Dataset Information

Latent Class Dynamic Mediation Model with Application to Smoking Cessation Data.

ABSTRACT: Traditional mediation analysis assumes that a study population is homogeneous and the mediation effect is constant over time, which may not hold in some applications. Motivated by smoking cessation data, we propose a latent class dynamic mediation model that explicitly accounts for the fact that the study population may consist of different subgroups and the mediation effect may vary over time. We use a proportional odds model to accommodate the subject heterogeneities and identify latent subgroups. Conditional on the subgroups, we employ a Bayesian hierarchical nonparametric time-varying coefficient model to capture the time-varying mediation process, while allowing each subgroup to have its individual dynamic mediation process. A simulation study shows that the proposed method has good performance in estimating the mediation effect. We illustrate the proposed methodology by applying it to analyze smoking cessation data.

SUBMITTER: Huang J

PROVIDER: S-EPMC6594758 | biostudies-literature | 2019 Mar

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Latent Class Dynamic Mediation Model with Application to Smoking Cessation Data.

Huang Jing J Yuan Ying Y Wetter David D

Psychometrika 20190103 1

Traditional mediation analysis assumes that a study population is homogeneous and the mediation effect is constant over time, which may not hold in some applications. Motivated by smoking cessation data, we propose a latent class dynamic mediation model that explicitly accounts for the fact that the study population may consist of different subgroups and the mediation effect may vary over time. We use a proportional odds model to accommodate the subject heterogeneities and identify latent subgro ...[more]

PMID: 30607659

Similar Datasets

Project description:BackgroundEngagement with digital interventions is a well-known predictor of treatment outcomes, but this knowledge has had limited actionable value. Instead, learning why engagement with digital interventions impact treatment outcomes can lead to targeted improvements in their efficacy.ObjectiveThis study aimed to test a serial mediation model of an Acceptance and Commitment Therapy (ACT) smartphone intervention for smoking cessation.MethodsIn this randomized controlled trial, participants (N=2415) from 50 US states were assigned to the ACT-based smartphone intervention (iCanQuit) or comparison smartphone intervention (QuitGuide). Their engagement with the apps (primary measure: number of logins) was measured during the first 3 months, ACT processes were measured at baseline and 3 months (acceptance of internal cues to smoke, valued living), and smoking cessation was measured at 12 months with 87% follow-up retention.ResultsThere was a significant serial mediation effect of iCanQuit on smoking cessation through multiple indicators of intervention engagement (ie, total number of logins, total number of minutes used, and total number of unique days of use) and in turn through increases in mean acceptance of internal cues to smoke from baseline to 3 months. Analyses of the acceptance subscales showed that the mediation was through acceptance of physical sensations and emotions, but not acceptance of thoughts. There was no evidence that the effect of the iCanQuit intervention was mediated through changes in valued living.ConclusionsIn this first study of serial mediators underlying the efficacy of smartphone apps for smoking cessation, our results suggest the effect of the iCanQuit ACT-based smartphone app on smoking cessation was mediated through multiple indicators of engagement and in turn through increases in the acceptance of physical sensations and emotions that cue smoking.Trial registrationClinical Trials.gov NCT02724462; https://clinicaltrials.gov/ct2/show/NCT02724462.

Project description:Background and aimsPrevious evaluations of smoking cessation interventions in pregnancy have several limitations. Our solution to these limitations is the Economics of Smoking in Pregnancy (ESIP) model, which estimates the life-time cost-effectiveness of smoking cessation interventions in pregnancy from a National Health Service (NHS) and personal social services perspective. We aim to (1) describe how ESIP has been constructed and (2) illustrate its use with trial data.MethodsESIP links mothers' and offspring pregnancy outcomes to estimate the burdens of smoking-related disease they experience with different rates of smoking in pregnancy, both in pregnancy and throughout their life-times. Smoking rates are inputted by model users. ESIP then estimates the costs of treating disease burdens and also mothers' and offspring life-years and quality-adjusted life years (QALYs). By comparing costs incurred and healthy life following different smoking rates, ESIP estimates incremental cost-effectiveness and benefit-cost ratios for mothers or offspring or both combined. We illustrate ESIP use using data from a pragmatic randomized controlled trial that tested a smoking cessation intervention in pregnancy.ResultsThroughout women's and offspring life-times, the intervention proved cheaper than usual care, having a negative incremental cost of £38.37 (interquartile range = £21.46-56.96) and it improved health, demonstrating a 0.04 increase in incremental QALYs for mothers and offspring, implying that it is 'dominant' over usual care. Benefit-cost ratios suggested that every £1 spent would generate a median of £14 (interquartile range = £8-20) in health-care savings.ConclusionsEconomics of Smoking in Pregnancy is the first economic model to link mothers' and infants' costs and benefits while reporting cost-effectiveness in readily-comparable units. Using ESIP with data from a trial which reported only short-term economic analysis showed that the intervention was very likely to be cost-effective in the longer term and to generate health-care savings.

Dataset Information

Latent Class Dynamic Mediation Model with Application to Smoking Cessation Data.

Publications

Latent Class Dynamic Mediation Model with Application to Smoking Cessation Data.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets