Background. Despite high per capita health care expenditure, the United States has crude infant survival rates that are lower than similarly developed nations. Although differences in vital recording and socioeconomic risk have been studied, a systematic, cross-national comparison of perinatal health care systems is lacking.
Objective. To characterize systems of reproductive care for the United States, Australia, Canada, and the United Kingdom, including a detailed analysis of neonatal intensive care and mortality.
Design/Methods. Comparison of selected indicators of reproductive care and mortality from 1993–2000 through a systematic review of journal and government publications and structured interviews of leaders in perinatal and neonatal care.
Results. Compared with the other 3 countries, the United States has more neonatal intensive care resources yet provides proportionately less support for preconception and prenatal care. Unlike the United States, the other countries provided free family planning services and prenatal and perinatal physician care, and the United Kingdom and Australia paid for all contraception. The United States has high neonatal intensive care capacity, with 6.1 neonatologists per 10 000 live births; Australia, 3.7; Canada, 3.3; and the United Kingdom, 2.7. For intensive care beds, the United States has 3.3 per 10 000 live births; Australia and Canada, 2.6; and the United Kingdom, 0.67. Greater neonatal intensive care resources were not consistently associated with lower birth weight-specific mortality. The relative risk (United States as reference) of neonatal mortality for infants <1000 g was 0.84 for Australia, 1.12 for Canada, and 0.99 for the United Kingdom; for 1000 to 2499 g infants, the relative risk was 0.97 for Australia, 1.26 for Canada, and 0.95 for the United Kingdom. As reported elsewhere, low birth weight rates were notably higher in the United States, partially explaining the high crude mortality rates.
Conclusions. The United States has significantly greater neonatal intensive care resources per capita, compared with 3 other developed countries, without having consistently better birth weight-specific mortality. Despite low birth weight rates that exceed other countries, the United States has proportionately more providers per low birth weight infant, but offers less extensive preconception and prenatal services. This study questions the effectiveness of the current distribution of US reproductive care resources and its emphasis on neonatal intensive care.
The goal of reproductive health care is to maximize the health outcomes of mothers and newborns through the effective delivery of preconception, prenatal, peripartum, and neonatal care. National perinatal statistics, such as infant mortality rates and percentages of infants born of low birth weight, reflect the cumulative influences of these health care processes, as well as socioeconomic determinants. Over the past 20 years, as the fields of perinatal and neonatal medicine have made profound advances in knowledge and technology, infant mortality has significantly declined in all developed nations.1
Despite these encouraging secular trends, infant mortality and low birth weight rates in the United States have remained consistently higher than those of similarly developed nations, such as Australia, Canada, and the United Kingdom.2 The relatively poor overall US birth outcomes exist in a setting of comparatively greater financing of health care,1 a highly specialized neonatal workforce, and a growing number of neonatal intensive care units.3 Given these unrivaled resources, why are the outcomes of the US birth cohort inferior?
Comparing the US system of reproductive care with the United Kingdom, Canada, and Australia may offer insights into how health care structure affects mortality. Each of these countries has a distinct workforce, hospital capacity, system of health finance, and organization of medical care that affects reproductive care and, presumably, birth outcomes. Previous reports of national differences in the distribution of resources within reproductive care have been limited in scope and have failed to examine mortality within the context of reproductive health care.4–7 In this study, we analyze these 4 countries’ reproductive health care through a systematic comparison of their infrastructures and newborn mortality. Although neonatal care is indisputably vital to some infants’ survival, perhaps there is a threshold where additional neonatal resources yield little measurable benefit.8–10
Data Sources and Search Method
We conducted a systematic review of available information on reproductive and perinatal health care systems in the 4 largest English-speaking countries: Australia, Canada, the United Kingdom, and the United States. We searched 3 main sources of data: 1) scientific journal articles, 2) government and private publications, and 3) structured interviews of leaders in perinatal care and health policy. The initial phase of our research used Medline and HealthStar databases from the years 1993–2000, November week 2, to locate the significant sources in each country. The following medical subheadings were used: (“perinatal care,” or “intensive care, neonatal,” or “intensive care units, neonatal,” or “reproductive care, ” or “prenatal care,” or “pregnancy,” or “ infant mortality” or “the United States” or “the United Kingdom” or “Australia” or “Canada”) and (“ birth weight,” or “maternal welfare,” or “maternal health services,” or “infant mortality,” or “outcome and process assessment (health care),” or “fetal distress” or “comparative studies” or “cohort studies” or “ hospital mortality” or “risk assessment” or “health expenditures” or “life expectancy” or “models, econometric,” or “infant newborn”). From 892 articles, we excluded 5 non-English texts, and then examined the abstracts to identify pertinent articles, leading to 144 perinatal-specific publications. Excluded articles addressed primary experimental interventions or disease-specific outcomes (214), ethical discussions of treatment and patient management (121), nongeneralizable outcomes of discreet populations (189), long-term or neurodevelopmental outcomes (89), infectious disease trends (33), perinatal trends before 1995 (86), or unrelated topics (24). Not surprisingly, there are no known randomized trials of perinatal health care systems; all included studies were reviews, observational studies, or policy discussions. Editorials and letters to the editors were also included. Each article’s bibliography was reviewed to identify additional relevant publications, particularly those predating 1993.11 Lists of authors with multiple publications or seminal articles were tabulated for future contact.
Next, we identified organizations active in perinatal health within each country. We searched the World Wide Web using the key words listed above to locate government offices, professional academic units, and individuals dedicated to perinatal research and policy. Each country has a central department committed to perinatal or maternal and child health with readily available vital statistics and government reports, although some data were available only on a provincial or statewide basis. Academic units and private organizations also provided important information and databases.
In the third and final stage, we performed on-site, telephone, and e-mail interviews of perinatal health experts to gain access to unpublished data, reports, and raw data of published studies. We identified over 75 Web sites dedicated to perinatology and e-mailed over 200 individuals. In total, we contacted 8 centers each in Australia, the United Kingdom, and the United States, and 10 in Canada, and we conducted interviews with leaders in perinatology in each country. The 3 phases of our approach were repeated until considerable overlaps existed between these search techniques, indicating that we searched to the point of redundancy. Written permission was obtained for all personal communications and unpublished data.
Assessment of Endpoints
Reproductive Care Characteristics
Using the following criteria, we developed a list of key indicators with which to compare each country’s reproductive health care system. From evidence in the literature or by professional consensus, each indicator is thought to affect the availability and processes of care as well as perinatal outcomes. Furthermore, the data supporting the indicator must be available for most of the 4 countries, in identical or consecutive years. Some indicators had different but comparable statistics, such as neonatal intensive care bed capacity, as Canada and Australia provided data inclusive of both intensive and intermediate care beds, whereas the United Kingdom only had data on intensive care beds. Some desirable indicators were subsequently excluded, such as the percent of the health care budget spent on perinatal care, because comparable data were unavailable. Finally, indicators of practice patterns across the 4 countries had to measurably differ for inclusion. Innovations such as use of surfactant, for example, were excluded because they were adopted almost simultaneously across these countries so that there was no obvious time when the practice patterns were different. These responses were tabulated, summarized, and sent back to the experts within each country for accuracy.
We chose infant and neonatal mortality rates and percentages of low birth weight as our outcome measures. The infant mortality rate, defined as the number of deaths in the first year of life per 1000 live births, is extensively used in all countries to characterize reproductive outcomes and the status of maternal and child health.11 Crude rates are readily available from all the study countries, with close to 100% of the birth cohorts registered. However, the infant mortality rate lacks specificity because it combines outcomes of initial neonatal care with those of postneonatal events, like injuries, sudden infant death syndrome, and infections. The neonatal mortality rate, defined as the number of deaths within 28 days per 1000 live births, more closely reflects the pathophysiologic processes of the perinatal period and the outcomes of neonatal care. Collectively, these statistics characterize the mortality of a national birth cohort.
Birth weight is another readily accessible statistic because it is reliably recorded on birth certificates. In addition to being an excellent prognostic indicator for the infant, birth weight also reflects prenatal health. Maternal risk factors important to infant outcomes (like race, age, education, and income) affect newborn mortality primarily through their effects on birth weight. Low birth weight is defined as any birth <2500 g and is used as a proxy for gestational age. In this study, we calculated birth weight-specific mortality rates for infant and neonatal mortality. Death rates within specific birth weight categories were divided by its cohort of live births.
Unlike crude rates, birth weight specific mortality data were not nationally available for all countries in concurrent years. Specifically, Canadian infant and neonatal morality statistics, dated from 1996, exclude Ontario, where approximately 62% of all Canadian deliveries occur. As a point of comparison, the 1996 infant mortality rate for Ontario was 5.7 per 1000 live births, which is similar to Canada as a whole, at 5.6. Australian data were nationally available for neonatal mortality, but birth weight-specific infant mortality data are only tabulated for the State of Victoria. Victoria, whose largest city is Melbourne, has 24% of all Australian deliveries and has slightly better infant mortality statistics than the rest of Australia. Specifically, 1998 data show that Australia as a whole had a crude infant mortality of 5.0 per 1000 live births, whereas Victoria had 4.7. Data for the United Kingdom was limited to England and Wales.
Linkage of birth records to death records is not standard in all countries. Unlinked data for the US cohort was 1.6% of all births, 3% for Canada, 0.02% for Australia, and 0.13% for England and Wales.
Definitions of newborn mortality and low birth weight are uniform among Australia, Canada, the United Kingdom, and the United States. As defined by the World Health Organization, a live birth is defined in each of these countries as the “complete expulsion or extraction from mother of a production of conception, irrespective of duration of pregnancy, which after such separation, breathes or shows any other evidence of life.”12
This definition purposefully avoids any gestational age or birth weight exclusionary criteria. The United Kingdom, Canada, and United States data generally adhere to this definition. However, all data obtained from Australia exclude newborns weighing <400 g.
Statistical analysis was performed using χ2 technique, using Stata 6.0 (College Station, TX). We accept a level of significance of P < .05.
Table 1 illustrates the broad characteristics of each nation’s system of health care. The United States, the only country for which national insurance is not available, spends 104% more of its gross domestic product on the financing of its health care system than the United Kingdom, 65% more than Australia, and 51% more than Canada.1 The United States has a significant percent (15.5%) of citizens without insurance, with women (22.8%) and children (13.9%) disproportionately underinsured.13 In addition to national insurance, supplemental private insurance is common in Australia (41.2%),14 less frequent in the United Kingdom (10.9%),14,15 and unknown in Canada. The physician workforce is greatest in the United States, with 65% more physicians per capita than the United Kingdom, 33% more than Canada, and 12% more than Australia.1,16
National Policies on Reproductive Care
Table 2 illustrates selected policies that reflect each country’s approach to reproductive services. Australia and the United Kingdom offer comprehensive access to contraception services (D. Wheen, Family Planning, Australia, written communication, July 28, 2000; K. Horsley, Officer of the Department of Health, Sexual Health Service Policy, United Kingdom, August 18, 2000), whereas Canada covers physician services, (although private prescription plans are available to cover contraceptive devices).17 In the United States, most private plans cover surgical contraception, but only 50% cover reversible forms of contraception, like birth control pills.18 This may change, as a recent case, Erickson v. Bartell Drug Co, 141F.Supp. 2 days 1266 (W.D. Wash. 2001), ruled that an employer must cover prescription contraception in their health plans. In contrast, Medicaid provides comprehensive coverage, although it only insures 10.2% of the population.13 Australia and Canada, unlike the United States and the United Kingdom, lack national abortion policies, and instead leave this decision to individual states, provinces, or territories, where abortion is generally accessible.19 Government payment for abortion services, while difficult to document, varied: the National Health Service paid for 74% of all abortions in the United Kingdom (J. Roe, Abortion Law Reform Association, written communication, August 1, 2000), whereas the US government covered 14%,20 and Canadian payment varied from full to none, depending on the province.17 US policy provides the least support for maternity, paternity, and family leaves, as it supports 12 weeks of unpaid maternity leave through the Family and Medical Leave Act of 1993 (S. B. Kamerman, Centennial Professor, School of Social Work, Columbia University, written communication, July 31, 2000).
