Maternal and Child Health Journal

, Volume 17, Issue 6, pp 1016–1024

Health Care Utilization in the First Year of Life among Small- and Large- for-Gestational Age Term Infants

Authors

    • Division of Reproductive Health, National Center for Chronic Disease Prevention and Health PromotionCenters for Disease Control and Prevention
  • Joanne H. Rizzo
    • The Center for Health ResearchKaiser Permanente Northwest
  • Lucinda J. England
    • Division of Reproductive Health, National Center for Chronic Disease Prevention and Health PromotionCenters for Disease Control and Prevention
  • William M. Callaghan
    • Division of Reproductive Health, National Center for Chronic Disease Prevention and Health PromotionCenters for Disease Control and Prevention
  • Kimberly K. Vesco
    • The Center for Health ResearchKaiser Permanente Northwest
  • F. Carol Bruce
    • Division of Reproductive Health, National Center for Chronic Disease Prevention and Health PromotionCenters for Disease Control and Prevention
  • Joanna E. Bulkley
    • The Center for Health ResearchKaiser Permanente Northwest
  • Andrea J. Sharma
    • Division of Reproductive Health, National Center for Chronic Disease Prevention and Health PromotionCenters for Disease Control and Prevention
  • Mark C. Hornbrook
    • The Center for Health ResearchKaiser Permanente Northwest
Article

DOI: 10.1007/s10995-012-1082-z

Cite this article as:
Dietz, P.M., Rizzo, J.H., England, L.J. et al. Matern Child Health J (2013) 17: 1016. doi:10.1007/s10995-012-1082-z
  • 225 Views

Abstract

The objective of the study was to assess if small- and large-for gestational age term infants have greater health care utilization during the first year of life. The sample included 28,215 singleton term infants (37–42 weeks) without major birth defects delivered from 1998 through 2007 and continuously enrolled at Kaiser Permanente Northwest for 12 months after delivery. Birth weight for gestational age was categorized into 3 levels: <10th percentile (SGA), 10–90th percentile (AGA), >90th percentile (LGA). Length of delivery hospitalization, re-hospitalizations and sick/emergency room visits were obtained from electronic records. Logistic regression models estimated associations between birth weight category and re-hospitalization. Generalized linear models estimated adjusted mean number of sick/emergency visits. Among term infants, 6.2 % were SGA and 13.9 % were LGA. Of infants born by cesarean section, SGA infants had 2.7 higher odds [95 % 1.9, 3.8] than AGA infants of staying ≥5 nights during the delivery hospitalization; of those born vaginally, SGA infants had 1.5 higher adjusted odds [95 % 1.1, 2.1] of staying ≥4 nights. LGA compared to AGA infants had higher odds of re-hospitalization within 2 weeks of delivery [OR 1.25, 95 % CI 0.99, 1.58] and of a length of stay ≥4 days during that hospitalization [OR 2.6, 95 % CI 1.3, 5.0]. The adjusted mean number of sick/emergency room visits was slightly higher in SGA (7.8) than AGA (7.5) infants (P < .05). Term infants born SGA or LGA had greater health care utilization than their counterparts, although the increase in utilization beyond the initial delivery hospitalization was small.

Keywords

Low birth weightFetal growth restrictionLarge for gestational ageMacrosomiaHealth care utilization

Introduction

Infants born small- (SGA) and large- for-gestational age (LGA) experience greater health complications shortly after birth compared to infants born appropriate for gestational age (AGA). SGA infants experience greater risk of seizures, respiratory distress, hypoglycemia, and hyperbilirubinemia [1]. On the other end of the birth weight spectrum, infants born LGA experience higher rates of birth trauma and jaundice [2, 3]. Previous studies have found that the greatest complication rates occur among those in the tails of the birth weight distribution. For example, among 82,361 term infants delivered at Parkland Hospital in Dallas between 1988 and 1996, term infants born at or below the 3rd birth weight percentile compared to infants born between the 26th and 75th birth weight percentiles had a neonatal mortality rate almost 10 times higher [4]; infants born in the 4–5th birth weight percentile had no statistically significant differences in outcomes. Another study found fetal death rates highest among infants born at the lowest birth weight percentile for gestational age, but found elevated fetal death rates for fetuses for the 10th through the 15th percentile [5]. Macrosomic infants with birth weight >4,500 g compared with infants with birth weights 3,500–3,999 g are at increased risk for stillbirth, neonatal mortality, birth injury, asphyxia and meconium aspiration; however, infants with births weights 4,000–4,499 g are not at increased risk of mortality or morbidity compared with infants with birth weights 3,500–3,999 g [3]. While previous studies have focused on morbidity recorded during the delivery hospitalizations, none have examined morbidity and health care utilization after delivery hospitalization during the entire first year of life. A comprehensive assessment of the burden of health care utilization in the first year of life may identify groups of infants with the greatest health care needs and potential preventable morbidities. We evaluated health care utilization patterns and morbidity during the first year of life among infants born SGA and LGA, specifically examining infants with the lowest and highest birth weights when numbers allowed. We hypothesized that being born SGA or LGA would be associated with higher health care utilization in the first year of life.

