Interactive Inpatient Asthma Education: A Randomized Controlled Trial Free

We conducted a prospective investigator-blinded, randomized controlled trial evaluating the effectiveness of 2 asthma education models at a large, urban, freestanding children’s hospital between October 2016 and June 2017. Eligibility criteria included being aged 2 to 16, diagnosis of asthma by a physician, and admission for asthma exacerbation. Patients with comorbid diseases (see Supplemental Information), those without a consenting caregiver available at the bedside or by phone, and those who did not speak English were also excluded. The Children’s Hospital institutional review board approved this study. This trial was registered with clinicaltrial.gov [registration code redacted for review].

Patients were randomized using stratified block randomization by asthma severity status (mild and moderate/severe; Fig 1). The randomization schedule was generated by a blinded biostatistician and implemented using Research Electronic Data Capture software.¹⁵  Caregivers were the primary recipients of the educational interventions, but children aged >10 were encouraged to participate, as well. Caregivers received gift cards for completing inpatient and 12 months postdischarge questionnaires.

Participants randomized to the control group received the current standard-of-care education provided at our institution: informational videos on asthma, which describe asthma triggers and symptoms, in-person didactic teaching by an asthma educator on metered-dose inhaler (MDI) use, and review of a prewritten standard AAP (written by the primary hospital physician). A trained health professional (a physician or a respiratory therapist) was assigned to provide control education during this study.

Although participants randomized to the interactive model also received the same informational videos on asthma, they actively developed a picture-based AAP with the help of an asthma educator. To do this, patients first reviewed all the medications on their AAP (written by the primary hospital physician) and then were handed a blank AAP and asked to place pictures of their medications and/or actions (eg, “Go to ED or call 911.”) in the correct zones of their AAP. MDI education used the validated Teach-to-Goal method (TTG).^16,17  This process and the use of the validated asthma checklist¹⁷  for both MDI with mask (score range 0–10) and MDI with mouthpiece (score range 1–12) administration, with a score >7 (mask) and >9 (mouthpiece), indicating competency.¹⁶  The critical steps of the checklists are further described in the Supplemental Information. A trained health professional (a physician or a registered nurse), different than those assigned to the control group, was assigned to provide interactive education during this study.

In both groups, educators were asked to record the start and stop times for their AAP- and MDI-technique teaching to assess the difference in the amount of time needed for each segment of the didactic and interactive educational interventions.

Upon entry into the study, caregivers were asked to complete several surveys: (1) Brief Health Literacy Screen,¹⁸  (2) Barriers to Asthma Care,¹⁹  (3) Newcastle Asthma Knowledge Questionnaire (NAKQ),²⁰  (4) Parent Asthma Management Self Efficacy (PAMSE) scale,¹⁹  and (5) the Patient-Reported Outcomes Measurement Information System (PROMIS) Parent Asthma version 2.0 survey.²¹  The Supplemental Information includes a complete description of the survey evaluation.

The primary outcome was the number of subsequent asthma ED visits for asthma exacerbation as recorded in the electronic medical record (EMR) at 6 and 12 months post-discharge. Secondary outcomes were hospitalizations at 6- and 12-months postdischarge (per the EMR); changes in PAMSE questionnaires between admission, discharge, and 12 months postdischarge; and changes in PROMIS Parent Asthma survey and MDI technique between admission and 12 months postdischarge. MDI technique was measured at baseline for all participants in the study by demonstration and at the 12-month follow-up by phone call, where participants were asked to list the steps for MDI administration. The 6- and 12-month time points were chosen to determine if the interactive model would impact asthma management skills for an extended period of time.

ED visit or hospitalization for asthma were compared between treatment groups using χ² tests for any visits and Fisher’s exact tests for the number of visits. A separate sensitivity analysis compared the number of ED visits and hospitalizations using the Wilcoxon rank sum test. Baseline Likert-scale responses to the Brief Health Literacy Screen questionnaire were compared using Wilcoxon rank sum tests. Differences among treatment groups for the change at baseline for the NAKQ, PAMSE, and PROMIS questionnaires at each time point were compared using separate repeated measures mixed-effects models and within-group change at baseline analyses were also evaluated in each model. TTG score at baseline and at the last assessment in the interactive group were compared using a 2-sample paired t test. Education times in total minutes were compared among treatment groups using linear regression models. All reported P values are 2-sided and CIs are reported at the 95% level.

To evaluate differences in the proportion of patients with ED visits between patients and caregivers receiving interactive education and patients and caregivers receiving didactic education, we performed a ×2 test for 2 independent proportions. We estimated 80% power with a sample size of 70 per group (140 total) to detect a difference of 22.5% in patients requiring >1 ED visit among groups at a 5% level of significance and assuming an anticipated attrition rate of 20% from loss to follow-up.⁸  The required sample size in each group was calculated with PASS 11 (NCSS Statistical Software, Kaysville, UT).