Reproductive Health Care Indicators
Available indicators of reproductive care (Table 3) reveal that the US population of expectant mothers is younger and probably less mature, experienced, and financially stable. The US has higher fertility rates12 and overall percentage of pregnancies aborted.21 Adolescent-specific rates, such as pregnancy, abortion rates, and the percentage of delivering mothers <20 years old are notably higher in the United States.22–28
Neonatal Intensive Care: Physician and Hospital Capacity
The United States has the largest neonatal workforce and hospital capacity (Table 4). The US number of neonatologists is 6.1 per 10 000 live births, which is 1.7 and 1.8 times higher than Australia and Canada, respectively, and 2.3 times greater than the United Kingdom.29–32 It might be expected that the United States needs more neonatologists, given its higher proportion of low birth weight infants, but adjusting for the higher low birth weight infants and very low birth weight infants still shows a 1.4 to 2.1 times greater US capacity. Similarly, the US pediatrician, obstetrician, and maternal-fetal specialty workforces are proportionately larger than the other 3 countries.29,33–35 Family practitioners are conversely more prevalent in Canada and Australia.29,33,34,36
The United States also has the highest hospital capacity for its newborn population. The number of neonatal intensive care beds is 4.9 times greater in the United States than the United Kingdom, and the combined number of intensive and intermediate beds is 2.0 times greater than both Australia and Canada.30,37,38 Like the US physician capacity, the bed capacity in the United States is 1.5 to 4.7 times greater than the other countries, even when stratified by low birth weight infants.
No country has consistently better mortality outcomes (Tables 5 and 6). As reported elsewhere,2 crude infant and neonatal mortality rates rank the United States behind the other 3 countries (US infant mortality rate [IMR] is 7.1 deaths per 1000 live births, compared with 4.3–5.5; US neonatal mortality rate [NMR] is 4.7 compared with 3.0–3.8).24,26,39–41 Stratifying these rates into birth weight specific categories significantly narrows these observed differences.
As shown in Table 5, all 4 countries have statistically similar birth weight-specific IMR for newborns <500 g.26,39–41 For infants above 500 g, statistically significant differences were observed, but no country had consistently better outcomes. The exception was for infants weighing above 2500 g, where the United States has statistically higher mortality (2.6 per 1000 live births) compared with all 3 countries (1.3–2.3).
NMR measures outcomes during the period where neonatal intensive care is most influential. Like the IMR data, (Table 6) no country can claim better survival.24,26,39,40 With the United States as the reference, (relative risk [RR]: 1), Australia has significantly better outcomes for infants born <1000 g, (RR: 0.84; 95% confidence interval [CI]: 0.75–0.94). In contrast, the Canadian infants seem to have a relative disadvantage for infants born between 1000 and 2499 g, (RR: 1.26; 95% CI: 1.10–1.45) and those infants born in England and Wales above 2500 g likewise have higher mortality (RR: 1.15; 95% CI: 1.07–1.24). Although the non-US countries administer neonatal care with fewer available physicians and neonatal hospital beds, their infants do not suffer greater mortality.
This report examines both reproductive care and mortality across 4 developed countries and finds that the United States commits greater resources to neonatal intensive care, although its infants do not have proportionately better survival. The other 3 countries’ health care systems, in contrast, place more emphasis on preconception and prenatal care. These findings suggest that additional expansion of neonatal intensive capacity would not result in proportionate benefit in mortality. Our data support this view for the following reasons: First, birth weight, a reflection of socioeconomic context and the quality of medical care before birth, is an important predictor of mortality that is not influenced by postnatal medical interventions including neonatal intensive care.6,42,43 Second, even for low birth weight infants, it does not seem that higher US neonatal care capacity is associated with improved survival.44 Finally, there are unmet needs in other areas of preconception and prenatal care that may affect birth weight and, in turn, neonatal mortality.43,45
Previous studies comparing international differences in newborn care have primarily examined regionalization of neonatal intensive care.46,47 All 4 countries in this study endorse this practice, with varying scales of implementation and regulation. However, the impact of regionalization is not easily measured or compared across different systems.7,48 Furthermore, each country’s regionalized structure continues to evolve, often without evidence of improved outcomes.49–53 In the United States, the policy of regionalization is often overshadowed by hospital competition for neonatal intensive care unit and obstetric services. This study, instead, examines where resources are concentrated in the health care system, with attention to antenatal and postnatal care. Although we are unable to identify which components of a country’s system has the greatest influence, our data show that the relative emphasis toward neonatal intensive care resources in the United States is not associated with better birth weight-specific survival.
