Introduction:

Birth of very low birth weight (VLBW) infants outside subspecialty perinatal centers increases risk for death and major morbidities.

Objective:

The purpose of this study is to evaluate barriers to utilizing a regional perinatal center for the birth of VLBW infants to mothers not living in the immediate vicinity of the center.

Methods:

We conducted a retrospective cohort study of VLBW infants residing in the catchment area of a community level II, Specialty Neonatal Unit (SN) admitted to a Regional Subspecialty Neonatal Intensive Care Unit (RC) between January 1999 and December 31, 2004. Maternal demographics and prenatal care as well as outcomes were compared by place of birth.

Results:

Out of 98 VLBW infants admitted to the RC, 49 (50%) were delivered outside the RC (out-born) and 49 (50%) were born at the RC (in-born). There was no statistical difference in insurance coverage, race, gestational age, severity of illness or maternal demographic factors between out-born and in-born infants. Less than adequate prenatal care rather than distance of maternal residence from the RC was associated with birth outside the RC. Adjusting for prenatal care, distance of residence from the RC increased the risk for delivering outside the center in the subset of mothers insured by Medicaid.

Conclusions:

Mothers of VLBW infants who received less than adequate prenatal care and did not live in the vicinity of a subspecialty center had an increased risk for delivery outside that center compared to those with adequate care. Appropriate place of birth for VLBW infants to low-income mothers may be influenced by the distance of their residence to an RC.

Introduction

Regionalization of perinatal care¹ has been shown to improve outcomes among high-risk infants.^{2, 3} Premature infants born at subspecialized perinatal centers have lower probability of death and major morbidities.^{4, 5, 6, 7, 8, 9} Very low birth weight (VLBW) infants transferred to the specialized centers after birth have greater risk for serious morbidities compared to infants born at the subspecialized centers after the transfer of their pregnant mothers.¹⁰

Healthy People 2010 proposed a goal that 90% of VLBW infants be delivered in subspecialty hospitals, and that 90% of pregnant women get early and adequate prenatal care.¹¹ Previous reports have described multiple barriers to adequate utilization of perinatal services related to population characteristics, geographic factors and system delivery elements.^{9, 12, 13, 14, 15, 16, 17} However, the US Department of Health and Human Services¹⁸ cited little progress in these population-based objectives, and identified geographic areas and underserved populations for targeted interventions.¹⁸

We conducted a study of the back transport process in a regionalized perinatal care system regarding utilization of health care services and found that about half of VLBW infants were out-born.¹⁹ We further analyzed place of birth for VLBW infants admitted to the same Regional subspecialty center (RC) over a longer period of time to identify potential barriers for the appropriate place of birth among these high-risk infants.

Materials and methods

This is a retrospective cohort study conducted after obtaining project approval from the Institutional Review Boards at the University of Michigan Health System (UMHS) in Ann Arbor, Michigan and Foote Hospital in Jackson, Michigan. We collected demographic and clinical data including the Score for Neonatal Acute Physiology, Version II (SNAP-II)²⁰ during care of infants treated in a 37-bed neonatal intensive care unit, part of a Regional subspecialty perinatal center (RC) and in a 10-bed Special Care Unit Nursery (SN) (level II) from January 1, 1999 to May 31, 2002 and from April 1, 2003 to December 31, 2004. Data were collected initially as part of an evaluation for intervention measures to improve the back transport process in the regionalized perinatal system. The intervention was implemented between June 1, 2002 and March 31, 2003, and included efforts to improve communication between neonatal and pediatric healthcare providers and RC patient education regarding the back transport process. These intervention measures did not include obstetric providers or pregnant women. Analysis of that data was reported previously.¹⁹ In this report, we combine the data collected from the periods of time before and after the intervention measures as place of birth would not be affected.

