Economic consequences of preterm birth and low birthweight
Correspondence: Dr S. Petrou, National Perinatal Epidemiology Unit, Institute of Health Sciences, University of Oxford, Old Road, Headington, Oxford OX3 7LF, UK.
The high rates of mortality and morbidity arising from preterm birth and low birthweight impose an immense burden on the health, education and social services and on families. This was evaluated in several economic studies published in the 1970s and early 1980s, but an ability to intervene effectively to increase survival has transformed neonatal care over the past two decades. In addition, the relative prices of resource inputs are vastly different. This paper surveys the scientific literature on the economic consequences of preterm birth and low birthweight, including: (1) studies, which estimate the economic costs of the conditions and (2) economic evaluations of primary and secondary prevention and treatment strategies. This paper summarises the scope and scientific quality of the published evidence, identifies gaps in our knowledge and considers the future research agenda in this area.
Although the numbers of live births registered in most industrialised nations have remained relatively stable in recent years, the incidence of preterm birth and low birthweight has increased. In England and Wales, for example, the percentage of liveborn babies weighing 2500 g and less increased from 6.79% in 1990 to 7.28% in 19961. The reported incidence of very small babies has increased at a faster rate. For example, the percentage of liveborn babies weighing less than 1500 g in England and Wales increased from 0.96% in 1990 to 1.22% in 1996 and that of liveborn babies weighing less than 1000 g from 0.34% to 0.49%1. Similar trends have been noted in other industrialised nations2. Furthermore, advances in perinatal practices, such as increased use of assisted ventilation in the delivery room and surfactant therapy, have improved the chances for survival of low gestational age and birthweight infants. A recent review of the world literature, revealed that at 23 weeks gestation, survival ranged from 2% to 35%. At 24 weeks gestation the range was 17% to 58%, and at 25 weeks gestation 35% to 85%3. The large variations in survival rates are largely explained by differences in population descriptors, in the initiation and withdrawal of treatment, and the duration of survival considered. Nevertheless, a consistent upward trend in survival rates has been noted, particularly during the 1990s3.
The increasing incidence and improved survival chances of infants of low gestational age and birthweight, combined with the diffusion of new technologies, have increased the demand for and cost of care provided to these infants during the neonatal period and in later life. This is reflected, in part, in national trends in resources for neonatal care1 and in macroeconomic estimates of conditions originating in the perinatal period2. This paper surveys the scientific literature on the economic consequences of preterm birth and low birthweight for the health services, for other sectors of the economy, and for families and carers. In so doing, it highlights gaps in our current knowledge of the topic and identifies requirements for further research in this area.
Consideration of the economic consequences of preterm birth and low birthweight requires elucidation of a number of methodological issues. The majority of studies that have estimated the economic consequences of preterm birth or low birthweight have tended to focus on their economic costs. In theory, this should involve considering the benefits or social sacrifices foregone by caring for these infants, referred to by economists as opportunity costs. However, as an operational standard, economists have tended to focus on the market prices of resource inputs entailed by the care process. The economic costs of the care process can broadly be divided into direct costs and indirect costs. Direct costs represent the resources purchased directly as a result of the healthcare activity. Indirect costs, on the other hand, represent the element of indirect consumption of resources in the production process, for example the value of lost earnings by patients or carers of patients who are unable to work as a result of the healthcare activity.
The process of measuring the economic costs of preterm birth and low birthweight can encompass a variety of methods of varying degrees of complexity. The most common alternative approaches are based on cost accounting methods, either using detailed information about resources used by individual patients (the ‘bottom up’ approach), or by allocation of total costs by unit workload (the ‘top down’ approach). Many studies, particularly those conducted in countries, such as the USA, where there is a comprehensive system of billing and fee-for-service payment of providers, cite charges for care rather than costs. However, it is worth noting that healthcare charges may include elements arising from corporate financial decisions and may therefore be poor proxies for the costs of caring for preterm or low birthweight infants4.
