| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, Health Outcomes, Policy and Economics (HOPE) Center,
Brain Institute, Miami Children's Hospital, 3100 S.W. 62 Ave., Miami, FL
33155.
2 Department of Radiology, Children's Hospital Medical Center, Cincinnati,
OH.
3 Department of Neurosurgery, Children's Hospital Medical Center, Cincinnati,
OH.
Received November 9, 2001;
accepted after revision January 2, 2002.
Address correspondence to L. S. Medina.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. A decision-analytic and cost-effectiveness model was constructed to compare three evaluation in strategies in children with suspected synostosis: no imaging, radiography (if abnormal, followed by three-dimensional CT [3D CT]), and 3D CT. Three risk groups were analyzed on the basis of the prevalence (pretest probability) of disease: low (completly healthy children; prevalence, 34/100,000), intermediate (healthy children with head deformity; prevalence, 1/115), and high risk (children with syndromic craniofacial disorders [i.e., Crouzon's syndrome or Apert's syndrome]; prevalence, 9-10/10). Test performance (sensitivity and specificity) of the evaluation strategies was obtained from the literature. Costs (not charge) estimates were obtained from the hospital cost-accounting database and from the Medicaid fee schedule.
RESULTS. In the low-risk group, the radiographic and 3D CT strategies resulted in a cost per quality-adjusted life year (QALY) gained of more than $560,000. In the intermediate-risk group, the radiographic strategy resulted in a cost per QALY gained of $54,600. Three-dimensional CT was more effective than the two other strategies but at a higher cost—hence, with a cost per QALY gained of $374,200. In the high-risk group, 3D CT was the most effective strategy with a cost per QALY gained of $33,800. Less experienced radiologists and poor-quality studies increased the evaluation cost per QALY gained for all of the risk groups because of decreased effectiveness.
CONCLUSION. Radiologic screening of completely healthy children (low risk) for synostosis is not warranted because of the high cost per QALY gained of the radiographic and 3D CT strategies. In healthy children with head deformity (intermediate risk), the radiographic strategy had a reasonable cost per QALY gained. Three-dimensional CT was more effective but had a high cost per QALY gained. In children with syndromic craniofacial disorders (high risk), 3D CT was the most effective strategy and had a reasonable cost per QALY gained. Selection of children with suspected craniosynostosis based on their risk group and use of the most appropriate evaluation strategy could maximize clinical and economic outcomes for these patients.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In children with Crouzon's, Apert's, or Pfeiffer's syndromes, synostosis is almost universally present [1,2,3,4]. The prevalence of synostosis in the general population ranges from 34 to 48 per 100,000 live births [5, 6]. In the general population, syndromic cases of synostosis are less common than nonsyndromic cases [1,2,3,4]. Sagittal followed by coronal synostosis is the most frequent type, accounting for 56% and 22% of the cases, respectively [6].
Since 1992, there has been an exponential increase in the number of infants seen with deformational posterior plagiocephaly (positional molding) [7]. The most likely explanations are the 1992 recommendation that infants sleep in the supine position to decrease the risk of sudden infant death syndrome and the increased awareness among pediatricians and other primary care providers of plagiocephaly [8,9,10,11]. With this new epidemic of deformational posterior plagiocephaly, appropriate management of posterior plagiocephaly requires differentiation of occipitoparietal flattening caused by lambdoid synostosis from that caused by deformational forces. In a 2.5-year prospective study of 115 infants presenting with unilateral posterior cranial flattening, only one child had synostotic posterior plagiocephaly (lambdoid synostosis), whereas 114 infants had deformation posterior plagiocephaly [7].
Craniosynostosis is usually suspected on the basis of the clinical findings [3]. However, skull radiographs, CT, or a combination of both, are necessary for diagnostic confirmation [3]. Several studies have shown a higher diagnostic sensitivity and specificity with three-dimensional CT (3D CT) than with skull radiographs in children with suspected craniosynostosis [12,13,14,15,16,17,18] (Table 1). However, the cost of CT is significantly higher than that of skull radiography (Table 2). Therefore, controversy exists on the most appropriate diagnostic imaging method for craniosynostosis when the important factors of diagnostic performance (sensitivity and specificity) and cost are considered.
