 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
PEDIATRICS Vol. 108 No. 2 August 2001, pp. 255-263
, and
From the * Neuroradiology, and Health Outcomes and Policy
Section, Department of Radiology, Children's Hospital Medical Center,
Cincinnati, Ohio; Objective. To assess the clinical and
economic consequences of 3 diagnostic strategies Materials and Methods. A decision-analytic Markov model
and cost-effectiveness analysis was performed incorporating the risk
group prior probability, MRI and CT sensitivity and specificity, tumor
survival, progression rates, and cost per strategy. Outcomes were based
on quality-adjusted life year (QALY) gained and incremental cost per
QALY gained.
Results. For low-risk children with chronic nonmigraine
headaches of >6 months' duration as the sole symptom (prior
probability of brain tumor 0.01%), no neuroimaging with close clinical
follow-up was less costly and more effective than the 2 neuroimaging
strategies. For the intermediate-risk children with migraine headache
and normal neurologic examination (prior probability of brain tumor 0.4%), CT-MRI was the most effective strategy but cost >$1 million per QALY gained compared with no neuroimaging. For high-risk children with headache of <6 months' duration and other clinical predictors of
a brain tumor such as an abnormal neurologic examination (prior probability of brain tumor 4%), the most effective strategy was MRI,
with cost-effectiveness ratio of $113 800 per QALY gained compared
with no imaging.
Conclusion. Our analysis suggests that MRI maximizes QALY
gained at a reasonable cost-effectiveness ratio in children with
headache at high risk of having a brain tumor. Conversely, the strategy
of no imaging with close clinical follow-up is cost saving in low-risk children. Although the CT-MRI strategy maximizes QALY gained in the
intermediate-risk patients, its additional cost per QALY gained is
high. In children with headache, appropriate selection of patients and diagnostic strategy may maximize quality-adjusted life expectancy and decrease costs of medical workup.
Department of Health Policy and Management,
Harvard School of Public Health, Boston, Massachusetts; and § Division
of Neuro-Oncology, Department of Neurology, Children's Hospital
Medical School, Boston, Massachusetts.
ABSTRACT
Top
Abstract
MaterialsMethods
Results
Discussion
Conclusion
References
magnetic resonance
imaging (MRI), computed tomography followed by MRI for positive results
(CT-MRI), and no neuroimaging with close clinical follow-upin the
evaluation of children with headache suspected of having a brain tumor.
Three risk groups based on clinical variables were evaluated.
Headache is a very common symptom among children and
adults, accounting for 18 million (4%) of the total outpatient visits in the United States each year.1 It has been estimated that 11.3 million people nationwide experience moderate to severe migraine headaches.2 In the United States, adolescent boys
and girls have a headache prevalence of 56% and 74% and a migraine
prevalence of 3.8% and 6.6%, respectively.1 In
Scandinavia, 38% of 7-year-olds have headaches, and 1.4% to 3% have
migraine.3,4 The direct and indirect annual cost of
migraine has been estimated at >$5.6 billion.5
A primary concern in children with headache is the possibility of a
brain tumor.6,7 Although brain tumors constitute the
largest group of solid neoplasms in children and are second only to
leukemia in overall frequency of childhood cancers, the annual
incidence is low at 3 per 100 000.6 Several studies have
demonstrated improved survival in patients who have totally resectable
brain neoplasm and lower tumor stage compared with patients with more
advanced disease.8-11 Brain tumors also may produce mass
effect and bleed, increasing the risk of brain
herniation.12 Early detection may improve patient outcome
because late diagnosis of a brain tumor may affect brain tumor
resectability, neoplasm stage, and risk of cerebral herniation.
However, because the overall incidence of brain tumors among children
with headache is low, clinicians may be trapped between ordering
imaging studies with some inherent risks (sedation, contrast use, and
false-positive rate) and early detection of intracranial tumors.