It could be argued that neonatal health care capacity in the United States is not excessive, but geographically maldistributed with regions of higher newborn risk lacking sufficient intensive care resources to optimize outcomes. Although a recent study reveals greater than fourfold regional variation in neonatal intensive capacity, greater resources are not located in areas with higher neonatal illness rates.37 Related analyses showed that the areas with “very low” compared with “low” neonatal intensive care capacity have worse risk-adjusted neonatal mortality rate but that mortality rates did not improve further even with “very high” capacity.44 These results suggest that a few regions in the United States may have inadequate access to neonatal intensive care, but that most regions already have adequate resources. The sufficiency of supply is further supported by the finding that up to 60% of US neonatologists care for normal newborns, indicating that neonatologists are currently not overburdened with high-risk infants.54 These reports, coupled with the findings of this study, shed doubt on the need for continued expansion in neonatal intensive care capacity.
Several limitations of this study should be noted. First, there are factors that contribute to infant survival that are not reliably measured in some or all of the countries we studied. For example, race is an important determinant of infant outcome, with very low birth weight black infants experiencing lower mortality in comparison to white infants, and conversely higher mortality at higher birth weights.55,56 Adjusting for these racial differences in our international comparisons, however, does not alter the conclusion that none of the countries has consistently better birth weight-specific outcomes. Other factors affect neonatal survival by primarily influencing the incidence of low birth weight. These include individual characteristics such as ethnicity, gestational age, intrauterine growth retardation, maternal smoking, and education.56–58 Only in the United States are these data routinely linked to infant mortality and, even in the United States, the measurement of these risk factors is not always accurate. In addition, other differences that affect infant outcomes but are not reliably measured include poverty, local interpretations of the “level” of intensive care provided, intensive care unit volume and practice style, and place of delivery.52,59–66 Although a multivariate analysis using all of these known risk factors would enhance our understanding of infant mortality to some extent, the fact remains that birth weight is the major predictor of infant outcome.
A second possible limitation is that the mortality rates might vary between countries because of different definitions of “live birth,” “ stillbirth,” and “abortion.”67–72 Although registration standards differ across regions and nations, these 4 countries generally follow the definitions of the World Health Organization.61,71 Australia, in contrast to the other countries, does exclude newborns <400 g, perhaps accounting for their better NMR observed at the smallest weight category. In the 1995 US cohort, 51% of registered infants <500 g actually weighed <400 g.37 Sensitivity analysis reveals that adjusting for this lack of registration in Australia would result in mortality rates for infants <1000 g that are comparable with the United States. However, even with this adjustment, this potential bias does not challenge our primary finding that the US birth weight-specific mortality is not better than the other countries.
We should also note that roles of physicians differ in these countries, and this may confound our comparisons of workforce capacity. For example, the involvement of pediatricians in neonatal intensive care units varies. Furthermore, available data do not fully account for differences in clinical effort. The US data include only full-time clinicians, the most conservative number available; the measures for the other countries include all clinicians, teachers and researchers. Mid-level providers, (eg, neonatal nurse practitioners) are present in small numbers in the other countries, but provide substantial care to neonates in the United States, possibly leading to an underestimation of the US neonatal intensive care workforce.
One unexpected finding of our study is that the US infant mortality rate for infants greater than 2500 g is significantly higher than those of the other 3 countries. If the United States were able to reduce mortality in this group to the rate in Canada, (2.6–2.3 per 1000 live births) this would prevent almost 3000 deaths. Neonatal mortality rates (<28 days) were similar, indicating that the higher mortality risk for the normal weight infants occurs after the first month of life. Only 23% of postneonatal mortality is attributable to perinatal causes,73 and perinatal events are less important in these normal birth weight newborns. More frequent causes of mortality (eg, sudden infant death, injuries, and infections) are not directly influenced by neonatal intensive care, but may instead reflect social disadvantage or access to medical care.74 These findings merit additional investigation.