Methods

We analyzed data from Kaiser Permanente Northwest (KPNW), a large nonprofit prepaid, federally certified, Joint Commission-accredited, group practice HMO with approximately 475,000 members in western Oregon and Washington State as of January 1, 2011. Members include individuals and families covered by commercial group and individual self-pay health plans, Washington State Basic Health Plan (subsidized, Washington State only), Medicare Advantage, and Medicaid (Oregon and Washington State).

The data for this study came from various KPNW individual-level clinical and administrative data systems, including electronic medical records, enrollment, hospital discharge, outpatient encounters, emergency department, outside claims and referrals, imaging, laboratory, and pharmacy. We adapted a computerized algorithm that links indicators and dates of pregnancies and pregnancy outcomes to create pregnancy “episodes” [6]. Validation of this algorithm with experienced medical records technicians showed 100 % agreement for live birth outcomes between episodes identified by the algorithm and those identified through medical record reviews. After pregnancy episodes were identified, mother and infant records were readily linked as mothers’ health insurance IDs are typically entered into the infants’ medical and administrative records at birth. KPNW infant records were then matched to live birth certificate records, providing additional information on race/ethnicity, other demographic variables, and infant birth weight. A probabilistic method was used to link birth certificate records with KPNW records using a scored matching system based on mother’s name (maiden and married), date of birth, and address, and infant’s name, date of birth, and facility of delivery/birth. For this study, 85.9 % of live births were matched to a live birth certificate. This study was approved by the Centers for Disease Control and Prevention and the KPNW Institutional Review Boards.

We identified 46,807 singleton term (37–42 weeks gestational age) infants born between January 1, 1998, and December 31, 2007, who were matched to a mother’s record. Infants were not eligible for this study if they died in the first year of life (n = 70), or were not enrolled in KPNW for their entire first year of life (n = 13,198). Infants with a major birth defect (structural or genetic) (n = 3,126), or whose mother had a diagnosed drug or alcohol dependency (n = 642) were also excluded due to potential effects of these conditions on pregnancy outcomes and health care utilization. Among infants who were eligible (n = 29,771), we excluded 1,122 who had unreliable gestational age information, 338 with no data on the delivery hospitalization, and 96 without an available birth weight. We analyzed 28,215 deliveries to 23,371 mothers.

Infant birth weight was taken primarily from the hospital discharge record, and birth weight from the birth certificate was used to verify it. Infant gestational age in completed weeks was based on the physician’s best obstetric estimate recorded in the prenatal medical record. Infant birth weight for gestational age was categorized into three mutually exclusive categories: <10th percentile (SGA), 10–90th percentile (AGA), >90th percentile (LGA). When numbers allowed, we also looked at the infants in the extreme tail of the birth weight distribution, <3rd percentile for gestational age and macronomic infants ≥4,500 g. All 2004 US singleton term births were used as the birth weight reference population and cut points were stratified by infant sex and race/ethnicity (see Appendix 1 for cut points). Gestational age in the reference population was based on clinical or obstetric estimate as these measures perform better than using last menstrual period [7].

