The Behavioral Risk Factor Surveillance System (BRFSS) survey is a state-based, ongoing, random-digit-dialed telephone survey of civilian, noninstitutionalized adults aged ≥18 years residing in the United States. Since 1984, BRFSS has been collecting data on risk behaviors and the use of preventative health practices that affect health status (14). The BRFSS Asthma Call-back Survey (ACBS) is a survey that has been implemented as a follow-up survey to the BRFSS survey since 2005. The BRFSS survey contains a Random Child Selection module and the Child Asthma Prevalence module, which were used in participating states to identify households with a child who had asthma to administer the Child ACBS.

The BRFSS survey respondents (i.e., adults or children) with an affirmative response to the question “Have you {has the child} ever been told by a doctor, nurse, or other health professional that you {the child} had asthma?” were asked to participate in the adult or child ACBS within 2 weeks to provide more information about their or the child's asthma. Only one adult or one child per household could participate in the ACBS, which obtains in-depth information about asthma symptoms and episodes/attacks, self-management education, healthcare utilization and access, medication use, comorbidities, and environmental allergens and irritants (15). An adult family member served as a proxy respondent for children. A sampling weight was assigned to each survey respondent in the ACBS sample to adjust for selection probabilities and estimates of the age–sex–race distribution of adults and children in the participating state each year. More information on weight calculation can be found in the 2006–2008 ACBS Summary Data Quality Report at http://www.cdc.gov/brfss/acbs/2007/2007.htm. The median ACBS response rates for adults were 50.6% in 2006 and 54.3% in 2007; the median response rates for children (via adult proxies) were 49.1% in 2006 and 51.5% in 2007.

Respondents were considered to have “active asthma” if they reported, during the previous 12 months, any of the following: having talked to a doctor or other health professionals about asthma, taken any asthma medications, or experienced asthma symptoms (16, 17).

A total of 31,278 persons (26,047 adults and 5231 children) were in the combined 2006 and 2007 BRFSS ACBS from the District of Columbia and 34 states (Alaska, Arizona, California, Colorado, Connecticut, Florida, Georgia, Hawaii, Illinois, Indiana, Iowa, Kansas, Maine, Maryland, Massachusetts, Michigan, Mississippi, Missouri, Montana, Nebraska, Nevada, New Hampshire, New Mexico, New York, Ohio, Oklahoma, Oregon, Pennsylvania, Texas, Utah, Vermont, Washington, Wisconsin, West Virginia) (15). Of the 31,278 persons who ever had asthma, 3953 (75.8%) children and 19,723 (72.8%) adults were classified as having active asthma and were included in this analysis.

Demographic characteristics (e.g., age, sex, race/ethnicity, and education) of survey respondents who met the definition of active asthma were defined as shown in Table 1. Asthma health outcome measures (e.g., “asthma episode” [one or more asthma episodes in the past 12 months] and asthma control status) and healthcare access and use were assessed for both adult and child populations with active asthma. For adults, health risk factors such as obesity, smoking status, and comorbid conditions (e.g., physician-diagnosed chronic obstructive pulmonary disease [COPD], emphysema, chronic bronchitis, and depression) were included in the analysis. Following the guidelines in the National Asthma Education and Prevention Program (NAEPP) Expert Panel Report 3, we classified asthma control as well controlled, not well controlled, and very poorly controlled by using three parameters: daytime symptoms, nighttime symptoms, and use of short-acting β₂-agonists (3, 16).

For adults, body mass index (BMI) was categorized as obese (30.0 kg/m² and above) and not obese (less than 30.0 kg/m²), and smoking status was classified as current smoker (adults who smoked at least 100 cigarettes in their lifetime and currently smoke, either once in a while or everyday) and nonsmoker (persons who do not currently smoke). For children, data for smoking status and comorbid conditions were not collected and the BMI was not calculated. The use of BMI to screen for overweight and obesity in children aged 0–2 years is not recommended and for children aged 2–19 years, the interpretation of BMI varies by age and sex (18).

Healthcare access and use was determined by the following measures: health insurance (had any type of health insurance), routine care visit (saw a doctor for a routine asthma care), urgent care visit (had urgent care visit in the past 12 months because of an asthma episode/attack), ED visit (visited an ED in the past 12 months because of their asthma), hospitalization (stayed overnight in a hospital because of their asthma during the past 12 months), and cost as a barrier (being unable to see a primary care physician or specialist for asthma care or unable to buy medication for asthma in the past 12 months).