This report finds that Australia, Canada, and the United Kingdom provide more consistent access to preconception and prenatal care than the United States, and although the United States has more neonatal intensive care capacity, birth weight-specific mortalities are no better. These findings question the marginal benefit of allocating additional resources to neonatal intensive care. Because decreasing neonatal intensive care resources may have unintended consequences, a possible alternative for the United States is to maintain level resources in neonatal intensive care while improving funding for preconception and prenatal care.10,47,75 Unintended pregnancies in particular remain an important determinant of infant outcomes that merit policy interventions.76 Furthermore, many authors endorse the promotion of prenatal care as a means to reduce low birth weight and improve infant mortality,4,42,45,77–80 although research is needed to identify which prenatal interventions are most effective.81
The recognition that improved infant outcomes requires consistent care from preconception through early childhood should stimulate policies directed toward a more equitable distribution of resources across the domains of reproductive care. The 1976 document, Toward Improving the Outcome of Pregnancy (or TIOP I), provided a conceptual framework for the present hospital-based regionalized system of care and may have facilitated much of the earlier improvements in perinatal outcomes.46 The report of a second committee on perinatal health (or TIOP II) in 1993, made additional policy proposals that focused beyond hospital based care into preconception and perinatal care, public education, data management, accountability and financing.47 These recommendations have not been effectively implemented in the United States. As the current study shows, the outcomes of the total US birth cohort lag behind similarly developed countries, despite the best-funded system of neonatal intensive care in the world.
This paper was funded by the Agency for Healthcare Research and Quality, through a National Research Service Award (T32 HS00070), and the Robert Wood Johnson Foundation.
We thank the many contacts that we made in Canada, the United Kingdom, Australia, and the United States for their patience and participation in this manuscript. In particular, we thank Russell Wilkins at Statistics Canada, Nicola Cooper at the Office for National Statistics in the United Kingdom, and James King, chair of the Consultative Council of Obstetric and Pediatric Mortality and Morbidity in Victoria, for supplying us with special analyses of perinatal mortality. Furthermore, Alexander C. Allen, at the Department of Neonatal Paediatrics in Halifax, and Doug McMillan facilitated our Canadian research. In Australia, the help provided by Paul Gavel at the Department of Health in New South Wales and by Deborah Donoghue and the Australian and New Zealand Neonatal Network was invaluable. We are also thankful for the assistance of Martin Staniforth, the Deputy Director of Human Resources at the National Health Service, Janet Tucker of the UK Neonatal Staffing Study at the University of Aberdeen, and Andrew Wilkinson of the Neonatal Unit at the John Radcliffe Hospital in Oxford. Finally, we thank Alexander Thorngren for his thoughtful review of this manuscript.
In addition, we would like to call attention to the publication of Dr Tucker and the UK Neonatal Staffing Group (Tucker J. The UK Neonatal Staffing Study Group. Patient volume, staffing and workload in relation to risk-adjusted outcomes in a random stratified sample of UK neonatal intensive care units: a prospective evaluation. Lancet. 2002;359:99–107). Our manuscript was already in press at the time of its release.
- ↵Organisation for Economic Co-operation and Development, Health Policy Unit. A Comparative Analysis of 29 Countries, OECD Heath Data 99. CD-ROM ed. Paris, France: OECD; 1999
- ↵March of Dimes Birth Defects Foundation. International Comparisons of Infant Mortality Rates, 1996. Health Library. Available at: http://www.modimes.org Accessed April 10, 2002
- ↵Miller CA. Maternal and infant care: comparisons between Western Europe and the United States. Int J Health Ser.1993;23 :655– 664
- Swyer PR. Organisation of perinatal/neonatal care. Acta Paediatr.1993;385(suppl) :1– 18
