BACKGROUND. The Early Treatment for Retinopathy of Prematurity trial demonstrated that peripheral retinal ablation of eyes with high-risk prethreshold retinopathy of prematurity (early treatment) is associated with improved visual outcomes at 9 months' corrected gestational age compared with treatment at threshold disease (conventional management). However, early treatment increased the frequency of laser therapy, anesthesia with intubation, treatment-related systemic complications, and the need for repeat treatments.
OBJECTIVE. To determine the cost-effectiveness of an early treatment strategy for retinopathy of prematurity compared with conventional management.
DESIGN/METHODS. We developed a stochastic decision analytic model to assess the incremental cost of early treatment per eye with severe visual impairment prevented. We derived resource-use and efficacy estimates from the Early Treatment for Retinopathy of Prematurity trial's published outcome data. We used a third-party payer perspective. Our primary analysis focused on outcomes from birth through 9 months' corrected gestational age. A secondary analysis used a lifetime horizon. Parameter uncertainty was quantified by using probabilistic and deterministic sensitivity analyses.
RESULTS. The incremental cost-effectiveness of early treatment was $14200 per eye with severe visual impairment prevented. There was a 90% probability that the cost-effectiveness of early treatment would be less than $40000 per eye with severe visual impairment prevented and a 0.5% probability that early treatment would be cost-saving (less costly and more effective). Limiting early treatment to more severely affected eyes (eyes with “type 1 retinopathy of prematurity” as defined by the Early Treatment for Retinopathy of Prematurity trial) had a cost-effectiveness of $6200 per eye with severe visual impairment prevented. Analyses that considered long-term costs and outcomes found that early treatment was cost-saving.
CONCLUSIONS. Early treatment of retinopathy of prematurity is both efficacious and economically desirable. Because of the high lifetime costs of severe visual impairment, the early treatment strategy provides long-term cost savings.
Retinopathy of prematurity (ROP) remains the second leading cause of childhood blindness and results in >500 cases of severely limited vision or blindness each year.1,2 In 1988, the Multicenter Trial of Cryotherapy for Retinopathy of Prematurity demonstrated improved outcomes with peripheral retinal ablation of eyes developing threshold disease (corresponding to eyes with a 50% risk of retinal detachment without intervention).3,4 Even with treatment, however, visual acuity at 15 years was 20/200 or less in 45% of these eyes.5 More recently, the Early Treatment for Retinopathy of Prematurity (ETROP) trial reported that earlier peripheral retinal ablation at high-risk prethreshold ROP is associated with a significant decrease in unfavorable structural and functional visual outcome.6
Despite these improvements, there were several potentially important cost implications associated with the early treatment (ET) strategy. ET was associated with a higher rate of eyes requiring repeat laser or cryotherapy after initial treatment. In addition, a higher proportion of the ET cases were performed under general anesthesia. There were also more systemic complications associated with ET, including apnea, bradycardia, cyanosis, and need for reintubation.6,7 Furthermore, the ETROP study protocol called for more frequent screening eye examinations than was standard practice and resulted in the treatment of some eyes in which the ROP would likely have spontaneously regressed. Although each of these differences is well within the range of what would be considered clinically tolerable given the observed difference in outcomes, collectively they raise the possibility that ET may be an expensive intervention. We undertook this economic evaluation, based on published data from the ETROP randomized trial, to assess the cost-effectiveness of ET of ROP.
Study Design and Model Specification
We developed a decision analytic model based on the ETROP study design and published results using TreeAge Pro Suite software (TreeAge Software Inc 2006, Williamstown, MA). The model presented in Fig 1, included resource use, direct medical costs, and visual outcome associated with ET for ROP.
To parallel the ETROP trial, we compared ET at “high-risk prethreshold” ROP with continued monitoring and treatment only with progression to threshold ROP (conventional management). High-risk prethreshold disease was defined as a 15% risk of unfavorable retinal outcome at 3 months without treatment. Our analysis considered all possible combinations of the following events for both ET and conventional management protocols: (1) the need for anesthesia with intubation for treatment, (2) the need for reintubation after treatment, (3) the need for repeat treatment, (4) the need for future vitrectomy, and (5) expected visual acuity outcomes. The relative frequency of these events was estimated by using data derived from the ETROP study. When spontaneous regression occurred in eyes in the conventional management arm, the model considered only expected visual acuity outcomes.
