Efficacy and feasibility of the BREATHE asthma intervention with American Indian children: a randomized controlled trial

Trial design

This study aimed to evaluate the efficacy and feasibility of the 24-month long (range 15–35 months) BREATHE asthma intervention involving parental and child education, during a 3-year follow-up period for American Indian children (n = 108), using a randomized controlled study design. The duration of the BREATHE asthma intervention varied in duration due to funding deadlines. The ratio of intervention to control participants was 1:1 and there were no changes to the intervention or trial outcomes during the study period.

Setting and participants

The FIPA Study was conducted among an American Indian population in a rural region in north-central United States, where tribal members primarily rely on Indian Health Service (IHS), and the tribal health department for healthcare. To receive specialty care for asthma from a pulmonologist or allergist, families travel 2–3 h. Challenges to access healthcare in this low income community, and suboptimal housing conditions contribute to excessive burden from asthma².

The definitive diagnosis of asthma for both research and clinical purposes is difficult. We chose a case definition that is relatively conservative, practical in application, and more objective than patient/parent self-report of asthma. All participants met 2 of the 3 criteria for inclusion based on a chart review: (1) a diagnosis of asthma on at least 2 occasions by more than one provider during the past 2 years, (2) refills of asthma treatment medications on at least 2 occasions during the past 2 years, or (3) improvement in forced expired volume at one second (FEV1) of at least 20% with a short-acting bronchodilator. Exclusionary criteria were (1) neonatal ventilator treatment, (2) hospitalization at birth lasting more than 15 days, (3) congenital heart anomaly requiring surgery, (4) diagnosis of cystic fibrosis, (5) congenital lung, diaphragm, chest wall, or airway anomaly, (6) diagnosis of pneumonia, pertussis, or tuberculosis within the past year, or (7) congenital muscular disorder.

Cases were randomly assigned to either the BREATHE arm described above, or to the control Non-Intervention arm, consisting only of the child’s usual and customary clinical care and a packet of written material about asthma management provided by study staff at enrollment. After consent and assent was obtained by research staff, participants were randomized without replacement by allowing the participant to blindly draw one of two different colored balls from a container for an equal chance to be in either group. The research staff recorded the color chosen and the participant signed an acknowledgement of the color. A total of 108 participants with asthma were enrolled; 53 were randomly selected to receive BREATHE, 53 were randomly selected to be in the Non-Intervention group, and 2 did not consent to participate in the RCT portion of the study so they were assigned to the Non-Intervention group. Staff were aware of participants’ assignment to intervention or non-intervention arms. Additional details on setting and participants were previously described^2,4,5.

The research protocol was approved by the Institutional Review Boards at the Collaborative Research Center for American Indian Health, Sanford Research, Sioux Falls, SD; the Great Plains Indian Health Service area office in Aberdeen, SD, and the local tribal government. All research was performed in accordance with relevant guidelines/regulations. Participants’ parents gave informed consent in writing and children provided assent. The study was not preregistered, but it was registered on 05/10/2017 with the US National Library of Medicine on clinicaltrials.gov as the “Factors Influencing Pediatric Asthma Study, identifier NCT03302962.

Recruitment

Children with asthma were identified through an automated query of the IHS and tribal health department electronic health records using the International Classification of Diseases, ninth edition (ICD-9) code 493.9 within the ages of 6 and 17 years. Additional potential participants were identified through local providers at other healthcare facilities. These recruitment methods resulted in over 700 children who were identified as potentially meeting inclusion criteria, of which about 450 were excluded because they lacked a clinical provider diagnosis of asthma (rather than a pharmacy diagnosis automatically assigned to those with a bronchodilator prescription), or lacked repeated asthma medication refills. Approximately 130 of the remaining 250 were able to be contacted—of those, some declined the medical record review to determine eligibility, and others consented to the medical record review, but were ineligible, leaving 108 cases interested and eligible to participate in the study.

Because this study used a convenience sample in a small community with a limited number of children that would meet inclusion criteria, no sample size or power calculations were used.

