Modifiable Risk Factors for Suboptimal Control and Controller Medication Underuse Among Children With Asthma
OBJECTIVES. Our aims were (1) to describe rates of suboptimal control and controller medication underuse in a diverse population of children with asthma and (2) to identify potentially modifiable parental behaviors and beliefs associated with these outcomes.
METHODS. We conducted telephone interviews with parents of 2- to 12-year-old children with persistent asthma, in a Medicaid plan and a large provider group. Suboptimal control was defined as ≥4 symptom days, ≥1 symptom night, or ≥4 albuterol use days in the previous 2 weeks. Controller medication underuse was defined as suboptimal control and parent report of <6 days/week of inhaled steroid use. Multivariate analyses identified factors that were independently associated with suboptimal control and controller medication underuse.
RESULTS. Of the 754 study children, 280 (37%) had suboptimal asthma control; this problem was more common in Hispanic children (51%) than in black (37%) or white (32%) children. Controller medication underuse was present for 133 children (48% of those with suboptimal asthma control and 18% overall). Controller medication underuse was more common among Hispanic (44%) and black (34%) children than white (22%) children. In multivariate analyses, suboptimal control was associated with potentially modifiable factors including low parental expectations for symptom control and high levels of worry about competing household priorities. Controller medication underuse was associated with potentially modifiable factors including parental estimation of asthma control that was discordant with national guidelines and no set time to administer asthma medications.
CONCLUSIONS. Deficiencies in asthma control and controller medication use are associated with potentially modifiable parental beliefs, which seem to mediate racial/ethnic and socioeconomic disparities in suboptimal control and controller medication underuse.
Asthma, one of the most prevalent chronic childhood conditions, is the most frequent cause of hospitalization among children.1 Evidence-based national guidelines for asthma care suggest that asthma morbidity can be prevented with appropriate medications and outpatient management practices.2 However, ∼40% of children with asthma have suboptimal symptom control, and a similar proportion do not use daily controller medications, as recommended by national guidelines.3 Underuse of inhaled corticosteroids is documented to increase the risk of preventable morbidity, including emergency department visits and hospitalizations.4–7
The reasons why so many children experience suboptimal asthma control and controller medication underuse are not well understood. Previous studies observed that, in the United States, individuals who have racial/ethnic minority or lower socioeconomic status are at increased risk for controller medication underuse.8–11 There remain questions regarding whether the observed racial and ethnic differences reflect differences in socioeconomic status. Limited research exists regarding risk factors for asthma undertreatment that are potentially modifiable with appropriately tailored interventions, such as parental beliefs and behaviors. Studies of relatively small groups have suggested that parental expectations about treatment and concerns about medications may be barriers to controller medication use.12,13 However, no large study has simultaneously evaluated potentially modifiable factors, such as parental perceptions of asthma control, beliefs, behaviors, and competing household priorities, identified as potentially relevant from the literature and our previous qualitative research.14 Our aims were to (1) describe the prevalence of both suboptimal control and controller medication underuse in a diverse population of children with asthma, (2) to identify potentially modifiable parental behaviors and beliefs associated with these problems, and (3) to compare the relative influences on these problems of these potentially modifiable risk factors and the more extensively studied socioeconomic factors.
Study Design and Data Collection
This was a cross-sectional study of parents of children 2 to 12 years of age with persistent asthma. Children were identified from the populations of either Neighborhood Health Plan, a large, Medicaid-predominant, health maintenance organization, or Harvard Vanguard Medical Associates, a large multispecialty group serving both privately insured and Medicaid-insured patients in the Boston, Massachusetts, area. Children with probable persistent asthma were identified from computerized claims records and electronic medical records by using criteria similar to those of the Health Plan Employer Data Information Set.15 Patients were potentially eligible if they had an International Classification of Diseases, Ninth Revision, code for asthma (493.xx) at a hospitalization, emergency department visit, or outpatient visit plus any of the following during the previous 12 months: (1) ≥4 prescriptions for any asthma medication, (2) ≥1 asthma-related hospitalization or emergency department visit, or (3) ≥4 asthma-related outpatient visits and ≥2 prescriptions for asthma medications.