Data on the studied infants and characteristics of their mothers were obtained from the UMHS records. Data on prenatal care utilization were obtained from the Michigan Department of Community Health (MDCH) vital statistics records. We assigned level of perinatal care following the perinatal care guidelines of the American Academy of Pediatrics.¹

Population and groups

Very low birth weight infants whose parents resided in the SN catchment area, received prenatal and/or postnatal care in the vicinity of the SN, and were admitted to the RC during the study period were study subjects. The SN catchment area was defined as the county (Jackson County) where the SN is located, as well as the zip codes of areas in the surrounding counties that the community hospital business office considers to be their service area. We further analyzed data on place of birth and clinical course of VLBW infants within this population.

Clinical definitions and measurements

Very low birth weight was birth weight 1500 g. An infant was out-born if the birth site was outside the RC. Infant's Medicaid insurance was used as a marker of socio-economic status of the infant's family. We used the Kessner index collected by the MDCH to describe prenatal care utilization.²¹ Time of initiation, frequency and gestational age were factored in the calculation of the index and reported as adequate, intermediate or inadequate. We defined less than adequate prenatal care as intermediate, inadequate or no prenatal care. Necrotizing entero-colitis (NEC) was diagnosed if patients were treated for possible or radiologically proven NEC when treatment included interruption of feeding and administration of antibiotics for at least a week. Severe retinopathy of prematurity was defined as requiring surgical or laser treatment. A major surgery was defined as any of the following surgeries: ophthalmology surgery, neurosurgery, tracheotomy, surgery for congenital diaphragmatic hernia, heart surgery, extracorporeal membrane oxygenation and other abdominal surgeries except inguinal hernia repair. Web-based geographic files supported by 2005 NAVTEQ Tele Atlas (NAVTEQ Corp., Chicago, IL and Tele Atlas Corp, Lebanon, NH) were used to measure distance from a mother's residence to the RC. The shortest road distance in miles was estimated from the centroid of the maternal residence zip code to the actual address of the RC.

Analysis

The Wilcoxon rank sum test was used to compare continuous data and the ² test was used to compare categorical data. Multivariate logistic regression using a backward stepwise approach was used to determine the association between birth outside the RC and potential effect modifiers and confounders, including maternal race, maternal age, maternal education, insurance, place of birth, antenatal treatment with steroids, previous live births now dead, SNAP II Score and distance to RC (in miles). An interaction term between distance to the RC and insurance was included in the full model to assess effect modification. Evaluation of an interaction term indicated that distance to the RC modified the effect of insurance status on place of birth. Factors were sequentially removed from the model until all variables that remained were either significant at the 0.05 level or were included to ensure good model fit as assessed by the Hosmer–Lemeshow goodness of fit test and the likelihood ratio test.²² Point estimates and 95% confidence intervals (CI) of the adjusted odds ratios (AOR) for the final model are presented. Statistical significance was defined as P0.05. Data were managed and analyzed using SAS statistical software (SAS Institute Inc., Cary, NC).

Results

Among 98 VLBW infants in the study population, 49 (50%) infants were in-born and 49 (50%) infants were out-born and transported to the RC within the first day of life. Demographics, quality of antenatal care, maternal social characteristics and level of education are presented in Table 1. There were no significant differences in insurance coverage, race, gestational age, Apgar Scores, SNAP II Scores, maternal characteristics or distance of residence to the RC between the groups of infants delivered in or outside the RC. However, mothers of out-born infants more frequently had less than adequate prenatal care (82 vs 28%, P<0.0001) and less frequently received prenatal steroid treatment (37 vs 82%, P<0.0001) compared to mothers who delivered at the RC.

Table 1 - Selected infant and maternal characteristics.

Full table

There was no statistical difference in the frequency of mothers who previously gave birth (para gravida 2) to out-born and in-born infants. However, in a subset of mothers who had given birth previously, (n=54), those who had a prior infant death (n=9) more frequently gave birth in the RC compared to mothers who had not experienced an infant death (in-born; 98 vs 11%, P=0.024).

