Although few studies have tabulated the medical care costs of preterm birth in early childhood, fewer still have analyzed the medical costs beyond early childhood for those born prematurely. Lifetime estimates of cost, however, have been made for individuals with certain conditions and developmental disabilities associated with preterm birth and LBW, such as specific birth defects (Waitzman et al., 1996), CP (CDC, 2004c; Honeycutt et al., 2003; Waitzman et al., 1996) and MR, HL, and VI (CDC, 2004c; Honeycutt et al. 2003). Prevalence estimates of developmental disabilities in a Centers for Disease Control and Prevention (CDC) study (CDC, 2004c) were drawn from MADDSP, which CDC established in 1991 to identify children with these developmental disabilities (Yeargin-Allsopp et al., 1992). The CDC study, together with unpublished tabulations of differences in the prevalence of each developmental disability by ges-tational age from MADDSP (see Table 12-8), permitted the assessment of the incremental lifetime direct medical and special education costs as well as the indirect costs of lost household and labor market productivity for those born preterm with the four developmental disabilities presented here.
Cost estimates from the CDC study that served as the basis for cost estimates provided here were based on cross-sectional data on age-specific average service utilization multiplied by average cost or on labor market productivity multiplied by average compensation for those with these developmental disabilities relative to the population as a whole. These cross-sectional data were then applied to a synthetic cohort, based on prevalence data from the MADDSP and on survival estimates from the literature. Costs were discounted back to birth at a 3% rate.
Service utilization, labor market participation rates, and cost estimates were made from national databases. The primary sources for disability-specific service utilization were the 1994 and 1995 National Health Interview Survey-Disability Supplement (NHIS-D) (inpatient, prescription medications, therapy and rehabilitation, long-term care and the 1994 and 1995 National Health Interview Survey (NHIS) (physician visits). Pricing of outpatient medical care services relied primarily on the 1987 National Medical Expenditure Survey (NMES), whereas inpatient service cost was based on charge data from the 1995 Healthcare Cost and Utilization Project adjusted to costs using Medicare cost-to charge ratios. The receipt of special education services by developmental disability was based on the MADDSP, whereas special education placement category for those receiving services by disability was taken from the NHIS-D. Average special education costs were then estimated based on the incremental price of placements from Moore et al (1988). Disability-specific work limitation estimates were based on the NHIS-D, while earnings losses associated with such limitations were based on the Survey of Income and Program Participation (SIPP) (Honeycutt et al., 2003). Costs taken from that study for this report were inflated to 2005 based on Medicare reimbursement indices for medical care, weighted by type of service, and on the employee compensation index for public education (special education services costs) and the general employee compensation index (lost productivity estimates).
Estimates of lifetime medical care costs per individual with one or more of the four developmental disabilities, regardless of gestational age, ranged from $23,209 for those with HL to $123,205 for those with MR, expressed in 2000 dollars (Honeycutt et al., 2003) (Table 12-9). Although long-term care comprises a large proportion of medical costs for those with developmental disabilities, inpatient hospitalization and physician visits also contribute significant costs. Long-term care costs comprised about 44 percent of the medical care costs for individuals with MR, but physician visits and inpatient hospitalization accounted for more than 40 percent of the costs (Table 12-9). For those with CP, HL, and VI, physician visits and inpatient hospitalization accounted for more than two-thirds of the medical care costs (Honeycutt et al., 2003). These estimates of long-term care costs are particularly conservative, as such care for children under age 18 years and for those in institutionalized settings were not included. Long-term care costs for those with CP, for example, were found to be more than 63 percent of the lifetime medical care costs in earlier work, based on a cohort of individuals with CP born in California, which included estimates for those institutionalized in that state's developmental centers (Waitzman et al., 1996).
Excess cases of the four developmental disabilities associated with preterm birth were based on unpublished tabulations from MADDSP on the prevalence of the four developmental disabilities by gestational age for the cohort born from 1981 to 1991 surviving to age 3 years. With the exception of HL among infants born at 33 to 36 weeks of gestation, among whom the difference in prevalence of HL relative to that among infants born at term was of only marginal significance, the prevalence of all four developmental disabilities was significantly higher among preterm infants than among term infants. The prevalence exhibited a distinct inverse gradient by gestational age (Table 12-8). Although births at less than 28 weeks of gestation comprised 0.8 percent of all births, for example, extreme prematurity accounted for 19, 6, 7.5, and 17.4 percent of all cases of CP, MR, HL, and VI, respectively (Table 12-8). In addition, although extremely preterm births comprised just 6.9 percent of all preterm births in the MADDSP sample of survivors to age 3 years, children born extremely preterm accounted for more than 40, 23, 23, and 47 percent of all individuals born preterm with CP, MR, HL, and VI, respectively. The ratios of the rates of CP, MR, VI, and HL among preterm infants to those among term infants surviving to age 3 years were 6.5, 2.1, 2.1, and 4.3, respectively. Prevalence differences, given in the final column of Table 12-8, were used to estimate the excess number of cases of each developmental disability among survivors born preterm. Prevalence by birth weight category demonstrated similarly sharp gradients (Table 12-10).