Our evaluation was undertaken in 3 phases. For the primary analysis, we sought to minimize assumptions and to optimize consistency between the ETROP trial and our economic evaluation. This model therefore examined the cost-effectiveness of ET for all patients with high-risk prethreshold ROP. The time horizon for this analysis was from birth through 9 months' corrected gestational age, reflecting the time period for data collection to reach the primary end point in the ETROP trial.6 In the second analysis, we examined the cost-effectiveness of restricting ET to infants with a particular subtype of ROP (“type 1 ROP”), as recommended by the ETROP cooperative group based on a posthoc analysis of the 6-month retinal structural outcomes. We designed the third analysis to determine the cost-effectiveness of ET over a lifetime. Details of these models are given later.
Description of Primary Analysis
The primary model closely paralleled the ETROP trial. Estimates of treatment efficacy and health resource use were drawn from the published ETROP study findings. The main outcome measure was the additional, or incremental, cost (ΔC) incurred by the ET strategy expressed relative to the decrease in unfavorable visual outcome (ΔE). The resulting ratio (ΔC/ΔE) was the incremental cost-effectiveness of ET expressed in terms of the cost per eye with severe visual impairment prevented. The analysis was performed from the perspective of a third-party payer. As such, only direct medical costs were included. Productivity losses, lost future earnings, and out-of-pocket expenses were not considered. As in the ETROP trial, a 9-month time horizon was used.
Model Inputs: Effectiveness
Effectiveness in our model was measured, as in the clinical trial, as the decrease in the proportion of eyes with an unfavorable visual outcome, defined as visual acuity testing at <1.85 cycles per degree (or >4 SDs below the mean) on the Teller acuity card procedure.6 It is expected that eyes with this outcome will eventually meet criteria for legal blindness (<20/200 vision corrected).
In the ETROP trial, ET was associated with a 27% relative reduction in unfavorable functional outcome at 9 months' corrected gestational age (14.3% vs 19.8%; P < .005). There was also a significant decrease in unfavorable structural outcome (9.0% vs 15.6%; P < .001).
Model Inputs: Resource Use and Costs
A summary of the resource-use estimates used in the base-case analysis is presented in Table 1. All of the eyes in the ET group underwent peripheral retinal ablation compared with only 66% of the conventionally managed eyes. Among treated eyes, ET was associated with higher rates of anesthesia with intubation (36.6% vs 30.9%), reintubation (11.1% vs 5.1%), and need for repeat treatments (13.9% vs 11.0%). However, only 6.4% of the eyes treated early required a vitrectomy compared with 12% of the eyes randomized to conventional management.6
Where specific information related to resource use was not available from the published ETROP results, we incorporated assumptions based on consultation with experts in the field. We assumed that no additional eye examinations were required for the ET strategy compared with conventional management. Because nearly all peripheral retinal ablations are performed in the NICU, the cost of anesthesia with intubation was considered to be equal to the incremental cost associated with 1 day of mechanical ventilation in the NICU. Patients who required reintubation after treatment were assumed to remain intubated for 1 to 2 days (with an average of 1.5 days). For repeat treatments, the rate of anesthesia with intubation was considered to be the same in both study arms, and equal to the rate of anesthesia with intubation observed with initial treatments in the conventional management group. Finally, vitrectomy or scleral buckle was modeled as a 1.5-hour procedure performed in the operating room, after which patients were either admitted to the PICU for 1 day (50% of patients) or spent 30 minutes in the postanesthesia care unit and then were admitted to the general pediatrics ward for 1 day (50% of patients). In addition, the primary model assumed that 20% of these eyes would require a repeat surgical procedure.
Total costs were calculated as the product of the resources used and the unit prices associated with those resources. The unit costs are presented in Table 2. Hospital costs were derived from charges, converted using department-specific cost-to-charge ratios. The expense for a day of intubation was calculated as the difference in the median summed daily charges for a day of NICU admission between intubated and nonintubated infants at 10 to 12 weeks of life (the mean age of treatment), converted to costs using department-specific cost-to-charge ratios.8 Costs for physician care were derived from the average commercial insurance reimbursement rates for equivalent services at major metropolitan tertiary care institutions. Of note, according to current fee structures, ophthalmology fees are typically bundled charges in which all services provided from the time of an initial procedure (ie, laser therapy or vitrectomy) until 90 days after the procedure are incorporated in the initial fee. Because repeat treatments to an eye almost always occur within this 90-day window, there was no ophthalmology reimbursement considered for these repeat services. However, patients were charged for other resource use associated with these repeat procedures. Patients were also charged for procedures performed on the contralateral eye in accordance with usual billing and reimbursement practices.