In our population, mothers of infants insured by Medicaid (n=64) were also younger (median (quartile range): 23 (20 to 25) vs 28 (24 to 31) years, P<0.0001), more frequently non-white (33 vs 3%, P<0.001), less than 21 years old (31 vs 9%, P=0.013), not married (75 vs 9%, P<0.0001) and more likely to have less than 13 years of education (83 vs 53%, P=0.004) compared to the mothers of infants with other insurance. However, there was no statistical difference in the site of delivery (out-born; 53 vs 44%, P=0.527) or adequacy of prenatal care (less than adequate antenatal care; 61 vs 47%, P=0.281) between the Medicaid and non-Medicaid insured infants. Distance of residence from the RC was greater for Medicaid insured infants (median (quartile range): 38.5 miles (36.45 to 38.5) vs 36.2 miles (35.8 to 38.5); P=0.028) compared to those with other insurance.

The final model of logistic regression analysis, with place of birth as the dependent variable, included prenatal care, antenatal steroid use, SNAP II Score, insurance, distance to the RC and the interaction term of insurance and distance to the RC. Mothers who had less than adequate prenatal care were more likely to deliver outside the RC compared to mothers who have adequate prenatal care after adjusting for all other variables in the model (AOR 6.78, 95% CI: 2.14 to 21.50). Distance of residence to the RC was predictive of birth outside the RC only for the mothers of infants insured by Medicaid (AOR 1.33, 95% CI: 1.01 to 1.75) but not for the non-Medicaid insured infants (AOR 0.98, 95% CI: 0.87 to 1.10). Mothers who were given antenatal steroid treatment were less likely to deliver outside the RC compared to mothers who were not treated (AOR 0.19, 95% CI: 0.06 to 0.62).

Table 2 shows clinical outcome by site of birth. There were no significant differences in mortality, length of stay or complications related to site of birth, although length of mechanical ventilation tended to be longer in out-born infants (P=0.08).

Table 2 - Infant clinical outcomes.

Full table

Discussion

Half of VLBW infants in this study were born outside an RC, and only 43% of their mothers received adequate prenatal care, far below the recommendations of Healthy People 2010.¹¹ The factor most strongly associated with birth in the RC was receipt of adequate prenatal care. Among infants insured by Medicaid, increased distance of maternal residence from the RC was associated with greater risk of birth at a lower level of care; however, distance was only statistically important for the subset of Medicaid insured infants. Our data suggest that targeted programs are needed to increase adequacy of prenatal care to very high-risk mothers to diminish the risk of extreme prematurity.

The high frequency of out-born infants in this report is similar to findings reported by Samuelson et al.¹⁵ They found that patients living away from subspecialized centers in Georgia had a higher risk for delivering their VLBW infants outside these centers, potentially subjecting the infants to additional morbidity and mortality.¹⁵ Distance of the maternal residence from a subspecialty center was reported to be associated with birth of VLBW infants in level I nurseries in other population based studies.^{15, 17, 23} Obviously, distance to care may present a larger impediment to women with less social capital (e.g. low income, transportation or community resources). Our report, of residents from mainly one county, showed an independent effect for increased distance of maternal residence to the RC and the place of birth among a subset of infants insured by Medicaid.

Type of maternal insurance coverage, maternal age, marital status, level of education and race were reported as factors associated with less reliable use of perinatal services and birth outside the subspecialty centers.^{5, 12, 13, 15, 16} We used infants' qualification for Medicaid insurance as a marker for maternal financial status. Insurance coverage and these other markers of perinatal care under utilization, in our study, were not associated with birth hospital level. We did not demonstrate the higher mortality and major morbidities in the out-born group of our study as reported in larger studies.^{15, 8, 10, 7} Our findings could reflect a study size limitation (underpowered to detect the differences) or alternatively may be a sign of partially effective measures taken by the local healthcare system to improve access to prenatal care for the high-risk subgroups. Such success was reported among women of low socioeconomic status following Medicaid expansions for pregnant women.²⁴