The economic costs of preterm birth and low birthweight may be felt beyond the health services. Parents of sick infants, for example, may require support from social service departments upon their discharge from hospital. They may have to forego other productive activities (paid or unpaid work) in order to spend time with them; their transport costs to and from the hospital may be considerable, and care for other children may have to be arranged. In addition, the birth of a preterm or low birthweight infant may have an economic impact over many years. Preterm and low birthweight infants may, for example, have additional healthcare needs, which lead to hospital readmissions or increased contacts with general practitioners and other healthcare providers, different educational needs, and additional care requirements for everyday living. Consequently, it is incumbent upon analysts to recognise these broader and long-term impacts in their economic calculations.
Estimates of the economic costs of preterm birth and low birthweight can inform the planning of services and can provide a basis for assessing competing strategies for research and prevention. However, cost data alone cannot identify the most efficient allocation of finite perinatal resources. Rather, it is information on incremental costs and incremental health gains associated with particular healthcare activities that can identify the combination of human and material inputs that maximise health benefits. Economic evaluation provides a framework for assessing the costs and benefits of alternative approaches to caring for preterm and low birthweight infants. The different forms of economic evaluation are broadly similar in their approach to measuring and valuing economic costs. However, they differ in the manner in which benefits are measured and valued. Cost benefit analysis measures the costs and benefits of healthcare activities in the same monetary units and attempts to ascertain whether the activities can be justified per se, that is, whether the benefits of the activities exceed their costs. Cost effectiveness analysis also compares the costs of healthcare activities to their benefits, but the benefits are measured in natural or physical units, such as life years gained or hospital days saved. Finally, cost utility analysis is a refinement of cost effectiveness analysis in that it constructs a single index of outcomes that reflects preferences for each possible outcome. Preferences for different outcomes depend on prevailing ethical and moral standards about life, death and disability. Potential instruments for measuring health outcomes within a cost utility framework include the quality-adjusted life year (QALY)5, the healthy years equivalent (HYE)6 and the saved young life equivalent (SAVE)7.
Costs During the Neonatal Period
A number of studies in the scientific literature have estimated hospital service costs for preterm or low birthweight infants during the neonatal period. The results of these studies allow us to draw the following conclusions. First and foremost, an inverse relationship is apparent between gestational age or birthweight and hospital service costs during the neonatal period. This is illustrated in Table 1, which summarises the results of the key studies published in the scientific literature since the 1970s. Hospital service costs for babies born weighing less than 1000 g were 75% higher, on average, than those incurred by babies born weighing 1000–1499 g, and more than four times higher, on average, than those incurred by babies born weighing at least 1500 g. A similar relationship has also been noted between hospital service costs during the neonatal period and gestational age8,9. Many of these studies suffer from a number of methodological limitations, including a failure to provide detailed and disaggregated information on reported costs and a failure to explore the implications of uncertainty surrounding the values of key parameters. In addition, many of the older studies were conducted before the diffusion of effective perinatal practices, such as the use of antenatal corticosteroids, new modes of ventilation, exogenous surfactant and better parenteral nutrition. Nevertheless, an inverse relationship between gestational age or birthweight and hospital service costs during the neonatal period was observed, regardless of the methodological quality of the study and the date of publication.
Table 1. Neonatal care–hospital service cost per case (£ UK 1998 prices).
A second conclusion that can be drawn from this body of literature is that hospital service costs incurred during the neonatal period are likely to be related to the mortality rate of infants10. Some studies have estimated costs separately for infants who died, and the estimates for low gestational age or birthweight infants show that the cost is usually higher for survivors than for those who die10. The implication is that hospital service costs during the neonatal period will increase as a consequence of the improved survival chances of the smallest infants. A third conclusion that can be drawn from this body of literature is that hospital service costs during the neonatal period are related to the degree of surgical intervention performed on the infant and the level of assisted ventilation received by the infant11.
A preterm or low birthweight infant's stay in a neonatal care unit is also likely to have economic implications for sectors of the economy other than the health service. The travel expenses incurred by parents visiting their children in neonatal care units may be considerable if travel to a regional centre is entailed. Indeed, these travel expenses have been estimated at between £12012 and £78013 (1998 £ sterling). Other expenses incurred by families, such as the additional costs related to babysitting, child care, arrangements for siblings and accommodation expenses incurred by family members during the infant's neonatal stay, have also been reported in the scientific literature. Gennaro14 interviewed 224 families of low birthweight infants following the infant's discharge from the neonatal unit and estimated that out-of-pocket expenses incurred by families average US$95 dollars per week (price date not reported). Although further research is required to confirm these results, it is important that healthcare providers are sensitive to the financial constraints faced by parents on low incomes at a critical time in the parent–child relationship.
