Background
Although pediatric appropriate use criteria (AUC) for outpatient transthoracic echocardiography (TTE) are available, little is known about TTE utilization patterns before their release. The aims of this study were to determine the relation between AUC and TTE utilization and to identify patient and physician factors associated with discordance between the AUC and clinical practice.
Methods
A retrospective review of 3,000 initial outpatient pediatric cardiology encounters at six centers was performed. Investigator-determined indications were classified using AUC definitions. Concordance between AUC and TTE utilization was determined. Multivariate analysis was performed to identify patient and physician factors associated with TTE’s being performed for rarely appropriate and TTE’s not being performed for appropriate indications.
Results
Concordance between AUC and TTE utilization was 88%. TTE was performed for rarely appropriate indications in 9% and was associated with patient age < 3 months, indications of murmur, noninvasive imaging physician subspecialty, and physician volume. No TTE was ordered for appropriate indications in 3% and was associated with indications including prior test result (primarily abnormal electrocardiographic findings), older patients, and physician subspecialty other than generalist or imaging. There was high variability in TTE utilization among centers.
Conclusions
There was a reasonable degree of concordance between AUC and clinical practice before AUC publication. Several patient and physician factors were associated with discordance with the AUC. These findings should be considered in efforts to disseminate the AUC and in the development of future iterations. The causes for variation among centers deserve further exploration.
Highlights
- •
Little is known about baseline TTE utilization patterns and how well they align with the pediatric AUC.
- •
Concordance between pediatric AUC and TTE ordering was 88%.
- •
Among discordant ordering, the highest rates of TTE for rarely appropriate indications were in patients less than 3 months of age. An abnormal prior test result (primary abnormal electrocardiographic findings) was the indication most strongly associated with not ordering TTE for appropriate indications.
There was high variability in concordance between AUC and TTE utilization between the different participating centers.
Transthoracic echocardiography (TTE) use is widespread in the evaluation of pediatric patients for suspected heart disease. Although patient risk with TTE is low, it is relatively expensive, and ambiguous findings may lead to additional testing. Given the current emphasis on providing high value health care, appropriate use criteria (AUC) for TTE in outpatient pediatric cardiology were developed.
The development of the AUC used a modified Delphi method, developed by the RAND Corporation, which incorporated a writing group, a review panel, and an independent rating panel. Sparse data on outpatient pediatric TTE utilization were available to inform this process. Prior reports on outpatient TTE use have generally been limited to a single-center or a specific presenting symptom, such as murmur, chest pain, or syncope. Rates of outpatient TTE utilization according to AUC indication, and the patient and physician factors that influence TTE use, remain incompletely defined. Knowledge of these data would provide the foundation for future initiatives to refine the AUC and improve concordance with TTE utilization.
We therefore sought to determine TTE utilization rates for initial outpatient evaluation of pediatric patients and to assess patient and physician factors associated with variation from the AUC to (1) provide data on concordance between baseline physician practice and AUC before their release, (2) explore discordance between practice and AUC to identify areas in which the AUC may be refined or in which adherence to AUC may be improved, and (3) provide baseline data to serve as a comparison for investigations of TTE ordering after AUC release.
Methods
A multicenter retrospective cohort study was performed at six pediatric cardiology centers in the United States. The institutional review board at each center approved the investigation with waiver of the requirement to obtain informed consent. Data use agreements were obtained for each site. Children’s Hospital at Montefiore served as the core site. Patient selection was performed for first-time pediatric cardiology clinic visits in patients ≤18 years of age in encounters before September 29, 2014 (the date of online publication of the AUC). Site investigators started at this date and reviewed records in reverse chronologic order until a target of 500 from each site was achieved. All included patient visits were from within 1 year before AUC release. Patients were excluded if there was a history of preexisting structural, functional, or primary electrical cardiac disease (except for indications of abnormal electrocardiographic findings, prolonged QT interval, or ventricular preexcitation). Patients were also excluded if they were seen by a site investigator or if TTE was ordered at the visit but the results were unavailable. Data were entered by each site investigator into a Web-based Research Electronic Data Capture database hosted at the core site.