We used the Kessner index, also known as the Institute of Medicine (IOM) Index, reported in the vital statistics for the state of Michigan as a measure of the adequacy of prenatal care.²¹ The Kessner index might underestimate use of prenatal services for pregnancies of more than 36 weeks compared to two newer indices: the Adequacy of Prenatal Care Utilization (APNCU) Index and the R-GINDEX.^{25, 26} The population we studied was born at less than 36 weeks. Analysis of data accuracy from Ohio vital statistics showed that the reported prenatal care had good agreement with information recorded in the medical records and was similarly accurate for teaching and non-teaching hospitals.²⁷ Although such information is a useful benchmark, it may require validation. The decreased use of prenatal steroids in the out-born group in this report supports similar findings by other investigators^{7, 10} and may reflect lower quality of acute prenatal care. The obstetrical care providers' practices and their access to support from the RC may influence the adequacy of prenatal care; however, this could not be evaluated in this study. Means of transportation and level of available ambulance services available to mothers who do not live close to regional centers are other factors that may affect the place of birth of VLBW infants independently from the adequacy of prenatal care. These health care and community services and maternal resources should be considered in future evaluations of access for pregnant women to perinatal services.

This study models an effort that could be undertaken by other regional centers to identify geographic areas and populations who underutilize the perinatal system. Interventions to target high-risk mothers need to be developed to better ensure that women receive adequate prenatal care.