We examined 6 indicators of health care utilization for the infants during their first year of life, including length of stay for the delivery hospitalization, percent of infants who were re-hospitalized within 2 weeks of delivery, percent of infants re-hospitalized 3–52 weeks after delivery, and mean number of days for those re-hospitalizations, and the mean number of non-routine outpatient visits (sick and emergency room visits combined). International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) and Current Procedural Terminology, Fourth Edition (CPT-4) diagnosis and procedure codes were used to define health care utilization and diagnoses. Well child care visits were defined as visits with ICD-9-CM code V20.2 or CPT-4 procedure codes 99381 or 99391; all outpatient visits that were not identified as well visits were classified as sick visits. Sick and emergency room visits were combined into one measure, as the number of emergency room visits was small and the reasons for the two types of visits were similar. Total length of stay was defined as the number of days from admission to discharge. Morbidity data based on ICD-9-CM and CPT-4 diagnosis and procedure codes in the health care episodes were grouped into related conditions (e.g. all respiratory infections, jaundice, birth trauma). We examined the five most common diagnoses at hospitalizations and sick/emergency room visits. Codes were available for 99 % (28,027/28,215) of delivery hospitalizations, 99 % (503/509) of re-hospitalizations within 2 weeks of delivery, 94 % (1,125/1,196) of re-hospitalizations within 3–52 weeks, and 90 % (162,833/180,081) of sick/emergency room visits. See Appendix 2 for ICD-9-CM codes included in the grouped conditions.

We evaluated several maternal demographic variables as possible confounders, which were derived from birth certificate and KPNW electronic medical record (EMR) sources. The birth certificate was the primary source of mother’s educational attainment, and race/ethnicity, and the secondary source of pre-gestational diabetes and hypertension during pregnancy, while the EMR was the primary source for mother’s parity. The EMR also provided pregnancy complications, and maternal behaviors as possible confounders including diagnoses of maternal mental health conditions (depression, anxiety, and other mental health conditions); pre-gestational and gestational diabetes mellitus; hypertensive disorder during pregnancy; and smoking. Hypertensive disorders included chronic hypertension, gestational hypertension, and preeclampsia/eclampsia. Prenatal smoking was ascertained from the mother’s prenatal encounter records, and/or from the infant birth certificate. Infant tobacco exposure after delivery was ascertained from maternal health care encounter data (maternal smoking status).

Confounders were defined as covariates that changed the association between birth weight groups (SGA, AGA, and LGA) and the health care utilization measures by more than 10 % in bivariate analyses. Logistic regression with generalized estimating equations (GEE) (to account for correlations among siblings and repeated visits or hospitalizations) was used to assess the association between birth weight groups and re-hospitalization within 2 weeks and re-hospitalization between 3 and 52 weeks after delivery as well as length of hospital stay. Mean number of sick/emergency room visits were adjusted for confounding variables and differences were tested for statistical significance using GEE. Pairwise statistical comparisons were calculated, using AGA as the reference group for all comparisons. Statistical analyses were run in Statistical Analysis Software (SAS) version 9.2 (SAS Institute, Cary, NC).

Results

We began by examining differences between eligible infants who were included (n = 28,215) or excluded (n = 1,556) from the sample. Excluded infants were more likely to have mothers who were non-white, not enrolled in Medicaid, and who had <12 years of education. Mothers of included and excluded infants had similar distributions of age, parity, and percent with a diagnosis of hypertension or diabetes (data not shown). Among infants included in the study, 0.8 % were SGA < 3rd, 5.4 % were SGA 3rd-10th (total SGA 6.2 %), 79.9 % were AGA, and 11.3 % were LGA ≤ 4,500 g and 2.6 % were macrosomic (total LGA 13.9 %). During the study period, the proportion of infants born LGA decreased from 14.8 % in 1998 to 12.7 % in 2007 and the proportion of infants born SGA remained stable.

Maternal and Pregnancy Characteristics

Maternal and pregnancy characteristics varied by birth weight, with the exception of mental health diagnoses and Medicaid status. Compared with the other birth weight groups, SGA infants were more likely to be born at 37–38 weeks gestation, and have mothers who were nulliparous, diagnosed with hypertension during pregnancy, and smoked either during and/or after pregnancy (Table 1) Mothers of LGA infants were more likely to be older, white, multiparous, diagnosed with pre-gestational or gestational diabetes, deliver by cesarean section and non-smokers compared with those of other birth weight groups. When the sample was limited to only the first pregnancy episode per woman, the findings remained unchanged.
Table 1

Characteristics of mother/infant pairs of singleton term births by infant birth weight

Characteristic

SGA

N = 1,738 (%)

AGA

N = 22,546 (%)

LGA

N = 3,931 (%)

P value for Chi-square

Infant gestational age (weeks)

 37–38

32.3

23.5

19.9

<0.0001

 39–40

58.1

61.3

61.4

 

 41–42

9.6

15.2

18.6

 

Age (years)

 <18

3.9

3.1

1.5

<0.0001

 18–24

30.8

27.4

21.7

 

 25–29

29.8

31.7

33.7

 

