Sense2stop Micro-Randomized Trial Primary Analysis

https://clinicaltrials.gov/ct2/show/NCT03184389

0. About Sense2Stop

The aim of this research is to build systems that can recognize when people are stressed and then provide them with relaxation prompts in the moment to reduce their likelihood of being stressed, smoking, or overeating in the near future. This should help smokers be more effective in their attempts to quit by reducing their tendency to lapse when they are stressed or experiencing other negative moods or behaviors.

The Sense2Stop study evaluates whether an app and worn sensors can help smokers quit smoking and not relapse.

1. About This Repository

This repository contains code for performing analysis of the Sense2stop MRT data and documentation. Files corresponding to particular stages of the project are placed under the relevant header.

2. Documentation

2.1 Code: Creating Data Frames

• create_activity_df.py is a Python script used to create ~/Box/MD2K Northwestern/Processed Data/primary_analysis/data/pickle_jar/activity_df.pkl, a cleaned dataset corresponding to the classification of physical activity.
• create_log_dicts.py is a Python script used to create ~/Box/MD2K Northwestern/Processed Data/primary_analysis/data/pickle_jar/log_dict.pkl, a cleaned dataset corresponding to the phone log files, specifically at randomization times.
• create_quality_ecg_df.py is a Python script used to create ~/Box/MD2K Northwestern/Processed Data/primary_analysis/data/pickle_jar/quality_ecg_df.pkl, a cleaned dataset corresponding to ECG quality.
• create_quality_rep_df.py is a Python script used to create ~/Box/MD2K Northwestern/Processed Data/primary_analysis/data/pickle_jar/quality_rep_df.pkl, a cleaned dataset corresponding to REP quality.
• create_stress_episode_classification_df.py is a Python script used to create ~/Box/MD2K Northwestern/Processed Data/primary_analysis/data/pickle_jar/stress_episode_classification_df.pkl, a cleaned dataset corresponding to stress episode classification.

2.2 Code: Missing Data

• show_missing_data.py is a Python script used to show the extent of the missing data in the primary outcome. In addition, this script creates a data frame that is used to run one of the covariate analyses in order to predict missing episodes within the primary outcome.

2.2 Code: Availability condtion for randomization

• MD2Korg/mCerebrum-Configuration

2.3 Notes

• ema_emi_randomization_algorithm.pdf

• Exploratory_Analyses.pdf

• Availability_for_randomization

3. Primary Analysis

3.1 Missing Data

• The file Missing_Data_in_Sense2Stop.pdf provides detail on the missing data within the Sense2Stop MRT.

3.2 Covariate Analysis

Detailed documentation for these analyses is provided in Covariate_Analyses.pdf.

3.3 Primary Analysis Method

• The file Sense2Stop_Supplement.pdf provides the appendix for the design paper with detail on the primary analysis methodology.

• simple_estimating_equation_and_symptotics.pdf

• complex_estimating_equation_and_symptotics.pdf

3.4 Consistency Simulation Code

• Data generating model: R/dgm.r
• Estimator: R/estimator.r
• Simulation consistency: R/simulation_consistency.r

4. References

• Sarker, H., Tyburski, M., Rahman, M.M., Hovsepian, K., Sharmin, M., Epstein, D.H., Preston, K.L., Furr-Holden, C.D., Milam, A., Nahum-Shani, I. and Al'Absi, M., 2016, May. Finding significant stress episodes in a discontinuous time series of rapidly varying mobile sensor data. In Proceedings of the 2016 CHI conference on human factors in computing systems (pp. 4489-4501). ACM.

• Hovsepian, K., al'Absi, M., Ertin, E., Kamarck, T., Nakajima, M. and Kumar, S., 2015, September. cStress: towards a gold standard for continuous stress assessment in the mobile environment. In Proceedings of the 2015 ACM international joint conference on pervasive and ubiquitous computing (pp. 493-504). ACM.

• Liao, P., Dempsey, W., Sarker, H., Hossain, S.M., Al'Absi, M., Klasnja, P. and Murphy, S., 2018. Just-in-Time but Not Too Much: Determining Treatment Timing in Mobile Health. Proceedings of the ACM on interactive, mobile, wearable and ubiquitous technologies, 2(4), p.179.