Responses to seven questions were used to assess respondents' asthma education status: taught episode signs (Has a doctor or other health professional ever taught you how to recognize early signs or symptoms of an asthma episode?), taught episode response (Has a doctor or other health professional ever taught you how to respond to episodes of asthma?), taught peak flow meter (PFM) use (Has a doctor or other health professional ever taught you how to use a PFM to adjust your daily medications?), given an action plan (Has a doctor or other health professional ever given you an asthma management plan?), asthma management course (Have you ever taken a course or class on how to manage asthma yourself?), taught inhaler use (Did a doctor or other health professional show you how to use the inhaler?), and environment control advice (Has a doctor or other health professional ever advised you to change things in your home, school, or work to improve your asthma?). An asthma self-management education score with eight levels was created by counting the number of “yes” responses to the seven asthma education questions (from “yes” to all seven to “no” to all seven).

All analyses were conducted using SAS callable SUDAAN (v. 10.0) (v. 9.2, SAS Institute Inc., Cary, NC, USA.) to account for the complex survey design. Data from the states that participated in both years were reweighted to account for the differences in sample size in each year. Statistical significance was determined as p < .05 by a nondirectional z-test. Relative standard error (RSE = standard error/prevalence estimate) was used as a measure of an estimate's reliability (RSE of <0.30 indicates a “reliable” estimate) (19).

A multivariate (binary response) logistic regression model was used to determine the association between each component of asthma education and the explanatory factors. A multivariate (multinomial response) logistic regression model (proportional odds model) was used to assess the association between the asthma education score and the explanatory factors. The explanatory factors for all regression models were sociodemographic factors (e.g., age, sex, race/ethnicity for both children and adults, and education for adults only), asthma health outcomes (e.g., asthma episode and asthma control level), healthcare access and use (e.g., health insurance, cost as a barrier, routine care visit, ED visit, and hospitalization in the past 12 months), and health risk factors for adults only (e.g., obesity, smoking, COPD, and depression). The explanatory variables for the models shown in Tables 3 and 4 were selected using a backward selection process. All variables were included in the initial models. The least significant variable was removed and the model was refit until the resulting regression coefficients for all remaining explanatory variables were significant (p-values less than .05). The nonmodifiable variables (age, sex, and race) were kept in the models regardless of statistical significance status because of their importance in considering asthma self-management education (20, 21). As seen in Table 5, all explanatory variables were included in the proportional odds model to estimate the cumulative odds of asthma scores (sum of reported asthma education components).

Among the 3953 children with active asthma, 58.8% were boys, 51.2% were aged 0–4 years, and 59.6% were non-Hispanic White. About half (53.7%) reported an asthma episode in the past 12 months, 20.5% had not-well-controlled asthma, and 13.8% had very poorly controlled asthma. Almost all children had health insurance (96.8%), 73.8% reported a routine care visit, 35.1% an urgent care visit, 17.6% an ED visit, and 3.3% a hospitalization. Reporting cost as a barrier for care among children was lower (8.3%) than adults (17.0%). Reporting cost as a barrier was 3.5% for a primary care physician visit, 1.8% for a specialist visit, and 6.0% for buying medicine for their asthma during the past year (Table 1).

A majority of children with active asthma had at least one (98.0%) and approximately 80% had three or more of the seven asthma education components. The three most commonly reported asthma education components were being taught inhaler use (78.6%); being taught episode response (86.3%); and being taught episode signs and symptoms (82.0%). More than one-third of children were given an action plan (42.6%) and being taught PFM use (42.6%) (Table 2).

The results from the multivariate logistic regression analyses for children are presented in Tables 3 and 5. Sex, race and ethnicity, and health insurance status of children were not associated with any of the education components. Episode signs and symptoms instruction was associated with a routine care visit (adjusted risk ratio (ARR) = 1.2; 95% confidence interval (CI) = 1.1–1.3) and episode response instruction was associated with being hospitalized (1.1; 1.1–1.2). Children aged 12–17 years were 1.3 times (1.1–1.6) and children who reported a routine care visit were 1.3 times (1.1–1.5) more likely to report having PFM instruction than children aged 0–4 years or those who did not have a routine care visit, respectively. Having an action plan was associated with a routine care visit (1.4; 1.2–1.7) and an ED visit (1.3; 1.1–1.6). Having taken an asthma management course was associated with being hospitalized (2.4; 1.4–4.3). Instruction in inhaler use was associated with being aged 12–17 years (1.3; 1.2–1.4) and with having cost as a barrier for care (i.e., being unable to see a primary care physician or specialist for asthma care or unable to buy medication for asthma in the past 12 months) (1.2; 1.1–1.2). Children with not-well-controlled asthma were 1.4 times (1.2–1.6) more likely to report being given advice on environment control than children with well-controlled asthma (Table 3). Children who reported having multiple education components were associated with an asthma episode (Adjusted odds ratio (AOR) = 1.5; 95% CI = 1.1–2.0) and a routine care visit (1.7; 1.3–2.3) (Table 5).

