ARTICLE |
Division of Pediatric Cardiology and Perinatal Epidemiology Research Unit, Department of Pediatrics, IWK Health Centre, Dalhousie University, Halifax, Nova Scotia, Canada
| ABSTRACT |
|---|
|
|
|---|
METHODS. Two questionnaires were developed and mailed to a sample of noncardiologist physicians and pediatric cardiologists in Canada. The difference in the proportion of noncardiologist-responders who performed a "full screen" before and after the announcement of the Health Canada advisory was analyzed. Willingness to prescribe stimulant medications to children was assessed by using a series of clinical scenarios.
RESULTS. A total of 2326 questionnaires were distributed, 717 (31%) of which were returned. The proportion who performed a full screen increased for both noncardiologists and cardiologists after the advisory. There was also a dramatic increase in the use of the "modified screen" by the noncardiologist group and the pediatrician subgroup. There was a considerable decrease in the proportion of noncardiologists who were willing to prescribe stimulant medications to children with potential or actual cardiac issues after the advisory.
CONCLUSIONS. The Health Canada advisory has had an impact on clinical practice. It has resulted in an increase in the use of cardiovascular assessment tools before starting stimulants and a decrease in noncardiologists' willingness to prescribe stimulants to those with real or potential cardiac disease. These changes have occurred despite the lack of studies to address the cardiac risks of these medications. Because it will be difficult to assess the true cardiac risks of stimulants, consensus recommendations are needed to determine whether screening is required, what constitutes an appropriate screen, and which children can be treated cautiously with stimulant medications.
Key Words: ADHD cardiovascular screening congenital heart disease sudden death Health Canada
Abbreviations: ADHD—attention-deficit/hyperactivity disorder ECG—electrocardiogram FDA—Food and Drug Administration CHD—congenital heart disease SVT—supraventricular tachycardia PVC—premature ventricular contraction AHA—American Heart Association
Attention-deficit/hyperactivity disorder (ADHD) is the most common neurobehavioral disorder of childhood and is managed with behavioral interventions and pharmacologic agents.1,2 Stimulant medications, specifically methylphenidates and dextroamphetamines, are frequently prescribed and are effective in reducing the core symptoms of ADHD.2 Despite their efficacy and long history of use, the safety of these medications has recently come into question as the result of isolated reports of sudden death in children undergoing treatment. In May 2006, Health Canada released a statement advising health care professionals to be aware of cardiac risk factors before initiating therapy and to avoid stimulant use for all patients with "symptomatic cardiac disease" or "known structural cardiac abnormalities"3; however, the advisory lacked any specific recommendations or guidelines for practitioners, leaving much room for interpretation and variations in clinical practice. Although the advisory has sparked much debate about the true cardiac risks of stimulants and the justification for the advisory,4–8 it remains unchanged. To date, it is unclear whether physicians in practice are aware of the Health Canada advisory and whether the advisory has had an impact on their clinical practice.
The primary objective of this study was to assess the impact of the Health Canada advisory on physicians' cardiovascular assessment of all children with ADHD before starting stimulant medications. This study also sought to determine the impact of the advisory on the treatment of children with potential or real cardiac disease.
| METHODS |
|---|
|
|
|---|
The questionnaire had a total of 19 questions and was composed of 3 sections. The first section focused on the practitioner's experience with children with ADHD and/or congenital heart disease (CHD), his or her familiarity with the Health Canada Advisory, and his or her approach to the cardiovascular assessment of a patient with ADHD before and after the advisory. For the purpose of the survey, a "full screen" was considered to include all 6 assessment tools listed next. This is with the understanding that there is no validated screen studied or published in this population for assessing the risk for sudden death, and the use of the term "full screen" in our study is merely a term chosen to allow for comparison with the other combination of assessment tools, the "modified screen":
A modified screen differed from the full screen only in that the ECG was performed selectively for children with an abnormal examination. Although these 6 assessment tools do not constitute a comprehensive cardiovascular risk assessment, they were included because they have been deemed essential components of risk assessment for other conditions for which sudden death is a potential complication or outcome9–12 and represent important elements of history taking, physical examination, and basic investigations. The second section was composed of a series of clinical scenarios describing children with CHD of varying severity, symptoms, and family history. Questions in this section were designed to assess prescribing practices before and after the Health Canada advisory. The third section covered general demographic information. Although multiple choice and yes/no questions were used throughout the questionnaire, each practitioner had the opportunity to expand on his or her answers and provide free-text commentary. Each questionnaire was numbered to identify nonresponders. A copy of the survey is available on request.
