 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
PEDIATRICS Vol. 105 No. 3 March 2000, pp. 585-590
,, and
From the * Department of Medical Education, Arnold Palmer
Hospital for Children and Women, Nemours Children's Clinic, Orlando
Florida; and the Division of General Pediatrics, Boston University
School of Medicine, Boston Medical Center, Boston, Massachusetts.
 |
ABSTRACT |
|---|
Top
Abstract
Methods
Results
Discussion
References
|
|---|
Objectives. To determine what factors primary care pediatricians believe are important in establishing the initial diagnosis of childhood asthma and to identify variations in physicians' beliefs concerning this clinical decision.
Study Design. Massachusetts American Academy of Pediatrics Fellows were surveyed about their beliefs concerning the importance of 20 clinical factors associated with establishing the initial diagnosis of asthma.
Results. Most clinicians considered recurrent wheeze (96%), symptomatic improvement with a bronchodilator (90%), recurrent cough (89%), exclusion of alternative diagnoses (87%), and suggestive peak flow findings (82%) as important in diagnosing asthma. However, there was substantial heterogeneity among clinicians as to which combinations of factors they each considered relevant; for example, only 60% identified all 5 of the above factors to be necessary or important. Further, <50% identified presence of any of the 20 factors as necessary. Although national guidelines cite objective assessment of pulmonary function as essential, spirometry and peak expiratory flow testing were identified as necessary by only 8% and 10%, respectively. Two factors believed to contribute to establishing the asthma diagnosis contradicted the National Asthma Education and Prevention Program guidelines and expert opinion (age >2 years and absence of fever during episodes) and these beliefs were more likely held by those clinicians in practice for >5 years.
Conclusions. The majority of pediatricians believe several common clinical factors establish a diagnosis of childhood asthma, but disagree over what combinations of these factors are important. Some misconceptions persist despite wide dissemination of clinical practice guidelines. We believe that future asthma guidelines will need to organize diagnostic criteria in an easily understood format, like a decision tree, to facilitate early recognition of asthma in young children. Key words: asthma, Massachusetts, physician's practice patterns, practice guidelines, guideline adherence, children, health services research, knowledge, attitudes, practice.
Asthma affects an estimated 4.8 million
children1,2 and accounts for 2.2 million visits to
pediatricians each year.3 Furthermore, this illness is
cited as the most frequent cause of school absenteeism and
hospitalization of children.3 The reported prevalence and
severity of asthma among children has risen substantially throughout
the last few decades.24-8 Systematic changes in how the
diagnostic label is acquired9 and in how prevalence is
determined10 may contribute to this finding. Nevertheless,
although serial prevalence studies11,12 support the
presence of at least a modest rise in asthma throughout the last
decade, as many as 14% to 50% of children with asthma are
undiagnosed.13-15
Childhood asthma can be difficult to diagnose because of its
heterogeneous presentations and lack of pathognomonic
findings.16,17 Children with occasional symptoms of
wheezing triggered only by upper respiratory tract infections; children
with daily breathlessness, coughing, and restriction in activities; and
children with occasional night time coughing can share a common
diagnosis of asthma despite having quite different presentations.
Complicating the diagnosis is its episodic nature and frequent
co-morbidity with other common illnesses, such as upper respiratory
tract infections. Further, wheezing and related symptoms in young
children are common and often abate by 6 years of
age18 The Expert Panel Report 2: Guidelines for the Diagnosis and
Management of Asthma prepared for the National Heart, Lung, and Blood Institute's National Asthma Education and Prevention Program (NAEPP)19 suggests clinicians identify the presence of
episodic symptoms of airflow obstruction, find that airflow obstruction
can be at least partially reversible, and exclude alternative diagnoses
to make the diagnosis of asthma. The NAEPP guidelines contain a lengthy
list of possible diagnostic factors (symptoms, clinical signs, and
diagnostic tests). However, this clinical practice guideline does not
suggest a diagnostic algorithm nor does it weigh the factors based on
their relative importance. By virtue of their encyclopedic description
of factors related to the disease, these expert guidelines lack
clinical utility for the clinician attempting to apply them to a
specific patient. Guidelines prepared by the American Thoracic
Society;20 Joint Council of Allergy, Asthma, and
Immunology;21 and the World Health
Organization22 similarly fail to present diagnostic
criteria in a readily practical format for the clinician. For example,
whether wheezing, breathlessness, chest tightness, cough, among several
other features must all be present to establish an asthma diagnosis or
if a single factor suffices is unclear.
Substantial physician variation has been described in the treatment of
several conditions,23-28 in prescribing
activity,29,30 in likelihood to
hospitalize,31-33 and in the use of diagnostic
tests.34 We suspect that making an initial diagnosis of
asthma is a complex decision and may lead to physician variation, as
well as contribute to missed or delayed diagnoses. We undertook this
study 1) to determine what common factors primary care pediatricians
believe are important in establishing the initial diagnosis of
childhood asthma, and 2) to describe variations in physician beliefs
concerning this decision-making process.
In December 1997, a self-administered 5-page questionnaire was
sent to all active Fellows of the American Academy of Pediatrics (AAP)
living in Massachusetts based on their listing in the state section of
the AAP membership directory. Resident fellows, emeritus fellows, and
subspecialty fellows were excluded from the initial mailing of 862 questionnaires. Two reminder postcards and a second copy of the survey
were sent to nonresponders in January and February 1998. Based on
responses to selected questions, respondents were considered ineligible
if they 1) did not provide patient care, 2) performed administrative
tasks primarily, or 3) primarily practiced in nonclinical fields or in
fields without exposure to children with asthma, such as neonatology.
The study was approved by the Boston Medical Center's Institutional
Review Board for Human Research.
The questionnaire was divided into 3 sections: the first section listed
20 factors identified from expert guidelines, informal focus group, and
a series of piloted surveys as relevant to the asthma diagnosis in
children. An expert panel of 20 pediatric pulmonologists and allergists
living in New England and listed in the specialty section of the AAP
membership directory corroborated the relative importance of these
factors. Examples of factors included: the presence of recurrent
wheezing, the presence of recurrent coughing, and the absence of fever
during suspected episodes. Respondents were asked to designate the
importance of each factor in establishing the initial diagnosis of
asthma using a Likert scale. Possible responses included: 1) necessary,
2) important, 3) somewhat important, and 4) not important. Respondents were further instructed to list the 3 criteria among the 20 factors they considered most important and would use routinely. The second section included inquires about obstacles which may delay or discourage the diagnostic labeling of asthma such as concern about future problems
obtaining life/health insurance. The third section presented 4 vignettes that featured children presenting with symptoms of wheezing
and coughing. Respondents were asked to suggest differential diagnoses
and appropriate therapy for each child. The respondents' rating of the
importance of the 20 factors relevant to the diagnosis of asthma in
children is the focus of this analysis.
The following variables (physician characteristics) were examined for
possible contribution to respondent variation: physician gender,
practice location, practice duration, weekly patient volume, and NAEPP
guideline familiarity. We estimated that to detect a 20% difference
between suburban and urban physicians (assuming a 3:2 ratio) in
selection of 3 factors as critical, <200 study participants would be
required to demonstrate a significant difference at a 95% confidence
level with 1 Statistical analysis was performed using SAS for PC, version 6.12 (SAS
Institute Inc, Cary, NC). Questionnaires were sent to 862 Massachusetts AAP Fellows; 571 (66%) completed surveys. One hundred twenty study participants met
exclusion criteria leaving 451 (79%) study participants in the final
cohort. The response rate was 66%.37 The majority of
respondents were male, practiced in a suburban location, and were in
group practice (Table 1). Almost all
respondents spent the majority of the workweek providing patient care.
Respondents saw an average of 98 patients/week and had practiced a mean
of 15.5 years. The majority of respondents identified themselves as
often responsible for making the initial diagnosis of asthma in a
child. Most indicated an understanding of current asthma diagnosis and
treatment; 405 (91%) were familiar with the NAEPP guidelines, 276 (62%) attended a lecture on childhood asthma within 6 months, and 322 (72%) taught residents and/or medical students. Male pediatricians
were more likely than female pediatricians to practice in a suburban
location (RR = 1.24, 95% confidence interval [CI] = 1.05-1.47), have practiced >5 years (RR = 1.10, 95% CI = 1.03-1.17), and to have seen >100 patients/week (RR = 1.58, 95%
CI = 1.33-1.89). Further, suburban pediatricians were more likely
to have practiced >5 years (RR = 3.33, 95% CI = 1.60-6.94) and to have seen >100 patients/week (RR = 2.34, 95% CI = 1.80-3.04). The male pediatricians had practiced an average of 18.2 years compared with 12.2 years for female pediatricians
(P = .0001). No significant associations were
identified among demographic characteristics and familiarity with NAEPP
guidelines.
TABLE 1
possibly discouraging commitment to the diagnosis
of a chronic disease. In addition to varied clinical presentations,
conditions differing in pathophysiology such as allergen-mediated
asthma as compared with exercise-induced asthma will receive the same
diagnostic label of asthma.17
METHODS
Top
Abstract
Methods
Results
Discussion
References
= 0.80.35
2 tests (or Fisher's Exact
test when expected frequencies were small) and t tests for 2 independent groups were used to compare physicians based on practice
characteristics. An F test was performed to determine equal
variances before selecting the appropriate t test procedure. When comparing continuous variables between groups, sample size was
determined to be sufficiently large to insure the means were normally
distributed (central limit theorem). Relative risk RR) was calculated
with the Mantel-Haenszel method. Multiple logistic regression analysis
was performed to determine the relative importance of physician
characteristics in the preference for selected diagnostic factors. For
consistent reporting throughout this manuscript, conditional odds
ratios were converted to corrected RRs based on the formula by Zhang
and Yu.36 All statistical tests were performed at the .05 level of significance for a 2-tailed analysis.
RESULTS
Top
Abstract
Methods
Results
Discussion
References
Demographic Characteristics of General Pediatricians Group
Fewer than 50% of respondents identified presence of any of the 20 factors as necessary to the diagnosis of asthma (Table 2). However, >80% of respondents rated 5 factors as necessary or important in establishing the initial diagnosis of asthma. These factors included: 1) presence of recurrent wheezing (96%); 2) symptomatic improvement with a bronchodilator (90%); 3) presence of recurrent cough (89%); 4) exclusion of other possible diagnoses (87%); and 5) suggestive peak flow findings (82%). Despite support for these 5 factors, there was substantial heterogeneity among clinicians as to which combinations of factors they individually considered relevant. For example, only 60% identified all 5 of the above factors to be necessary or important, whereas 79% agreed on the importance of the 3 most common factors (presence of recurrent wheezing, symptomatic improvement with a bronchodilator, and presence of recurrent cough). Two factors were commonly cited that contradicted NAEPP guidelines and expert opinion: 27% reported a child's age (>2 years old) and 18% reported the absence of fever during suspected episodes were necessary or important. Although the NAEPP guidelines and our expert panel considers pulmonary function measures necessary to the diagnosis of asthma, 37% of clinicians did not find this factor necessary or important.
|
Only 2 factors were selected by a majority as among the 3 most critical factors a respondent would use in diagnosing asthma in children: presence of recurrent wheezing (79%) and symptomatic improvement of an acute episode with use of a bronchodilator (54%). Other factors considered critical were the presence of recurrent cough (32%), the exclusion of other possible diagnoses (31%), and the experience of >2 wheezing episodes in a child's lifetime (29%).
Practice duration and reported NAEPP guidelines familiarity had a substantial effect on a clinician's choice of factors that contradict NAEPP guidelines and/or expert opinion (child's age <2 years, absence of fever, and lack of pulmonary function testing) (Table 3). For example, in bivariate analysis, pediatric practice for >5 years was associated with both considering a child's age (>2 years) and the absence of a fever as necessary or important (RR = 1.77, 95% CI 1.17-2.67 and RR = 1.90, 95% 1.14-3.17, respectively). Clinicians who failed to consider pulmonary function testing as either necessary or important were in practice longer (16.75 ± 9.75 years) in contrast with those who placed a greater importance on this factor (13.24 ± 8.24 years; P = .0002). Inner city urban practitioners were more likely than suburban to consider a child's age (>2 years old) as necessary or important (RR = 1.67, 95% CI 1.03-2.70). Practice location was not associated with selecting a specific factor as critical.