|
|
Because craniosynostosis is a common pediatric disorder and early diagnosis and treatment are important, the role and cost-effectiveness of imaging in different synostosis risk groups need to be determined. A decision-analytic and cost-effectiveness model was constructed to compare three evaluation strategies in children with suspected synostosis: no imaging, radiography (if abnormal, followed by 3D CT), and 3D CT. Three risk groups were analyzed on the basis of the prevalence (pretest probability) of disease: low (completly healthy children; prevalence, 34/100,000), intermediate (healthy children with plagiocephaly; prevalence, 1/115), and high risk (children with syndromic craniofacial disorders [i.e., Crouzon's, Apert's, or Pfeiffer's syndrome]; prevalence, 9-10/10).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A decision-analytic and cost-effectiveness model was constructed to calculate the diagnostic yield and costs of three diagnostic strategies in children with suspected craniosynostosis. The three evaluation strategies included no imaging, radiography (if abnormal, followed by 3D CT), and 3D CT (Fig. 1). Children were divided into low-, intermediate-, and high-risk groups on the basis of clinical findings and on the estimated probability (pretest probability) of their having craniosynostosis. The model was constructed by using the computer program DATA (TreeAge Software, Williamstown, MA).
|
The cost-effectiveness analysis was performed from a societal perspective. Costs were expressed in 1999 United States dollars. Incremental cost-effectiveness ratios were calculated as the additional cost of one strategy divided by the additional effectiveness compared with the next most effective strategy [20]. According to the Panel on Cost-Effectiveness in Health and Medicine (U. S. Public Health Service), when the strategy under study is both more effective and less costly than the alternative, it is said to dominate the alternative strategy [20]. When the strategy under study has a lower incremental cost-effectiveness ratio and a greater effectiveness, it is said to have extended dominance over the alternative strategy [20]. Cost-effectiveness ratios were expressed as cost per QALY gained.
Data
Data sources from the published medical literature were used. These data
sources included medical publications from 1960 to 1999 reporting on
cranio-synostosis or craniostenosis. Model baseline values and ranges are
summarized in Table 1.
Prior Probability of Craniosynostosis
Pretest probability (prevalence) estimates of craniosynostosis were derived
from the medical literature (Table
1). Three risk groups were determined. Low-risk children were
healthy individuals with no craniofacial deformity or known underlying
syndrome (completely healthy children)
[5,
6,
23]. For this group, the
estimated prevalence (pretest probability) of craniosynostosis is 34 per
100,000 live births [5,
6]. Intermediate-risk patients
were healthy children with plagiocephaly with an estimated prevalence (pretest
probability) of craniosynostosis of one of 115
[7]. High-risk children were
those with syndromic craniofacial disorders such as Crouzon's, Apert's, or
Pfeiffer's syndrome
[1,2,3,4].
The pretest probability (prevalence) of craniosynostosis among high-risk
patients has been estimated at 9-10 of 10
[1,2,3,4].
Diagnostic Tests
Several studies have shown that 3D CT has a higher diagnostic performance
(sensitivity and specificity) than that of radiography
[12,13,14].
Sensitivity and specificity of 3D CT have been estimated at 96.4% and 100%,
respectively [14]. Sensitivity
and specificity of radiography have been estimated at 80% and 95%,
respectively [13]. Diagnostic
performance ranges are shown in Table
1. Pilgram et al.
[12] have shown a significant
decrease in the sensitivity and specificity of 3D CT and radiography if the
studies are of poor quality (Table
1). In addition, Vannier et al.
[14] have shown differences in
the diagnostic performance of 3D CT between experienced and less experienced
reviewers (Table 1). Their
study showed a decreased specificity of 3D CT (83.3%) among less experienced
reviewers [14].
Outcomes
Outcome was based on QALY gained by each of the three evaluation strategies
(Fig. 1). Bayes theorem was
used to calculate the final outcome on the basis of the sensitivity and
specificity of the different diagnostic tests
[22]
(Table 1). Cost-effectiveness
ratios were based on the cost per QALY saved.
The base-case patient was an infant without co-existent disease. The probability of dying of causes other than complications from craniosynostosis was obtained from the United States vital statistic data at the National Center for Health Statistics [24]. To determine the effect of imaging on outcome, the following parameters were incorporated into the model: the impact of delayed diagnosis of craniosynostosis in the quality of life [25], perioperative mortality rate of true-positive and false-positive imaging results for craniosynostosis, and mortality rate of imaging (CT) sedation (Table 1). We projected long-term outcomes such as QALYs gained using a conservative horizon of 20 years.
The quality-of-life weight (utility) was obtained from the treating physician's perspective [25]. A utility can vary from zero, representing death, to one, representing perfect health. The quality weight (utility) for a year of life spent in a more favorable state (e.g., early diagnosis and treatment of craniosynostosis) versus a more unfavorable state (e.g.,. delayed diagnosis and treatment of craniosynostosis) was incorporated into the Markov model. Patients with undiagnosed craniosynostosis were assumed to have a utility of 0.95.
Costs
The cost of all imaging studies was estimated using the 1999 Ohio Medicaid
fee schedule. In addition, direct and indirect costs were estimated from the
Cincinnati Children's Hospital Medical Center cost accounting system. The cost
of the imaging studies included the technical and professional fee. A summary
of the cost data for the different diagnostic studies is shown in
Table 2.