The value of neuroimaging in evaluating headache in children is
controversial. Some retrospective studies have concluded that neuroimaging studies have a limited role in children with
headache.13,14 These studies have been limited by a sample
size that yielded no brain tumors. Another study suggests dividing
children with headache into high- and low-risk groups according to
clinical predictors of surgical space-occupying lesions.15
This study suggests that neuroimaging should be reserved for high-risk patients.15 Furthermore, if neuroimaging is indicated, controversy exists about the choice of computed tomography (CT) or
magnetic resonance imaging (MRI). CT traditionally has been the
imaging study of choice because of its availability and lower cost
per case.16 However, CT has disadvantages, including posterior fossa beam hardening artifacts and potential iodinated contrast reaction.17,18 MRI with its high soft-tissue
characterization, multiplanar capability, and lack of ionizing
radiation has emerged as the technically optimal imaging modality.19 However, its added cost and time, higher sedation rate, and limited availability has hampered its universal use
as the first imaging modality.16
Our objective was to assess the clinical role and economic consequences
of 3 diagnostic strategies: MRI, CT followed by MRI for positive
results (CT-MRI), and no imaging with close clinical follow-up in
children with headache suspected of having an intracranial tumor.
Specifically, we sought to assess the effectiveness (quality-adjusted life years [QALY] gained), lifetime costs, and cost-effectiveness of
these 3 diagnostic strategies in children with headache at low,
intermediate, and high risk of having a brain tumor.
Model
Cost-effectiveness analysis (CEA) is a method used to evaluate
the outcomes and costs of interventions designed to improve health.20-23 Primary and secondary (medical publications)
data are used to construct a comprehensive decision-analytic model. The
results of an analysis are summarized in a series of cost-effectiveness
ratios that show the cost of achieving 1 unit of health outcome (eg,
the cost per QALY gained) for different risk groups and
strategies.20-23
A decision-analytic Markov model was constructed to calculate QALY
gained and long-term costs of 3 diagnostic strategies in children with
headache suspected of having a brain tumor.20-23 Brain
tumors were defined as brain neoplasms and vascular malformations such
as arteriovenous malformations (Tables 1
and 2). The model was constructed by
using the computer program Decision Analysis by TreeAge ([DATA];
TreeAge Software, Williamstown, MA) to examine and compare the
diagnostic performance (sensitivity and specificity) of the imaging
tests, the complications associated with these tests, and the
probability of detecting an early brain tumor. The 3 strategies
included: brain MRI, CT-MRI, and no imaging with close clinical
follow-up. Patients with MRI findings consistent with a brain tumor
were assumed to have a brain biopsy. Figure
1 shows a simplified representation of
the decision tree.
TABLE 1 TABLE 2
MATERIALS AND METHODS
Top
Abstract
MaterialsMethods
Results
Discussion
Conclusion
References
Summary of Data Used in the Analysis
Summary of Brain Tumor Data Used in the Analysis
View larger version (10K):
[in a new window]
Fig. 1.
Three strategies for studying children with headache suspected of
having a brain tumor. BX indicates biopsy; pos, positive; neg,
negative.
Children with headache were divided into low-, intermediate- and high-risk groups based on clinical predictors and on their estimated probability of having a brain neoplasm (Table 1). Markov (state-transition) models24 were used to calculate the lifetime cost and QALY gained for the 3 risk groups.
The CEA was performed from a societal perspective. Costs were expressed in 1997 US dollars. Costs and QALY gained were discounted at 3% per year.21 Incremental cost-effectiveness ratios were calculated as the additional cost of a strategy divided by the additional effectiveness compared with the next most effective strategy.23 According to the Panel on Cost-Effectiveness in Health and Medicine (US Public Health Service),23 when the strategy under study is both more effective and less costly than the alternative, it is said to dominate the alternative strategy.23 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.23 Cost-effectiveness ratios were expressed as cost per QALY gained.
Data
Primary and secondary data sources were used. Primary information included a hospital-based database of 315 children with headache who were imaged with MRI and CT.15 Secondary data sources included medical publications from 1962 to 1997 reporting on headache and brain tumors. Sensitivity analyses were performed to explore the effects of varying the model estimates over clinically plausible ranges and determine the robustness of the results.23 Model baseline values and ranges are summarized in Table 1.