|
Multivariate models were created to explore the relationship between
physician characteristics identified in the bivariate analysis and
child's age >2 years, absence of fever during suspected episodes, and
suggestive findings with pulmonary function tests. These models (Table
4) revealed clinicians with longer practice duration (>5 years) believed a child's age (>2 years) (corrected RR = 1.42, 95% CI 1.04-1.74) factored prominently in their decision-making. Clinicians with longer practice duration (>5
years) were more likely to believe that the absence of fever was
necessary or important to the diagnosis (corrected RR = 1.52, 95%
CI 0.94-2.10)although failing to reach statistical significance. Similarly, inner city practitioners tended to be more likely than suburban practitioners to consider a child's age (>2 years old) as
necessary or important (corrected RR = 1.48, 95% CI 0.97-2.03). The model exploring physician characteristics and selection of suggestive findings with pulmonary function tests failed to identify significant associations (Table 4).
|
| |
DISCUSSION |
|---|
|
Top
Abstract
Methods
Results
Discussion
References
|
|---|
The complexity in diagnosing pediatric asthma is reflected in the heterogeneity of responses among practitioners. We are unaware of any other data about how physicians make this diagnosis, particularly in young children. Our findings suggest that the decision processes used by individual clinicians vary; however, they share some common elements. There was general agreement that the following factors played a prominent role in establishing the diagnosis: presence of recurrent wheezing, symptomatic improvement with a bronchodilator, presence of recurrent cough, exclusion of other possible diagnoses, and suggestive peak flow findings. These factors are consistent with NAEPP guidelines instructing clinicians to consider the diagnosis on identifying the presence of episodic symptoms of airflow obstruction, finding the airflow obstruction can be at least partially reversible, and excluding alternative diagnoses.19
The NAEPP guidelines indicate "spirometry measurements (forced expiratory volume in 1 second (FEV1), forced vital capacity (FVC), FEV1/FVC) before and after the patient inhales a short-acting bronchodilator should be undertaken for patients in whom the diagnosis of asthma is being considered" but recognized "some children cannot conduct the maneuver adequately until after age 7".19 Few respondents reported pulmonary function tests as necessary in their decision-making and less than two thirds consider it important. Similarly few respondents identified peak expiratory flow testing as necessary, but they were more likely to consider it important. The relative importance placed by clinicians on peak flow versus spirometry findings probably reflects their experience with and access to instruments measuring pulmonary function. Adherence to the recommendation on spirometry is problematic,38 especially among pediatricians. For a pediatrician to perform spirometry, either appropriate training, additional equipment, and a longer office visit would be required or the primary care provider would need to refer the child to an outside facility and/or specialist for testing. Both sets of choices may delay the diagnosis and, in addition, referral threatens a clinician's autonomy in treatment of this child's illness. Although the procedure is useful in establishing the asthma diagnosis,39 it is virtually impossible to perform on young children and to interpret their results.40
Our findings suggest that pediatricians choose among multiple factors
listed in published guidelines to help them in determining if a patient
has asthma. Unfortunately, some pediatricians' beliefs seem to
contradict expert opinion41-43 and include factors not
addressed in the NAEPP guidelines.19 For example, many
would delay the diagnosis until suspected episodes occurred without
concurrent fever. Knowing that wheezing among young children is so
common, it is not surprising that so many clinicians may delay the
diagnosis until an older age. However, delaying the diagnosis may
result in persistence in troublesome symptoms, failure to meet
families' expectations, and most concerninga delay in appropriate
treatment and increased morbidity through chronic pulmonary
changes.44-48
The majority of respondents reported being familiar with the latest
NAEPP Clinical Practice Guidelines19 and having attended a
recent lecture on childhood asthma. Many have the responsibility to
teach residents and/or medical students. Nevertheless, there was
substantial variation among study participants concerning factors they
believed important togethereven among the key points listed in the
NAEPP Clinical Practice Guidelines (episodic symptoms of airflow
obstruction, finding the airflow obstruction can be at least partially
reversible, and excluding alternative diagnoses) and few identified
these factors individually or in composite as necessary or critical.
NAEPP guidelines familiarity did not seem to affect selection of
important factors or protect against selection of factors prone to
misconception. These results are predictable because the ability of
clinical practice guidelines to influence physician practice is
limited.49,50 Even following a deliberate educational
intervention using the first expert panel's report19 from
the National Heart, Lung, and Blood Institute in 1991, investigators51 found heterogeneity in physician attitudes
and behavior with regard to subsequent medication use and peak flow
monitoring.