Trained interviewers administered a structured, 30-minute telephone interview to parents in either English or Spanish. Surveys were mailed to families not reached after 12 telephone attempts. Participants received a $20 gift certificate after completion of the interview. On the basis of computerized data, 1594 patients met eligibility criteria. Of those, 547 were ineligible (449 were unreachable because of disconnected, nonworking, or wrong telephone numbers; 36 could not be interviewed in English or Spanish; and 62 denied having been diagnosed with asthma). Of the 1047 eligible families, 754 (72%) completed telephone interviews (690 families) or mailed surveys (64 families), whereas the remaining 293 declined (133 actively refused and 160 passively refused, on the basis of repeated nonanswers, answering machine pick-ups, or repeated requests for call-backs). The study was approved by the institutional review boards of Boston University Medical Center and Harvard Pilgrim Health Care.
Definitions of Outcomes and Predictors
This study evaluated 2 outcomes, that is, suboptimal asthma control and controller medication underuse. Our definitions of suboptimal control and controller medication underuse were consistent with the 2007 National Asthma Education and Prevention Program guidelines.2 Asthma symptom frequency, including wheeze, chest tightness, cough, shortness of breath, activity limitation, and nocturnal symptoms, and days of controller medication use were assessed for the 2 weeks before the interview, on the basis of parent report. Suboptimal control was defined as having 1 of the following: (1) ≥4 days of symptoms in the past 2 weeks, (2) ≥2 nights of symptoms in the past month, or (3) ≥4 days of albuterol use in the past 2 weeks. This definition encompasses children who would be classified as having either not well controlled or very poorly controlled asthma, according to the 2007 National Asthma Education and Prevention Program guidelines.
Parents also reported whether their child had ever been given prescriptions for controller medications, including inhaled corticosteroids, cromones, and leukotriene modifiers, and, if so, whether the medications were supposed to be used daily or only during certain periods, such as symptomatic periods. Controller medication underuse was defined operationally as both (1) having suboptimal control and (2) not using a controller medication ≥6 days/week.2,16
For the predictor variables, we collected data in the following domains: demographic characteristics, asthma-related clinical use, parental health beliefs about asthma, parental concerns about asthma medications, parental definitions of good asthma control and expectations for child's functioning, family asthma medication routines, competing family priorities, experience of racial or ethnic bias, and provider-parent communication. Each child's race/ethnicity was classified as black, Hispanic, white, or other (including those with >1 race identified), on the basis of parent report. Children identified as Hispanic were included in the Hispanic group regardless of race. Respondents also reported the language they were most comfortable speaking, because language spoken at home and limited English proficiency have been shown to be related to child health status and health care quality, even after adjustment for race/ethnicity.17–19 Computerized data on the numbers of asthma-related outpatient visits, emergency department visits, and hospitalizations and the number of prescriptions for orally administered steroids filled in the previous year were used to adjust for asthma morbidity.
Our measure of parental expectations for child's functioning was based on parents' level of agreement with 4 items developed by Yoos et al,13 as follows. (1) “I believe that my child can be symptom-free most of the time.” (2) “I expect that asthma will not affect my child's school/day care attendance.” (3) “I expect that my child can fully participate in gym, normal physical activity, and playing.” (4) “I expect that my child will have no emergency department visits or hospitalizations because of asthma.” Each item was scored on a scale of 0 to 4. Using a summary score of 0 to 16, based on the total of the 4 items, we defined low parental expectations as scores of 10 to 16, which resulted from disagreeing or strongly disagreeing with ≥2 of the 4 statements. Parents rated their child's asthma symptom control on a 5-point Likert scale, from excellent to poor. Parents' definition of good asthma control was assessed in 2 ways, based on their responses to the following questions. (1) “When your child's asthma is in good control, how many days per week does he/she have asthma symptoms?” (2) “If your child had asthma symptoms 2 days per week, how would you rate his/her control?” Parents' estimation of asthma control was defined as discordant with national guidelines when the child had frequent symptoms (≥4 days of asthma symptoms or albuterol use in the previous 2 weeks or 1 night of nocturnal symptoms in the same period) but parents rated the child's asthma control as excellent or very good. Parental concerns about asthma medication, measured with items based on previously developed scales, were combined into a summary score of 0 to 16, with high levels of medication concerns defined as scores of 10 to 16.20,21 We also assessed parents' perception of asthma as either a chronic or intermittent condition.