Prior Probability of Brain Tumor
Estimates of the prior probability of a brain tumor in children with headache were derived from the primary and secondary data sources (Table 1). Three risk populations were determined. Low-risk children were characterized as having had nonmigraine headaches for >6 months as the sole symptom and a normal neurologic examination.14,15 For this group, the estimated prevalence (prior probability) of a brain tumor is 0.01% (1/10 000).25,26 Intermediate-risk patients were characterized as having migraine headache and a normal neurologic examination with an estimated prevalence (prior probability) of brain tumor of 0.4% (4/1000).14,25 Migraine headaches were defined as recurrent headaches that are often pulsating, unilateral, and associated with symptoms such as nausea, light and sound sensitivity, and visual or neurologic symptoms.25 High-risk children were characterized by headaches for <6 months and at least 1 other clinical predictor of a surgical space-occupying lesion.15 These predictors included sleep-related headache, vomiting, confusion, absence of visual aura, absence of family history of migraine, and abnormal neurologic examination.15 The probability of a brain tumor among high-risk patients has been estimated at 4% (4/100).15 Ranges considered plausible for each risk group were studied with sensitivity analysis.
Diagnostic Tests
MRI and CT sensitivity for brain tumors has been estimated in the range of 82% to 100% and 65% to 100%, respectively.15,27,28 MRI and CT specificity has been estimated in the range of 81% to 100% and 72% to 100%, respectively.15,27 Because there are no large series assessing the joint diagnostic performance of MRI and CT, conditional independence between these tests was assumed for the case where a positive CT examination was followed by a MRI test (Fig 1).
The mortality and morbidity for the different diagnostic procedures were based on the use of contrast and the need for sedation (Table 1). Severe adverse reactions and death are higher with CT iodinated contrast than MRI gadolinium-based contrast.29-32 Because MRI tends to be a longer study than CT, sedation rates are higher with the former test. The potential risk of death from radiograph radiation was not included in the analysis of the CT strategy because the radiation dosage is relatively low and it is comparable to the risk of other normal activities in life.33
Long-Term Outcomes
The base-case patient was a 5-year-old child without coexistent disease. The probability of dying of causes other than brain tumor was obtained from the US vital statistic data (National Center for Health Statistics). Model parameters, including prevalence, stage distribution, 5-year survival data, and herniation risk of brain tumors were based on values obtained from the primary and secondary data sources used (Tables 1 and 2). We projected long-term outcomes such as life years (LYs) and QALY gained using a conservative horizon of 20 years.
In the pre-CT and pre-MRI era, the Childhood Brain Tumor Consortium study demonstrated 14% of children to have a 1-year delay in the diagnosis of brain tumor from the initiation of clinical findings.6 Five-year brain tumor survival data (Table 2) demonstrate differences between local versus metastatic medulloblastomas and total versus partial resection for pilocytic astrocytomas and ependymomas.8-1134-36 Therefore, for the 8 most common pediatric brain neoplastic types,34 we assumed that pilocytic astrocytoma, medulloblastoma, and ependymoma (60% of all tumors) can have a worse prognosis if diagnosed late (Table 2). For those tumors, based on the Brain Tumor Consortium results,6 we assumed a 14% probability of having a localized or resectable disease (favorable prognosis) progress to either metastatic or partially resectable disease (unfavorable prognosis) if not diagnosed at clinical presentation. To ensure a conservative model, the rest of the brain neoplasms were considered to have the same 5-year survival rate regardless of time of diagnosis (Table 2).8-11,35,36 Estimated population distributions of cancer stage at initial presentation were determined from the published literature (Table 2).8-11,35,36
Intracranial tumors may cause brain herniation because of mass effect (ie, large brain neoplasm) or bleeding (ie, hemorrhagic neoplasm or vascular malformation). The brain herniation mortality has been estimated to be between 0.2% and 1%.12
QALY gained was estimated using the Health Utilities Index, which is based on children with neoplasm (Table 1).37 The Health Utilities Index was used to determine the quality weight (utility) from the treating physician's perspective. A health-quality weight (utility) can range from 0 (death) to 1 (perfect health). The quality weight (utility) for a year of life spent in a favorable prognosis versus unfavorable prognosis tumor state was incorporated into the Markov model. Health utilities were subdivided into nonmedulloblastoma and medulloblastoma tumor populations. This was done because of the more aggressive and debilitating nature of the treatment and disease process in patients with medulloblastoma (Table 2).11 Quality-of-life weights and disease-specific mortalities were assumed to persist only for the first 5 years after diagnosis, after which time the child was assumed to be cured.
