Presented at the Pediatric Emergency Research Canada annual meeting, Platform Presentation, Quebec City, Quebec, January 2010; and the Pediatric Academic Societies annual meeting, Vancouver, British Columbia, Canada, May 2010.
North American Practice Patterns of Intravenous Magnesium Therapy in Severe Acute Asthma in Children
Version of Record online: 2 NOV 2010
© 2010 by the Society for Academic Emergency Medicine
Academic Emergency Medicine
Volume 17, Issue 11, pages 1189–1196, November 2010
How to Cite
Schuh, S., Macias, C., Freedman, S. B., Plint, A. C., Zorc, J. J., Bajaj, L., Black, K. J., Johnson, D. W. and Boutis, K. (2010), North American Practice Patterns of Intravenous Magnesium Therapy in Severe Acute Asthma in Children. Academic Emergency Medicine, 17: 1189–1196. doi: 10.1111/j.1553-2712.2010.00913.x
Funded by the Paediatric Consultants, The Hospital for Sick Children.
Supervising Editor: Richard Sinert, DO.
- Issue online: 2 NOV 2010
- Version of Record online: 2 NOV 2010
- Received March 16, 2010; revision received April 20, 2010; accepted April 20, 2010.
- acute asthma;
- IV magnesium;
- emergency department;
Objectives: Although intravenous (IV) magnesium (Mg) can decrease hospitalizations in children with severe acute asthma, its use is often limited to resistant disease, and disposition may be determined prior to its use. Since knowledge about practice patterns of IV Mg would enhance knowledge translation and guide future research, we surveyed pediatric emergency physicians with interest in clinical research to determine the frequency, indications, adverse events, and barriers to use of IV Mg in children with severe acute asthma.
Methods: A cross-sectional online survey of two national pediatric emergency physician associations in Canada and the United States was conducted using a modified Dillman technique.
Results: Response rates were 124 of 180 (69%) in Canada and 75 of 108 (69%) in the United States. Although 88% of participants report knowing that Mg is effective, only 14 of 199 (7%) give it to prevent hospitalizations and 142 of 199 (71%) give it to prevent admissions to the intensive care unit (ICU). Thirty-eight percent of respondents use Mg in < 5% of stable children with severe acute asthma, while 79% use it in 50% or more of the ICU candidates with concern about impending respiratory failure. Seventy-nine percent of the participants report that < 5% of their patients given Mg are discharged home from the ED. Twenty-four percent of respondents who administer Mg have encountered associated severe hypotension requiring therapy, and 2% have witnessed-related apnea. Factors affecting Mg use include concern about side effects expressed by 24% of physicians and a belief that IV therapy is not necessary, expressed by 31%.
Conclusions: Intravenous Mg appears to be uncommonly used in stable children with severe acute asthma and does not frequently play a role in reducing hospitalizations. Further research to justify its enhanced use and to better establish its true adverse effect profile is indicated.
ACADEMIC EMERGENCY MEDICINE 2010; 17:1189–1196 © 2010 by the Society for Academic Emergency Medicine
Despite extensive research and knowledge translation efforts, the emergency department (ED) management of severe acute asthma in children has not substantially improved patient outcomes. Over the past decade, hospitalizations have remained stable and account for almost half of the annual asthma care costs.1,2 As many as 30% of patients do not improve after inhaled β2 agonists due to multiple factors, which may include disease severity, pharmacogenomic variation, or a delayed response to therapies such as corticosteroids.3 Recently, two meta-analyses have shown that early intervention with intravenous (IV) magnesium sulfate (Mg) appears effective in children with severe disease not responding to initial bronchodilator therapy and significantly reduces hospitalization rates;4,5 this agent is therefore recommended for consideration in this subpopulation.4–8
Although Rowe and Camargo9 confirmed that IV Mg appears to be used appropriately in general EDs looking mainly after adults, it is unclear how frequently IV Mg is used in North America in the pediatric population. This intervention is invasive and requires careful monitoring due to its potential for severe side effects.10 Although the literature suggests that the addition of IV Mg may allow many children to be successfully discharged home,11 this effect may not be reflected in practice, since, anecdotally, the decision to hospitalize is often made prior to the institution of Mg therapy. Investigation of the frequency of use of IV Mg in children with severe disease is needed to establish under what circumstances this intervention is employed and what the therapeutic goals are, to identify the gaps that need to be addressed in future knowledge translation research. Thus, the main objective of this survey of academic pediatric emergency physicians (EPs) with a specific interest in multicenter research was to investigate the estimated frequency of use of IV Mg in children with stable severe acute asthma and those with impending respiratory failure. In addition, we aimed to determine the indications, therapeutic goals, observed toxicity, and potential barriers to the use of IV Mg.
Study Design and Population
Using a modified Dillman’s Total Design Method for contact and follow-up,12 in March 2009 we sent an electronic survey to all staff EPs who were members of the Pediatric Emergency Research in Canada (PERC), and the Pediatric Emergency Medicine–Collaborative Research Committee network (PEM-CRC) in the United States. These physicians were accessed using the member databases of these national organizations. Both networks contain comprehensive and readily accessible membership lists with e-mail addresses, which are important for subsequent comparison of responders and nonresponders, as well as enhancing the response rate. Physicians who were not staff-level employed or no longer practicing emergency medicine or whose ED practice did not include children were excluded, as were those who could not be reached by e-mail. Physicians were informed that their participation was voluntary and their responses confidential. Prior to survey administration, our research ethics board approved the study, as did the respective national associations.
Survey Content and Administration
Because we could not find any published relevant validated questionnaire, we have used the methods proposed by Streiner and Norman.13 The survey questions were derived using the evidence from the literature with respect to the efficacy, indications, dosing range, and potential toxicity of IV Mg.4,5,11,14–16 These were refined through consultation with an expert panel involving eight pediatric EPs on both sides of the border (listed authors) with survey and/or content expertise. Items were generated and subsequently reduced after revision and pretesting.
The survey infrastructure consisted of the following domains, with respective number of questions: participants’ demographics (six); their knowledge of the evidence of IV Mg efficacy (one); practice patterns of IV Mg administration in severe asthma exacerbations, such as the estimated frequency of its use in children with and without impending respiratory failure and the usual disposition goal of Mg therapy (home, ward, intensive care unit [ICU]; 12); barriers to the use of this agent such as concern for side effects, invasiveness, belief in the need for an IV in severe exacerbations, resource issues, or lack of familiarity with the drug (five); and previously encountered severe side effects of IV Mg such as hypotension requiring therapy, apnea, and heart block (three). Severe acute asthma was defined as severe suprasternal/supraclavicular retractions with severe tachypnea for age, since these are widely used indicators of the severity of respiratory distress in the acute care setting.8,17 To establish what proportion of physicians have observed severe side effects of this intervention, we asked how many times they have personally encountered a drop in blood pressure requiring treatment, apnea necessitating airway management, or heart block during Mg therapy and the frequency of these encounters.
Based on consensus, relevant response options included either estimated proportions of less than 5% (reflecting rare events), followed by increasing frequencies with comparable intervals of 5%–19%, 20%–34%, 35%–49%, and 50% or greater (reflecting frequent events) or a five-point Likert scale with the response options ranging from “strongly disagree” to “strongly agree.” The reported agreement results included those where physicians marked either a moderate or a strong agreement.
We have created the survey using the SurveyMonkey tool (http://www.surveymonkey.com/). This website has the advantage of an unlimited number of responses, data collection via e-mail and web link, forced responses, skip logic, and downloading to a spreadsheet. To minimize incomplete responses, multiple-item screens were incorporated into the design.18 After refining the survey, we e-mailed it with an introductory letter and a link to all eligible participants. Subsequently, we sent four reminder e-mails describing the intent of the study, approximate accrual rates, and request for participation. Both national organizations allowed access to individual e-mail addresses. To maintain confidentiality, the survey was sent by an independent party.
To obtain a 95% confidence interval (CI) with a 7% precision around an estimated use of IV Mg of 50% in less than 5% of the children with stable severe asthma, and a response rate of 60%, the required sample size was 186 analyzable respondents.19,20 The differences in relevant demographic proportions between Canadian and the U.S. participants were analyzed using the chi-square statistic. The association between the Mg users versus nonusers (defined as using it ≥ 5% of the time vs. < 5% of the time—to capture those with highly sporadic use—in a stable severe attack) and participants’ characteristics potentially associated with this outcome were determined using multivariable logistic regression analysis. Participant characteristics included full- versus part-time employment status, pediatric emergency medicine fellowship training, years of practice (less than 10 years vs. 10 or more), site of employment (pediatric ED vs. general ED), goal of Mg therapy (discharge or not), country (Canada vs. U.S.), and concern regarding toxicity of Mg. Wald and likelihood ratio tests were used to remove noncontributory variables from the model.21 The Hosmer-Lemenshow test was used to examine the goodness of fit of the final model. All statistics were performed with SAS version 9.12 (SAS Inc., Cary, NC). Available information on responders and nonresponders was extracted from the databases by a research assistant, combined for both countries, and analyzed by the chi-square statistic for relevant proportions and the Student’s t-test for continuous variables.
