Passive Nocturnal Physiologic Monitoring Enables Early Detection of Exacerbations in Children with Asthma. A Proof-of-Concept Study

Abstract

Rationale: Asthma management depends on prompt identification of symptoms, which challenges both patients and providers. In asthma, a misapprehension of health between exacerbations can compromise compliance. Thus, there is a need for a tool that permits objective longitudinal monitoring without increasing the burden of patient compliance.

Objectives: We sought to determine whether changes in nocturnal physiology are associated with asthma symptoms in pediatric patients.

Methods: Using a contactless bed sensor, nocturnal heart rate (HR), respiratory rate, relative stroke volume, and movement in children with asthma 5-18 years of age (n = 16) were recorded. Asthma symptoms and asthma control test (ACT) score were reported every 2 weeks. Random forest model was used to identify physiologic parameters associated with asthma symptoms. Elastic net regression was used to identify variables associated with ACT score.

Measurements and main results: The model on the full cohort performed with sensitivity of 47.2%, specificity of 96.3%, and accuracy of 87.4%; HR and respiratory parameters were the most important variables in this model. The model predicted asthma symptoms 35% of the time on the day before perception of symptoms, and 100% of the time for a select subject for which the model performed with greater sensitivity. Multivariable and bivariable analyses demonstrated significant association between HR and respiratory rate parameters and ACT score.

Conclusions: Nocturnal physiologic changes correlate with asthma symptoms, supporting the notion that nocturnal physiologic monitoring represents an objective diagnostic tool capable of longitudinally assessing disease control and predicting asthma exacerbations in children with asthma at home.

Keywords: asthma; asthma control; longitudinal monitoring; pediatric asthma.

Figure 1.

Top device output variables according to mean decrease in accuracy (MDA) from random forest algorithm. The MDA is the decrease in accuracy of model predictions when a variable is removed from the model by permuting its values so the values are unrelated to the outcome. Omitting the information from a variable that is not important would not change the model accuracy, and the MDA would be 0. Omitting the information from a variable important to the model would decrease the model’s accuracy substantially, and the MDA would be relatively greater than 0. Statistical measures of heart rate were the top five most important variables for prediction of loss of asthma control. HR = heart rate.

Figure 2.

Model prediction of asthma during the 14 days before reported asthma symptoms for all subjects with asthma symptoms. The random forest model was applied to data from days preceding report of asthma symptoms to evaluate ability of the model to predict asthma before it was reported by the subject. On the day before report of symptoms, the model predicted loss of asthma control in 35% of cases tested.

Figure 3.

Model prediction of asthma during the 14 days before symptoms reported for subject 3. The random forest model was applied to data from subject 3 alone from days preceding report of asthma symptoms to evaluate ability of the model to predict asthma before perception of symptoms. The model predicted loss of asthma control 100% of the time on the 2 days before the report of onset of asthma symptoms.

Figure 4.

Asthma control test (ACT) score and nocturnal heart rate (HR) over time in subject 16. The trend ACT score and median, 10th percentile, and 90th percentile HR over time for subject 16 are shown. ACT score is shown by the black line, median HR is shown by the dashed line, and the gray area reflects range of HR from 10th to 90th percentile. Decreases in ACT score, indicative of loss in asthma control, were mirrored by elevations in HR.

Figure 5.

Asthma control test (ACT) score and nocturnal respiratory rate (RR) over time in subject 16. The trend ACT score and median, 10th percentile, and 90th percentile RR over time for subject 16 are shown. ACT score is shown by the black line, median RR is shown by the dashed line, and the gray area reflects range of RR from 10th to 90th percentile. Decreases in ACT score, indicative of loss in asthma control, were mirrored by elevations in RR.