Asthma is the most common chronic illness in children and the third-leading cause of hospitalization of children under the age of 15. Asthma is a dynamic condition with variable severity. As such, monitoring for symptoms and changes in lung function is an integral part of asthma management.
Most cases of pediatric asthma are successfully controlled with a regimen of inhaled steroids with or without the addition of second line medication. However, a small proportion of children have asthma that is difficult to keep under control. In these children, symptoms are not alleviated through typical use of high doses of inhaled steroids plus additional asthma controllers. These patients may require chronic use of oral steroids.
Difficult-to-control asthma in children accounts for significant morbidity and health care costs. While some children have severe, treatment-resistant asthma, most uncontrolled asthma is a result of modifiable factors — the most common being poor adherence to therapy. For this reason, evaluating and improving adherence are key to successful outcomes in those patients.
At Mayo Clinic Children’s Center, patients with severe and difficult-to-control asthma receive comprehensive, multidisciplinary evaluation and treatment in the new Mayo Clinic Pediatric Difficult to Control Asthma Clinic, which opened fall 2021.
“Direct observed therapy (DOT) is one of the most effective interventions, ” according to Manuel Arteta, M.D., a pediatric pulmonologist at Mayo Clinic in Rochester, Minnesota, and the director of the Pediatric Difficult to Control Asthma Clinic. “Historically, DOT has been difficult and expensive, requiring in-person evaluation at a health facility or in the patient’s home, but technology has allowed for asynchronous virtual DOT, making this form of therapy more accessible and affordable for providers and patients alike.”
Asynchronous virtual DOT allows patients to record short, time-stamped video clips of themselves taking each dose of their prescribed inhalers. Patients send encrypted video to members of the health care team, who review it to confirm that the medication is being used as prescribed and with adequate technique. Finally, the team provides feedback to the patient.
“With virtual DOT, we’re able to correct and improve inhaler technique and more closely monitor adherence to the treatment plan. We find a collaborative approach including children, parents and our clinical team to be most effective in helping improve asthma symptoms and, ultimately, the child’s quality of life,” says Dr. Arteta.
Virtual assessment and feedback have other applications for this patient population as well.
Variable bronchial obstruction is a key characteristic of asthma. The presence of obstruction is a marker of uncontrolled asthma and a recognized risk factor for asthma exacerbations. Spirometry is the test of choice to evaluate for airway obstruction; it allows for assessment of asthma control and helps predict the risk of future exacerbation.
Unsupervised at-home spirometry in children often results in inadequate findings due to suboptimal technique. Newly developed technology allows for virtual, supervised measurement of pulmonary function at home and provides reliable results to guide management and decision-making. Patients at the Mayo Clinic Difficult to Control Asthma Clinic perform spirometry at home using their own cellphones or tablets under the supervision of a respiratory therapist familiar with home spirometry.
“Implementing at-home, asynchronous virtual DOT and spirometry has been a game changer for our patients with difficult-to-treat asthma,” according to Dr. Arteta. “And these are not the only advances in technology that are being used to improve asthma care for children.”
Artificial intelligence (AI) tools such as natural language processing (NLP) and AI-assisted clinical decision support systems also are being developed to help children with asthma, as cited in a 2022 study published in The Journal of Allergy and Clinical Immunology: In Practice.
A team at Mayo Clinic Children’s Center recently validated an AI-assisted clinical decision support system called the Asthma Guidance and Prediction System (A-GPS). The team was led by Young J. Juhn, M.D., professor of pediatrics and director of both the AI Program Promoting Adolescent and Childhood Health (APPROACH) and the Precision Population Science Lab A-GPS program. Dr. Juhn’s team uses multiple AI algorithms such as NLP, unlocking free text information embedded in electronic health records (EHRs) to provide clinicians with concise, easy-to-access summaries of the most relevant clinical information for asthma management from patients’ EHRs.
Providing clinicians with a concise, easily accessed summary of the most relevant clinical information for asthma management from the patient’s EHR will be a significant timesaver.
Also, if the AI algorithms work well to proactively identify patients whose asthma may be falling out of control, the algorithms could potentially intervene to help these patients get back to stable asthma management.
This information is in alignment with recommended guidelines for care from the National Asthma Education and Prevention Program (NAEPP) and includes the following: current asthma severity and control status; asthma care quality; risk factors for asthma and asthma exacerbations; scores on standardized measures for asthma outcomes (for example, Asthma Control Test); spirometry results; and frequency of health care utilization, both planned and unplanned.
The A-GPS tool also includes a machine learning algorithm to predict future risk of asthma exacerbation based on the collected information from the patient’s EHR. In their study published in PLOS One, Dr. Juhn’s team compared A-GPS with usual asthma care and found promising results regarding the feasibility of A-GPS implementation in asthma care in the primary care setting, including reduction in asthma exacerbations, a 70% decrease in the time clinicians spend reviewing the EHR, lower health care costs and improved clinician satisfaction with asthma care.
Despite growing interest in the use of AI and machine learning in asthma care, there remains a chasm between the development of these tools and deployment within health care settings. To help bridge this gap, Dr. Juhn’s research team will soon perform a pragmatic, randomized clinical trial to examine whether the implementation and use of the A-GPS tool combined with remote patient monitoring and self-management tools (at-home spirometry, asthma symptom monitoring and a digital asthma action plan) will increase clinician and parent and caregiver satisfaction with asthma care, reduce clinician workload and burden of EHR review, and improve asthma control for child participants compared with usual asthma care.
If successful, these tools may enable remote asthma management for patients with limited access to clinics or hospitals while reducing the burden on patients and their caregivers, clinicians and care teams. This study will be run as a decentralized clinical trial and take place in real-world pediatric clinical care settings through a collaboration with multiple partners including the Mayo Clinic Precision Population Science Lab, the Mayo Clinic Center for Digital Health, the AI program within Pediatrics and Adolescent Medicine, the Mayo Clinic Cures at Home program, and the Karolinska Institute (Stockholm, Sweden).
Khoury P, et al. A framework for augmented intelligence in allergy and immunology practice and research — A work group report of the AAAAI Health Informatics, Technology and Education Committee. The Journal of Allergy and Clinical Immunology: In Practice. 2022;10:1178.
Seol HY, et al. Artificial intelligence-assisted clinical decision support for childhood asthma management: A randomized clinical trial. PLOS One. 2021;16,8:e0255261.