Costs After the Neonatal Period
Relatively few studies have documented the economic costs of preterm birth or low birthweight following the infant's initial discharge from the neonatal unit. Brooten et al.15 and McCormick et al.16 report that preterm or low birthweight infants are significantly more likely to consume hospital and community health services during the early years of life than infants born at full term or at normal birthweight. The studies that have attached a monetary value to the additional healthcare resources consumed by these infants following their discharge from the neonatal unit are of varying methodological quality and differ with regard to the nature of their comparison and control groups, duration of follow-up, and the measurement and classification of outcomes. Rogowski17 limited her analysis to medical costs incurred during the first year of life and reported a 24-fold differential in costs between very low birthweight single live births (<1500 g) in the state of California Medicaid programme between 1986 and 1987 and all US births in 1989. In contrast, Stevenson et al.18 followed up a cohort of very low birthweight infants (<1500 g) born in 1979–1981 and recorded their use of hospital and family practitioner services up to age 8–9 years. When compared with a group of controls, matched for age, sex and school class, it was found that the low birthweight children used hospital and family practitioner services more intensively throughout the follow-up period. Indeed, there was a five-fold differential in mean total costs per child between low birthweight infants without disability and the control infants they were matched to. This differential increased to 16-fold amongst the lowest birthweight group (<1000 g). Amongst infants with disability, mean health service costs for the entire follow-up period were estimated at £14,510 for the lowest birthweight group (<1000 g), £12,051 for the intermediate birthweight group (1000–1500 g) and £7,178 for the highest birthweight group (>1500 g) (1998 £ sterling).
The long-term economic costs of preterm birth and low birthweight are not restricted to the health sector. Chaikind and Corman19 investigated the relationship between low birthweight, enrolment in special education and special education costs in the USA. Using a sample of approximately 8,000 children aged 6 to 15 who were in school, they calculated the probability of a child attending special education. Children who weighed less than 2500 g at birth were almost 50% more likely to be enrolled in any type of special education than children who were of normal weight at birth. This resulted in an estimated incremental cost to the USA education services of £322.9 million per year due to low birthweight (1998 £ sterling).
Other long-term economic implications of preterm birth and low birthweight are reported in the scientific literature. Many mothers of infants who intend to return to work after the birth either postpone doing so, reduce their hours or leave the workforce altogether to care for their child. This is usually associated with a reduction in family income; a 20% reduction is cited in one paper20, a 32% reduction in another14. Direct non-medical expenses incurred by families and carers as a result of additional healthcare contacts during the first year of life also remain substantial. Travel costs have been estimated at US$15 per month, childcare costs at US$47 per month and other expenses at US$62 per month (price date not reported)16. The studies by Papiernik and Kieth21, Javitt et al.22 and Gennaro14 described, but did not estimate, some of the intangible costs associated with caring for preterm or low birthweight infants. These include the emotional and physical energy required to care for the infant and the consequent isolation and restricted social contact that ensues.
Economic Evaluation of Neonatal Intensive Care
A review of the scientific literature since the 1970s, conducted by the author, revealed six studies that compared the costs and benefits of alternative organisational models of neonatal intensive care for the at-risk infant9,23–27. The studies are of variable quality in terms of the evidence for the effects of neonatal intensive care on mortality. One study from Canada by Boyle et al.23 and a study from Australia by Kitchen et al.26, were based on outcomes for population based cohorts before and after access to neonatal intensive care was expanded. For babies born at less than 1000 g, the studies estimated an additional cost per additional survivor at discharge of between £84,490 and £174,040 (1998 £ sterling) (Table 2). The cost effectiveness ratio fell to between £4,440 and £15,790 when the outcome of neonatal intensive care was measured in terms of an additional life year gained, and fell to between £4,190 and £38,030 when the outcome was measured in terms of an additional QALY gained. The most recent economic evaluation of neonatal intensive care extended the earlier work by Kitchen et al.26 and showed that the introduction of exogenous surfactant has led to a reduction in the incremental cost effectiveness ratio, however this is measured27. Therefore, despite fears to the contrary, the efficiency of neonatal intensive care has been improved by the introduction of exogenous surfactant28.
