Dynamic models of stress-smoking responses based on high-frequency sensor data

Abstract

Self-reports indicate that stress increases the risk for smoking; however, intensive data from sensors can provide a more nuanced understanding of stress in the moments leading up to and following smoking events. Identifying personalized dynamical models of stress-smoking responses can improve characterizations of smoking responses following stress, but techniques used to identify these models require intensive longitudinal data. This study leveraged advances in wearable sensing technology and digital markers of stress and smoking to identify person-specific models of stress and smoking system dynamics by considering stress immediately before, during, and after smoking events. Adult smokers (n = 45) wore the AutoSense chestband (respiration-inductive plethysmograph, electrocardiogram, accelerometer) with MotionSense (accelerometers, gyroscopes) on each wrist for three days prior to a quit attempt. The odds of minute-level smoking events were regressed on minute-level stress probabilities to identify person-specific dynamic models of smoking responses to stress. Simulated pulse responses to a continuous stress episode revealed a consistent pattern of increased odds of smoking either shortly after the beginning of the simulated stress episode or with a delay, for all participants. This pattern is followed by a dramatic reduction in the probability of smoking thereafter, for about half of the participants (49%). Sensor-detected stress probabilities indicate a vulnerability for smoking that may be used as a tailoring variable for just-in-time interventions to support quit attempts.

The Digital Therapeutics Real-World Evidence Framework: An Approach for Guiding Evidence-Based Digital Therapeutics Design, Development, Testing, and Monitoring

“…For example, optimization trials can be used to support the evidence-based selection of DTx components (factorial trials) [ 64 - 68 ], refinement of components, particularly those used across time (microrandomized trials) [ 35 , 69 , 70 ], and refinement of adaptation algorithms to match the provision of support to context, individual differences, and timing (microrandomized trials [ 35 , 71 ], and sequential multiple assignment randomized trials [ 72 - 75 ]). Note also that optimization trials could also be conducted that are explicitly used to support algorithm development (eg, system identification experiments [ 76 - 78 ] or more data-driven algorithm development from RWD [ 79 , 80 ]). Like phase I, proof-of-concept and optimization trials can be used iteratively.…”

“…If the optimization criteria are met, then this can often justify movement to phase III. Plausible optimization trials for this phase could include but are not limited to: A/B testing (as used in the technology industry for improving usability) [ 85 - 87 ], factorial trials as used in MOST [ 64 , 65 , 67 , 68 ], sequential multiple assigned randomized trials [ 72 - 75 ], microrandomized trials [ 35 , 36 , 69 - 71 ], system identification experiments [ 76 - 78 ], studies explicitly designed to support algorithm development [ 79 , 80 ], and control optimization trials [ 37 , 88 ]. Nahum-Shani et al [ 40 ] provide guidance on when to use common optimization trial designs.…”

Multidimensional digital biomarker of cognitive health: unobtrusive and continuous monitoring of cognitive changes using smartphones

“…Computational models mathematically characterize processes (or systems) that repeatedly output specific outcomes in response to repeated inputs 131 . Mechanistic models belong to the family of computational models.…”

“…Importantly, no modeling approach is universally better than others, and the mechanistic-computational approach is no exception. Instead, it is often advised to, when appropriate, combine mechanistic modeling with machine learning and statistical modeling as such a hybrid approach allows one to take advantage of the benefits of each modeling methodology while balancing their pros and cons [131][132][133][134][135][136][137] .…”