 30–34

22.3

25.6

28.6

 

 35–39

11.0

10.5

12.6

 

 ≥40

2.2

1.8

1.9

 

Race/ethnicitya

 White

72.3

75.1

77.3

0.0003

 Black

3.6

3.8

3.6

 

 Hispanic

9.0

8.6

8.6

 

 Asian

12.3

10.5

8.7

 

 Other/unknown

2.8

1.9

1.7

 

Education (years)b

 <12

24.9

23.3

20.8

0.001

 12

27.8

26.4

26.3

 

 >12

47.3

50.3

53.0

 

Medicaid or Washington basic health plan

12.4

11.3

10.6

0.14

Parityc

 0

54.0

42.9

30.6

<0.0001

 1

27.4

33.4

39.1

 

 2

11.3

14.6

17.7

 

 ≥3

7.3

9.1

12.6

 

Cesarean delivery

23.0

21.0

31.4

<0.0001

Gestational diabetes mellitusd

3.1

3.0

4.9

<0.0001

Pre-gestational diabetes mellitusd

0.9

1.1

2.9

<0.0001

Hypertensive disorders during pregnancyd

17.7

11.4

13.2

<0.0001

Tobacco use

 During and/or after pregnancy

18.9

12.8

9.4

<0.0001

 After pregnancy only

9.4

8.5

7.9

 

 No tobacco use

60.9

65.7

68.4

 

 Tobacco use unknown

10.8

13.0

14.3

 

Mental health diagnosesd

 During and after

5.8

5.0

4.3

.16

 During pregnancy only

3.6

3.0

3.2

 

 After pregnancy only

8.4

9.3

9.3

 

 None

82.2

82.7

83.2

 

There were 28,215 deliveries to 23,371 mothers

Birth weight for gestational age: <10th percentile (SGA), 10–90th percentile (AGA), >90th percentile (LGA)

aMissing race/ethnicity = 230

bMissing education = 4,468

cMissing parity = 1,346

dSee Appendix 2 for definition

Delivery Hospitalization

Length of delivery hospitalization was stratified by infants delivered vaginally and by cesarean section. Among infants born by cesarean section, SGA infants had 2.7 higher odds [95 % 1.9, 3.8] than AGA infants of staying 5 or more nights during the delivery hospitalization; among those born vaginally, SGA infants had 1.5 higher adjusted odds [95 % 1.1, 2.1]. LGA infants had a similar length of stay as AGA infants (Table 2). When examining length of stay among infants born SGA < 3rd (data not shown, n = 233), among infants delivered by cesarean section, SGA < 3rd infants had 6.8 times higher adjusted odds [95 % CI 3.6, 12.7] of staying 5 or more nights during the delivery hospitalization than AGA infants. Among infants born vaginally, SGA < 3rd infants had 4.5 higher adjusted odds [AOR 4.5, 95 % CI 2.7, 7.6] of staying 4 or more nights during the delivery hospitalization than AGA infants (data not shown). Diagnoses occurring during the delivery hospitalization varied by birth weight group, with a greater percent of LGA infants diagnosed with birth trauma and with tachypnea compared with AGA infants.
Table 2

Delivery hospitalization by birth weight for gestational age among term singleton infants

Characteristic

SGA

N = 1,738

AGA

N = 22,546

LGA

N = 3,931

Vaginal delivery

 LOS(%)

   

  ≤1 day

28.3

34.0

33.8****

  2 days

59.8

57.8

55.9

  3 days

8.2

5.9

8.0

  ≥4 days

3.7

2.2

2.3

Adjusted odds ratiob LOS ≥ 4 days [95 % CI]

1.5 [1.1, 2.1]

Reference

1.1 [0.9, 1.5]

Cesarean delivery

   

 LOS (%)

   

  ≤3 days

63.2

72.6

67.9****

  4 days

25.1

22.7

26.7

  ≥5 days

11.8

4.8

5.4

Unadjusted odds ratio LOS ≥5 days [95 % CI]

2.7 [1.9, 3.8]

Reference

1.1 [0.9, 1.5]

 Diagnoses during hospitalization (%)c

  Jaundice

10.6

9.1

10.2*

  Respiratory conditions

4.1

3.8

4.4

  Birth trauma

1.8

2.5

3.8****

  Tachypnea

1.6

2.0

3.9****

  Fever

2.3

1.7

1.6

Birth weight for gestational age: <10th percentile (SGA), 10–90th percentile (AGA), >90th percentile (LGA)