Respondents in practice >5 years were more likely than those in practice fewer years to mistakenly consider a child's age and presence of fever as important, and to ignore pulmonary function measures in the decision-making process. These findings are consistent with observations that awareness and adoption of expert recommendations deteriorates with time elapsed from medical school.52-54 This observation has important implications for how we maintain competence in caring for patients when faced with ever expanding growth in medical information. Developing habits of life-long learning, use of evidence-based practice guidelines and secondary publications of structured abstracts, and shift in traditional continuing medical education activities to more effective modalities seem critical to reversing this trend.
This study has several limitations. Firstly, one-third of potentially eligible pediatricians did not respond. However, among nonresponders, a sample of 51 clinicians that refused to participate provided simple demographic information. There was no significant difference between responders and nonresponders based on gender, location, and setting. But, nonresponders were in practice 3 years longer than those who participated in our study (P = .03). Secondly, as suggested by the high frequency of teaching among respondents, Massachusetts has many academic centers. Further, our sample included only those pediatricians who are active members of the AAP; hence, the results may not be generalizable to clinicians in more remote practices, those without AAP membership, or those practicing in other states. Thirdly, as with other studies reliant on self-administered questionnaires, this study is subject to limitations common to this research methodology. Respondents may have reported their beliefs inaccurately. If they under reported more extreme views or failed to report actual practice, the size of effects could be reduced and some real associations may have gone undetected. A longitudinal study of actual physician practice would address these concerns.
With a lengthy list of possible indicators and variable presentations, general pediatricians rely on combinations of findings to aide them in establishing the diagnosis of asthma. The combinations are individually determined and vary among practitioners. It is possible that this variation in turn reflects how and when pediatricians institute therapy for children with recurrent wheezing and cough. Most clinicians choose factors consistent with published guidelines. Some rely on less accurate factors like age limitations. Consistent and practical diagnostic criteria relevant to young children need to be agreed on and integrated into continuing medical education efforts. A decision tree or specific set of criteria (perhaps weighted as major and minor) would be helpful to the practicing clinician responsible for making the initial determination of childhood asthma. In addition, effective means to facilitate its adoption among clinicians need to be developed. Such a scheme is lacking in published clinical practice guidelines and we believe could substantially improve the timely diagnosis and effective treatment of childhood asthma.
| |
ACKNOWLEDGMENTS |
|---|
This study was supported in part by a Health Resources and Service Administration grant. Key Pharmaceuticals provided for some materials.
We are grateful for the contributions of Colleen Pearson and Loretta Beasley in the processing of materials, data collection, and data entry.
| |
FOOTNOTES |
|---|
Received for publication Jun 22, 1999; accepted Aug 31, 1999.
Reprint requests to (L.N.W.) Department of Medical Education, Arnold Palmer Hospital for Children and Women, Nemours Children's Clinic, 86 W Underwood St, Suite 202, Orlando, FL 32806. E-mail: lwerk{at}nemours.org
| |
ABBREVIATIONS |
|---|
NAEPP, National Asthma Education and Prevention Program; AAP, American Academy of Pediatrics; RR, relative risk; CI, confidence interval; FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity.
| |
REFERENCES |
|---|
|
Top
Abstract
Methods
Results
Discussion
References
|
|---|
United States, 1980-1993.
MMWR Morb Mortal Wkly Rep
1996;
45:350-353 [Medline]United States, 1960-1995. MMWR Morb Mortal Wkly Rep. 1998;47 No. SS-1 CDC Surveillance Summariestime for a change of name? Commentary.
Arch Dis Child.
1997;
77:64-65 time for a change of name?
Arch Dis Child.
1997;
77:62-65 This article has been cited by other articles:
 |
|
 |
|
  L. M. Graham Classifying Asthma Chest, July 1, 2006; 130(1_suppl): 13S - 20S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. P. David and D. S. Greer Social Marketing: Application to Medical Education Ann Intern Med, January 16, 2001; 134(2): 125 - 127. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. Finkelstein, P. Lozano, R. Shulruff, T. S. Inui, S. B. Soumerai, M. Ng, and K. B. Weiss Self-Reported Physician Practices for Children With Asthma: Are National Guidelines Followed? Pediatrics, October 1, 2000; 106(4): 886 - 896. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2000 American Academy of Pediatrics. All rights reserved. |
||
|
||
