The system described herein collects patient data passively and non-passively via onboard and external sensors, and combines the data with past clinical history to generate digital biomarkers. The collected data can also be further combined with other data generating systems to more accurately predict disease exacerbations. The system monitors the digital biomarkers in real-time, and can detect a change in the disease state prior to clinical decompensation and suggest pre-emptive intervention. The system enables a patient to be treated early in the clinical timeline when the disease exacerbation is at the subclinical level rather than waiting until the disease exacerbation reaches the clinical level. Acting when the exacerbation is at the subclinical level enables preemptive treatment rather than reactive treatment, which is often more cost effective while improving clinical outcomes. The system is able to make the predictions by detecting subclinical changes in digital biomarkers that are generated from respiratory, cardiac, patient reported symptoms, user behaviors, and environmental triggers.