Several of the infants displaying one of the four index developmental disabling conditions had multiple index disabling conditions, and aggregation of the costs for all these disabling conditions therefore required that such cooccurrences be taken into account to avoid double counting. The cooccurrence rates generated from MADDSP permitted such aggregation (Year-gin-Allsopp et al., 1992). Because per-case costs were not estimated separately for each permutation of multiple conditions, an algorithm was established to assign cases to conditions. The algorithm was adopted by assigning cases hierarchically to the condition with the highest cost among each case with multiple conditions. The implicit assumption was that a case with multiple disabling conditions costs, on average, at least as much as the mean cost of its highest-cost disabling condition. The resulting hierarchical order of conditions, from highest to lowest cost, was MR, CP, VI, and HL. An individual born with both MR and HL, for example, was assigned the average cost of all individuals with MR. On the basis of the data reported by MADDSP, 64 percent of survivors to age 3 years with CP had MR, 73 percent of those with VI had either MR or CP, and 23 percent of those with HL had MR, CP, or VI. Medical care costs were adjusted for inflation to 2005 dollars by using a price index weighted by Medicare medical price and employment cost indices, based on the percentage contributions of inpatient services; outpatient services; and therapy, rehabilitation, and long-term care to the overall treatment of these developmental disabilities.
Because medical care costs for all premature infants were reported above for those through age 7 years, medical care cost estimates for these four developmental disabilities through age 5 years were subtracted from the total medical care costs in the disability cost study to avoid double counting. The net effect was to reduce the total lifetime medical costs associated with these conditions by less than 5 percent. Because the age categories used in the developmental disability cost study made it difficult to accurately carve out medical care costs for 6- and 7-year-olds separately, cost estimates for those 6- and 7-year-olds were subtracted from the overall medical care costs for preterm birth provided above to avoid double counting. Estimates of the costs at each age were discounted at a rate of 3 percent to the year of birth. Adjustment for the prevalence differences between term and preterm birth by condition and application of the algorithm for multiple disabilities discussed above yielded a total of $976 million in lifetime incremental medical care costs associated with preterm birth after age 5 years, or nearly $2,000 per preterm birth. The addition of these costs to the national costs of preterm birth from birth to age 5 years yielded a total national cost of medical care for preterm birth of $16.86 billion in 2005.
The estimate of national medical care cost associated with preterm birth is conservative, in the sense that it includes the costs associated with all premature infants only through the first 5 years of life. The costs beyond age 5 years, as demonstrated by these cost estimates for the four developmental disabilities associated with preterm birth, are quite substantial. Although these conditions associated with preterm birth are among the most disabling, there are several others, including autism and certain birth defects (Rasmussen et al., 2001). Furthermore, the medical care cost estimates for even these four developmental disabilities are conservative, because long-term care provided for the small subset of the population who required institutionalized settings and therapy and rehabilitation services provided for adults were not included.
In summary, extremely preterm birth contributes disproportionately to the medical care costs of prematurity, not only because early medical care for such infants is so expensive but also because survivors have disproportionately high rates of disabling conditions that generate high lifetime medical care costs. Very preterm and moderately preterm infants, however, also have significantly elevated rates of developmental disabilities relative to those for term infants. Although the prevalence of disabilities is lower among this group than among extremely preterm infants, the disproportionate number of births in this category yields a significant contribution to the national cost. Moderately preterm infants, for example, comprise over 50 percent of preterm cases with MR.
Treatment for reactive airway disease, for which all premature infants are at risk, and treatment of infection may contribute to higher average outpatient costs among those born pre-term. The relatively high outpatient medical care costs in the upper tail of the medical care cost distribution for those born at 32 to 36 weeks of gestation noted earlier is likely explained, in part, by the costs incurred by individuals with developmental disabilities. LBW is clearly a risk factor for developmental disabilities and their associated costs and one that is correlated with gesta-tional age. The extent to which these risk factors contribute independently to these developmental disability costs is not fully understood. Future studies should investigate costs for other disabling conditions which were not available for these analyses but are prevalent among preterm infants such as asthma and attention deficit hyperactivity disorder.
Was this article helpful?
If you suffer with asthma, you will no doubt be familiar with the uncomfortable sensations as your bronchial tubes begin to narrow and your muscles around them start to tighten. A sticky mucus known as phlegm begins to produce and increase within your bronchial tubes and you begin to wheeze, cough and struggle to breathe.