From October 2016 to June 2017, 330 patients were assessed for study participation (Fig 1); 182 were excluded on the basis of established criteria, 17 declined to participate, and 8 were consented and randomized but were discharged before completing asthma education; this yielded 140 children and caregivers in our final sample (69 in the interactive group, 71 in the control group). There were no differences in mean age, sex, race, insurance status, or asthma control among the groups (Table 1). Uncontrolled asthma (57% vs 42%; P = .096) and smoking exposure (25% vs 13%; P = .25) were more prevalent in the interactive group, but these differences did not reach statistical significance.

ED visits among the 2 groups were not significantly different at 6 months postdischarge (20.3% for the interactive group versus 25.4% for the control group; P = .50); however, fewer participants in the interactive group had been hospitalized at 6 months (10.1% vs 22.5% in the control group; P = .048). This difference was also present when evaluating the number of hospitalizations during that 6-month period (89.9% in the interactive group versus 77.5% in the control group had 0 hospitalizations, 8.7% vs 16.9% had 1 hospitalization, 0% vs 5.6% had 2 hospitalizations, and 1.5% vs 0% had 3 hospitalizations; P = .03). There was no significant difference in ED visits or hospitalizations at 12 months after discharge (Table 2).

Baseline MDI technique for patients has been fully reported in a previously published article.⁵  In the interactive group, the asthma checklist score increased (n = 55; mean change of 4.07 [95% CI: 3.21–4.94]) directly after TTG education. MDI demonstration data immediately after education is not available for the control group because this was not part of the control group didactic education. At the 12-month follow-up, which took place for both groups, the interactive group demonstrated a significant 1-point increase in MDI technique compared with baseline (P = .03); the control group had an increase of 0.71 points compared with baseline (P = .10; Table 3).

Baseline health literacy was high in both the interactive and control groups. Asthma knowledge (NAKQ) showed similar baseline scores in both the interactive and control groups (13.65 vs 13.16, respectively; max score of 18) and was slightly increased in both groups (0.46 interactive [95% CI: −0.16–1.08] vs 0.47 control [95% CI: −0.16–1.11]; P = .97). Parental confidence and self-reported management skills (PAMSE) were high at baseline in both the interactive and control groups (mean 52.6 vs 54.4; max score of 65). Both groups showed an improvement in PAMSE after asthma education was completed (least square mean change + 7.37 in the interactive group versus + 7.46 in the control group; P = .90) and this improvement persisted for those who completed the 12-month survey (mean change + 5.92 [n = 36 interactive group] versus + 6.50 [n = 39 control group]; P = .60). Modified PROMIS scores were similar in both the interactive and control groups at baseline (19.9 vs 20.62, lowest [best] possible score 8). At the 12-month follow-up, they were significantly more improved in the interactive group (least square mean change of −3.52 in the interactive group versus −1.75 in the control group; P < .01), with improvements in both asthma symptoms and asthma quality of life (Table 4).

Our primary outcome data were obtained from the EMR; thus, we had complete data for all participants with regard to subsequent asthma-related ED visits and hospitalizations. However, for the 12-month follow-up surveys, we had a 46% completion rate. Response rates did not differ by treatment group (45.1% control group versus 47.8% interactive education group, P = .87). Comparison of those who did and did not complete the 12-month follow-up showed no differences in age (7.02 vs 6.6; P = .50), sex (40% vs 37.5% female; P = .80), race (18.9% vs 20.3% White; 40.5% vs 42.2% African American; P = .98), or asthma severity (60% vs 63.1% mild; P = .70).

The interactive group had a mean education time of 12.3 minutes (SD 5.9; n = 56) for picture-based AAP development and education compared with the control AAP education mean time of 8.05 minutes (SD 6.6; n = 60). TTG MDI education in the interactive group had a mean time of 7.6 minutes (SD 4.08; n = 53), whereas control group MDI education had a mean time of 5.44 minutes (SD 4.48; n = 59). Overall education (AAP and MDI teaching) was on average 6.33 minutes longer in the interactive group compared with the control group.

To our knowledge, this is the first study to evaluate the clinical effectiveness of different teaching models in pediatric inpatient asthma education. Although we did not find a difference in ED visits for asthma between our intervention and control group, there was a decrease in hospitalizations in the subsequent 6 months after discharge in the interactive education group. Our interactive education protocol was distinct from the standard-of-care didactic education in 2 ways: 1) it included the codevelopment of a picture-based AAP; and 2) it used TTG demonstrations to teach MDI use.

The observed reduction in subsequent hospital admissions may be because of a combination of more effective home management and/or improved administration of inhaled medications as a consequence of the interactive education received. It is unclear at this time why there were no changes in ED visits and only in 6 months posteducation hospitalizations. It is possible that the interactive education intervention prompted caregivers to seek ED care earlier or reduced the overall risk of more severe exacerbations requiring hospital admission. Thus, these children may not have needed advanced interventions (eg, continuous albuterol or intravenous magnesium), which often prompt hospital admission. This is supported by a previous study by Cabana and Bardach,²²  which found that improved asthma education led to greater health care utilization.