Sensitivity Analysis
Sensitivity analyses were performed to explore the effects of varying the
model estimates over clinically plausible ranges and to determine the
robustness of the results [20]
(Table 1). The effect of these
changes on outcomes, costs, and cost-effectiveness ratios was analyzed.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
In the low-risk group, both the radiographic and CT strategies had cost per QALY gained of more than $560,000. In the intermediate-risk group, the radiographic strategy had a cost per QALY gained of $54,600. Three-dimensional CT was more effective than the two other strategies but at a higher cost, hence, with a high marginal cost per QALY gained of $374,200. In the high-risk group, 3D CT was the most effective strategy with a cost per QALY gained of $33,800.
Sensitivity Analysis
Two-way sensitivity analysis for the variable pretest probability of
craniosynostosis is shown in Figure
2. This analysis showed a significant decrease in the cost per
QALY gained for 3D CT and radiography as the pretest probability (prevalence)
for craniosynostosis increased. The decrease in the cost per QALY gained was
most marked as the study population moved from the intermediate- to high-risk
group. The decrease in the cost per QALY gained was more marked for 3D CT than
that for radiography because of the higher cost of the former.
|
Sensitivity analysis for the diagnostic performance (sensitivity and specificity) of radiography and 3D CT in the intermediate group is shown in Table 4. As the sensitivity of radiography decreases (i.e., poor quality study, Table 1), the cost per QALY gained of radiography increases. As the sensitivity of radiography increases, the cost per QALY gained of radiography decreases, and the cost per QALY gained of 3D CT increases. The higher the sensitivity of radiography, the steeper is the increase in the cost per QALY gained of 3D CT. A similar effect is seen with the specificity of radiography in the intermediate-risk group (Table 4).
|
Table 4 reveals that as the sensitivity of 3D CT decreases (i.e., poor-quality study or less experienced reviewer, Table 1), the cost per QALY gained of 3D CT increases. In addition, a decrease in the 3D CT sensitivity (Fig. 4) causes an increase in the conventional radiographic cost per QALY gained but to a lesser degree than the 3D CT cost per QALY gained, because the abnormal findings on conventional radiography are followed by 3D CT (Fig. 1). One-way sensitivity analysis shows that when the 3D CT specificity is equal to or less than 95%, radiography becomes the preferred strategy because it dominates 3D CT. However, the lower the 3D CT specificity, the higher is the radiographic cost per QALY gained because the abnormal findings on conventional radiography are followed by 3D CT (Fig. 1).
|
Two-way sensitivity analysis of the conventional radiographic and 3D CT costs for the intermediate risk group is shown in Figures 3 and 4, respectively. As the cost of radiography increases, the cost per QALY gained of radiography increases while the cost per QALY gained of 3D CT decreases. The break-even point (cost per QALY gained of radiography = cost per QALY gained of 3D CT) is reached when the cost of radiography reaches $194.
|
As the cost of 3D CT decreases, the cost per QALY gained of 3D CT decreases while the cost per QALY gained of radiography increases slightly (Fig. 4). The higher the cost of 3D CT, the steeper the increase in the cost per QALY gained of 3D CT becomes (Fig. 4). The break-even point (cost per QALY gained of 3D CT = cost per QALY gained of radiography) is reached when the cost of 3D CT is $51.
Sensitivity analysis for the quality weight (utility) was performed. For children with undiagnosed craniosynostosis, the utility range analyzed is between 0.8 and 0.97 (Fig. 5). As the utility decreases (worse quality of life because of undiagnosed synostosis), the cost per QALY gained decreases for the radiographic and 3D CT strategies. The decrease is more marked for 3D CT than for radiographs (Fig. 5).
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In the high-risk group of patients with syndromic craniofacial disorders, CT was the most effective strategy with a cost per QALY gained of $33,800. The rationale behind this result is that craniosynostosis has a high prevalence in the population with syndromic craniofacial disorder and that patients are better managed by going directly to 3D CT for final evaluation and surgical planning. Furthermore, 3D CT also allows evaluation of the brain parenchyma for other potential abnormalities. Although the incidence of associated hydrocephalus tends to be low at 4% [26] in patients with craniosynostosis, most incidences of associated hydrocephalus tend to affect children with syndromic craniofacial disorders such as Crouzon's and Pfeiffer's syndromes [26]. In addition, 3D CT can depict commonly associated facial and base of the skull anomalies seen in patients with syndromic craniofacial disorders [27].
In the intermediate-risk group, the radiographic strategy had a cost per QALY gained of $54,600. Three-dimensional CT was more effective than the two other strategies (Table 3) but at a higher cost, hence, with a high marginal cost per QALY gained of $374,200.