We assessed family medication routines, including whether there was a set time for taking asthma medications.22–24 Competing family priorities were assessed on the basis of parental ratings of frequency of worry during the past 2 months in the following areas: home or neighborhood safety, job, personal/family relationships, income/making ends meet/keeping up with bills, parents' own health, and other family member's health (K. Volpp, MD, PhD, unpublished data, 2004). The experience of racial bias was defined as a positive response to either of the following questions, modified from previously developed items.25,26 “Was there ever a time when doctors paid less attention to you or your child because of your race or ethnicity?” “Was there ever a time when your child would have gotten better care if she/he had belonged to a different racial or ethnic group?” We also assessed parents' ratings of communication and trust in clinicians, including how often the child's doctor listened carefully and spent enough time with the parent and child.27
We conducted bivariate analyses to identify predictors associated with each of the 2 outcomes of interest (suboptimal control and controller medication underuse), using the χ2 test of association for categorical variables and t tests for continuous variables. For analyses of controller medication underuse, we compared children who had underuse (suboptimal control and not using controller medications ≥6 days a week) with those who used controller medications ≥6 days a week. For each outcome, we created preliminary multivariate models that included age, gender, race, and the variables that were significant at the P ≤ .20 level in bivariate analyses. In the final models, we included age, gender, race, and predictor variables found to be significant at P ≤ .10 in the initial multivariate models. We created generalized linear mixed models to adjust for clustering of patients according to provider.28
The odds ratios from multivariate models sometimes do not provide an intuitive picture of effect sizes, especially when an outcome is relatively common or when an independent variable is measured on a continuous scale. To illustrate how changes in each independent variable influenced the probability of having each outcome, we constructed examples with hypothetical patients at medium risk of each outcome by setting all variables at medium-risk levels. We then calculated the different predicted probabilities of the outcome resulting from varying the value of each variable in the model while holding the other variables at their medium-risk levels and inserting the covariate values of interest into the regression equation and inverting the logistic function.29 Analyses were conducted with SAS 9.1.3 (SAS Institute, Cary, NC).
Study Population and Outcomes
Of the 754 study children, 26% were black, 21% Hispanic, and 42% white; 11% were classified as “other” (Table 1). Among the Hispanic children, 52% were Puerto Rican, 16% were Dominican, and 6% were Central American; the rest were from other Hispanic subgroups, with Mexican and Cuban children representing <1% each. One fourth of parents had no education beyond high school, and 36% had household incomes of $30000 or less. Most parents reported that their child's doctor had prescribed controller medications to be used either daily (39%) or only when the disease was symptomatic or during certain periods (34%), whereas 13% indicated that controller medications had been prescribed in the past and 14% that these medications had never been prescribed (Table 1).
On the basis of parent reports, 280 children (37%) had suboptimal asthma control. Among the children with suboptimal asthma control, 133 children (48%) were classified as having controller medication underuse (Table 1).
Parental Perceptions of Asthma Symptoms, Control, and Medications
Most parents (62%) rated their child's asthma symptom control as either excellent or very good (Table 2). However, 12% of parents' ratings of symptom control were discordant with national guidelines, in that they reported persistent symptoms but rated the child's control as excellent or very good. Approximately one third (35%) of parents perceived their child's asthma as an intermittent, rather than chronic, condition.
On the basis of the summary measure of expectations, 27% of parents had low expectations for their child's functioning with asthma (Table 2). In addition, 60% of parents reported that, if their child had symptoms 2 days/week, they would consider that to be excellent, very good, or good control. When asked to define good asthma symptom control, 22% of parents reported that this would be having symptoms ≥2 days/week. Thirty-five percent of parents reported that their family had no set time each day for the child to take asthma medications. On the basis of the summary measure of medication concerns, 39% of parents had a high level of concern.
