Factors Predicting Recovery From Asthma Exacerbations

Application of Machine Learning to Support Self-Management of Asthma with mHealth

While there have been several efforts to use mHealth technologies to support asthma management, none so far offer personalised algorithms that can provide real-time feedback and tailored advice to patients based on their monitoring. This work employed a publicly available mHealth dataset, the Asthma Mobile Health Study (AMHS), and applied machine learning techniques to develop early warning algorithms to enhance asthma self-management. The AMHS consisted of longitudinal data from 5,875 patients, including 13,614 weekly surveys and 75,795 daily surveys. We applied several well-known supervised learning algorithms (classification) to differentiate stable and unstable periods and found that both logistic regression and naïve Bayes-based classifiers provided high accuracy (AUC > 0.87). We found features related to the use of quick-relief puffs, night symptoms, frequency of data entry, and day symptoms (in descending order of importance) as the most useful features to detect early evidence of loss of control. We found no additional value of using peak flow readings to improve population level early warning algorithms.

Dyspnea Perception During Induced Bronchoconstriction Is Complicated by the Inhaled Methacholine in Children With Clinical Asthma

Purpose

Dyspnea is not widely utilized as an indicator of asthma provocation despite its universal presentation. We hypothesized that dyspnea severity was proportionate with the lung function decline, methacholine dose-step, and the degree of bronchial hyperresponsiveness (BHR).

Methods

We retrospectively analyzed 73 children's bronchial provocation test data with an assessment of dyspnea at every dose-step. Dyspnea severity was scored using a modified Borg (mBorg) scale. A linear mixed effect analysis was performed to evaluate the relationship between the mBorg scale, the percentage fall in the forced expiratory volume in 1 second (FEV1) (ΔFEV1%), the methacholine dose-step, and the degree of BHR (BHR grade).

Results

Subjects were divided into 5 BHR groups based on their last methacholine dose-steps. The mBorg scores did not differ significantly among BHR groups (P=0.596, Kruskal-Wallis test). The linear mixed effect analysis showed that ΔFEV1% was affected by the methacholine dose-step (P<0.001) and BHR grade (P<0.001). The mBorg score was affected by the dose-step (P<0.001) and BHR grade (P=0.019). We developed a model to predict the mBorg score and found that it was affected by the methacholine dose-step and ΔFEV1%, elevating it by a score of 0.039 (χ² [1]=21.06, P<0.001) and 0.327 (χ² [1]=47.45, P<0.001), respectively. A significant interaction was observed between the methacholine dose-step and ΔFEV1% (χ² [1]=16.20, P<0.001).

Conclusions

In asthmatic children, inhaled methacholine, as well as the degree of BHR and lung function decline, may affect dyspnea perception during the bronchial provocation test. If we wish to draw meaningful information from dyspnea perception, we have to consider various complicating factors underlying it.