Among the 19,723 adults with active asthma, 62.2% were female, 32.3% were aged 18–34 years, 75.3% were non-Hispanic White, and approximately 10% were not high school (HS) graduates. About one-fifth were current smokers (20.3%), 36.8% were obese, 33.7% were with COPD (reported having physician-diagnosed emphysema, chronic bronchitis, or COPD), while 35.3% reported having physician-diagnosed depression (Table 1).

About half (50.9%) reported an asthma episode in the past 12 months, 26.4% had not-well-controlled asthma, and 22.9% had very poorly controlled asthma. A majority (86.7%) of adults reported having health insurance, 52.5% reported a routine care visit, 23.0% an urgent care visit, 11.2% an ED visit, and 3.0% a hospitalization. Among adults overall 17.0% reported cost as a barrier for care. Reporting cost as a barrier was 5.3% for a specialist visit, 9.6% for a primary care physician visit, and 14.2% for buying asthma medicine during the past year (Table 1).

The majority (95.2%) of adults with active asthma had at least one of the seven asthma education components and approximately 71% of adults had three or more. The three most commonly reported asthma education components included being taught inhaler use (89.8%); being taught episode response (74.6%); and being taught episode signs and symptoms (64.4%). About one-fourth (26.8%) of adults with active asthma reported being given an action plan, 43.5% reported being taught PFM use, and 43.2% reported being given advice on environment control (Table 2).

The results from the multivariate logistic regression analyses for adults are presented in Tables 4 and 5. Each explanatory factor was associated with one or more of the education components except for obesity and cost as a barrier for care, which were not a predictor for any education component.

For instruction in asthma episode signs and symptoms, the prevalence rate ratios were lower for adults aged 65+ years (0.8; 0.7–0.9) than for adults aged 18–34 years and for adults with less formal education (not a HS graduate) (0.8; 0.7–0.9) than for those with a college degree or more. However, for instruction in asthma episode signs and symptoms, the ratios were higher (1.2; 1.1–1.3) for other-race adults than whites and higher (1.4; 1.3–1.5) for those who reported a routine care visit than for those who did not (Table 4).

For instruction in episode response, the prevalence rate ratios were lower for adults aged 65+ years (0.8; 0.8–0.9) than for adults aged 18–34 years and for adults who were not HS graduates (0.8; 0.7–0.9) than for those with a college degree or more. Adults who reported having an asthma episode (1.2; 1.1–1.2), those with health insurance (1.1; 1.1–1.3), and those who reported a routine care visit (1.2; 1.2–1.3) were more likely to report having instruction in episode response than the corresponding reference groups (Table 4).

PFM instruction was less likely for smokers (0.8; 0.7– 0.9) and more likely for Blacks (1.3; 1.1–1.4), adults who reported having asthma episodes (1.1; 1.1–1.3), a routine care visit (1.5; 1.4–1.7), or being hospitalized (1.4; 1.2–1.6) (Table 4).

Adults who reported an asthma episode (1.3; 1.2–1.5), a routine care visit (1.8; 1.6–2.0), or being hospitalized (1.4; 1.2–1.7) were more likely to be given an action plan compared with those who did not (Table 4).

Having taken an asthma management course was associated with several characteristics of survey respondents, whereas instruction in inhaler use was associated only with having a routine care visit (1.1; 1.1–1.1). Being older (aged 45–54 years: 1.6; 1.1–2.2 and aged 55–64 years: 1.5; 1.1–2.1) and being a member of other race (1.6; 1.1–2.4); having very poorly controlled asthma (1.5; 1.2–1.9); an asthma episode (1.4; 1.1–1.7); a routine care visit (1.4; 1.1–1.8); or ED visits (1.5; 1.1–1.9) were associated with taking an asthma management course (Table 4).

The prevalence rate ratios for advice on environment control were lower for older adults aged 55–64 years and 65+ years than for adults aged 18–34 years (0.8; 0.7–0.9 and 0.5; 0.5–0.6, respectively) and lower for adults who were not HS graduates (0.7; 0.5–0.8) than for those with college degrees or more, whereas the odds were higher for adults having COPD (1.2; 1.1–1.3) and reporting a routine care visit (1.3; 1.2–1.4) than corresponding rate ratios for adults who did not (Table 4).

Several personal characteristics of adults were associated with the asthma education score (number of reported asthma education components). Older adults (aged 55–64 years: AOR = 0.7; 95% CI = 0.6–0.9 and aged 65+ years: 0.4; 0.4–0.6), adults who did not graduate from HS (0.5; 0.4–0.7), and those who were smokers (0.7; 0.6–0.8) were less likely to report multiple asthma education components compared with those who did not have each characteristic. Adults who were female (1.2; 1.1–1.5), those reporting an asthma episode (1.5; 1.3–1.8), having health insurance (1.6; 1.2–2.0), reporting a routine visit (2.5; 2.2–2.9), and those who were hospitalized (1.7; 1.2–2.5) in past year were more likely to report multiple asthma education components compared with those who did not have those characteristics (Table 5).