Parents rated communication with clinicians highly, with 76% reporting that their child's doctor always or often listened carefully to them, gave clear instructions about how to care for their child's symptoms, and spent enough time with them and their child. Almost all parents (92%) reported that their child's doctor had excellent or very good knowledge of their child's medical history and of him or her as a person. With regard to reports of discrimination, 10% of parents reported that they ever felt that they or their child had been judged unfairly by doctors or medical staff members or treated with disrespect (Table 2).
Predictors of Suboptimal Asthma Control
Suboptimal control was more common in Hispanic children (51%) than in black (37%) or white (32%) children (P < .001). Other demographic factors associated with suboptimal control in bivariate analyses included lower household income (P < .001), lower parental education (P < .01), parental unemployment (P < .05), and speaking Spanish at home (P < .05). In addition, parents with low expectations for asthma control, those who defined symptoms 2 days/week as good control, and those with more competing family priorities were significantly more likely to have children with suboptimal control (Table 2). There was no association between measures of clinician communication and suboptimal control in bivariate analyses.
In preliminary multivariate analyses, the association between race/ethnicity and suboptimal asthma control was diminished after adjustment for household income and parental employment. The relationship between Hispanic ethnicity and suboptimal control remained borderline significant (adjusted odds ratio: 1.6; 95% confidence interval: 1.0–2.5). In the final multivariate analysis, low parental expectations for asthma control and competing family priorities were the 2 domains significantly associated with suboptimal control (Table 3). With these factors included in the model, the associations between demographic variables, including race/ethnicity, and suboptimal control were not significant, which suggests that parental expectations and competing priorities may mediate the association of these demographic variables with asthma control.
Patterns and Predictors of Controller Medication Underuse
Among parents of the 133 children with controller medication underuse, 50 (38%) reported that they had been instructed by their child's doctor to use controller medications on a periodic or as-needed basis. Another 38 (29%) said that controller medications had never been prescribed for the child, whereas 24 (18%) said controller medications had been prescribed in the past but were not currently and 20 (15%) said they were supposed to be using controller medications daily.
Controller medication underuse was more common among Hispanic (44%) and black (34%) children than among their white counterparts (22%; P = .001). Lower household income, lower parental education, single parenting, and being most comfortable speaking Spanish were significantly associated with controller medication underuse in bivariate analyses.
Children with controller medication underuse were significantly more likely to have parents whose estimation of asthma control was discordant with national guidelines (32%), compared with children without underuse (15%; P < .0001) (Table 2). Children with controller medication underuse also were significantly more likely to have parents who viewed asthma as an intermittent condition, defined symptoms 2 days/week as adequate control, had more concerns about asthma medications, and did not have a set time each day for the child to take medications (Table 2). There was no association between measures of clinician communication and trust and controller medication underuse in bivariate analyses.
In preliminary multivariate analyses, the association between race/ethnicity and controller medication underuse decreased after adjustment for parental education. In the final multivariate analysis, controller medication underuse was significantly associated with discordant parent estimation of asthma control, no set time to take asthma medications, and concerns about asthma medications (Table 3). When these factors were included in the model, the associations between demographic variables, including parental education, and controller medication underuse were no longer significant, which suggests that parental perceptions may mediate the association between parental education and controller medication underuse.
Examples Illustrating the Impact of Modifiable Risk Factors
Tables 4 and 5 illustrate the impact of individual factors on the probability of suboptimal control and controller medication underuse by changing the value of each variable from its lowest-risk level to its highest-risk level while holding all other variables in the final model constant at their medium-risk levels. For suboptimal control, the hypothetical medium-risk patient had a predicted probability of 52% of this outcome, when all variables were set to their medium-risk values (Table 4). For race/ethnicity, there was a 15% increase in probability of suboptimal control when this variable was changed from its lowest-risk to highest-risk value. This relatively modest difference reflects the modest independent effect of race on the risk of suboptimal asthma control. In contrast, the largest increase in predicted probability (37%) occurred when the parental expectation variable was varied from its lowest-risk to highest-risk value.