References

1. American Academy of Pediatrics and the American College of Obstetrics and Gynecology. Guidelines for Perinatal Care, 5th edn. American Academy of Pediatrics: Elk Grove Village, IL; American College of Obstetrics and Gynecology: Washington, DC, 2002.
2. Horbar JD, Badger GJ, Carpenter JH, Fanaroff AA, Kilpatrick S, LaCorte M et al. Trends in mortality and morbidity for very low birth weight infants, 1991–1999. Pediatrics 2002; 110: 143–151. | Article | PubMed | ISI |
3. Bode MM, O'Shea TM, Metzguer KR, Stiles AD. Perinatal regionalization and neonatal mortality in North Carolina, 1968–1994. Am J Obstet Gynecol 2001; 184: 1302–1307. | Article | PubMed | ISI | ChemPort |
4. Cifuentes J, Bronstein J, Phibbs CS, Phibbs RH, Schmitt SK, Carlo WA. Mortality in low birth weight infants according to level of neonatal care at hospital of birth. Pediatrics 2002; 109: 745–751. | Article | PubMed | ISI |
5. Sanderson M, Sappenfield WM, Jespersen KM, Liu Q, Baker SL. Association between level of delivery hospital and neonatal outcomes among South Carolina Medicaid recipients. Am J Obstet Gynecol 2000; 183: 1504–1511. | Article | PubMed | ISI | ChemPort |
6. Chien LY, Whyte R, Aziz K, Thiessen P, Matthew D, Lee SK. Improved outcome of preterm infants when delivered in tertiary care centers. Obstet Gynecol 2001; 98: 247–252. | Article | PubMed | ISI | ChemPort |
7. Warner B, Musial MJ, Chenier T, Donovan E. The effect of birth hospital type on the outcome of very low birth weight infants. Pediatrics 2004; 113: 35–41. | Article | PubMed | ISI |
8. Menard MK, Liu Q, Holgren EA, Sappenfield WM. Neonatal mortality for very low birth weight deliveries in South Carolina by level of hospital perinatal service. Am J Obstet Gynecol 1998; 179: 374–381. | Article | PubMed | ISI | ChemPort |
9. Hogue CJ, Vasquez C. Toward a strategic approach for reducing disparities in infant mortality. Am J Public Health 2002; 92: 552–556. | PubMed | ISI |
10. Palmer KG, Kronsberg SS, Barton BA, Hobbs CA, Hall RW, Anand KJ. Effect of inborn versus outborn delivery on clinical outcomes in ventilated preterm neonates: secondary results from the NEOPAIN trial. J Perinatol 2005; 25: 270–275. | Article | PubMed |
11. Healthy People 2010. US Department of Health and Human Services: Washington, DC, 2000, http://www.healthypeople.gov/d
    ocument/.
12. D'Ascoli PT, Alexander GR, Petersen DJ, Kogan MD. Parental factors influencing patterns of prenatal care utilization. J Perinatol 1997; 17: 283–287.
13. Sinkin RA, Fisher SG, Dozier A, Dye TD. Effect of managed care on perinatal transports for the publicly funded in upstate New York. J Perinatol 2005; 25: 79–85. | Article | PubMed |
14. Adams EK, Gavin NI, Benedict MB. Access for pregnant women on Medicaid: variation by race and ethnicity. J Health Care Poor Underserved 2005; 16: 74–95. | PubMed | ISI |
15. Samuelson JL, Buehler JW, Norris D, Sadek R. Maternal characteristics associated with place of delivery and neonatal mortality rates among very-low-birthweight infants, Georgia. Paediatr Perinat Epidemiol 2002; 16: 305–313. | Article | PubMed | ISI |
16. Bronstein JM, Capilouto E, Carlo WA, Haywood JL, Goldenberg RL. Access to neonatal intensive care for low-birthweight infants: the role of maternal characteristics. Am J Public Health 1995; 85: 357–361. | PubMed | ISI | ChemPort |
17. Gould JB, Sarnoff R, Liu H, Bell DR, Chavez G. Very low birth weight births at non-NICU hospitals: the role of sociodemographic, perinatal, and geographic factors. J Perinatol 1999; 19: 197–205. | Article | PubMed | ChemPort |
18. Carmona R. Access to Quality Health Services. Progress Review. US Department of Health & Human Services, 2002, http://www.healthypeople.gov/d
    ata/2010prog/focus01/.
19. Attar MA, Lang SW, Gates MR, Iatrow AM, Bratton SL. Back transport of neonates: effect on hospital length of stay. J Perinatol 2005; 25: 731–736. | Article | PubMed |
20. Richardson DK, Corcoran JD, Escobar GJ, Lee SK. SNAP-II and SNAPPE-II: simplified newborn illness severity and mortality risk scores. J Pediatr 2001; 138: 92–100. | Article | PubMed | ChemPort |
21. Kessner DM, Kalk CE, Singer J. Assessing health quality — the case for tracers. N Engl J Med 1973; 288: 189–194. | PubMed | ISI | ChemPort |
22. Hosmer DW, Lemeshow S. Applied Logistic Regression. John Wiley & Sons Inc.: New York, 1989.
23. Howell EM, Vert P. Neonatal intensive care and birth weight-specific perinatal mortality in Michigan and Lorraine. Pediatrics 1993; 91: 464–469. | PubMed | ISI | ChemPort |
24. Dubay L, Joyce T, Kaestner R, Kenney GM. Changes in prenatal care timing and low birth weight by race and socioeconomic status: implications for the Medicaid expansions for pregnant women. Health Serv Res 2001; 36: 373–398. | PubMed | ISI | ChemPort |
25. Kogan MD, Martin JA, Alexander GR, Kotelchuck M, Ventura SJ, Frigoletto FD. The changing pattern of prenatal care utilization in the United States, 1981–1995, using different prenatal care indices. JAMA 1998; 279: 1623–1628. | Article | PubMed | ISI | ChemPort |
26. Misra DP, Guyer B. Benefits and limitations of prenatal care: from counting visits to measuring content. JAMA 1998; 279: 1661–1662. | Article | PubMed | ISI | ChemPort |
27. DiGiuseppe DL, Aron DC, Ranbom L, Harper DL, Rosenthal GE. Reliability of birth certificate data: a multi-hospital comparison to medical records information. Matern Child Health J 2002; 6: 169–179. | Article | PubMed |

Acknowledgments

This work was financially supported by a grant from Michigan Department of Community Health.