Paul Crawford
Research Subjects
Research Interests
Paul Crawford has a total of 160 co-authors. Here are the top co-authors by number of shared publications:
- Christy J. W. Ledford
Medical College of Georgia, Augusta University
6 shared publications Carla L. Fisher
University of Florida, College of Journalism and Communications, P.O. Box 118400, Gainesville, FL, USA
5 shared publicationsChristy J. W. Ledford
4 shared publications- Dean A. Seehusen
Dean A. Seehusen, MD, MPH, CAQ-HALM, is a Professor and Chair, Department of Family and Community Medicine, Medical College of Georgia, Augusta University, Augusta, Georgia, USA
4 shared publications - Jeremy T. Jackson
Henry M. Jackson Foundation, Bethesda, MD
4 shared publications Brian Crownover
3 shared publications
Latest Publications
- 29 November 2025
- Published by SAGE Publications in Acupuncture in Medicine
- Vol. 44 (1) , 25-35
Effect of needle selection on battlefield acupuncture tolerance and pain reduction: an exploratory randomized controlled trial
Introduction: There has been limited research comparing the different types of needles used for auricular battlefield acupuncture (BFA). The aim of this exploratory study was to determine the feasibility of examining whether there are differences in needle tolerance and/or pain reduction between different needle types. Methods: We conducted an unblinded, exploratory, randomized controlled trial in human subjects with complaints of acute or chronic pain. We randomized 30 patients to receive BFA with ASP needles, disposable needles of 15 mm length or Pyonex needles (n = 10 per group). Differences in Defense and Veteran Pain Rating Scale (DVPRS) scores and needle tolerance questionnaire responses by treatment group and follow-up time point (baseline, 10 min, 24 h and 1 week) were assessed using random effects regression and ordinal logistics models. Results: A difference in needle tolerance by treatment group was observed for pain at the site(s) of needling (p = 0.002), which was lower for 15-mm needles (odds ratio (OR) = 0.12, p = 0.005) than for ASP needles, but not significantly different for Pyonex versus ASP (OR = 1.45, p = 0.525). There were no difference in overall patient satisfaction, dizziness, bruising, fatigue or headaches between the three BFA needle groups (p = 0.190). No statistically significant associations were observed between treatment group and clinical characteristics (all p > 0.05). DVPRS scores significantly declined after treatment (interaction beta coefficient (β) = −1.40, p < 0.001) across all treatment groups. Declines were also observed post treatment for DVPRS questions 1 (β = −1.23, p = 0.003), 3 (β = −1.23, p < 0.002) and 4 (β = −1.10, p = 0.002). A decline over time after treatment was also observed for DVPRS question 1 (β = −0.15, p = 0.014), 3 (β = −0.12, 0.049) and 4 (β = −0.13, p = 0.016) but was not DVPRS overall score. No significant interactions by treatment group were observed. Conclusion: Testing BFA needle types against each other was feasible and the interventions were acceptable to subjects. Through a post hoc power calculation, this study provides a greater understanding of the number of patients that would be needed to more definitively assess for potential differences in effectiveness (pain reduction) between various types of needles in BFA. Trial registration number: NCT04464954 (ClinicalTrials.gov)
- 29 November 2025
- Published by SAGE Publications in Acupuncture in Medicine
- Vol. 44 (1) , 25-35
Effect of needle selection on battlefield acupuncture tolerance and pain reduction: an exploratory randomized controlled trial
Introduction: There has been limited research comparing the different types of needles used for auricular battlefield acupuncture (BFA). The aim of this exploratory study was to determine the feasibility of examining whether there are differences in needle tolerance and/or pain reduction between different needle types. Methods: We conducted an unblinded, exploratory, randomized controlled trial in human subjects with complaints of acute or chronic pain. We randomized 30 patients to receive BFA with ASP needles, disposable needles of 15 mm length or Pyonex needles (n = 10 per group). Differences in Defense and Veteran Pain Rating Scale (DVPRS) scores and needle tolerance questionnaire responses by treatment group and follow-up time point (baseline, 10 min, 24 h and 1 week) were assessed using random effects regression and ordinal logistics models. Results: A difference in needle tolerance by treatment group was observed for pain at the site(s) of needling (p = 0.002), which was lower for 15-mm needles (odds ratio (OR) = 0.12, p = 0.005) than for ASP needles, but not significantly different for Pyonex versus ASP (OR = 1.45, p = 0.525). There were no difference in overall patient satisfaction, dizziness, bruising, fatigue or headaches between the three BFA