All costs were expressed in 2005 US dollars; conversions from other dates were made by using the medical care component of the consumer price index.9 Because the time horizon for the primary analysis is <1 year, discounting was not undertaken.
We assessed sampling uncertainty, or uncertainty in the reproducibility of the rates of clinical outcomes observed in the ETROP clinical trial, through probabilistic sensitivity analysis. A computer simulation was performed in which 1000 cohorts were run through the original model. A series of β distributions, based on the number of patients at risk for a given outcome and the number of outcomes observed, were used to generate probability distributions at each of the model's chance nodes. For each cohort, the appropriate probability distribution was sampled to provide a value to be used at each chance node within the model. This process generated 1000 estimates of the incremental cost-effectiveness ratio, which were then used to create cost-effectiveness acceptability curves and scatter plots. We performed a probabilistic sensitivity analysis on the base-case, best-case, and worst-case models.
Uncertainty in cost and resource assumptions was assessed through deterministic sensitivity analyses, where a specific input to the model was changed within a plausible range and the effect on the incremental cost-effectiveness ratio determined. We performed deterministic sensitivity analyses for all cost assumptions, for the primary efficacy result of the trial, and for the following key resource assumptions: (1) the number of days of mechanical ventilation in patients requiring reintubation; (2) the operating room time required to perform a vitrectomy; (3) the percentage of patients who required a repeat procedure after their initial vitrectomy; (4) the percentage of patients requiring admission to the PICU after a vitrectomy; (5) the percentage of patients receiving laser photocoagulation under anesthesia with intubation; and (6) the percentage of patients requiring reintubation related to laser photocoagulation. We also performed a “best-case” analysis, in which each input was chosen at the extreme of its plausible range favoring ET, and a “worst-case” analysis, in which each input was chosen at the extreme of its plausible range favoring conventional management.
Description of Analysis of Restricting Early Treatment to Infants With Type 1 ROP
In a secondary analysis of the ETROP trial data, the investigators proposed that fewer eyes could be treated with similar results by restricting treatment to eyes with type 1 ROP (defined as Zone I, any stage ROP with plus disease; Zone I, stage 3 ROP with or without plus disease; Zone II, stage 2 or 3 ROP with plus disease).6 As peripheral retinal ablation of type 1 ROP has been widely adopted as clinical practice, we developed a model to examine the incremental cost-effectiveness of treatment of type 1 ROP compared with conventional management (ie, treatment of threshold ROP). The risk of developing type 1 ROP among those with high-risk prethreshold ROP was determined from the ETROP published data.6
With the exception of restriction of treatment to eyes with type 1 ROP, this model was similar to the primary analysis. Specifically, the same 9-month time horizon, third-party payer perspective, and effectiveness measure were used.
Because data regarding the rate of anesthesia with intubation, reintubation, and repeat treatment among those with type 1 ROP were not available, we assumed these rates would the same as in the ET group with high-risk prethreshold disease. Furthermore, all of the patients in the ET arm who required a vitrectomy were assumed to have type 1 ROP. These conservative assumptions were designed to bias these analyses toward a higher incremental cost-effectiveness ratio.
Description of Analysis of Long-term Cost-effectiveness
The long-term analysis differed from the primary analysis in several respects. To facilitate comparison with other important social programs, we performed the analysis from the societal perspective by extending the third-party payer perspective to include all associated costs, regardless of the parties to whom they accrue, over the course of a lifetime. We therefore included the expected lifetime cost of a poor visual outcome, estimated at $615723 per person10 after discounting at 3% per annum. To make the results more easily interpretable, we evaluated costs and outcomes per person rather than per eye. Therefore, the costs for those with bilateral high-risk prethreshold ROP were assumed to be doubled, because both eyes require treatment; and those infants expected to have a good visual outcome in at least 1 eye were considered to have an overall good outcome. For the base-case analysis, we assumed 100% concordance between eyes. This assumption was then varied in a sensitivity analyses.