Survey Population
For detection of a minimum 10% difference in the proportion of noncardiologists who performed a full screen before and after the Health Canada advisory with an
of .05 and a power of 80%, a total of 660 respondents were calculated to be necessary. On the basis of an expected questionnaire return rate of 30%, a minimum of 2200 questionnaires were calculated to be required. Although the expected return rate used to perform the power calculation was low, this number is in accordance with the response rate of other physician surveys.13–18
The survey groups were selected primarily to represent the types of practitioners who are likely to assess and treat children with ADHD. The pediatric cardiologists were included as a comparative group to help define the current state of practice in assessing children with ADHD.
Those selected for the mailed survey included all family practitioners and pediatricians throughout the provinces of Nova Scotia, New Brunswick, and Prince Edward Island and all pediatric cardiologists, child psychiatrists, pediatric neurologists, and pediatric developmental specialists across Canada. A total of 2243 noncardiologist and 83 pediatric cardiologists were identified through their respective colleges, subspecialty organizations, and Internet searches. A total of 1763 were family practitioners, and 480 were general pediatricians (n = 83) or pediatric subspecialists (n = 397).
The surveys were mailed in January 2007 and March 2007. To improve response rate, 2 separate mailings occurred; a self-addressed, stamped envelope was provided; and a short letter summarizing the importance of the project was included.19 Before the second mailing, efforts were made to locate physicians when the first survey was returned undelivered. Return of the answered survey was taken as implied consent. All surveys were coded to protect confidentiality. Returned surveys were excluded when the practitioners did not treat children, were retired, refused to participate, or could not be located.
Analysis
Data were entered into a database (Excel [Microsoft Corporation, Redmond, WA]), and statistical analysis was performed by using SPSS 14 (SPSS Inc, Chicago, IL). All data were categorical in nature. Open-ended responses were grouped according to theme. Descriptive data were used for demographic information.
The primary analysis examined the difference in the proportion of noncardiologists who performed a full screen before, as compared with after, the Health Canada advisory. The secondary analysis examined the difference in stimulant-prescribing practices before versus after the advisory, for both the noncardiologists and cardiologists, and the differences in cardiovascular risk assessment and prescribing recommendations between these 2 groups. Subgroup analysis was also performed specifically to examine the general pediatricians' responses, because ADHD is largely managed by primary care pediatricians in Canada. Individuals who did not provide responses to a specific question were excluded from analysis of that particular question. Practitioners who indicated that they referred all patients to pediatrics were included in the analysis, and their answers were entered as "no" before and after the advisory for both screening evaluation and clinical scenarios.
The difference in the proportion of various responder groups was analyzed by using the McNemar,
2 test, or Fisher's exact test when appropriate. Univariate regression analysis followed by a multiple regression analysis was used to investigate the impact of demographic factors on awareness of the warning. The study was approved by the institutional research ethics board at the IWK Health Centre (Halifax, Nova Scotia, Canada).
| RESULTS |
|---|
|
|
|---|
|
Premedication Cardiovascular Risk Screening
Table 2 describes the cardiovascular assessment tools analyzed and the percentage of the noncardiologists and pediatric cardiologists who used the various assessment tools to assess cardiovascular risk. Quantitatively, there was an increase in the use of all of the individual assessment tools by noncardiologists (P < .0001) after the advisory. The proportion of noncardiologists who performed a full screen and a modified screen also increased significantly from 0.2% to 15.1% (P < .0001) and 7.4% to 34.5% (P < .0001), respectively. Examination of the pediatric cardiologists' responses revealed that there was a statistically significant change in the proportion who used a full screen (54.9% to 68.6%; P = .016) but no change in the proportion who used a modified screen (7.8% to 5.9%). A higher proportion of pediatric cardiologists used the full screen both before and after the advisory when compared with the noncardiologists (Table 3).