Measures

The primary outcome measure was the number of Indian Health Service ED visits for respiratory complaints during the observation period from 9/1/13 through 5/25/17. The electronic medical records, including dates of visit, for all participants were available for the entire observation period, thus there was no “loss to follow-up” for our primary outcome, although it is recognized that some participants may have accessed other facilities, resulting in records that were unavailable to us. This was analyzed both as a mean number of visits per participant or dichotomized as none vs ≥ 1.

Secondary outcomes were measured following consent and randomization procedures using a self-administered questionnaire completed by the participant and a guardian depending on the child’s age at the initial exam and final exam. Using the Check Your Asthma I.Q. Quiz¹³, we assessed current asthma knowledge of the child if age 10–17, or the parent if the child was age 5–9, in which each point represents the number of correct responses out of 12 points. Severity of disease was assessed using the Childhood Asthma Control Test™¹⁴, for parents and children ages 5–11 (27 points possible; scores ≥ 20 indicate well controlled asthma), or the Asthma Control Test™¹⁴, for children ages 12–18 (25 points possible; scores ≥ 20 indicate well controlled asthma). Present quality of life was assessed using the Asthma Life Quality Test¹⁵, in which each point represents the number of quality of life items impacted by asthma out of 20 possible points, so a greater score indicates worse quality of life. Parents and children of all ages responded to the Asthma Life Quality Test independently, but in the presence of each other with the child answering before the parent.

A saliva Accutest NicAlert™ (Jant Pharmacal Corp, Encino, CA) assay was used to assess salivary levels of cotinine; categorical scores range from 0 which identifies a non-tobacco user, to scores of 1, 2, 3, 4, 5, or 6 which identifies tobacco users or individuals highly exposed to environmental tobacco smoke (ETS) to varying degrees¹⁶.

Treatment conditions

Following the baseline questionnaire and exam, participants randomized to the BREATHE arm received comprehensive information from a single asthma counselor about asthma and the management of the disease was discussed with caregivers involved in the care of the participant. Materials from the American Lung Association’s “Controlling Asthma: What You Need to Know” flipbook was used for education¹⁷. The asthma counselor emphasized medication use and compliance, a customized asthma action plan, and the deleterious effects of smoking. The asthma counselor had no previous experience delivering asthma education, but was trained by an experienced certified asthma educator to deliver the intervention and had a 4-year medical degree. Control participants were given an educational packet and encouraged to maintain regular contact with their usual provider; no case management calls or questionnaires were attempted between baseline and final exam.

After the baseline education, standardized case management procedures were conducted by research staff. Two to four weeks after education, contact was attempted via phone to repeat the questionnaire, discuss questions or concerns, and coordinate care if needed. Case management was attempted every 3 months for the first year, every 6 months during the second year, for a total of 6 possible case management sessions between the baseline and final exam, which ranged from 15 to 35 months. Content and questionnaires were the same for all educational and case management sessions. Initial exams that included the BREATHE Intervention or the control conditions began in September 2013 and were concluded in March 2015. Final exams occurred 13 to 33 months after the baseline exam, from January 2016 to December 2016.

Statistical analysis

Kaplan–Meier analysis was undertaken to analyze our primary outcome of the differences in the number of days to first ED visit or hospitalization for respiratory problems between the BREATHE arm and the Non-Intervention control arm. The “time to event” was calculated from the participant’s enrollment exam (synchronous with the start of intervention) and ended with the defined endpoints of first ED visit or hospitalization, as noted above. We also compared the rate of ED visits or hospitalization over the entire study period using Poisson regression. Analysis for secondary outcomes included a t-test for independent samples to test for group mean differences in quantitative variables, whereas chi-square tests were used to evaluate differences in discrete variables. A paired t-test or a McNemar’s chi-square test was used to compare an individuals’ change from baseline to final contact. Quantitative comparisons were two-tailed. All data analysis was pre-specified.

Data was entered into Microsoft Excel, and STATA 14.2 (Stata Corp, College Station, TX, USA) was used to identify outliers to be verified or corrected. All other analyses were conducted on SPSS 13.0 (SPSS Inc., Chicago, IL, USA) software.