The two national organizations have a total of 360 members (PERC, 207; PEM-CRC, 153). There were 58 repeatedly failed deliveries, and 14 physicians were not staff-level employed or did not treat children. This left 288 eligible physicians, and 199 responded to the survey (61.9%). The response rate in Canada was 124 of 180 (68.8%) and in the United States 75 of 108 (69.4%). Table 1 summarizes the demographics of the participants. Overall, significantly fewer of our U.S. colleagues see exclusively children in their EDs, but more are PEM-trained, compared to their Canadian counterparts.
|Characteristic||Overall, N = 199 (%)||Canada, n = 124 (%)||United States, n = 75 (%)||p-value|
|Years in practice|
|<5 (%)||47 (23.6)||30 (24.2)||17 (22.7)||0.056|
|5–9||40 (20.1)||31 (25.0)||9 (12.0)|
|10–14||41 (20.6)||19 (15.3)||22 (29.3)|
|15–19||40 (20.1)||23 (18.6)||17 (22.7)|
|20 or more||31 (15.6)||21 (16.9)||10 (13.3)|
|Full-time (≥ 50% of time in ED)||170 (85.4)||104 (83.9)||66 (88.0)||0.420|
|Children ≤ 20 years of age as a proportion of ED patients|
|100%||149 (74.1)||107 (86.3)||42 (56.0)||<0.0001|
|50%–99%||41 (20.6)||8 (6.5)||33 (44.0)|
|5%–49%||11 (5.5)||9 (7.3)||2 (2.7)|
|Highest level of training|
|Board certified in pediatrics||36 (18.1)||34 (27.4)||2 (2.7)||<0.0001|
|Board certified in pediatrics and PEM||130 (65.3)||64 (51.6)||66 (88.0)|
|Board certified in EM||11 (5.5)||10 (8.1)||1 (1.3)|
|Board certified in EM and PEM||7 (3.5)||3 (2.4)||4 (5.3)|
|Other||15 (7.5)||13 (10.5)||2 (2.7)|
|University-affiliated pediatric ED||115 (57.8)||69 (55.7)||46 (61.3)||0.470|
|University-affiliated general ED||76 (38.2)||51 (41.1)||25 (33.3)|
|Other||8 (4.0)||4 (3.2)||4 (5.3)|
Since the PERC and PEM-CRC databases do not contain identical parameters or parameter criteria, the comparison of responders and nonresponders was limited to five characteristics, and the source of the data on responders was therefore either the database or the survey. The responders and nonresponders were comparable with respect to the full-time status (103/176 or 58.5% vs. 47/72 or 65.3%), median year of completion of training (1999 for both), the mean number of years in practice (10.4 vs. 12.2), and PEM fellowship training (137/199 or 68.8% vs. 34/54 or 63%). A total of 101 of 121 or 83.5% of the Canadian respondents were employed in a pediatric ED versus 42 of 56 (75.0%) of their Canadian abstaining colleagues. In the United States, 100% of the participants spent at least some working time in the pediatric ED versus 98.2% of nonparticipants.
Knowledge of Efficacy of IV Mg and Its Use
The vast majority of EPs, in both Canada and the United States, agree that the available evidence shows benefit of IV Mg in children with severe acute asthma (110/124 or 89% and 66/75 or 88%, respectively) and that this agent should be considered in patients who do not respond to the initial therapy. However, when managing stable patients with severe distress, the majority of physicians in both countries use Mg in less than 20% of this population, and almost 40% use it in less than 5% (Table 2). Rather, PEM physicians tend to resort to Mg when they are concerned about impending respiratory failure and when the patient is a likely candidate for the ICU (Table 2). Likewise, while most physicians employ IV Mg in children requiring ICU care after stabilization in the ED, this intervention appears to be used infrequently in children with severe asthma prior to an admission to a pediatric ward (Figure 1), with comparable proportions on both sides of the border. The time of Mg administrations appears disparate. While 66 of 99 (33%) of our colleagues initiate IV Mg soon after initial bronchodilator therapy, another 54 of 199 (27%) wait until 4–6 hours after systemic corticosteroids have been administered. Additionally, the Mg dose employed varies widely, with 36 of 199 (18%) of physicians administering 20–29 mg/kg, 9 of 199 (5%) giving 30–39 mg/kg, 47 of 199 (24%) using 40–49 mg/kg, 49 of 199 (25%) using 50–59 mg/kg, and 9 of 199 (5%) giving 60 mg/kg or more. A total of 15 of 199 physicians use other doses, 18 of 199 do not remember the dose, and 16 of 199 do not use IV Mg.