LOS length of stay

bAdjusted for parity

cSee Appendix 2 for definitions

P < .05; **** P < .0001

Re-hospitalizations Within 2 Weeks After Delivery

Overall, 1.6 % of all SGA infants, 1.7 % of all AGA infants and 2.2 % of all LGA infants were re-hospitalized within 2 weeks of delivery (Table 3). SGA was not statistically associated with an increased odds of re-hospitalization, however, LGA infants had a 20 % increased odds of re-hospitalization [95 % CI 0.99, 1.58] compared to AGA infants. None of the available variables confounded this association. Re-hospitalized LGA infants were more likely to have an extended length of stay compared to re-hospitalized AGA infants [OR for ≥4 days length of stay 2.6, 95 % CI 1.3, 5.0]. Diagnoses during the re-hospitalizations did not vary by birth weight group. Jaundice was the most common diagnosis during this hospitalization.
Table 3

Non-delivery hospitalization within 2 weeks of delivery by infant birth weight among term singleton infants

 

SGA

AGA

LGA

N = 1,738

N = 22,546

N = 3,931

Re-hospitalization within 2 weeks of delivery % (n)

1.6 (27)

1.7 (393)

2.2 (86)

Unadjusted odds ratio hospitalized [95 % CI]

0.9 [0.6, 1.3]

Reference

1.2 [1.0, 1.6]

LOS (%)

 ≤1 day

48.2

50.1

37.2

 2 days

29.6

31.6

33.7

 3 days

c

10.7

11.6

 ≥4 days

c

7.6

17.4

Unadjusted odds ratio LOS ≥ 4 [95 % CI]

0.5 [0.1, 3.5]

Reference

2.6 [1.3, 5.0]

Diagnoses during hospitalization (%)b

 Jaundice

48.1

51.8

57.3

 Feeding problems

c

25.9

24.4

 Upper and non-upper respiratory infections

c

17.3

23.2

 Respiratory conditions

c

8.4

15.8

 Dehydration

c

16.2

19.5

Birth weight for gestational age: <10th percentile (SGA), 10–90th percentile (AGA), >90th percentile (LGA)

LOS length of stay

bSee Appendix 2 for definitions

cCell size ≤5

Re-hospitalizations Within 3–52 Weeks After Delivery

A total of 1,075 infants (3.8 %) were re-hospitalized between 3 and 52 weeks after delivery (Table 4): 91.3 % of re-hospitalized infants had one re-hospitalization during the time period, 6.7 % had two, and 2.0 % had three or more (data not shown). Infant birth weight was not associated with being re-hospitalized between 3 and 52 weeks after delivery or with any specific morbidity.
Table 4

Non-delivery hospitalization within 3–52 weeks of delivery by infant birth weight among term singleton infants

 

SGA

AGA

LGA

N = 1,738

N = 22,546

N = 3,931

Re-hospitalization 3–52 weeks after delivery % (n)

4.3 (75)

3.7 (844)

4.0 (156)

Unadjusted odds ratio for hospitalized [95 % CI]

1.2 [0.9, 1.5]

Reference

1.1 [0.9, 1.3]

LOS (%)

 ≤1 day

42.7

41.8

41.0

 2 days

25.3

28.1

28.2

 3 days

12.0

13.5

14.1

 ≥4 days

20.0

16.6

16.7

Unadjusted odds ratio LOS ≥ 4 days [95 % CI]

1.2 [0.7, 2.2]

Reference

1.0 [0.6, 1.6]

Diagnoses during hospitalization (%)b

 Upper respiratory infections

36.4

36.0

37.1

 Non-upper respiratory infections

20.8

25.3

29.3

 Respiratory conditions

22.1

16.5

12.0

 Dehydration

9.1

16.2

13.2

 Otitis

c

12.8

15.0

Birth weight for gestational age: <10th percentile (SGA), 10–90th percentile (AGA), >90th percentile (LGA)

LOS length of stay

bSee Appendix 2 for definitions

cCell size ≤5

Sick/Emergency Room Visits

The adjusted mean number of sick/emergency room visits was slightly higher in SGA < 10th (7.8) than AGA (7.5) infants (P < .05) but no difference was found for LGA infants (Table 5). There were minor differences in the distributions of common diagnoses by birth weight groups listed at sick and emergency room visits.
Table 5

Sick/emergency room visits during first 12 months by birth weight and gestational age among term singleton infants