Table 2. Economic evaluations of neonatal intensive care (£ UK 1998 prices).
|Additional cost per additional survivor|
| <1000 g||1964–9 and 1973–7||Boyle et al. (1983)23||Canada||174,040|
| ||1979–80 and 1985–7||Kitchen et al. (1993)26||Australia||84,490|
| 1000–1499 g||1964–9 and 1973–7||Boyle et al. (1983)23||Canada||101,030|
|Gestational age group|
| 24–28 weeks||1971–4 and 1977–83||Doyle et al. (1989)9||Australia||56,080|
| ||1977–83 and 1984–6||Doyle et al. (1989)9||Australia||89,680|
|Additional cost per additional life year gained|
| <1000 g||1964–9 and 1973–7||Boyle et al. (1983)23||Canada||15,790|
| ||1979–80 and 1985–7||Kitchen et al. (1993)26||Australia||4,440|
| ||1985–7 and 1991–2||Victorian Study Group (1997)27||Australia||2,250|
| 1000–1499 g||1964–9 and 1973–7||Boyle et al. (1983)23||Canada||4,920|
|Additional cost per additional QALY gained|
| <1000 g||1964–9 and 1973–7||Boyle et al. (1983)23||Canada||38,030|
| ||1979–80 and 1985–7||Kitchen et al. (1993)26||Australia||4,190|
| ||1985–7 and 1991–2||Victorian Study Group (1997)27||Australia||2,990|
| 1000–1499 g||1964–9 and 1973–7||Boyle et al. (1983)23||Canada||5,430|
Studies that attempted to weigh up the overall lifetime social costs and benefits of neonatal intensive care in monetary terms have concluded that there would be a net economic loss for babies born at less than 1000 g, but that neonatal intensive care could provide a net gain for bigger babies23–25. The results of these cost benefit studies have been contentious, partly because of the methods for valuing survival. This was done by estimating the long-term care costs and subtracting them from potential earnings of survivors. Even if one accepts this as a good measure of social costs and benefits, the studies were based on very limited information about the life expectancy and long-term disability for low birthweight survivors, and about likely lifetime care needs and earning patterns. These comparisons of costs and effects of increased access to neonatal intensive care were all based on historical comparisons. It is therefore likely that the estimated change in effects may be biased.
Economic Evaluation of Prevention and Treatment of Respiratory Distress Syndrome
A number of economic evaluations in the scientific literature report the cost effectiveness of preventing or treating respiratory distress syndrome (RDS) in the at-risk infant. Table 3 summarises the results of those studies considered by the author to be sufficiently transparent and robust in their methodology to allow comparisons to be made. The table shows that surfactant treatment for RDS may have a neutral effect on costs. Among heavier babies with RDS, surfactant treatment may even lead to a reduction in overall costs, as the savings from less severe disease outweigh the costs of the treatment itself, the costs of administration and the increased care costs, which can be attributed to increased survival29. Similarly, the use of antenatal corticosteroids appears to reduce overall healthcare costs30–32. The relatively low acquisition cost of the drug and the ancillary administration and monitoring costs are outweighed by reductions in neonatal care costs. The prophylactic use of surfactant, on the other hand, appears to increase overall healthcare costs with economic studies estimating an additional cost per additional survivor of between £15,05430 and £149,36933 (1998 £ sterling). Recent trials comparing early with delayed surfactant treatment favour early use34. However, the only study to date to estimate the economic impact of early surfactant treatment suggests that it may result in a net additional cost per additional survivor35. Further economic studies are required to confirm the cost effectiveness of early surfactant treatment for newborns intubated for respiratory distress within the first two hours of life.