|Characteristic||Overall (N = 199)||Canada (n = 124)||United States (n = 75)|
|Frequency of use without impending respiratory failure|
|<5%||75 (37.7)||48 (38.7)||27 (36.0)|
|5%–19%||39 (19.6)||24 (19.4)||15 (20.0)|
|20%–34%||21 (10.6)||14 (11.3)||7 (9.3)|
|35%–49%||22 (11.1)||12 (9.7)||10 (13.3)|
|≥50%||42 (21.1)||26 (21.0)||16 (21.3)|
|Frequency of use with impending respiratory failure|
|<5%||20 (10.1)||15 (12.1)||5 (6.7)|
|5%–19%||11 (5.5)||6 (4.8)||5 (6.7)|
|20%–34%||1 (0.5)||1 (0.8)||0|
|35%–49%||10 (5.0)||5 (4.0)||5 (6.7)|
|≥50%||157 (78.9)||97 (78.2)||60 (80.0)|
|Impact on decision to use magnesium (% moderately/strongly agree)|
|Concern about impending respiratory failure||162 (81.8)||97 (78.9)||65 (86.7)|
|Persistence of severe distress without respiratory failure||141 (71.2)||91 (74.0)||50 (66.7)|
Disposition Goals of IV Mg Therapy
Overall, the vast majority of EPs do not feel that IV Mg plays an important role in reducing hospitalizations of children with severe asthma in their practice (Table 3), and most report that < 5% of their patients who receive IV Mg are discharged to home from the ED (Figure 2). Only a minority of physicians claim that their goal of this intervention is to prevent admission to hospital, with most using it to avoid admission to an ICU or an ICU step-down unit (Table 3).
|Characteristic||Overall (N = 199)||Canada (n = 124)||United States (n = 75)|
|IV Mg plays important role in reducing severe asthma hospitalizations in my practice (% agree/strongly agree)||30 (15.1)||18 (14.5)||12 (16.0)|
|Goal of Mg therapy is to prevent hospitalization (% agree/strongly agree)||14 (7.0)||10 (8.1)||4 (5.3)|
|Goal of Mg therapy is to prevent ICU admission (% agree/strongly agree)||142 (71.4)||91 (73.4)||51 (68.0)|
|Estimated discharge home after IV Mg|
|<5%||158 (79.4)||97 (78.2)||61 (81.3)|
|5%–9%||16 (8.0)||8 (6.5)||8 (10.7)|
|10%–19%||7 (3.5)||4 (3.2)||3 (4.0)|
|20%–29%||1 (0.5)||1 (0.8)||0|
|≥ 30%||2 (1.0)||2 (1.6)||0|
|N/A, do not use Mg||15 (7.5)||12 (9.7)||3 (4.0)|
Who Is More Likely to Use IV Mg?
All clinically significant variables considered are depicted in Table 4. The physicians using IV Mg in stable severe acute asthma more than 5% of the time are significantly more likely to be PEM fellowship trained (odds ratio = 2.06, 95% CI = 1.1 to 3.9; Table 4). The assumptions for logistic regression models, such as a binary-dependent variable, goodness of fit, and absence of multicollinearity, were met. The Hosmer-Lemenshow test showed a nonsignificant p-value of 0.64, indicating no evidence to reject the fit of the model.