- ↵Goodman DC, Little GA. General pediatrics, neonatology, and the law of diminishing returns. Pediatrics.1998;102 :396– 399
- ↵Rosenblatt RA. The perinatal paradox: doing more and accomplishing less. Health Aff.1989;8 :158– 168
- ↵Prentice T. World Health Report: Annex . Basic Indicators for all Member States. Geneva, Switzerland: World Health Organization; 1999
- ↵Mills RJ. Health Insurance Coverage: 2000. Current Population Reports. Washington, DC: US Census Bureau; 2001
- ↵Private Health Insurance Administration Council. Australian Supplemental Health Insurance. Deakin ACT, Australia; 2000. Available at: http://www.phiac.org Accessed April 10, 2002
- ↵Department of Health, National Health Service. United Kingdom; 2000. UK Supplemental Health Insurance. Available at: http://www.doh.gov.uk Accessed April 10, 2002
- ↵Randolph SB, Pasko T. Physician Characteristics and Distribution in the US. Chicago, IL: American Medical Association; 1997
- ↵Statistics Canada. Therapeutic Abortions. Ottawa, Canada: Canadian Institute for Health Information; 1995
- ↵The Alan Guttmacher Institute. Uneven and Unequal: Insurance Coverage and Reproductive Health Services. New York, NY: The Alan Guttmacher Institute; 1994
- ↵Rahman A, Katzive L, Henshaw SK. A global review of laws on induced abortion, 1985–1997. Int Fam Plann Perspect.1999;24 :56– 64
- ↵The Alan Guttmacher Institute. Facts in Brief: Induced Abortion. New York, NY: The Alan Guttmacher Institute; 2000
- ↵Henshaw S, Singh S, Haas T. The incidence of abortion worldwide. Int Fam Plann Perspect.1999;25(suppl) :S30– S38
- ↵Ventura SJ, Martin JA, Curtain SC, Mathews TJ, Park MM. Births: Final Data for 1998. National Vital Statistics Reports. Vol. 48. No. 3. Hyattsville, MD: National Center for Health Statistics; 2000
- ↵Day P, Sullivan EA, Ford J, Lancaster P. Australia§ Mothers and Infants-1997. Perinatal Statistics Series, Number 9. Sydney, Australia: Australian Institute of Health and Welfare, National Perinatal Statistics Unit; 1999
- Statistics Canada. Births and Deaths, 1997. Ottawa, Canada: Health Statistics Division; 1998
- ↵Office for National Statistics. Mortality Statistics: Childhood, Infant and Perinatal. London, England: The Stationery Office; 1998
- ↵Macfarlane A, Mugford M. Birth Counts: Statistics of Pregnancy and Childbirth. 2nd ed. London, England: The Stationery Office; 2000
- ↵Statistics Canada. Canadian Institute Heath Information. Manitoba, Canada: Canadian Perinatal Society; 1995
- ↵Pasko T, Seidman B. Physician Characteristics and Distribution in the US. Chicago, IL: American Medical Association; 1999
- ↵Donoghue D. Australia and New Zealand Neonatal Network 1996–7. Sydney, Australia: Australian Institute of Health and Welfare, National Perinatal Statistics Unit; 1999
- ↵Hall J. University Department Subspecialty Pediatrician FTE Data 1999–2000. Vancouver, Canada: UBC and BC§ Children§ Hospital; 2000
- ↵Royal College of Paediatric and Child Health. Census of Consultant and Non Consultant Career Grade Paediatricians in the UK. London, England: Royal College of Paediatric and Child Health; 1999
- ↵Canadian Institute for Health Information. Supply, Distribution and Migration of Canadian Physicians, 1998. Ottawa, Canada: Canadian Institute for Health Information; 1999
- ↵Australian Institute of Health and Welfare. National Medical Labor Force Survey. Sydney, Australia: Australian Institute of Health and Welfare; 1998
- ↵American Collge of Obstetrics and Gynecology Diplomate. Subspecialty certification: maternal-fetal specialists. www.acog.org (members only side): 1999. Numbers 21–26
- ↵Royal College of General Practitioners; http://www.rcgp.org.uk/; 1999
- ↵Goodman DC, Fisher ES, Little GA, Stukel TA, Chang C. Are neonatal intensive care resources located where need is greatest? Regional variation in neonatologists, beds, and low birth weight newborns. Pediatrics.2001;108 :432– 437
- ↵Tucker J. UK neonatal intensive care services in 1996. Arch Dis Child Fetal Neonatal Ed.1999;80 :F233– F234
- ↵Mathews T, Curtin SC, MacDorman MF. Infant Mortality Statistics from the 1998 Period Linked Birth/Infant Death Data Set. Vol. 48. Atlanta, GA: Centers for Disease Control, National Center for Health Statistics; 2000:12