Effectiveness was measured as quality-adjusted life-years (QALYs). We converted expected visual acuity outcomes into utilities using the following formula: Utility = 0.374(x) + 0.514, where x is the Snellen visual acuity in the better seeing eye in decimal form (eg, 20/40 = 0.67).11 This conversion was derived by interviewing patients with visual loss of different etiologies using time trade-off techniques.11 Life-expectancy was based on the 2001 National Vital Statistics Report.12 For infants surviving to 9-months' corrected gestational age (∼1 year of life), the average life expectancy was 77.6 years. Life expectancy was assumed to be unaffected by visual acuity outcomes. We calculated the QALYs associated with each treatment pathway as the product of the visual acuity-dependent utility value and life expectancy, discounted at a rate of 3% per annum. The main outcome measure for this ling-term analysis was therefore expressed as cost per QALY gained.
Primary Analysis: Treating All High-Risk Prethreshold ROP
The results of our primary model indicate that although ET is more costly, it is more effective than conventional treatment strategies. The median cost and effectiveness of ET calculated by the model are shown in Table 3. The incremental cost-effectiveness of ET of eyes with high-risk prethreshold disease compared with treatment at threshold is $14200 per eye with severe visual impairment prevented.
The results of the probabilistic sensitivity analysis of the base case are presented in Fig 2. Each point represents the results of a single run of the simulation model. In this primary analysis, 97.4% of the model runs yielded cost-effectiveness ratios that lie in the northeast quadrant of the cost-effectiveness plane, indicating that the intervention is more costly, but more effective (see Fig 2). There is little chance that ET is both less costly and more effective (“dominant” = 0.5%), or that it is more costly but less effective (“dominated” = 2.1%). These same results are expressed as a cost-effectiveness acceptability curve in Fig 3. This relates a given willingness-to-pay to the probability that the cost of the intervention falls below that level. For example, if a policy maker were willing to pay $40000 to avert 1 additional eye with severe visual impairment, examination of the graph indicates a 90% probability that this intervention would be considered cost-effective.
Deterministic sensitivity analyses, in which the cost and resource assumptions used in the base-case analysis were varied, resulted in the cost-effectiveness estimates presented in Table 4. We found that the cost-effectiveness of ET was most sensitive to the number and cost of eye examinations required for ET. Decreasing the number of preprocedure eye examinations by 2 in the patients in the ET group improved the incremental cost-effectiveness ratio to $7300 per eye with severe visual impairment prevented; while increasing the number of examinations by 4 leads to an additional cost of $28000 per additional eye with severe visual impairment prevented. The incremental cost-effectiveness of ET was also sensitive to the cost of laser therapy with a range of $3200 to $19100 per eye with severe visual impairment prevented based on the range of procedure costs shown in Table 2. The incremental cost-effectiveness ratio was minimally sensitive to all other individual cost and resource assumptions, falling between $10000 and $20000 per eye with severe visual impairment prevented.
In addition, we performed deterministic sensitivity analyses varying the primary efficacy result of the trial. As expected, the incremental cost-effectiveness ratio was quite sensitive to the estimated effectiveness of ET. However, even if ET were only half as effective as predicted by the ETROP trial (or a 2.75% risk difference between ET and conventional management), the cost-effectiveness of ET would be just $28500 per eye with severe visual impairment prevented (see Table 4).
In the “best-case” analysis, in which we concurrently varied all assumptions in a direction favoring ET, we found an 89% chance that ET would be cost-saving or dominant (both less costly and more effective) and only a 0.1% chance that it would be dominated (more costly, but less effective). At a willingness-to-pay of just $10000, there was a 97% chance that ET would be considered cost-effective. This best-case analysis assumed no difference in the number of eye examinations; lowest costs for laser therapy, anesthesia with intubation, and reintubation; and modeled vitrectomy as a 3-hour procedure with physician fees at highest reimbursement rate, postoperative admission to the PICU, and a 30% rate of repeat operative procedures.
The worst-case analysis varied all assumptions in a direction favoring conventional treatment. Here we assumed 4 additional eye examinations associated with ET; highest costs for anesthesia with intubation and reintubation; ophthalmology fees for repeat laser or cryotherapy equal to the fees for initial treatment; and modeled vitrectomy as a 1-hour procedure with physician fees at the lowest reimbursement rate, postoperative admission to the postanesthesia care unit followed by overnight observation on the general medical service, and a 10% rate of repeat operative procedures. With these assumptions, the expected incremental cost per additional eye with severe visual impairment prevented was $50500. The probabilities that ET would be considered cost-effective at a willingness-to-pay of $40000, $60000, and $80000 were 31%, 63%, and 77%, respectively. There was a 2% chance that ET was dominated (more costly and less effective).