Table 3. Cost–effectiveness of prevention and treatment of respiratory distress syndrome (£ UK 1998 prices).
|Surfactant treatment for RDSaversus no surfactant|
|25–29 weeks||Maniscalo et al. (1989)46||USA||No difference|
|Preterm||Tubman et al. (1990)47||UK||9,667b|
|>1350 g||Schumacher et al. (1991)29||USA||Net saving|
|>650 g||Phibbs et al. (1993)48||USA||No difference|
|>1250 g||Backhouse et al. (1994)49||USA||No difference|
|700–1350 g||Mauskopf et al. (1995)50||USA||No difference|
|Antenatal corticosteroids versus none|
|<35 weeks||Mugford et al. (1991)30||UK||Net saving|
|<31 weeks||Egberts (1992)31||Netherlands||Net saving|
|<2000 g||Simpson and Lynch (1995)32||USA||Net saving|
|Prophylactic surfactant versus no surfactant|
|700–1100 g||Sell (1991)51||USA||57,505|
|<35 weeks||Mugford et al. (1991)30||UK||15,054|
|700–1350 g||Phibbs et al. (1993)48||USA||77,484|
|600–1250 g||Soll et al. (1993)52||USA||17,580b|
|<1500 g||Virtanen et al. (1993)53||Finland||47,744|
|<1500 g||Schwartz et al. (1994)33||USA||149,369b|
|800–1100 g||Egberts (1995)35||Netherlands||31,594b|
|Prophylaxis versus treatment with surfactant|
|800–1100 g||Egberts (1995)35||Netherlands||17,773b|
Economic Implications of Other Aspects of Perinatal Care
Economic evaluations of a range of other perinatal interventions targeted at the at-risk infant are also reported in the scientific literature. These include economic evaluations of antibiotic therapy36, neonatal ECMO37, septic work-ups38 and neonatal surgery39.
Two important studies have estimated the cost of birth defects40,41, which are disproportionately experienced by preterm and low birthweight infants, and one study has estimated the cost of bronchopulmonary dysplasia42, a common complication of neonatal intensive care. Waitzman et al.40 used cost of illness methodology to estimate the direct and indirect costs of 18 birth defects to the Californian economy. For a number of birth defects, such as spina bifida, Down syndrome and cerebral palsy, the costs extended beyond the first year of life and included special education costs and costs of developmental services. For all birth defects, the costs were considerable, ranging from £58,210 per case of small intestine atresia to £404,760 per case of cerebral palsy (1998 £ sterling).
Several studies document the increased risk of preterm birth or low birthweight, and consequent neonatal requirements, associated with antenatal cocaine exposure43, maternal substance abuse44 and maternal smoking45. The general conclusion of this body of literature is that neonatal costs could be reduced substantially by identifying family and social problems that mothers face antenatally and by delivering effective anti-smoking and anti-drug abuse advice.
This paper has surveyed the published evidence regarding the economic consequences of preterm birth and low birthweight. In addition to health-service costs incurred during the initial hospital stay, the paper has revealed that preterm birth and low birthweight result in substantial costs to the health services following the infant's discharge from the neonatal unit.
Furthermore, the conditions impose a substantial financial burden on special education services and on the families and carers of the infants. Other long-term consequences of preterm birth and low birthweight that require evaluation from an economic perspective include institutionalised care for the physically and medically handicapped, day-care services and respite care, and costs borne by local authorities and voluntary organisations.
Estimates of the economic costs of preterm birth and low birthweight can be informative to decision-makers. However, cost data alone cannot identify the most efficient allocation of finite perinatal resources. Rather, it is information on incremental costs and incremental health gains associated with particular healthcare activities that can identify the combination of human and material inputs that maximise health benefits. Early economic evaluations of strategies for caring for the at-risk infant focussed on neonatal intensive care. By the second half of the 1980's, neonatal intensive care had become widely established, and economic analysts focussed on the cost effectiveness of the introduction of exogenous surfactant and increased use of antenatal corticosteroid therapy. Over the coming years, it is likely that economic analysts will evaluate the impact of recent developments in perinatal care, such as head cooling, whole body cooling, liquid ventilation, intravenous immunoglobulin therapy and high frequency oscillation ventilation.
The National Perinatal Epidemiology Unit receives core funding from the Department of Health, England. The views expressed by the author do not necessarily reflect those of the funding body.