Although everyone talks about the importance of doing cost analyses related to early hearing detection and intervention, terms are frequently misused. It is important to define what we mean by some of the most frequently used terms. As shown in slide #61, three distinctions are important. First, it is important to determine how much early hearing detection and intervention programs cost completely independent of the effects of such programs. Methodologies for estimating costs are well developed and relatively straight forward to apply.

Cost effectiveness is a second type of cost analysis. By definition, the cost effectiveness of a program can only be determined if one program is being compared to another. It is inappropriate to say that program A is cost effective. Instead, we must evaluate whether program A is more cost effective than program B. In other words, if we examine both the costs and effects of two different programs, which program yields the most effects for every unit of cost? In a cost-effectiveness comparison of two programs, the costs of each program are analyzed in the same way as the first type of cost analysis described above. However, the effects of each program are also calculated according to variables like the number of children per thousand identified, the age at which those children are fitted with amplification, scores on various measures of developmental functioning, etc. The results of a cost-effectiveness analysis might conclude that the cost of identifying a child with hearing loss in program A is $7,294, while the cost of identifying a child with hearing loss in program B is $8,492. In this hypothetical example, program A would be more cost effective than program B.

A cost-benefit study can be either comparative or applied to a single program. Costs are computed in exactly the same way as in the first two types of cost analyses. However, in this case, the benefits associated with newborn hearing detection and intervention must be translated into monetary values. A program is said to be cost beneficial if the amount of money spent on the program is less than the monetary worth of the benefits resulting from that program. As you can imagine, good cost-benefit studies are very, very difficult to do, because of the difficulty of assigning monetary value to outcomes such as a year and a half worth of reading gain or the monetary value to a parent of being able to communicate more effectively with their child.

A number of different kinds of cost studies related to early hearing detection and intervention have been reported in the literature. Most of these are either sample cost estimates or cost-effectiveness analyses, and most suffer from serious weaknesses. For example, in most studies the estimates of costs are based on hypothetical assumptions or unverified self-reports. Such cost estimates are often incomplete in that they ignore costs, such as fringe benefits, indirect costs, and costs to parents. Other studies only include a part of the detection and intervention process (e.g., costs for follow up and tracking may be excluded), and standard economic analysis concepts, such as discounting, sensitivity analysis, and robustness estimates, are frequently not used.

The only area where we have some credible data is related to the costs of newborn hearing screening. One of the first such studies was done with the Rhode Island Hearing Assessment Program, in which a complete cost analysis was done using actual expenditures, instead of self- report data, for the entire screening and tracking process, but diagnosis and intervention costs were explicitly not included. As shown in slide #63, the cost of a two-stage screening, including scheduling and tracking the babies into a diagnostic evaluation, but not including the cost of the diagnosis itself, was approximately $26 per baby. This did not include the cost of parents' time for participating in the screening activities.

A similar study was done at Logan Regional Hospital as reported by Weirather and her colleagues and shown in slide #64. Because of the way this program was organized, the cost per baby was substantially cheaper ($7.42 per baby), even though the analysis included all of the same activities and was just as complete as the RIHAP study.

Recently, CDC, under the direction of Scott Grosse, did a multi-center study in which the costs of newborn hearing screening were estimated in six different centers in six states (slide #65). Three of those centers were AABR-based sites and three were otoacoustic-emission-based sites. Cost estimates were based on self-report questionnaires, and site visits were made to four of the six sites to gather confirmatory data wherever possible. Although the actual time devoted to various activities was not tracked, as was the case in the studies done in Rhode Island and Utah, the analysis did use standardized techniques for including the costs of fringe benefits, equipment, supplies, and overhead. As shown in slide #66, the cost per baby ranged from about $18 to $26, with TEOAE sites being less expensive.

A final study of costs of screening was reported by John Stevens and his colleagues (slide #67) for ten different hearing screening programs in Great Britain. Five of these programs did targeted screening with high-risk infants, three were universal newborn hearing screening programs, and two were home-visitor programs. Results ranged from an average of about $8 per baby born for the high-risk programs to $22 per baby for the universal newborn hearing screening programs to $32 per baby for the home-visitor programs. Although not as much detail about the procedures used in the cost analysis were included in this report, the results are quite similar to what has been found by studies done in the U.S. It is also important to note that in the ten programs evaluated by Stevens, there was reasonable consistency within the various types of programs, which makes the results more believable.

It is possible to do some very rough estimates of some of the more obvious monetary benefits associated with newborn hearing detection and intervention programs by using information from other sources about the prevalence and costs of screening for various diseases among newborns. As shown in slide #68, the prevalence of permanent congenital hearing loss at 3 per thousand is substantially higher than the prevalence of PKU, hyperthyroidism, or sickle cell anemia, which are required for screening in every state. All three of those diseases are screened for using the same blood test, which for purposes of this comparison was estimated to cost $25 per test. Because the prevalence of those other diseases is so low, the cost for identifying a child with any one of those diseases is approximately $41,000 per child, compared to a cost of $8,683 to identify a child with permanent congenital hearing loss (PCHL).

Is it worth it to spend almost $9,000 to identify a child with PCHL during the first few months of life? Based on data presented about the benefits of early identification, it appears that children identified early will have better cognitive skills, social skills, and language skills, which are all the types of things that contribute to the child being capable of being educated in a regular mainstream classroom or a self-contained classroom, instead of a residential program. In other words, the data we have suggests that if children are identified early, they will be capable of being appropriately educated in a less restrictive and less expensive environment. (Note that these data should not be used to suggest that states could save money by inappropriately moving children from more restrictive to less restrictive environments. All educational systems must be available for the child, and the most appropriate setting should be determined based on the capabilities of the child and the preferences of the parent and child.) As shown in slide #70, the U.S. Department of Education estimates that for hearing-impaired children, the annual costs of education in a regular mainstream classroom in 1990 was $3,383, while the annual costs for a hearing-impaired child in a self-contained classroom or residential placement was $9,689 and $35,780, respectively. Thus, over the educational lifetime of a child, substantial amounts of money would be saved if, as a result of early identification and intervention, the most appropriate educational setting for the child is a regular mainstream classroom instead of a self- contained classroom or a self-contained classroom instead of a residential program. In fact, if only 2% of the children identified with a hearing loss were educated in a self-contained classroom instead of a residential program, it would more than pay for the cost of the newborn hearing screening program in which all of the children were identified. Although there have not been empirical studies of the number of children who would be more appropriately educated in less restrictive environments as a result of newborn hearing screening programs, based on the data about the benefits of early intervention and the costs of early identification programs, it is very plausible to expect many early identified children to be educated in less restrictive environments. As a result, at least this much money, and probably much more, would be saved.

Back | Intro | Next