Little information is available about the transportation costs incurred from the missed prenatal diagnosis of congenital heart disease (CHD). The objectives of the present study were to analyze the costs of emergency transportation related to the postnatal diagnosis of major CHD and to perform a cost/benefit analysis of additional training for ultrasound technicians to study the implications of improved prenatal detection rates. The 1-year costs incurred for emergency transportation of pre- and postnatally diagnosed infants with CHD in Northern California and North Western Nevada were calculated and compared. The prenatal detection rate in our cohort (n = 147) was 30.6%. Infants postnatally diagnosed were 16.5 times more likely (p <0.001) to require emergency transport. The associated emergency transportation costs were US$542,143 in total for all patients with CHD. The mean cost per patient was $389.00 versus $5,143.51 for prenatally and postnatally diagnosed infants, respectively (p <0.001). Assuming an improvement in detection rates after 1-day training for ultrasound technicians, the investment in training cost can be recouped in 1 year if the detection rate increased by 2.4% to 33%. Savings of $6,543,476 would occur within 5 years if the detection rate increased to 50%. In conclusion, CHD diagnosed postnatally results in greater costs related to emergency transportation of ill infants. Improving the prenatal detection rates through improved ultrasound technician training could result in considerable cost savings.
It is important for payers, insurers, health service providers, and government to understand congenital heart disease (CHD) economics when formulating policy around prenatal detection. Although CHD prenatal detection rates can be improved, the potential economic effect has not been defined. We hypothesized that a prenatal diagnosis leads to reduced costs associated with emergent transport of infants with major CHD. Using data from our initial study, our objective was to determine the incremental costs incurred in Northern California and North Western Nevada because of a missed prenatal diagnosis of major CHD, focusing on the costs of emergency transportation to a tertiary care center after delivery. In addition, a cost/benefit analysis was conducted to investigate the potential effects of a 1-day training program for ultrasound technicians (UTs) on costs.
Methods
The present cost analysis was performed at the Hospital for Sick Children and University of Toronto (Toronto, Ontario, Canada) using data collected prospectively at 3 referral centers in Northern California (Lucile Packard Children’s Hospital at Stanford University, University of California, San Francisco, and University of California, Davis). The study group consisted of infants <6 months old with major CHD (i.e., requiring or expected to require intervention), admitted to 1 of these referral centers, during a 1-year period (2004 to 2005). The infants were either born in a referral center, transferred by air or ground ambulance from a regional hospital after birth, or transported by personal motorized vehicle within 6 months after birth. In some instances, the infants were transferred more than once. For example, some infants were first transferred to a regional center and then onward to the referral center. The infants were categorized into those diagnosed prenatally and those diagnosed postnatally from the parent’s response to questionnaires and information from prenatal care providers. Only infants with CHD expected to be detected by prenatal ultrasonography were included. We, therefore, excluded patients with isolated ventricular septal defect, atrial septal defect, or minor valve lesions. For the purposes of the cost analysis, and to better define the cost parameters using a conservative approach, we excluded infants transported >230 miles, because they could have been transported to closer tertiary hospitals.
The transport of infants with CHD requires a multidisciplinary team, the potential use of mechanical ventilation, prostaglandin therapy in potentially ductal dependent lesions, and a ground or air transportation vehicle. The distances and associated costs were estimated for all emergency medical transfers to the tertiary care centers using the 2009 University of California San Francisco Children’s Hospital, private United States air ambulance service providers’ fees and transportation guidelines ( Table 1 ). The cost of a fetal echocardiogram used in the sensitivity analysis was $3,642.
| Services | Costs | Transport Guidelines (miles) |
|---|---|---|
| Ground transport (ambulance) | $500 ($8/mile after first 20 miles) | <75 |
| Air transport | 75–149 | |
| Rotary wing | $1,500 start up, then $35/mile | |
| Fixed wing | $8,000–$12,000 (depending on distance) | 150–230 |
| Registered nurse and physician team | $3,000/hour | |
| Prostaglandin E 1 | $225/ampoule/hour | |
| Mechanical ventilation | $60/hour |
Categorical data were compared using the chi-square test and the mean using the Student t test. The costs incurred by each group were aggregated and the average costs per patient calculated. For the calculation of the mean costs, if the total cost for a population can be extrapolated, parametric tests are appropriate. However, in acknowledgment of the non-normal distribution of the cost data, the median values were also calculated and compared using the Wilcoxon rank-sum test. There is debate regarding the appropriateness of parametric tests for cost data, and some investigators have suggested modeling the data before analysis. However, for practical situations, where the true distribution is unknown, the sample mean remains the estimator of choice, particularly for a small sample size.