Costs
Costs (not charges) of the radiologic and nonradiologic (ie, hospitalization, surgery, chemotherapy, radiation therapy, laboratories, and pharmacy) procedures were estimated from the Boston Children's Hospital cost accounting system. Fixed, variable, and overhead costs were analyzed. Cost data for the different diagnostic studies, favorable versus unfavorable brain tumor treatment, sedation, and adverse contrast media reaction are summarized in Table 3
|
The cost of the radiologic studies included the technical and professional fee. Costs incurred by patients with favorable and unfavorable prognosis brain tumors were obtained. We used an estimate of $32 000 for the first-year cost of a favorable prognosis brain tumor and an annual cost of $4000 for subsequent years (Table 3). We used an estimate of $45 000 for the first-year cost of an unfavorable prognosis brain tumor and an annual cost of $8000 for subsequent years (Table 3). These costs included hospitalization, surgery, chemotherapy, radiation therapy, laboratories, pharmacy, imaging, and follow-up visits.
Sensitivity Analysis
One-way sensitivity analysis was performed for each parameter of the model over a clinically plausible range of values (Table 1). The effect of these changes on outcomes, costs, and cost-effectiveness ratios were analyzed.
| |
RESULTS |
|---|
|
Top
Abstract
MaterialsMethods
Results
Discussion
Conclusion
References
|
|---|
Baseline Analysis
Results for the low-,intermediate-, and high-risk groups (hypothetical cohorts) are shown in Table 4.
|
Low-Risk Children The most effective diagnostic strategy for the low-risk cohort was no imaging with close clinical follow-up. This strategy was also associated with the lowest cost when compared with CT and MRI. Given the very low prevalence (prior probability) of a brain tumor in this group of 0.01% (1/10 000), the risk from sedation, contrast, and false-positive test results plus the added costs of CT and MRI outweigh the benefit of detecting a rare brain tumor early.
Intermediate-Risk Children The most effective diagnostic strategy for children with headache in this group was CT-MRI. Given the low number of brain tumors in this group at 0.4% (4/1000), the CT-MRI strategy was more effective than MRI alone because of the fewer false-positive MRI test results with the former. The fewer false-positive MRI test results in the CT-MRI strategy resulted in less unnecessary brain biopsies, thereby decreasing morbidity, mortality, and cost from this invasive procedure. Although the CT-MRI strategy is the most effective strategy, its cost-effectiveness ratio is high, at >$1 million per QALY saved compared with no imaging.
High-Risk Children The most effective strategy for high-risk children was MRI. MRI was also less costly than the CT-MRI strategy, given the CT false-positive rate, which leads to further unnecessary workup with MRI. The cost-effectiveness ratio of MRI compared with no imaging was $113 800 per QALY gained.
Sensitivity Analysis
An increase in the number of predictors has been highly correlated with an increased risk of brain tumor.15 Assessment of this variable revealed that the cost-effectiveness ratio decreased from $113 800 to $67 000 per QALY saved when the pretest probability increased from 4% to 8%, respectively (Table 4). Lower cost-effectiveness ratios were noted as the increasing number of brain tumor predictors increases the prior probability (Table 5).
|
One-way sensitivity analyses for high-risk patients are shown in Table 5. An increase in the probability of having a favorable prognosis tumor progress to an unfavorable prognosis tumor if not diagnosed at clinical presentation decreases the cost-effectiveness ratio of MRI. The same applies to the percentage of death caused by brain herniation. As the percentage of patients' risk of dying from brain tumor-induced herniation increases, the MRI strategy cost-effectiveness ratio decreases.
MRI was ruled out by dominance or extended dominance if its specificity was lower than 97%. The greater the MRI specificity, the lower the cost-effectiveness ratios for the MRI strategy became. Likewise CT was no longer dominated if its specificity was greater than 97%. Results were not significantly sensitive to the sensitivity of MRI or CT.
The MRI diagnostic strategy in high-risk patients showed some sensitivity to changes in brain biopsy mortality. An increase in intracranial biopsy mortality resulted in a higher cost-effectiveness ratio for MRI. This was explained on the basis of decreased effectiveness and increased costs in patients with false-positive MRI results undergoing unnecessary brain biopsies with a higher mortality risk.
Extending the time horizon of the study 3 times the conservative assumption of 20 years showed a decrease in the cost-effectiveness ratio of MRI to $65 000 per QALY gained. When only LYs saved without quality adjustment was considered in the analysis for high-risk patients, MRI was still the preferred strategy but with a higher cost-effectiveness ratio (Table 5).