Analysis of Restricting Early Treatment to Infants With Type 1 ROP
On the basis of a posthoc analysis of the ETROP results, the ETROP Executive Committee recommended restricting ET to eyes with type 1 ROP. This strategy is expected to achieve similar outcomes with a reduction in the number of eyes treated. Our models suggest that treatment at type I ROP is even more cost-effective than ET of all eyes with high-risk prethreshold ROP. Treatment at type I ROP compared with conventional management resulted in an incremental cost-effectiveness ratio of $6200 per eye with severe visual impairment prevented. There was a 95% chance that this strategy is cost-effective at a willingness-to-pay of $40000, a 12% chance that treatment of type 1 ROP is cost-saving, and a 3% chance that it is dominated (more costly and less effective).
Long-term Cost-effectiveness of ET
In the analysis of long-term outcomes, ET provided overall cost-savings. Our model results suggest that treating 1 child at high-risk prethreshold ROP (compared with treatment at threshold ROP), will save $27000 and result in a gain of 0.284 QALYs. In a probabilistic sensitivity analysis, there was a 97% chance that ET would lead to both lower costs and improved outcomes (see Fig 3). In deterministic sensitivity analyses, ET remained cost-saving under several conservative assumptions including: (1) outcomes in the eyes of those with bilateral high-risk prethreshold ROP were completely independent of 1 another; and (2) QALYs gained by ET that were only half those calculated for the base-case analysis. Even under the unlikely assumption that no additional long-term costs are associated with a poor visual outcome, ET remained a cost-effective intervention, costing just $4900 per QALY gained.
We report a decision analytic-based economic evaluation of the ET strategies used in the ETROP trial. Our results indicate that a policy of treating ROP at high-risk prethreshold disease, compared with conventional management, is expected to cost $14200 per additional eye with severe visual impairment prevented and would have a 90% probability of costing less than $40000 per additional eye with severe visual impairment prevented. Restricting ET to those eyes with type 1 ROP, as recommended by the ETROP Cooperative Group, is an even more cost-effective strategy, with an incremental cost-effectiveness of $6200 per eye with severe visual impairment prevented. Most importantly, over the course of a lifetime, ET is expected to provide overall cost-savings; that is, it is less costly and improves overall quality of life.
On the basis of these results, we conclude that ET of ROP is both efficacious and has a cost-effectiveness that compares favorably with other accepted medical interventions. Similar to cryotherapy for threshold ROP, which was previously estimated to cost an additional $3600 per QALY gained compared with no intervention in the short-run and to be cost-saving over a lifetime, we found that earlier treatment of ROP cost an additional $14200 per additional eye with severe visual impairment prevented in the short-run and was cost saving over a lifetime.13 Such “cost-saving” interventions are quite infrequent in health care economic evaluations because it is almost always necessary to expend resources to improve health. For example, extracorporeal membrane oxygenation for severe respiratory failure in mature neonates costs 13385 pound sterling per life-year gained,14 inhaled nitric oxide for persistent pulmonary hypertension of the newborn costs $33200 per life saved,15 and universal newborn hearing screening costs $44000 per case of deafness diagnosed by 6 months of age.16
Because ETROP did not include a prospective ancillary economic evaluation, our analysis was based on a decision analytic model using aggregate estimates of efficacy and resource use from the published trial data. There are certain advantages to this approach. For example, a decision analytic approach increases generalizability, permits modeling of changes in the probabilities of various events, and allows for projection of the cost-effectiveness of an intervention over the long-term. There are, however, also disadvantages, including the increased need to make assumptions and the possibility of introducing bias as a result of the choice of model inputs. Although we based our model as closely as possible on the ETROP trial design and results, our approach did require us to make assumptions regarding the types of resources used, the variance in the use of those resources, and the associations between resource use and outcomes. However, even in a “worst-case” analysis, ET cost only an additional $50500 per additional eye with severe visual impairment prevented. Our conclusions were robust to testing of these assumptions in sensitivity analyses, suggesting that use of the underlying clinical trial data would not have resulted in different policy recommendations.
