Text Message Analysis Using Machine Learning to Assess Predictors of Engagement With Mobile Health Chronic Disease Prevention Programs: Content Analysis
BackgroundSMS text messages as a form of mobile health are increasingly being used to support individuals with chronic diseases in novel ways that leverage the mobility and capabilities of mobile phones. However, there are knowledge gaps in mobile health, including how to max...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
JMIR Publications
2021-11-01
|
| Series: | JMIR mHealth and uHealth |
| Online Access: | https://mhealth.jmir.org/2021/11/e27779 |
| _version_ | 1797735599138930688 |
|---|---|
| author | Harry Klimis Joel Nothman Di Lu Chao Sun N Wah Cheung Julie Redfern Aravinda Thiagalingam Clara K Chow |
| author_facet | Harry Klimis Joel Nothman Di Lu Chao Sun N Wah Cheung Julie Redfern Aravinda Thiagalingam Clara K Chow |
| author_sort | Harry Klimis |
| collection | DOAJ |
| description |
BackgroundSMS text messages as a form of mobile health are increasingly being used to support individuals with chronic diseases in novel ways that leverage the mobility and capabilities of mobile phones. However, there are knowledge gaps in mobile health, including how to maximize engagement.
ObjectiveThis study aims to categorize program SMS text messages and participant replies using machine learning (ML) and to examine whether message characteristics are associated with premature program stopping and engagement.
MethodsWe assessed communication logs from SMS text message–based chronic disease prevention studies that encouraged 1-way (SupportMe/ITM) and 2-way (TEXTMEDS [Text Messages to Improve Medication Adherence and Secondary Prevention]) communication. Outgoing messages were manually categorized into 5 message intents (informative, instructional, motivational, supportive, and notification) and replies into 7 groups (stop, thanks, questions, reporting healthy, reporting struggle, general comment, and other). Grid search with 10-fold cross-validation was implemented to identify the best-performing ML models and evaluated using nested cross-validation. Regression models with interaction terms were used to compare the association of message intent with premature program stopping and engagement (replied at least 3 times and did not prematurely stop) in SupportMe/ITM and TEXTMEDS.
ResultsWe analyzed 1550 messages and 4071 participant replies. Approximately 5.49% (145/2642) of participants responded with stop, and 11.7% (309/2642) of participants were engaged. Our optimal ML model correctly classified program message intent with 76.6% (95% CI 63.5%-89.8%) and replies with 77.8% (95% CI 74.1%-81.4%) balanced accuracy (average area under the curve was 0.95 and 0.96, respectively). Overall, supportive (odds ratio [OR] 0.53, 95% CI 0.35-0.81) messages were associated with reduced chance of stopping, as were informative messages in SupportMe/ITM (OR 0.35, 95% CI 0.20-0.60) but not in TEXTMEDS (for interaction, P<.001). Notification messages were associated with a higher chance of stopping in SupportMe/ITM (OR 5.76, 95% CI 3.66-9.06) but not TEXTMEDS (for interaction, P=.01). Overall, informative (OR 1.76, 95% CI 1.46-2.12) and instructional (OR 1.47, 95% CI 1.21-1.80) messages were associated with higher engagement but not motivational messages (OR 1.18, 95% CI 0.82-1.70; P=.37). For supportive messages, the association with engagement was opposite with SupportMe/ITM (OR 1.77, 95% CI 1.21-2.58) compared with TEXTMEDS (OR 0.77, 95% CI 0.60-0.98; for interaction, P<.001). Notification messages were associated with reduced engagement in SupportMe/ITM (OR 0.07, 95% CI 0.05-0.10) and TEXTMEDS (OR 0.28, 95% CI 0.20-0.39); however, the strength of the association was greater in SupportMe/ITM (for interaction P<.001).
ConclusionsML models enable monitoring and detailed characterization of program messages and participant replies. Outgoing message intent may influence premature program stopping and engagement, although the strength and direction of association appear to vary by program type. Future studies will need to examine whether modifying message characteristics can optimize engagement and whether this leads to behavior change. |
| first_indexed | 2024-03-12T13:00:24Z |
| format | Article |
| id | doaj.art-620a6513cf5c48249b6c45c9d7c94fe6 |
| institution | Directory Open Access Journal |
| issn | 2291-5222 |
| language | English |
| last_indexed | 2024-03-12T13:00:24Z |
| publishDate | 2021-11-01 |
| publisher | JMIR Publications |
| record_format | Article |
| series | JMIR mHealth and uHealth |
| spelling | doaj.art-620a6513cf5c48249b6c45c9d7c94fe62023-08-28T19:45:10ZengJMIR PublicationsJMIR mHealth and uHealth2291-52222021-11-01911e2777910.2196/27779Text Message Analysis Using Machine Learning to Assess Predictors of Engagement With Mobile Health Chronic Disease Prevention Programs: Content AnalysisHarry Klimishttps://orcid.org/0000-0002-3635-421XJoel Nothmanhttps://orcid.org/0000-0003-1497-3842Di Luhttps://orcid.org/0000-0001-8669-5824Chao Sunhttps://orcid.org/0000-0002-1428-2649N Wah Cheunghttps://orcid.org/0000-0001-6323-8006Julie Redfernhttps://orcid.org/0000-0001-8707-5563Aravinda Thiagalingamhttps://orcid.org/0000-0002-7763-7806Clara K Chowhttps://orcid.org/0000-0003-4693-0038 BackgroundSMS text messages as a form of mobile health are increasingly being used to support individuals with chronic diseases in novel ways that leverage the mobility and capabilities of mobile phones. However, there are knowledge gaps in mobile health, including how to maximize