For controller medication underuse, the hypothetical medium-risk patient had a predicted probability of this outcome of 53% (Table 5). For race/ethnicity, there was a 12% increase in probability of underuse when this variable was changed from its lowest-risk to highest-risk value. This relatively small difference reflects a modest independent effect of race/ethnicity on the probability of controller medication underuse. In contrast, variables measuring parental perceptions had larger effects on the probability of medication underuse. In particular, the largest increase in probability of controller medication underuse (56%) occurred when the medication concern variable was changed from its lowest to highest value. The impacts of discordant parental estimation of asthma control and having a set time for taking medication were also substantial, with changes in probability of medication underuse of 28% and 26%, respectively.
Our findings revealed that several modifiable factors were stronger and more-proximal correlates of suboptimal asthma control and medication underuse than were other, nonmodifiable factors, including race/ethnicity and socioeconomic status. This article presents novel results indicating that low parental expectations for asthma outcomes and competing family priorities were independent and important correlates of suboptimal asthma control. Parental estimation of asthma control that is discordant with guidelines, lack of family routines for taking medication, and parental concerns about medications were significantly associated with controller medication underuse. This study was unique in our use of a large diverse population to evaluate simultaneously the independent influences of many potentially modifiable factors on suboptimal asthma control and controller medication underuse.
Racial/Ethnic and Socioeconomic Disparities
Our study found significant racial and ethnic differences in asthma symptom control and patterns of controller medication use. Hispanic children were more likely to have suboptimal control and both Hispanic and black children were more likely to underuse controller medications than were their white counterparts. However, multivariate analyses indicated that these differences were mediated by differences in parental socioeconomic status, and these in turn were mediated by the potentially modifiable parental beliefs and behaviors we identified. We also found that parents who reported being most comfortable speaking Spanish were more likely to have children who underused controller medications. Language use issues in health care settings are likely more than solely a proxy of acculturation and may reflect availability, quality, and use of interpreter services, as well as availability and fluency of multilingual providers.17 In related analyses, we found that parents of black and Hispanic children had lower expectations for functioning than did white parents, even after controlling for age, gender, and household income.30 These modifiable factors may help explain residual racial/ethnic and socioeconomic disparities found in previous studies.
Findings on Suboptimal Asthma Control
Our results are consistent with an earlier study of 133 parents that suggested that low expectations for treatment outcomes may contribute to nonadherence for controller medications.13 Our study found that low expectations for functioning were associated with suboptimal asthma control. We hypothesize that low expectations may stem from a family's previous experiences of poor asthma control, which lead parents to think that good control is not achievable. This belief, in turn, may diminish their interest in and daily use of controller medications, creating a cycle we term the “hegemony of low expectations.” These results are consistent with a 2001 study that concluded that symptomatic children with asthma are often not brought to medical attention.31 Therefore, clinician interventions to improve asthma control may need to focus on identifying families with low expectations for functioning with asthma by proactively probing for gaps between what parents think is achievable and clinician ideals for optimal asthma control. Clinicians can then work with parents to raise their expectations, to break the cycle of low expectations and poor asthma control. Targeted prompting of providers about asthma symptoms and recommended practices may be one effective way to enhance preventive care.32
Our finding that competing family priorities were associated with suboptimal control of asthma is novel and lends credence to the idea that families may have relatively finite “worry budgets” (K. Volpp, MD, PhD, unpublished data, 2004). Previous work identified competing priorities as important barriers to the use of preventive health care services for children but did not connect them to specific health outcomes as we did in this study.33 It seems that family worries, such as food insecurity, housing instability, or job problems, can overshadow asthma control as priorities for parental attention. Clinicians can incorporate screening for these issues as part of their comprehensive social history and then refer their patients to appropriate support services in the practice setting or in the community.34 Emerging models of innovative multidisciplinary partnerships in the clinical setting address these types of competing priorities by ensuring that families' basic needs are met.35
Findings on Controller Medication Underuse
Our results suggest that intentional nonadherence to provider recommendations may not be the primary driver of controller medication underuse in children with persistent asthma symptoms. Among symptomatic children whose parents said they were not currently using controller medications daily, only 15% reported that they had been told to use them daily and almost one half reported either never having been given a prescription for such medications or not having a current prescription. In this group, more than one third of parents said they had been told to use controller medications only on a periodic or symptomatic basis. These findings should be interpreted with caution, because another study suggested that parents often do not report that their child has used a controller medication even when a physician has prescribed one.20
This study also found that parental estimation of control discordant with national guidelines was independently associated with controller medication underuse. This suggests that the patterns of underprescribing described above are likely the shared responsibility of parents and physicians. Physicians may be unaware of patients' asthma status if parents do not think their children have symptoms warranting treatment. Other studies suggest that providers may also underestimate symptom severity.36
Research in adults suggested that as-needed use of inhaled steroids may be an effective alternative to daily use for patients with mild persistent asthma.16 If such an approach is found to be practical in children, then it will be even more important to ensure that parents are able to evaluate their children's asthma control accurately, so that they and their physicians can make appropriate judgments about when to start inhaled steroid treatment.