needle groups (p = 0.190). No statistically significant associations were observed between treatment group and clinical characteristics (all p > 0.05). DVPRS scores significantly declined after treatment (interaction beta coefficient (β) = −1.40, p < 0.001) across all treatment groups. Declines were also observed post treatment for DVPRS questions 1 (β = −1.23, p = 0.003), 3 (β = −1.23, p < 0.002) and 4 (β = −1.10, p = 0.002). A decline over time after treatment was also observed for DVPRS question 1 (β = −0.15, p = 0.014), 3 (β = −0.12, 0.049) and 4 (β = −0.13, p = 0.016) but was not DVPRS overall score. No significant interactions by treatment group were observed. Conclusion: Testing BFA needle types against each other was feasible and the interventions were acceptable to subjects. Through a post hoc power calculation, this study provides a greater understanding of the number of patients that would be needed to more definitively assess for potential differences in effectiveness (pain reduction) between various types of needles in BFA. Trial registration number: NCT04464954 (ClinicalTrials.gov)
- 23 May 2023
- Published by Oxford University Press (OUP) in Family Practice
- Vol. 41 (3) , 277-282
Migraine care practices in primary care: results from a national US survey
Background: Primary care clinicians play a critical role in diagnosis and treatment of migraine, yet barriers exist. This national survey assessed barriers to diagnosis and treatment of migraine, preferred approaches to receiving migraine education, and familiarity with recent therapeutic innovations. Methods: The survey was created by the American Academy of Family Physicians (AAFP) and Eli Lilly and Company and distributed to a national sample through the AAFP National Research Network and affiliated PBRNs from mid-April through the end of May 2021. Initial analyses were descriptive statistics, ANOVAs, and Chi-Square tests. Individual and multivariate models were completed for: adult patients seen in a week; respondent years since residency; and adult patients with migraine seen in a week. Results: Respondents who saw fewer patients were more likely to indicate unclear patient histories were a barrier to diagnosing. Respondents who saw more patients with migraine were more likely to indicate the priority of other comorbidities and insufficient time were barriers to diagnosing. Respondents who had been out of residency longer were more likely to change a treatment plan due to attack impact, quality of life, and medication cost. Respondents who had been out of residency shorter were more likely to prefer to learn from migraine/headache research scientists and use paper headache diaries. Conclusions: Results demonstrate differences in familiarity with migraine diagnosis and treatment options based on patients seen and years since residency. To maximise appropriate diagnosis within primary care, targeted efforts to increase familiarity and decrease barriers to migraine care should be implemented.
- 23 May 2023
- Published by Oxford University Press (OUP) in Family Practice
- Vol. 41 (3) , 277-282
Migraine care practices in primary care: results from a national US survey
Background: Primary care clinicians play a critical role in diagnosis and treatment of migraine, yet barriers exist. This national survey assessed barriers to diagnosis and treatment of migraine, preferred approaches to receiving migraine education, and familiarity with recent therapeutic innovations. Methods: The survey was created by the American Academy of Family Physicians (AAFP) and Eli Lilly and Company and distributed to a national sample through the AAFP National Research Network and affiliated PBRNs from mid-April through the end of May 2021. Initial analyses were descriptive statistics, ANOVAs, and Chi-Square tests. Individual and multivariate models were completed for: adult patients seen in a week; respondent years since residency; and adult patients with migraine seen in a week. Results: Respondents who saw fewer patients were more likely to indicate unclear patient histories were a barrier to diagnosing. Respondents who saw more patients with migraine were more likely to indicate the priority of other comorbidities and insufficient time were barriers to diagnosing. Respondents who had been out of residency longer were more likely to change a treatment plan due to attack impact, quality of life, and medication cost. Respondents who had been out of residency shorter were more likely to prefer to learn from migraine/headache research scientists and use paper headache diaries. Conclusions: Results demonstrate differences in familiarity with migraine diagnosis and treatment options based on patients seen and years since residency. To maximise appropriate diagnosis within primary care, targeted efforts to increase familiarity and decrease barriers to migraine care should be implemented.