engagement. ObjectiveThis study aims to categorize program SMS text messages and participant replies using machine learning (ML) and to examine whether message characteristics are associated with premature program stopping and engagement. MethodsWe assessed communication logs from SMS text message–based chronic disease prevention studies that encouraged 1-way (SupportMe/ITM) and 2-way (TEXTMEDS [Text Messages to Improve Medication Adherence and Secondary Prevention]) communication. Outgoing messages were manually categorized into 5 message intents (informative, instructional, motivational, supportive, and notification) and replies into 7 groups (stop, thanks, questions, reporting healthy, reporting struggle, general comment, and other). Grid search with 10-fold cross-validation was implemented to identify the best-performing ML models and evaluated using nested cross-validation. Regression models with interaction terms were used to compare the association of message intent with premature program stopping and engagement (replied at least 3 times and did not prematurely stop) in SupportMe/ITM and TEXTMEDS. ResultsWe analyzed 1550 messages and 4071 participant replies. Approximately 5.49% (145/2642) of participants responded with stop, and 11.7% (309/2642) of participants were engaged. Our optimal ML model correctly classified program message intent with 76.6% (95% CI 63.5%-89.8%) and replies with 77.8% (95% CI 74.1%-81.4%) balanced accuracy (average area under the curve was 0.95 and 0.96, respectively). Overall, supportive (odds ratio [OR] 0.53, 95% CI 0.35-0.81) messages were associated with reduced chance of stopping, as were informative messages in SupportMe/ITM (OR 0.35, 95% CI 0.20-0.60) but not in TEXTMEDS (for interaction, P<.001). Notification messages were associated with a higher chance of stopping in SupportMe/ITM (OR 5.76, 95% CI 3.66-9.06) but not TEXTMEDS (for interaction, P=.01). Overall, informative (OR 1.76, 95% CI 1.46-2.12) and instructional (OR 1.47, 95% CI 1.21-1.80) messages were associated with higher engagement but not motivational messages (OR 1.18, 95% CI 0.82-1.70; P=.37). For supportive messages, the association with engagement was opposite with SupportMe/ITM (OR 1.77, 95% CI 1.21-2.58) compared with TEXTMEDS (OR 0.77, 95% CI 0.60-0.98; for interaction, P<.001). Notification messages were associated with reduced engagement in SupportMe/ITM (OR 0.07, 95% CI 0.05-0.10) and TEXTMEDS (OR 0.28, 95% CI 0.20-0.39); however, the strength of the association was greater in SupportMe/ITM (for interaction P<.001). ConclusionsML models enable monitoring and detailed characterization of program messages and participant replies. Outgoing message intent may influence premature program stopping and engagement, although the strength and direction of association appear to vary by program type. Future studies will need to examine whether modifying message characteristics can optimize engagement and whether this leads to behavior change.https://mhealth.jmir.org/2021/11/e27779 |
| spellingShingle | Harry Klimis Joel Nothman Di Lu Chao Sun N Wah Cheung Julie Redfern Aravinda Thiagalingam Clara K Chow Text Message Analysis Using Machine Learning to Assess Predictors of Engagement With Mobile Health Chronic Disease Prevention Programs: Content Analysis JMIR mHealth and uHealth |
| title | Text Message Analysis Using Machine Learning to Assess Predictors of Engagement With Mobile Health Chronic Disease Prevention Programs: Content Analysis |
| title_full | Text Message Analysis Using Machine Learning to Assess Predictors of Engagement With Mobile Health Chronic Disease Prevention Programs: Content Analysis |
| title_fullStr | Text Message Analysis Using Machine Learning to Assess Predictors of Engagement With Mobile Health Chronic Disease Prevention Programs: Content Analysis |
| title_full_unstemmed | Text Message Analysis Using Machine Learning to Assess Predictors of Engagement With Mobile Health Chronic Disease Prevention Programs: Content Analysis |
| title_short | Text Message Analysis Using Machine Learning to Assess Predictors of Engagement With Mobile Health Chronic Disease Prevention Programs: Content Analysis |
| title_sort | text message analysis using machine learning to assess predictors of engagement with mobile health chronic disease prevention programs content analysis |
| url | https://mhealth.jmir.org/2021/11/e27779 |
| work_keys_str_mv | AT harryklimis textmessageanalysisusingmachinelearningtoassesspredictorsofengagementwithmobilehealthchronicdiseasepreventionprogramscontentanalysis AT joelnothman textmessageanalysisusingmachinelearningtoassesspredictorsofengagementwithmobilehealthchronicdiseasepreventionprogramscontentanalysis AT dilu textmessageanalysisusingmachinelearningtoassesspredictorsofengagementwithmobilehealthchronicdiseasepreventionprogramscontentanalysis AT chaosun textmessageanalysisusingmachinelearningtoassesspredictorsofengagementwithmobilehealthchronicdiseasepreventionprogramscontentanalysis AT nwahcheung textmessageanalysisusingmachinelearningtoassesspredictorsofengagementwithmobilehealthchronicdiseasepreventionprogramscontentanalysis AT julieredfern textmessageanalysisusingmachinelearningtoassesspredictorsofengagementwithmobilehealthchronicdiseasepreventionprogramscontentanalysis AT aravindathiagalingam textmessageanalysisusingmachinelearningtoassesspredictorsofengagementwithmobilehealthchronicdiseasepreventionprogramscontentanalysis AT clarakchow textmessageanalysisusingmachinelearningtoassesspredictorsofengagementwithmobilehealthchronicdiseasepreventionprogramscontentanalysis |