To assess the robustness of the results, sensitivity analyses on the total costs, training costs, costs of follow-up fetal echocardiograms, and primary and improved secondary detection rates were performed. To account for variations in fuel prices, aircraft types, and transport team composition, a sensitivity analysis using ±20% of the costs was performed. We applied a high and low estimate of the UT training costs. Although we methodologically limited our analysis to the costs derived from our primary data set, which accrued solely by emergency transport, we assessed the effect of 1 or 2 follow-up fetal echocardiograms for patients with a prenatal diagnosis. The reported prenatal detection rates have varied widely. To account for the complete spectrum of possible prenatal detection rates, the sensitivity analysis included primary and secondary detection rates between 0 and 1 in intervals of 0.1. Statistical analysis used SAS, version 9.1 (SAS Institute, Cary, North Carolina), and p <0.05 was considered significant.
The cost/benefit analysis was conducted from a payer’s perspective to study the implications of UT training and improved detection rates on future incremental costs of postnatal CHD diagnosis. A 1- and 5-year prospective period was evaluated assuming the technology would be mostly unchanged throughout these periods. The total savings in the future was not discounted but calculated to today’s present value assuming that the time–value of money would offset the discount factor. The calculations were based on 2009 costs without adjustment for future inflation. The cost/benefit analysis was adjusted for the average population growth and birth rate in California (population growth 0.008%/year, extrapolating 15.1 births/1,000) and Nevada (population growth 0.024%/year, extrapolating 15.2 births/1,000) from 2006 to 2008. A full day UT program at the 2010 American Institute of Ultrasound in Medicine convention in San Diego cost $300.00–$600.00 per person. A similar course at New York University costs $150 for pediatric cardiologists and $75 for UTs. As such, we adopted average 1-day training costs of $500 per person in our analysis and $250 and $750 for the sensitivity analysis. In 2006, there were 1,650 UTs employed in Northern California and North Western Nevada. Our analysis considered the cost of training of 1,815 UTs, an additional 10% of the 2006 number assuming 5% population growth and a 5% cushion to mitigate against a conservative estimate. Finally, we determined the minimally required detection rate improvement for a positive net benefit after 1 year and calculated the net benefit at 5 years for a broad range (10% to 100% in steps of 10%) of new detection rates after training. We considered improvements of 19.4% to 50% in the detection rates.
Results
The study population is shown in Figure 1 . The CHD diagnostic categories for the prenatally and postnatally diagnosed groups and their requirement for transport are shown in Figure 2 . A total of 45 infants (30.6%) with a prenatal diagnosis and 102 (69.4%) with a postnatal diagnosis were studied. In aggregate, infants with a prenatal diagnosis were transported 243 miles (mean 5.4) at a cost of $17,505.00 (mean US$389/infant). Infants diagnosed postnatally were transported a total of 7,181 miles (mean 70.4) at a cost of $524,638.00 (mean US$5,143.51/infant).