It is important to note that the measure of effectiveness for the base-case and type 1 ROP analyses in our study was visual outcome per eye. This was done to maintain consistency with the ETROP trial. However, in most infants there is a high degree of correlation between the course of ROP in each eye, and therefore visual outcomes are expected to be similar. In fact, at the time of randomization in the ETROP trial, 79% of patients had bilateral high-risk prethreshold disease.6 This limitation did not apply to the long-term analysis, which considered visual acuity outcomes per patient, rather than per eye, and converted these visual acuity outcomes to QALYs.
This cost-effectiveness analysis is based on a clinical trial. Although we expect a high degree of external validity, it is possible that changes in standard management or reimbursement practices outside the trial setting could affect the cost-effectiveness estimate. For example, as ETROP becomes standard of care, the proportion of eyes requiring repeat treatments is expected to be higher than that previously seen. Therefore, we expect that either the cost for initial laser therapy will increase or that the practice of bundling ophthalmology charges will change. This would increase our estimate of the incremental cost-effectiveness of ET. Conversely, anesthesia with intubation is becoming a less common practice during peripheral retinal ablation therapy than in previous years. Fewer procedures performed under anesthesia would decrease our estimate.
On the basis of current evidence, ET of high-risk prethreshold retinopathy of prematurity, as undertaken in the ETROP protocol, is a highly economically desirable health policy. The ETROP Cooperative Group is currently conducting the 6-year long-term follow-up of visual outcomes in this cohort of patients. We look forward to the availability of this information about more long-term visual outcomes to validate and refine our cost-effectiveness estimates.
This project was supported by a grant from the Institute for Health Technology Studies.
We gratefully acknowledge the expert assistance of Drs Deborah Vanderveen, Peter Hovland, and Tatsuo Hirose regarding current care of infants with retinopathy of prematurity.
- Accepted April 22, 2008.
- Address correspondence to Karen L. Kamholz, MD, MPH, Boston Medical Center, Division of Neonatology, One Boston Medical Center Place, Boston, MA 02118. E-mail:
The authors have indicated they have no relationships relevant to this article to disclose.
What's Known on This Subject
Peripheral retinal ablation of eyes with high-risk prethreshold ROP (ET) is associated with improved visual outcomes. However, ET increases the frequency of laser therapy, anesthesia, systemic complications, and repeat treatments. The cost implications of ET have not been established.
What This Study Adds
We used published results of the ETROP trial to generate estimates for the cost-effectiveness of this therapy, showing that early treatment is both efficacious and economically desirable.
- ↵Good WV, Hardy RJ, Dobson V, et al. The incidence and course of retinopathy of prematurity: findings from the early treatment for retinopathy of prematurity study. Pediatrics.2005;116 (1):15– 23
- ↵Wheatley CM, Dickinson JL, Mackey DA, Craig JE, Sale MM. Retinopathy of prematurity: recent advances in our understanding. Arch Dis Child Fetal Neonatal Ed.2002;87 (2):F78– F82
- ↵McBride J, Parad R, Davis J, Allred L, Zupancic J. Cost analysis on the use of recombinant human superoxide dismutase (rhSOD) at birth in preterm infants to improve pulmonary outcome. J Perinatol. In press
- ↵Consumer Price Index (Medical Care, All Urban Consumers, Northeast Urban). National Bureau of Labor Statistics; 2006. Available at: www.bls.gov/cpi. Accessed March 1, 2007
- ↵Javitt J, Dei Cas R, Chiang YP. Cost-effectiveness of screening and cryotherapy for threshold retinopathy of prematurity. Pediatrics.1993;91 (5):859– 866
- ↵Petrou S, Bischof M, Bennett C, Elbourne D, Field D, McNally H. Cost-effectiveness of neonatal extracorporeal membrane oxygenation based on 7-year results from the United Kingdom Collaborative ECMO Trial. Pediatrics.2006;117 (5):1640– 1649
- ↵Lorch SA, Cnaan A, Barnhart K. Cost-effectiveness of inhaled nitric oxide for the management of persistent pulmonary hypertension of the newborn. Pediatrics.2004;114 (2):417– 426
- ↵Keren R, Helfand M, Homer C, McPhillips H, Lieu TA. Projected cost-effectiveness of statewide universal newborn hearing screening. Pediatrics.2002;110 (5):855– 864
- Copyright © 2009 by the American Academy of Pediatrics