The other predictors of controller medication underuse we identified, family routines and concerns about medications, have been described in other studies.13,22,23 This study's new finding is that each is independently associated with underuse, which suggests that underuse is a multifactorial phenomenon not explained by a single factor. Many families in our study reported that they had not discussed medication concerns with their clinician. The clinical implications of these findings suggest that, to increase appropriate controller medication use, physicians may need to elicit proactively and to respond to concerns about controller medication use, as well as helping some families establish workable routines for administering these medications.
This study population consisted of children who were publicly or privately insured; therefore, the findings may not be generalizable to populations of uninsured children. The study relied on parent reports and 2-week recall of asthma symptoms and medication use, which do not have perfect accuracy but remain the basis for most national surveys.37,38 Symptoms reported from the previous 2 weeks could misrepresent the child's overall level of control (for example, if the child had an acute upper respiratory infection). However, such an effect probably would have decreased the precision of our measurement of suboptimal control in general, rather than creating systematic biases that would change our findings. Because parents tend to overreport asthma controller medication use dramatically,39 our results probably overestimate the actual prevalence of controller medication use. However, parents who say their children are not using controller medications are usually reporting accurately. Although overreporting of medication use would tend to bias against our ability to identify factors associated with underuse, the factors we identified are likely valid. Our definition of controller medication was intentionally broad, including inhaled corticosteroids, cromones, and leukotriene modifiers. Therefore, these findings likely underestimate the underuse of inhaled steroids, the preferred first-line treatment for persistent asthma.2 In addition, we did not attempt to evaluate the influence of specific beliefs or behaviors regarding identification and reduction of environmental or other asthma triggers on asthma symptom control.
This study suggests that suboptimal asthma control and controller medication underuse are highly associated with potentially modifiable risk factors, especially low parental expectations for functioning and symptom control, discordant estimation of asthma control, lack of routines for administering medication, and concerns about asthma medications. These factors also seem to mediate racial/ethnic and socioeconomic disparities in suboptimal control and medication underuse. Interventions to improve asthma control and to increase appropriate controller medication use, especially among vulnerable populations, should include specific strategies, a few of which have been suggested here, to identify and to address the modifiable factors identified in this study. Such interventions have the potential to improve asthma care and outcomes generally and to address persistent disparities in asthma morbidity.
This work was supported by a grant from the National Institute of Child Health and Human Development (grant R01 HD044070). Dr Bokhour's effort was supported in part by the Department of Veterans Affairs, Health Services Research and Development Service. Dr Lieu's effort was supported in part by a Mid-Career Investigator Award in Patient-Oriented Research from the National Institute of Child Health and Human Development (grant K24 HD047667).
We offer our appreciation to Jack Lasche, MD and the pediatric clinicians of Harvard Vanguard Medical Associates and Pauline Sheehan, MD and the pediatric clinicians of Boston Medical Center for their support of this work. We are grateful to Charlene Gay for her capable assistance in coordinating this project and to Elizabeth Harrison and Erica Tobey for their help in preparing the manuscript. We appreciate the support and advice from James Glauber, MD, MPH, and David Yang, MS, on the Neighborhood Health Plan population. We also extend our thanks to our research assistants, Heather Brymer, Stephanie Love, Amarylis Perez, Tamaris Salerna, Patti Steele, Susan Swords, Elizabeth Suda, and Marlene Taveras, and to the patients and providers who contributed to this study. We gratefully acknowledge Dana Safran, ScD, and Anita Stewart, PhD, for their expert advice on patient-reported measures of care and Leanne Yinusa-Nyahkoon, MS, for her contribution regarding family medication routines.