Sensitivity analysis of the cost variables revealed that MRI became more cost-effective (lower cost-effectiveness ratio) as its cost equaled the cost of doing a CT examination. Decreases in the favorable prognosis brain neoplasm costs and increases in the unfavorable neoplasm costs reduced the MRI strategy cost-effectiveness ratio. Sensitivity analysis of the imaging contrast and sedation variables revealed few changes in our base-case results.
| |
DISCUSSION |
|---|
|
Top
Abstract
MaterialsMethods
Results
Discussion
Conclusion
References
|
|---|
The value of neuroimaging in evaluating headache in the child suspected of having a brain tumor is controversial.6,7,14,15 Some studies conclude that neuroimaging studies have a very limited role in low-risk children with chronic headaches,14 whereas other studies emphasize the importance of imaging children with nonchronic headache and clinical predictors of a surgical brain tumor such as an abnormal neurologic examination.7,15 Changes in our approach to health care emphasize the need to consider long-term outcomes and costs in evaluating medical diagnostic strategies. Accordingly, we performed a decision analytic Markov model and CEA of 3 risk groups incorporating the quality-of-life adjustments and costs associated with early versus late brain tumor diagnosis and treatment.
The low-risk group with baseline pretest probability of 0.01% (1/10 000) consisted of patients with the sole finding of chronic headaches of >6 months' duration.25,26 The most effective strategy was no imaging with close clinical follow-up. Furthermore, this strategy was cost-saving when compared with the CT-MRI and MRI strategies because of its greater effectiveness at a lower cost. Therefore, the risks of false-positive neuroimaging results, contrast use, and sedation outweigh the benefit of early diagnosis of a rare brain tumor in this risk group (Table 4).
The intermediate-risk group with baseline pretest probability of 0.4% (4/1000) consisted of children with migraine headaches and a normal neurologic examination.14,25 CT-MRI was a slightly more effective diagnostic strategy than MRI alone. However, the cost-effectiveness ratio of the CT-MRI strategy was >$1 million per QALY saved.
The high-risk group with baseline prevalence (prior probability) of 4% consisted of children with headache of <6 months' duration and at least 1 other clinical predictor suggestive of a surgical brain tumor such as an abnormal neurologic examination.15 MRI maximized QALY gained compared with CT-MRI or no imaging in this risk group. An increase in the number of predictors has been highly correlated with an increased risk of brain tumor.15 Assessment of this variable revealed a decrease in the cost-effectiveness ratio from $113 800 to $67 000 per QALY saved when the prior probability was increased from 4% to 8% (Table 4).
The cost-effectiveness ratios allow comparisons with alternative health care programs and may assist in resource allocation decisions.23 The high-risk group cost-effectiveness ratios compared favorably with other well-accepted diagnostic procedures. For example, annual mammography for women 55 to 64 years old costs $110 000 per LY saved (updated to 1993 US dollars),38 annual cervical cancer screening for women beginning age 20 costs $220 000 per LY saved (updated to 1993 US dollars),38,39 and colonoscopy for colorectal cancer screening for people older than 40 years costs $90 000 per LY saved (updated to 1993 US dollars).38,40
In the high-risk group we found our analysis to be sensitive to the percentage of favorable prognosis tumors that progressed to unfavorable prognosis because of delayed diagnosis and treatment, and increased risk of death from brain herniation. That is, the incremental cost-effectiveness ratio for MRI was lower for patients who had a greater probability of progression because of delayed tumor diagnosis and treatment, and greater risk of dying from brain herniation (Table 5).
We also found in our high-risk group that the diagnostic performance of the neuroimaging tests affected our base case-results. MRI became a dominated strategy if its specificity was <97%. CT was no longer dominated if its specificity was greater than 97%. A decrease in the MRI test sensitivity revealed an increase in the cost-effectiveness ratio of the MRI strategy (Table 5). Therefore, the exact diagnostic performance (sensitivity and specificity) of MRI and CT for a specific institution or individual may affect the diagnostic strategy of choice and its cost-effectiveness.15,27,28
Results were somewhat sensitive to changes in the MRI to CT cost ratios. For example, if CT and MRI costs were equal, the MRI cost-effectiveness ratio for the high risk group would decrease to 88 000 per QALY gained (Table 5). The impact of the MRI to CT cost ratios on the cost-effectiveness ratio should be kept in mind in countries with different neuroimaging costs and equipment availability.16
In the high-risk group, our analyses were not very sensitive to the percentage of patients needing sedation or contrast (Table 5). This is explained by the low risk of radiologic sedation, MRI contrast, and CT nonionic contrast media in this risk group.29-32,41
These results are summarized in Fig 2 as a suggested decision tree for use in children with headache. The decision tree reinforces the primary importance of carefully acquiring a medical history and performing a thorough examination, including a detailed neurologic examination. For patients in the high-risk group, neuroimaging with MRI is suggested, whereas in the intermediate-risk group CT-MRI may be indicated. For children in the low-risk group or those with negative findings from neuroimaging studies, close clinical follow-up with periodic reassessment is recommended. In children with headache, appropriate selection of patients for imaging based on these risk groups may maximize QALY gained by reducing potential risks associated with false-positive results, use of contrast material and sedation, and decrease costs of medical workup.