|Characteristic||N||Mg User (n = 124)||Mg Nonuser* (n = 75)||Difference in Mg Users||95% CI for the Difference†|
|Full-time employment||170||112 (66)||58 (34)||−24.5||−0.44 to −0.05|
|Part-time employment||29||12 (41)||17 (59)|
|PEM fellowship||130||90 (69)||40 (31)||−19.9||−0.34 to −0.06|
|No PEM fellowship||69||34 (49)||35 (51)|
|Years in practice (<10)||87||60 (69)||27 (31)||−11.9||−0.25 to 0.01|
|Years in practice (≥10)||112||64 (57)||48 (43)|
|Canada||124||76 (61)||48 (39)||2.7||−0.17 to 0.11|
|United States||75||48 (64)||27 (36)|
|Goal is hospitalization prevention||14||12 (86)||2 (14)||−24.8||−0.44 to −0.05|
|Goal is not hospitalization prevention||185||113 (61)||72 (39)|
|Pediatric ED||115||74 (64)||41 (36)||−4.8||−0.19 to 0.09|
|General ED||85||50 (60)||34 (41)|
Previous Encounters of Mg-related Side Effects
Of the 199 respondents, 179 (90%) have used Mg. Of these, 42 (24%) have personally observed at least one episode of a severe drop in blood pressure during Mg infusion that required treatment. Twenty-two have witnessed this event on multiple occasions. Apnea necessitating airway stabilization was seen by four physicians (once by three and three times by one). Heart block was not encountered by any participant. There appears to be no significant association between observing side effects and the dose of Mg given (p = 0.17).
Barriers to IV Mg Use
The perceived barriers are itemized in Table 5. Concern regarding severe side effects of IV Mg influences its use in almost a quarter of participants, and a perceived lack of need for an IV plays a role in approximately a third. Of the responders who do not use IV Mg in severe acute asthma (n = 20), the most common reasons cited included a lack of familiarity with the drug (10 physicians) and a belief that there exists an insufficient amount of evidence of benefit (nine respondents).
|Characteristic||Overall (N = 199)||Canada (n = 124)||United States (n = 75)|
|My decision to use Mg is influenced by the following (% agree/strongly agree):|
|IV invasiveness||26 (13.1)||16 (13.0)||10 (13.3)|
|Need for skilled nursing staff familiar with Mg infusion||22 (11.1)||14 (11.4)||8 (10.7)|
|Concern re: hypotension, apnea, or heart block||46 (23.2)||31 (25.2)||15 (20.0)|
|Most severe asthmatics recover without IV||61 (30.8)||36 (29.3)||25 (33.3)|
Almost all responding pediatric EPs are aware of the evidence supporting the use of IV Mg in severe acute asthma in children. However, our study shows that they rarely use this agent in children with stable severe exacerbations. The usual goal of Mg therapy in clinical practice is to prevent ICU admissions, rather than to avoid hospitalizations. The reluctance to use it more often partly stems from concern about side effects, since associated hypotension was reported by almost 25% of participating physicians.
In a brief physician survey and a Multicenter Airway Research Collaboration chart review, Rowe and Camargo9 found that general EDs use IV Mg rarely, and its use is associated with disease severity, lack of response to initial therapy, and use of additional pharmacointerventions. However, the authors did not address goals of therapy, barriers to the use of this agent, or observed toxicity. Furthermore, the participating EDs treat mainly adults, and the results may not be generalizable to children.
Gaps between published evidence and its incorporation into practice have previously been documented in other acute pediatric respiratory disorders, such as the use of metered dose inhalers, systemic corticosteroids,22–24 and the evidence-based use of pharmacotherapy in croup.25,26 Indeed, recent literature suggests that a substantial proportion of all patients do not receive evidence-based care27,28 and that various pediatric asthma guidelines developed in the past two decades have not decreased hospitalizations, ED visits, or deaths from this disease.29,30 Effective steps to narrow the gap include enhancing physicians’ belief that the intervention will lead to improved outcomes in a safe manner,31 as well as having a local devoted champion to undertake change.32–35 However, sufficiently strong evidence is required prior to the undertaking of any knowledge translation efforts. Of the four pediatric studies that have addressed the benefit of IV Mg in children, many patients were enrolled by a single author who determined study eligibility by the peak expiratory flow rate.11,36–39 Furthermore, the studies were small and excluded preschoolers. Because many physicians do not routinely employ pulmonary function testing due to its lack of accuracy in children with severe asthma, the clinical relevance of this entry criterion remains questionable. The severity of asthma among the children enrolled in these trials may differ greatly from those treated with IV Mg in everyday life, which may partially explain the relatively low reported use of this medication. Most hospitalized children are relatively stable, and some might have been successfully sent home from the ED following a timely use of IV Mg and observation. Of note, the study by Scarfone and colleagues39 using the Becker pulmonary index score did not show benefit of a high dose of IV Mg, possibly due to inclusion of children with moderate disease. Knowledge translation research may enhance generalizability of study results, improve physicians’ comfort with the use of IV Mg,40 and help define parameters for treatment by employing simple validated treatment criteria such as the Paediatric Respiratory Assessment Measure.41,42
Reports of drug-related toxicities usually rely on case reports and reports of observed adverse events to the relevant national organizations. The side effects of IV Mg reported in the literature have been very minor and infrequent,42 and severe events such as those we have specifically enquired about have not been published. Our findings show that severe side effects of IV Mg may not be as rare as previously believed, and concern about them appears to represent a valid reason for many clinicians to be reluctant to use IV Mg. Although it is difficult to ascertain from our research if these adverse events are related to deviations from suggested rates of Mg administration, future tracking of Mg side effects in the course of research and in registries would be important, as it would shed further light on this issue. In the meantime, this observation underscores the importance of careful patient monitoring during IV Mg therapy.