|
|
Prescribing Practices
Table 4 describes the willingness to prescribe stimulant medications to children with various severities of cardiac symptoms or clinical conditions. For each scenario, there was a significant decrease in the proportion of noncardiologists who would prescribe stimulant medications after the advisory for all clinical scenarios. In contrast, pediatric cardiologists reported very few changes in their recommendations. As outlined in Table 4, a significant difference in the proportion of general pediatricians who were willing to prescribe stimulants after the advisory was also seen in the following clinical scenarios: family history of sudden death, history of supraventricular tachycardia (SVT) or palpitations, a diagnosis of a hemodynamically insignificant ventricular septal defect, Fontan circulation, or a repaired tetralogy of Fallot with premature ventricular contractions (PVCs). Fewer than half of the general pediatricians would prescribe stimulants to those with a family history of sudden death; a history of SVT or palpitations; or a diagnosis of a Fontan circulation, repaired tetralogy of Fallot with PVCs, and long QT on a β-blocker. Comparing noncardiologists with pediatric cardiologists, there were statistically significant differences in willingness to prescribe stimulants after the advisory for all clinical scenarios, except for a patient with long QT syndrome on a β-blocker (Table 4). The similarities and differences between the general pediatricians' recommendations and those of the cardiologists are shown in Table 4.
|
| DISCUSSION |
|---|
|
|
|---|
Our study has demonstrated that the Health Canada advisory has reached a variety of Canadian practitioners who treat children with ADHD. Despite the level of awareness, it is apparent that a smaller proportion of noncardiologists, including the pediatric subgroup, are actually using all of the tools outlined by a full or a modified screen when compared with the pediatric cardiologists. Although the 5 history and physical examination tools do not encompass a complete cardiovascular assessment on their own, they are important components of risk screening. The Health Canada advisory lacks specific guidelines or recommendations for screening children for cardiovascular risk factors. In other patient populations at risk for sudden death, such as athletes, some believe that a thorough and focused cardiovascular history and physical examination may be the most cost-effective screening strategy, whereas others advocate for the addition of an ECG.9,10,20–22 Because the cardiac risk of stimulant medications in the general pediatric population is still not well defined, it remains unclear whether a universal screening approach is necessary for children with ADHD and whether either of these approaches would be effective in identifying children who are at risk for sudden death.
The changes in clinical practice observed in this study occurred despite a lack of consensus about the true cardiovascular risks of stimulant medications. The Health Canada advisory was based primarily on voluntary case reports available through the FDA.23 A systematic search of the FDA's database, between January 1992 and February 2005, revealed 31 cases of sudden death in individuals who were younger than 18 years, and, of these, 13 had associated cardiovascular findings.23 Although, to date, the FDA holds the best information available, the actual risk for sudden death cannot be accurately calculated from these data alone because underreporting is a common problem. Attempts to calculate the risk for sudden death from these data, by using the number of prescriptions filled as a substitute for the number of children who are treated with stimulants, revealed a risk of 0.2 to 0.5/100000 person-years.23 Although these numbers are lower than those quoted for sudden death in the general population, the method of calculation is flawed. Recently, Winterstein et al24 retrospectively reviewed 10 years of data from Medicaid claims and the Vital Statistics Death Registry to assess the incidence of cardiac outcomes of children with a diagnosis of ADHD. During this 10-year period, the risk for death for children with ADHD was 58.4/100000 patient-years; however, there were no reported cases of death associated with stimulant use. Despite that the study by Winterstein et al has its limitations, it is the first systematic evaluation of the cardiac risk of stimulants. Although our study does not address the risk of stimulant medications, it does provide information about the clinical impact of an advisory that is based on a theoretical risk.
The advisory has also influenced physicians' willingness to prescribe stimulants to children with ADHD and cardiac abnormalities. Since the advisory, noncardiologists are less likely to treat children with stimulants when there are any possible cardiac risk factors. This change seems to be independent of the seriousness of the cardiac condition. In contrast, the pediatric cardiologists noted very little change in their recommendations and are more likely than noncardiologists to recommend treatment of children with a variety of cardiac issues. These striking differences likely reflect a difference in risk perception, with cardiologists perceiving children as being less "ill" than their noncardiologist counterparts. This also reflects that the advisory is written in a manner that suggests that all cardiac conditions have similar risk.