|
Future studies are needed. First, our CEA demonstrated differences in the diagnostic strategy of choice and its cost-effectiveness ratio based on the diagnostic performance (sensitivity and specificity) of the tests. Prospective studies comparing the diagnostic performance and receiver operator characteristic curve of experienced and less experienced readers for CT, MRI, and CT-MRI are needed.
Second, because the spectrum of children with headache is so broad, other clinical risk groups must be studied and their prevalence (prior probability) of brain tumor determined. Large prospective cohort studies in children with other well-defined headache disorders are needed for this purpose. Such studies might ultimately form the basis for valuable evidence-based practice guidelines for the large population of children evaluated for headache each year in the United States and other countries.
| |
CONCLUSION |
|---|
|
Top
Abstract
MaterialsMethods
Results
Discussion
Conclusion
References
|
|---|
Our analysis suggests that MRI maximizes quality-adjusted life expectancy at a reasonable cost-effectiveness ratio in children with headache at high risk of having a brain tumor. Conversely, the strategy of no imaging with close clinical follow-up is cost saving in low-risk children. Although the CT-MRI strategy maximizes quality-adjusted life expectancy in the intermediate-risk patients, its additional cost per QALY gained is high. In children with headache, appropriate selection of patients and diagnostic strategy may maximize quality-adjusted life expectancy and decrease costs of medical work up.
| |
ACKNOWLEDGMENTS |
|---|
This work was supported in part by a research grant from the Society for Pediatric Radiology.
We thank William S. Ball, Jr, MD; Kerry R. Crone, MD; and Joel Tsevat, MD, MPH, for their valuable comments. We also thank Teri Haas and Kathleen Joiner for secretarial assistance.
| |
FOOTNOTES |
|---|
Received for publication May 10, 2000; accepted Dec 14, 2000.
Reprint requests to (L.S.M.) Department of Radiology, Miami Children's Hospital, 3100 SW 62nd Ave, Miami, FL 33155. E-mail: smedina{at}post.harvard.edu
| |
ABBREVIATIONS |
|---|
CT, computed tomography; MRI, magnetic resonance imaging; CT-MRI, CT followed by MRI for positive results; QALY, quality-adjusted life year; CEA, cost-effectiveness analysis; LY, life year.
| |
REFERENCES |
|---|
|
Top
Abstract
MaterialsMethods
Results
Discussion
Conclusion
References
|
|---|
This article has been cited by other articles:
 |
|
 |
|
  W. D. Graf, H. R. Kayyali, J. J. Alexander, S. D. Simon, and M. C. Morriss Neuroimaging-Use Trends in Nonacute Pediatric Headache Before and After Clinical Practice Parameters Pediatrics, October 6, 2008; (2008) peds.2008-1159v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. C. Reulecke, C. G. Erker, B. J. Fiedler, T.-U. Niederstadt, and G. Kurlemann Brain Tumors in Children: Initial Symptoms and Their Influence on the Time Span Between Symptom Onset and Diagnosis J Child Neurol, February 1, 2008; 23(2): 178 - 183. [Abstract] [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]
|
|
|
|
 |
|
 D. W. Lewis, S. Ashwal, G. Dahl, D. Dorbad, D. Hirtz, A. Prensky, and I. Jarjour Practice parameter: Evaluation of children and adolescents with recurrent headaches: Report of the Quality Standards Subcommittee of the American Academy of Neurology and the Practice Committee of the Child Neurology Society Neurology, August 27, 2002; 59(4): 490 - 498. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2001 American Academy of Pediatrics. All rights reserved. |
||
|
||