While the PERC database includes almost all pediatric EPs in Canada, many U.S.-based pediatric EPs are not members of the PEM-CRC. Therefore, our findings may not be generalizable to all U.S.-based pediatric EPs; opinions from other physician specialties such as general EPs and hospitalists may also not be captured by this survey’s results.
The greatest limitation is recall bias, as the frequency of IV Mg use, associated side effects, and usual disposition are subject to recall and are not represented by hard numbers. Our survey may be also limited by a self-selection bias. Although the respondents and nonrespondents were comparable, it is possible that a relatively higher proportion of responders were interested in asthma and use Mg and had therefore encountered more related adverse events compared to their abstaining colleagues. This in turn may have overestimated the true use of this medication and its side effects. Although this study suffers the usual limitations of survey design, there is no other practical method to obtain this important information. However, it is reassuring that our response rate is well above that of 52% for similar surveys,43,44 and that virtually all results were comparable on both sides of the border. These cross-border similarities could in turn facilitate future knowledge translation research.
Intravenous magnesium appears to be uncommonly used in stable children with severe acute asthma by pediatric emergency physicians with interest in clinical research. Although the literature supports the fact that intravenous magnesium may allow many children to be discharged home, most of the physicians we surveyed do not currently use it for this purpose. The frequency of severe side effects attributed to intravenous magnesium seems greater than reported previously, which may represent a valid barrier to the use of this agent. A large randomized controlled trial in select children followed by knowledge translational research of this intervention with side effect tracking appear warranted to assess the true effect on hospitalization rates in severe asthma and related side effects.
The authors thank Martin Pecaric for assistance with the survey and Catherine Wong for administrative assistance with the manuscript.
- 6British Thoracic Society. British Guideline on the Management of Asthma. British Thoracic Society/Scottish Guidelines Intercollegiate Network. Available at: http://www.sign.ac.uk/guidelines/fulltext/101/index.html. Accessed Aug 10, 2010.
- 7Canadian Asthma Consensus Group. Management of patients with asthma in the emergency department and in hospital. Can Med Assoc J. 1999; 161(11 Suppl):S53–59.
- 8Cincinnati Children’s Hospital Medical Center. Evidence Based Clinical Practice Guideline for Managing an Acute Exacerbation of Asthma. Available at: http://www.cincinnatichildrens.org/svc/alpha/h/health-policy/ev. Accessed Nov 11, 2009.
- 10Magnesium Sulphate: Adverse Effects. London, UK: The Pharmaceutical Press, 1989, p 1033..
- 12Mail and Internet Surveys: The Tailored Design Method. New York, NY: Wiley, 1999..
- 13Health Measurement Scales: A Practical Guide to Their Development and Use, 2nd ed. Oxford, UK: Oxford University Press, 1998., .
- 17Paediatric Society of New Zealand. Management of Asthma in Children Aged 1–15 Years. Available at: http://www.paediatrics.org.nz/index.asp?pageID=2145864725. Accessed Aug 10, 2010.
- 19Sample Size Methodology. New York, NY: Academic Press, 1990., .
- 20Statistical Intervals. New York, NY: John Wiley & Sons, 1991., .
- 21Applied Regression Analysis and Other Multivariable Methods, 3rd ed. Pacific Grove, CA: Duxbury Press, 1998., , , .
- 30The state of childhood asthma, United States, 1980–2005. Adv Data. 2006; 381:1–24..
- 43Response rates to surveys published in medical journals. J Clin Epidemiol. 1997; 10:1229–36., , .