Although there are differences between the 2 physician groups, the study also demonstrates that there are certain cardiac problems for which the majority of respondents from both groups would not recommend treatment of ADHD with stimulants. This is demonstrated by the fact that <40% of all practitioners would recommend stimulant use for children with long QT syndrome, SVT, or tetralogy of Fallot with PVCs. In addition to these 3 scenarios, <40% of noncardiologists would recommend stimulants for individuals with ADHD and a family history of sudden death, a Fontan circulation, and a history of palpitation or a ventricular septal defect.
The variation in recommendations and the significant change in noncardiologists' willingness to prescribe stimulants are particularly concerning for children with CHD. Evidence indicates that children with CHD are at risk for neurodevelopmental problems, including hyperactivity and inattention.25–31 The risk for sudden death associated with stimulants in this population has not been clearly defined; however, the long-term morbidity and mortality of ADHD is widely recognized and often devastating; ADHD puts children and adolescents at risk for academic and job failure, peer rejection, low self-esteem, substance abuse, and other high-risk behaviors.32–35 Taking these factors into account, the risk for potential sudden death must be weighed against the long-term morbidity and mortality of untreated ADHD in the process of considering stimulant therapy.
After the initial acceptance of this manuscript, the American Heart Association (AHA) released a scientific statement, based primarily on consensus opinion, outlining their recommendations for cardiovascular screening of children with ADHD.36 These guidelines suggest that all children with ADHD, before stimulant therapy, should undergo a thorough history and physical examination, and the physician should consider ordering an ECG,36 elements that we used to define our full screen. Like the Health Canada advisory, these recommendations are likely to reach a large number of practitioners and have a substantial impact on clinical practice. In considering how to optimize physicians' uptake of the AHA guidelines, our study showed that certain physician groups, such as family practitioners, are less aware of guidelines in this area than others. Targeting awareness campaigns and continuing education events around this topic for these groups in particular may increase awareness and adherence; however, because the AHA recommendations are based on expert opinion, it remains unclear whether they will ultimately prevent or "create" disease. In considering the medical community's reaction to the AHA guidelines, we anticipate an even more dramatic increase in the use of ECG in the pediatric ADHD population. This may result in a delay in treatment for some patients and a need for additional investigations to clarify both true-positive and false-positive ECG findings in others. This has the potential to cause significant stress and anxiety for at least a portion of these individuals.
Our study has several limitations. We were unable to assess the reliability of our survey because there is no gold standard instrument for comparison. We attempted to optimize reliability and validity with local pilot testing of the instrument before sending it to the study population. Although our response rate was low despite the use of multiple tools to improve our returns, this is in keeping with other physician surveys that focused on pediatric cardiology and pediatric primary care.13–18 The low response rate does introduce the issue of responder bias. We could not assess this because it was impossible to determine the characteristics of the nonresponders; however, analysis of the pediatric subgroup, whose response rate was high, also showed changes in clinical practice since the advisory. Despite these limitations, we believe that this study is important because it is the first study to address the impact of a government advisory on stimulant medications.
| CONCLUSIONS |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
Address correspondence to Jennifer Conway, MD, IWK Children's Heart Centre, 5850/5980 University Ave, PO Box 9700, Halifax, Nova Scotia, Canada B3K 6R8. E-mail: jennifer.conway{at}iwk.nshealth.ca
The authors have indicated they have no financial relationships relevant to this article to disclose.
| What's Known on This Subject To date, no published studies have addressed the clinical impact of Health Canada's advisory on stimulant medications and cardiac risks.
|
| What This Study Adds Our survey shows that Health Canada's advisory on stimulant medications has impacted clinical practice as physicians have increased their use of cardiovascular screening tools prior to prescribing stimulants and are less willing to treat children who have ADHD and real or potential cardiac disease.
|
| REFERENCES |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||