- Accepted December 21, 2007.
- Address correspondence to Lauren A. Smith, MD, MPH, Commonwealth of Massachusetts Department of Public Health, 250 Washington St, 2nd Floor, Boston, MA 02108. E-mail:
The views expressed in this article are those of the authors and do not necessarily represent the views of the Department of Veterans Affairs.
Dr Smith's current affiliation is the Massachusetts Department of Public Health.
The authors have indicated they have no financial relationships relevant to this article to disclose.
What's Known on This Subject
Suboptimal control of asthma symptoms and underuse of controller medications are common, especially among minority children, and the underlying factors are not well understood.
What This Study Adds
Our findings revealed that several modifiable parental beliefs and expectations were stronger and more-proximal correlates of suboptimal asthma control and medication underuse than were other, nonmodifiable factors, including race/ethnicity and socioeconomic status.
- ↵Merrill CT, Elixhauser A. Top 10 principal diagnoses by age group. In: HCUP Fact Book 6: Hospitalization in the United States. AHRQ Publication No. 05-0056, June 2005. Rockville, MD: Agency for Healthcare Research and Quality; 2005. Available at: www.ahrq.gov/data/hcup/factbk6/. Accessed June 1, 2007
- ↵National Asthma Education and Prevention Program. Expert Panel Report 3: Guidelines for the Diagnosis and Management of Asthma: Full Report 2007. Bethesda, MD: National Institutes of Health; 2007
- Homer CJ, Szilagyi P, Rodewald L, et al. Does quality of care affect rates of hospitalization for childhood asthma? Pediatrics.1996;98 (1):18– 23
- ↵Ordoñez GA, Phelan PD, Olinsky A, Robertson CF. Preventable factors in hospital admissions for asthma. Arch Dis Child.1998;78 (2):143– 147
- Lieu TA, Lozano P, Finkelstein JA, et al. Racial/ethnic variation in asthma status and management practices among children in managed Medicaid. Pediatrics.2002;109 (5):857– 865
- ↵Mansour ME, Lanphear BP, DeWitt TG. Barriers to asthma care in urban children: parent perspectives. Pediatrics.2000;106 (3):512– 519
- ↵Riekert KA, Butz AM, Eggleston PA, Huss K, Winkelstein M, Rand C. Caregiver-physician medication concordance and undertreatment of asthma among inner-city children. Pediatrics.2003;111 (3). Available at: www.pediatrics.org/cgi/content/full/111/3/e214
- ↵Riekert KA, Thompson R, Butz AM, et al. Assessing medication health beliefs of caregivers of high-risk inner-city children with asthma. Am J Respir Crit Care Med.2003;167 (7):A155
- ↵Commonwealth Fund, Princeton Survey Research Associates. Survey on Disparities in Quality of Health Care: Spring 2001. New York, NY: Commonwealth Fund; 2001. Available at: www.commonwealthfund.org/usr_doc/qualitysurvey_2001_questionnaire.pdf?section=4056. Accessed June 21, 2007
- ↵Agresti A. Categorical Data Analysis. 2nd ed. Hoboken, NJ: Wiley; 2002
- ↵Kenyon C, Sandel M, Silverstein M, Shakir A, Zuckerman B. Revisiting the social history for child health. Pediatrics.2007;120 (3). Available at: www.pediatrics.org/cgi/content/full/120/3/e734
- ↵Zuckerman B, Sandel M, Smith L, Lawton E. Why pediatricians need lawyers to keep children healthy. Pediatrics.2004;114 (1):224– 228
- ↵Edwards SW, Winn DM, Kurlantzick V, et al. Evaluation of National Health Interview Survey diagnostic reporting. Vital Health Stat 2.1994;(120):1– 116
- Copyright © 2008 by the American Academy of Pediatrics