- ↵Statistics Canada. Infant and Neonatal Mortality by Birth Weight, Canada 1996. Special tabulations from the Canadian Linked Birth and Infant Death File. Ottawa, Canada: Health Analysis and Modeling Group; 2001
- ↵The Consultative Council on Obstetric and Pediatric Mortality and Morbidity. Annual Report for the Year 1998: Incorporating the 37th Survey of Perinatal Deaths in Victoria. Melbourne, Australia: The Consultative Council on Obstetric and Paediatric Mortality and Morbidity; 2000
- ↵Goodman DC, Fisher ES, Little GA, Stukel TA, Chiang-hua C, Schoendorf K. The relation between the availability of neonatal care and neonatal mortality. N Engl J Med.2001. In press
- ↵Committee on Perinatal Health (TIOP I). Toward Improving the Outcome of Pregnancy: Recommendations for the Regional Development of Maternal and Perinatal Health Services. White Plains, NY: March of Dimes Birth Defects Foundation; 1976
- ↵Committee on Perinatal Health (TIOP II). Toward Improving the Outcome of Pregnancy: The 90s and Beyond. White Plains, NY: March of Dimes Birth Defects Foundation; 1993
- ↵International Neonatal Network, Scottish Neonatal Consultants, Nurses Collaborative Study Group. Risk adjusted and population based studies of the outcome for high-risk infants in Scotland and Australia. Arch Dis Child Fetal Neonatal Ed.2000;82 :F118– F123
- Hein HA, Lofgren MA. The changing pattern of neonatal mortality in a regionalized system of perinatal care: a current update. Pediatrics.1999;104 :1064– 1069
- ↵Little GA, Merenstein GB. Toward improving the outcome of pregnancy, 1993: perinatal regionalization revisited. Pediatrics.1993;92 :611– 612
- ↵Pollack LD, Ratner IM, Lund GC. United States neonatology practice survey: personnel, practice, hospital, and neonatal intensive care unit characteristics. Pediatrics.1998;101 :398– 405
- ↵Carmichael SL, Iyasu S. Changes in the black-white infant mortality gap from 1983 to 1991 in the United States. J Prev Med.1998;15 :220– 227
- ↵Demissie K, Rhoads GG, Ananth CV, et al. Trends in preterm birth and neonatal mortality among blacks and whites in the United States from 1989 to 1997. J Epidemiol.2001;154 :307– 315
- ↵Lightwood JM, Phibbs CS, Glantz SA. Short-term health and economic benefits of smoking cessation: low birth weight. Pediatrics.1999;104 :1312– 1320
- ↵Bird ST, Bauman KE. The relationship between structural and health services variables and state-level infant mortality in the United States. J Public Health.1995;85 :26– 29
- Bird ST, Bauman KE. State-level infant, neonatal, and postneonatal mortality: the contribution of selected structural socioeconomic variables. Int J Health Serv.1998;28 :13– 27
- ↵Lee SK, McMillan DD, Ohlsson A, et al, and the Canadian NICU Network. Variations in Practice and Outcome in the Canadian NICU Network: 1996–1997. Pediatrics.2000;106 :1070– 1079
- The ECSURF (Economic evaluation of Surfactant) Collaborative Study Group. Limited comparability of classifications of levels of neonatal care in UK units. Arch Dis Child Fetal Neonatal Ed.1998;78 :F179– F184
- Sepkowitz S. Why infant very low birth weight rates have failed to decline in the United States vital statistics. Int J Epidemiol.1994;23 :321– 326
- ↵Wegman ME. Infant mortality: some international comparisons. Pediatrics.1996;98 :1020– 1027
- ↵Phelan ST, Goldenberg R, Alexander G, Cliver SP. Perinatal mortality and its relationship to the reporting of low-birth weight infants. m J Public Health.1998;88 :1236– 1239
- ↵National Center for Health Statistics. Vital Statistics of the United States, Vol. II, Mortality, Part A. Washington, DC: Public Health Service; 1996
- ↵DiLiberti JH. The relationship between social stratification and all-cause mortality among children in the United States: 1968–1992. Pediatrics. 2000;105(1). Available at: http://www.pediatrics.org/content/full/105/1/e2
- ↵Wise P. Poverty, technology and recent trends in the United States infant mortality rate. Pediatr Perinatal Epidemiol.1990;4 :391– 401
- ↵Institute of Medicine, Committee on Unintended Pregnancy. The Best Intentions: Unintended Pregnancy and the Well-Being of Children and Families. Washington, DC: National Academy Press; 1995
- ↵Institute of Medicine, National Academy of Sciences, Committee to Study the Prevention of Low Birth Weight. Preventing Low Birth Weight. Washington, DC: National Academy Press; 1985
- Copyright © 2002 by the American Academy of Pediatrics