Characteristic

SGA

N = 1,738

AGA

N = 22,546

LGA

N = 3,931

Unadjusted mean number visits (±SE)

6.6 (0.1)

6.3 (0.04)

6.4 (0.1)

Adjusted mean (±SE)a

7.8 (0.2)*

7.5 (0.1)

7.6 (0.1)

Diagnoses during visits (%)b

 Upper respiratory infections

24.2

25.3

24.9**

 Otitis

18.2

19.5

20.7****

 Non-upper respiratory infections

11.3

12.3

12.3**

 Respiratory conditions

7.7

7.1

6.7**

 Feeding problems

5.7

4.2

4.2****

Birth weight for gestational age: <10th percentile (SGA), 10–90th percentile (AGA), >90th percentile (LGA)

aAdjusted for medicaid status and mental health diagnosis

bSee Appendix 2 for definitions

P < .05 compared to AGA group

** P < .05; **** P < .0001

Discussion

The study results supported our hypothesis that SGA and LGA infants compared to AGA infants experience marginally greater health care utilization in the first year of life. The amount of increased utilization was relatively small, and it was greater for LGA infants than for SGA infants. After the delivery hospitalization, SGA infants had only a slight increase in one health care utilization measure, the mean number of sick/ER visits in the year. LGA infants experienced higher utilization in two costlier measures: re-hospitalizations within 2 weeks of delivery and length of stay at this re-hospitalization within 2 weeks of delivery. Not surprising, among all infants, jaundice was the most common reason for hospitalizations within the first 2 weeks of delivery, a finding consistent in the literature [8]. The KPNW Region has a “Mother Baby Program” that is structured to evaluate mothers and newborns in an outpatient setting one to 2 days after the delivery discharge with the specific goal of identifying any concerns, such as jaundice or poor feeding, in the immediate postpartum period. Additional home visits and/or return to medical office visits are prescribed as needed and portable light boxes for home treatment are available. Continued close monitoring of infants at risk for hyperbilirubinemia by parents and medical staff may reduce re-hospitalizations for this condition.

In comparison to numerous studies that have examined health care utilization in the first year of life among preterm infants [911], there is a dearth of studies focused on SGA or LGA term infants. Consistent with previous studies, we found that term infants born SGA had longer mean length of stay at delivery hospitalization and that a greater proportion of LGA infants experienced birth trauma compared to AGA infants [13] Also consistent with a previous study, we found SGA infants below the 3rd birth weight percentile for gestational age had longer length of delivery stays than infants between the 3rd and 10th percentile [4].

Unexpectedly, health care utilization after delivery did not differ greatly among infants who were born SGA compared to infants whose weight was appropriate for gestational age. This may reflect that fetuses suspected of fetal growth restriction are delivered before 37 weeks due to close prenatal observation of women in the KPNW system where there is uniform access to health care. In such a system, one might expect that suspected intrauterine growth restriction would be detected prior to term as long as mothers maintain recommended/prescribed prenatal visits. This could result in a group of term SGA infants who, while smaller than their cohorts, are less dramatically affected by placental insufficiency, and thus are constitutionally small but otherwise healthy.

This analysis has several limitations, primarily related to generalizability and the use of administrative data. Our findings are not necessarily generalizable to all US infants as our study population included pregnant women and infants enrolled in a managed care health plan located in the Pacific Northwest United States, and was composed of higher percentages of white women in employed families than the general US population. KPNW health care practices also may not reflect those of other providers. Practices regarding the timing of delivering term infants and length of stay for a delivery hospitalization likely vary among clinicians and medical institutions. In addition, miscoded administrative data could lead to misclassification of clinical conditions. However, as administrative data systems of managed health care plans are designed to monitor the use of resources, we consider substantial errors in coding or underreporting to be unlikely [6]. Even with a relatively large number of pregnancies, births at the extremes of the birthweight distribution were rare and this limited our ability to look at first-year utilization in the very smallest and largest of infants. Finally, we were unable to control for some potential confounders, namely breastfeeding and daycare attendance, as these were not available in the electronic medical record data.

Among singleton, term infants enrolled in a managed health care system for the entire first year of life, infants born SGA or LGA had greater health care utilization than infants born normal weight. After delivery hospitalization, the increase in health care utilization was relatively small and greater for LGA infants than for SGA infants.

Acknowledgments

The findings and conclusions in this article are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention or Kaiser Permanente.

Copyright information

© Springer Science+Business Media, LLC (outside the USA)  2012