In the low-risk group for craniosynostosis, both the radiographic and 3D CT strategies had a cost per QALY gained of more than $565,000 (Table 3). Therefore, if we were to screen completely healthy children with radiography or 3D CT, these strategies not be warranted because of the high cost per QALY gained of these two diagnostic strategies.
Our studies also showed an impact of study quality and level of expertise of the reviewer in the cost per QALY gained of the different diagnostic strategies. Sensitivity analysis of the diagnostic sensitivity and specificity of radiography revealed that as the diagnostic performance parameters decrease (poor-quality study), the cost per QALY gained of radiography increases (Table 3). In addition, as the sensitivity of 3D CT decreases (i.e., poor-quality study or less experienced reviewer), the cost per QALY gained of CT increases significantly (Table 3). Several studies have shown an impact on diagnostic performance based on level of expertise. A study by deDombal [28] in patients with appendicitis revealed an increased in diagnostic performance if the operator was aided by a computer system. A study by Eng et al. [29] showed higher diagnostic performance in the diagnosis of emergency department cases by experienced radiologists when compared with those of emergency room staff.
Sensitivity analysis of the cost of the diagnostic strategies in the intermediate-risk group showed that as the cost of radiography increases, the cost per QALY gained of radiography increased while the cost per QALY gained of 3D CT decreases (Fig. 3). The break-even point (i.e., the cost per QALY gained of radiography = the cost per QALY gained of 3D CT) was reached when the cost of radiography was $194. This break-even point of radiography was relatively high compared with the Ohio Medicaid reimbursement of $38 and the Cincinnati Children's Hospital Medical Center cost accounting system total cost of $76.
Sensitivity analysis of the intermediate-risk group revealed that as the cost of 3D CT decreases, the cost per QALY gained of 3D CT decreases while the cost per QALY gained of radiography increases slightly (Fig. 4). The higher the cost of 3D CT, the steeper the increase in the cost per QALY gained of 3D CT becomes (Fig. 4). The break-even point (cost per QALY gained 3D CT = cost per QALY gained of radiography) is reached when the cost of 3D CT is $51. This break-even point for 3D CT is relatively low when compared with the Ohio Medicaid reimbursement of $261 and the Cincinnati Children's Hospital Medical Center cost accounting system total cost of $312.
The cost-effectiveness ratios allow comparisons with alternative health care programs and may assist in resource allocation decisions [16]. The high-risk group 3D CT and intermediate-risk group radiography cost-effectiveness ratio of $33,800 and $54,600 per QALY gained, respectively, compared favorably with other well-accepted diagnostic strategies. For example, annual mammography and breast examination for women 40-49 years [30] costs $62,000 per life year saved, annual cervical cancer screening for women beginning at age 21 years costs $50,000 per life year saved [30], and colonoscopy for colorectal cancer screening for people older than 40 years costs $90,000 per life year saved [30].
Future studies are clearly needed. Because the spectrum of children with calvarial deformity is so broad, other clinical risk groups should be studied and their pretest probability (prevalence) of craniosynostosis determined. Large prospective cohort studies in children with other well-defined craniosynostotic disorders are required for this purpose. Such studies might ultimately form the basis for valuable evidence-based practice guidelines for the large population of children evaluated for calvarial deformity each year in the United States and other countries.
In summary, radiologic screening of completely healthy children for craniosynostosis (low risk) is not warranted because of the high cost per QALY gained of the radiographic and 3D CT strategies. In healthy children with a head deformity (intermediate risk), the radiographic strategy had a reasonable cost per QALY gained. CT was more effective but had a high cost per QALY gained. In children with syndromic craniofacial disorders (high risk), 3D CT was the most effective strategy and had a reasonable cost per QALY gained. Selection of children with suspected craniosynostosis based on their risk group and use of the most appropriate evaluation strategy could maximize clinical and economic outcomes for these patients.
Acknowledgments
We thank Susan DeBusk for important secretarial assistance.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  H. J. Otero, F. J. Rybicki, D. Greenberg, and P. J. Neumann Twenty Years of Cost-effectiveness Analysis in Medical Imaging: Are We Improving? Radiology, December 1, 2008; 249(3): 917 - 925. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Griebsch, J. Coast, and J. Brown Quality-Adjusted Life-Years Lack Quality in Pediatric Care: A Critical Review of Published Cost-Utility Studies in Child Health Pediatrics, May 1, 2005; 115(5): e600 - e614. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Burstein Differential Diagnosis of the Trapezoid-Shaped Head AAP Grand Rounds, November 1, 2004; 12(5): 62 - 63. [Full Text] [PDF]
|
|
|
|
|
|
B. P. Wood Which Misshapen Head Needs a Workup? AAP Grand Rounds, October 1, 2002; 8(4): 43 - 44. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
= conventional radiography, [UNK] =
CT.