The importance of inpatient asthma education has been highlighted in many studies, providing support for the NHLBI’s recommendation that education occur at every opportunity across the health care continuum, including hospitalizations.¹⁰  The first studies to evaluate the importance of inpatient education^8,22,23  found that some asthma education (compared with no education) led to fewer subsequent asthma hospitalizations. Since then, several pediatric inpatient quality improvement initiatives have been published.^7,9,24–31  These studies focused on increasing education and AAP completion rates,^26,32  and/or evaluating the type of educator or number of learners (individual versus classes).^9,27  Furthermore, many inpatient studies^7,30,33,34  primarily compared longer, more intensive educational programs to “usual care” abbreviated interventions that provide significantly less instruction. These studies helped to define what information should be included in standard inpatient asthma education, such as review of asthma symptoms, triggers, and medication administration.¹⁰  Notably, in our study, the control and interactive groups did not differ in the amount of information they received, the type of educator, or the educational setting; furthermore, all families received education on asthma symptoms, triggers, and medications. The primary goal was to determine if the educational approach had an impact on asthma management and health care utilization, addressing an important gap in the literature about how inpatient asthma education should be conducted to improve clinical outcomes.

Effective educational models for chronic diseases such as asthma are critical. Even as the therapies for asthma have improved, knowledge and adherence to recommended treatment plans remain problematic. Although traditional educational models depend upon instruction, with an active teacher and passive learner, both educational psychology literature and systematic reviews in adults with obstructive lung disease have found didactic education to be far less effective than interactive models, particularly in retaining information over time.^35–37  The TTG teaching approach to MDI training used in our study used an interactive teaching approach by requiring participants to repeatedly demonstrate their MDI administration and was found to improve MDI technique both immediately after the intervention and at 12 months, albeit to a lesser degree. This finding supports the NHLBI expert panel recommendation for periodic reinforcement of education.¹⁰  Furthermore, previous studies have shown that as many as half of US adults lack basic literacy skills³⁸  and how this affects health care utilization,³⁹  which is highly relevant to efforts to improve outcomes for pediatric asthma. Health literacy-informed communication and best practices, such as TTG and picture-based AAP, address the underrecognized health disparity of inadequate health literacy.^16,40,41 

Studies in the outpatient and ED settings have also shown a beneficial effect of asthma education on health care utilization, caregiver knowledge and confidence, and improved follow-up^42–44 ; however, many of the interventions tested were specific to the clinical setting. For example, outpatient asthma education programs are built on frequent revisits to the clinic and ED education interventions are limited by the time available in this setting.⁴⁵  Inpatient asthma education likewise has its own benefits and limitations, including more time to allow patients and/or caregivers to repeatedly demonstrate their ability to manage asthma symptoms under the supervision of health care professionals. Furthermore, children hospitalized with asthma are a high-priority group for education given their risk for asthma-related morbidity and mortality. The educational interventions described in this study can be provided alongside clinical care in the hospital by leveraging the teachable moments during routine hospital care.

Hospitalization accounts for nearly 50% of the medical costs of asthma in the United States each year.^1,33,46  In Illinois, the cost of pediatric ED visits among children insured by Medicaid was the sixth highest in the nation.⁴⁷  Our study reinforces findings from the Cochrane Database review of asthma education programs that asthma education can improve the clinical outcomes of children with asthma.⁴⁸  From this perspective, investment in efficient and evidence-based inpatient asthma education programs is a cost-effective approach to decreasing subsequent hospitalization. Consequently, health care systems will be able to reserve more costly interventions, such as home visits and specialty clinic services, for patients with severe disease and/or other barriers that prevent appropriate home management.

Although our randomized controlled study design conferred rigor, there were important limitations with our study. First, it was conducted among English-speaking patients only. However, with the emphasis on pictures and demonstrations, it is very likely that the inclusion of Spanish-speaking patients may have increased the impact of our intervention. Second, while conducting the study, we discovered that many families could not participate in our study during daytime hours when asthma educators were available. These caregivers may have different education levels regarding their child’s asthma, such that our results could be biased. However, the direction of the bias is likely toward the null: inclusion of this group may have shown a greater impact of the interactive education. A third consideration is that in-person education could be a challenge to implement in hospitals where resources for educators are not available. Press et al have begun to study a “virtual Teach-To-Goal” intervention that could be adapted for use among children.^49–51  A fourth limitation is that our primary health care utilization outcomes were based on EMR data at our institution and do not reflect health care utilization at other institutions. However, although some visits may not have been accounted for with this approach, there is no reason missed health care utilization would have occurred more in 1 group than another in a randomized controlled model. A fifth limitation is that the survey completion rate for the 12-month follow-up survey was 46%. Those who returned their surveys may have reported different outcomes than those who did not return their surveys; however, response rates did not differ by treatment group, age, sex, race, or asthma severity. We acknowledge that strong conclusions cannot be made on the basis of this response rate and consider these results to be hypothesis-generating. Lastly, this was a single-center study conducted in an urban inner-city population and the results may not be applicable in different populations.

Although the interactive asthma education model did not impact asthma-related ED visits, the intervention did decrease hospitalizations 6 months after discharge. These findings indicate that an asthma educational approach that actively engages caregivers is one important part of improving clinical outcomes.

We thank Tina Carter for her efforts as a research coordinator and Laura Shreffler and Felisa Ray for their contributions as educators.