JMIR Publications

ABSTRACT

Background:

Background:

Ecological momentary assessment (EMA) provides insight into the effectiveness and feasibility of smoking-related interventions.

Objective:

To assessed adherence to an EMA protocol and compared EMA-derived responses to survey data.

Methods:

Data were from a two-arm pilot randomized clinical trial among adult smokers with a chronic condition (n=109). EMA data were collected via SMS prompts sent to participants four times daily for 12 weeks. Convergent validity between survey- and EMA-reported measures was evaluated using Pearson correlation and paired differences. Cigarettes per day (CPD) was modeled using negative binomial regression. Relative rates (RR) of 50%, 75%, and 100% CPD reduction were calculated.

Results:

The majority of participants were non-Hispanic White (58%) with median age of 60 (IQR 54-65). Median weekly EMA response rate was high (week 1, 98%; week 12, 89%). EMA and survey CPD measurements were positively correlated (r = 0.73; CI 0.60, 0.82) as were measures of craving (r = 0.38; CI 0.17, 0.56). No significant paired difference in CPD was observed. A significant effect of time on CPD EMA data (Incidence rate ratio (IRR) 1-week change = 0.93, p<0.01) and survey data was found (IRR 12-week change = 0.36, p < 0.01).

Conclusions:

EMA is a suitable method to collect recall-based smoking data. Though results from mixed effect modeling and RR comparisons were similar using EMA or survey data, EMA provides unique advantages, namely greater granularity in the time and the capability to detect switching patterns in near real-time.