparing performance of neonatal intensive care units. The International Neonatal Network [published erratum appears in Lancet 1993;342: 626], Lancet1993;342:193–8.
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      Newman TB, Xiong BX, Gonzales VM, et al. Prediction and prevention of extreme neonatal hyperbilirubinemia in a mature health maintenance organization. Arch Pediatr Adolesc Med2000;154:1140–7.
      OpenUrlCrossRefPubMedWeb of Science
    28. ↵
      Localio AR, Berlin JA, Ten Have TR, et al. Adjustments for center in multicenter studies: an overview. Ann Intern Med2001;135:112–23.
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      Harrell FE Jr, Lee KL, Califf RM, et al. Regression modelling strategies for improved prognostic prediction. Stat Med1984;3:143–52.
      OpenUrlCrossRefPubMedWeb of Science
    30. ↵
      SAS. Statistical Analysis Software, 6th edn. Carey, NC: SAS Institute 1989.
    31. ↵
      Ventura SJ, Martin JA, Curtin SC, et al. Births: final data for 1999. Natl Vital Stat Rep2001;49 (1) :1–100.
      OpenUrlPubMed
    32. ↵
      Newman TB, Liljestrand P, Escobar GJ. Infants with total serum bilirubin levels ⩾30 mg/dl in a large managed care organization. Pediatrics2003;111:1303–11.
      OpenUrlAbstract/FREE Full Text
    33. ↵
      Slater L . Home versus hospital phototherapy for term infants with hyperbilirubinemia: a comparative study. Pediatrics1984;73:515–19.
      OpenUrlAbstract/FREE Full Text
    34. Rogerson A . 14 Years of experience with home phototherapy. Clin Pediatr1986;25:296–9.
    35. ↵
      Grabert B . Home phototherapy. An alternative to prolonged hospitalization of the full-term, well newborn. Clin Pediatr1986;25:291–4.
    36. ↵
      Meropol SB, Luberti AA, De Jong AR, et al. Home phototherapy: use and attitudes among community pediatricians. Pediatrics1993;91:97–100.
      OpenUrlAbstract/FREE Full Text
    37. ↵
      Poland RL. Home phototherapy: Not seeing the light. Pediatrics1993;91:147.
      OpenUrlAbstract/FREE Full Text
    38. ↵
      Atkinson LR. Phototherapy use in jaundiced newborns in a large managed care organization: do clinicians adhere to the guideline? Pediatrics2003;111:e555–61.
      OpenUrlAbstract/FREE Full Text
    39. ↵
      Plastino R . Impact of eligibility criteria on phototherapy program size and cost. Pediatrics1990;85:796–800.
      OpenUrlAbstract/FREE Full Text
    40. ↵
      Newman TB, Maisels MJ. Evaluation and treatment of jaundice in the term newborn: a kinder, gentler approach. Pediatrics1992;89:809–18.
      OpenUrlAbstract/FREE Full Text
    41. ↵
      Spinner S . Earlier discharge of infants from neonatal intensive care units: a pilot program of specialized case management and home care. Clin Pediatr1998;37:353–7.
      OpenUrlCrossRefPubMed
    42. ↵
      Escobar GJ, Braveman PA, Ackerson L, et al. A randomized comparison of home and hospital-based group follow-up visits after early postpartum discharge. Pediatrics2001;108:719–27.
      OpenUrlAbstract/FREE Full Text
    43. ↵
      Lieu TA, Braveman PA, Escobar GJ, et al. A randomized comparison of home and clinic follow-up visits after early postpartum hospital discharge. Pediatrics2000;105:1058–65.
      OpenUrlAbstract/FREE Full Text
    44. ↵
      Green M . ed. Bright futures. Guidelines for health supervision of infants, children and adolescents. Arlington, VA: National Center for Education in Maternal and Child Health, 1994.
    45. ↵
      Escobar GJ, Fischer A, Li DK, et al. Score for neonatal acute physiology: validation in three Kaiser Permanente neonatal intensive care units. Pediatrics1995;96 (5 pt 1) :918–22.
      OpenUrlAbstract/FREE Full Text
    46. ↵
      Kotagal UR, Atherton HD, Eshett R, et al. Safety of early discharge for Medicaid newborns. JAMA1999;282:1150–6.
      OpenUrlCrossRefPubMedWeb of Science
    47. ↵
      Phibbs CS, Bronstein JM, Buxton E, et al. The effects of patient volume and level of care at the hospital of birth on neonatal mortality. JAMA1996;276:1054–9.
      OpenUrlCrossRefPubMedWeb of Science

    Supplementary materials

    • Web-only Appendix

      The appendix is available as a downloadable PDF (printer friendly file).

      If you do not have Adobe Reader installed on your computer,
      you can download this free-of-charge, please Click here

       

      Files in this Data Supplement:

    Footnotes

    • Competing interests: none declared

    • Presented at a poster session at the May 2002 meeting of the Society for Pediatric Research

    Linked Articles

    • Perspectives
      Readmission of neonates
      R Scott-Jupp
      Archives of Disease in Childhood 2005; 90 111-112 Published Online First: 21 Jan 2005. doi: 10.1136/adc.2004.052563