Patient data included age, sex, race, insurance status, and type and practice setting (hospital-based clinic vs community). Each consultation was mapped by the site investigator to one of the 113 AUC indications listed in the document after review of clinical notes. Appropriateness ratings corresponding to the AUC indication were assigned as appropriate, rarely appropriate, or may be appropriate. Some indications were not included in the AUC document and were considered unclassifiable. If a patient had more than one AUC indication, the one with the higher appropriateness rating was chosen. Because of the potential subjective nature of retrospectively assigning AUC indications, guidelines for determination of indications, on the basis of definitions from AUC document, were provided to site investigators. Any ambiguity was resolved through adjudication by the first and senior author.
Physician data included age, sex, years in practice since completion of fellowship, area of subspecialization, and number of outpatients seen per year. Physician sex and age were not collected at one site, as the data were considered potentially identifiable.
If TTE was ordered, the results were recorded as normal, normal with a patent foramen ovale, incidental finding, or pathologic finding. Pathologic findings were categorized by severity as per previously published criteria : minor: findings that may require follow-up, but no intervention is anticipated; moderate: findings that alter patient management but do not require urgent intervention; severe: findings that require urgent hospitalization or intervention.
Concordance with AUC was defined as ordering TTE for an indication rated appropriate and not ordering TTE for an indication rated rarely appropriate. Discordance was defined as not ordering TTE for an appropriate indication and ordering TTE for a rarely appropriate indication. Visits with AUC ratings of may be appropriate, or those that were unclassifiable, were excluded when determining concordance or discordance.
Statistical Analysis
Descriptive analysis was performed on patient and physician variables. Bivariate associations of all variables were performed for overall concordance versus discordance, TTE for rarely appropriate versus no TTE for rarely appropriate, and no TTE for appropriate versus TTE for appropriate by logistic regression with generalized estimating equations to account for the correlation among patients under the care of the same physician. Because of the different distribution of AUC indication categories in those with rarely appropriate versus those with appropriate indications, different categorical groupings were used for the separate analyses. Groupings were arranged as follows: for concordance versus discordance: palpitations/arrhythmia/syncope, murmur, chest pain/systemic disorders, prior test results, and others; for TTE for rarely appropriate versus no TTE for rarely appropriate: palpitations/arrhythmia, syncope, chest pain, murmur, and others; and for no TTE for appropriate versus TTE for appropriate: murmur, chest pain, prior test results, systemic disorders, and others. Multivariate logistic regression with generalized estimating equations was conducted to identify factors associated with TTE for rarely appropriate and no TTE for appropriate. A backward-stepwise approach with likelihood ratio tests was considered for variable selection from the full model with all potential predictors in multivariate analysis. Physician age and sex were excluded from bivariate and multivariate analysis, as they were not available from all centers. The bivariate association of a pathologic finding by TTE with AUC indication was assessed by multinomial logistic regression with generalized estimating equations. P values < .05 were considered to indicate statistical significance. Data were analyzed using SAS version 9.4 (SAS Institute, Cary, NC).
The sample size justification was based on comparison of rates of no TTE for appropriate and TTE for rarely appropriate between physicians with different experience levels. To detect a difference in rates of no TTE for appropriate between a presumed 3% in physicians with <5 years’ experience and 10% in those with >10 years’ at a power of 80% and a significance level of .05, 972 patients were needed. The same number was needed to detect a difference in rates of TTE for rarely appropriate between a presumed 20% in physician with <5 years’ experience and 10% in those with >10 years’ experience. With 972 patients each with indications of rarely appropriate and appropriate, and assuming a 20% rate of may be appropriate, the total number of patients required was 2,430 (1,944 from rarely appropriate plus appropriate and 486 from may be appropriate).
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