Infants with a postnatal diagnosis were more likely to require emergency transport (relative risk 16.5, 95% confidence interval 4.25 to 64.44; p <0.001). The mean distances and costs, including a ±20% sensitivity analysis for costs and follow-up fetal echocardiograms, are listed in Table 2 . The mean emergency transport costs and distances traveled were significantly different between the groups ($389 vs $5,143.51; p <0.0001, t test, and p <0.0001, Wilcoxon test). The emergency transportation costs for postnatally diagnosed infants were 13.2 times greater than the costs for the prenatally diagnosed group. The total emergency transportation costs for all study subjects, in the 1-year study period, were US$542,143 (mean US$3,688.05/infant).
| Resource per Patient | Prenatal Diagnosis (n = 45) | Postnatal Diagnosis (n = 102) | Mean Incremental Difference | p Value |
|---|---|---|---|---|
| Emergency transfer distance (miles) | 5.40 ± 25 | 70 ± 74 | 65 | <0.0001 |
| Costs (US$) | 389 ± 1,824 | 5,144 ± 4,602 | 4,755 | <0.0001 |
| Sensitivity analysis of costs | ||||
| +20% | 467 ± 2,189 | 6,172 ± 5,522 | 5,705 | <0.0001 |
| −20% | 311 ± 1,459 | 4,115 ± 3,682 | 3,804 | <0.0001 |
| Including 1 follow-up echocardiogram | 4,031 ± 1,824 | 5,144 ± 4,602 | 1,113 | 0.12; 0.59 |
| Including 2 follow-up echocardiograms | 7,673 ± 1,824 | 5,144 ± 4,603 | −2,530 | 0.0005; 0.09 |
The effect of the varying detection rates along a continuum from 10% to 100% on the predicted total 1-year costs in Northern California and North Western Nevada are shown in Figure 3 . The costs of prenatal detection increased by a relatively small degree with an increasing detection rate but with a greater decrease in the transport costs with prenatal detection.
In 2004 to 2005, when the study cohort was born, there were an estimated 220,000 births in Northern California and North Western Nevada. Using our cohort of 147 infants, we obtained a 0.68/1000 birth rate with major CHD (likely underestimated secondary to excluding common CHD types, estimates of birth rates, and possible referral of infants to other centers). In 2009, the population of Northern California and North Western Nevada was estimated at 14.9 million. The 5-year prospective cost/benefit analysis accounted for population growth and birthrate (although not migration such as might occur in agricultural areas) and the incidence rate of CHD requiring intervention from our study sample (0.68/1,000 births). The cost/benefit analysis results are shown in Figure 4 . In the worst-case scenario (with no follow-up echocardiograms), assuming an improvement in prenatal detection from 30.6% to 40% and $750.00 training costs for 1,815 UTs ($750.00 × 1,815 UT = $1,361,250.00), would result in a cost benefit within 1 year and would save US$21.3 million within 5 years. Assuming a best-case scenario with high detection rates of 70%, as reported previously, and using $250 training costs, the cost benefit would be positive within 1 year and would be US$94.5 million within 5 years.
Because prenatal detection usually leads to a fetal echocardiogram, as well as a variable number of follow-up fetal echocardiograms, we assessed their effect on the cost/benefit analysis ( Table 3 ). The results are also given for the entire spectrum of prenatal detection from 10% to 100%, stratified by the number of fetal echocardiograms and training costs ( Table 4 ). According to our study institutions’ practice, prenatal detection led to a fetal echocardiogram with 1 follow-up echocardiogram. This would affect the net benefit after 1 and 5 years as follows: with training costs of $500, the net benefit would be positive after 1 year, ranging from US$4.4 and 38.2 within 5 years. For training costs of $250 and $750, the net benefit would be positive after 1 year, with savings at 5 years of ≤US$37.8 million. For 2 follow-up fetal echocardiograms, the net benefit becomes positive only after 1 year at an improved detection rate of 71% for training costs of $250 and 75% for training costs of $750 ( Table 3 ). In this optimistic scenario of high prenatal detection rates, after 5 years, the investment of UT training at $250 would be cost beneficial at an improved detection rate of 70%, yielding savings of US$0.3 million, and at an improved detection rate of 80%, savings of US$12.22 million ( Table 4 ).
| Training Costs | Follow-Up Echocardiograms for Prenatally Diagnosed Patients | ||
|---|---|---|---|
| None | One | Two | |
| $250 | 0.32 | 0.35 | 0.71 |
| $500 | 0.32 | 0.39 | 0.73 |
| $750 | 0.33 | 0.43 | 0.75 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