- Open Access
- 6 September 2022
- Published by Society of Teachers of Family Medicine in Family Medicine
- Vol. 54 (8) , 657-658
Military Family Medicine Readiness for National Pandemic Hospital Support
- 6 September 2022
- Published by Society of Teachers of Family Medicine in Family Medicine
- Vol. 54 (8) , 657-658
Open AccessMilitary Family Medicine Readiness for National Pandemic Hospital Support
- Open Access
- 1 January 2021
- Published by Elsevier in Patient Education and Counseling
- Vol. 104 (1) , 143-148
Toward a model of shared meaningful diagnosis
- 1 January 2021
- Published by Elsevier in Patient Education and Counseling
- Vol. 104 (1) , 143-148
Open AccessToward a model of shared meaningful diagnosis
- 1 September 2020
- Published by Elsevier in Obesity Research & Clinical Practice
- Vol. 14 (5) , 443-448
Wearable and interactive technology to share fitness goals results in weight loss but not improved diabetes outcomes
- 1 September 2020
- Published by Elsevier in Obesity Research & Clinical Practice
- Vol. 14 (5) , 443-448
Wearable and interactive technology to share fitness goals results in weight loss but not improved diabetes outcomes
- Open Access
- 26 May 2020
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 22 (5) , e17968
Enhancing Patient Activation and Self-Management Activities in Patients With Type 2 Diabetes Using the US Department of Defense Mobile Health Care Environment: Feasibility Study
Background: Past mobile health (mHealth) efforts to empower type 2 diabetes (T2D) self-management include portals, text messaging, collection of biometric data, electronic coaching, email, and collection of lifestyle information.Objective: The primary objective was to enhance patient activation and self-management of T2D using the US Department of Defense’s Mobile Health Care Environment (MHCE) in a patient-centered medical home setting.Methods: A multisite study, including a user-centered design and a controlled trial, was conducted within the US Military Health System. Phase I assessed preferences regarding the enhancement of the enabling technology. Phase II was a single-blinded 12-month feasibility study that randomly assigned 240 patients to either the intervention (n=123, received mHealth technology and behavioral messages tailored to Patient Activation Measure [PAM] level at baseline) or the control group (n=117, received equipment but not messaging. The primary outcome measure was PAM scores. Secondary outcome measures included Summary of Diabetes Self-Care Activities (SDSCA) scores and cardiometabolic outcomes. We used generalized estimating equations to estimate changes in outcomes.Results: The final sample consisted of 229 patients. Participants were 61.6% (141/229) male, had a mean age of 62.9 years, mean glycated hemoglobin (HbA1c) of 7.5%, mean BMI of 32.7, and a mean duration of T2D diagnosis of 9.8 years. At month 12, the control group showed significantly greater improvements compared with the intervention group in PAM scores (control mean 7.49, intervention mean 1.77; P=.007), HbA1c (control mean −0.53, intervention mean −0.11; P=.006), and low-density lipoprotein cholesterol (control mean −7.14, intervention mean 4.38; P=.01). Both groups showed significant improvement in SDSCA, BMI, waist size, and diastolic blood pressure; between-group differences were not statistically significant. Except for patients with the highest level of activation (PAM level 4), intervention group patients exhibited significant improvements in PAM scores. For patients with the lowest level of activation (PAM level 1), the intervention group showed significantly greater improvement compared with the control group in HbA1c (control mean −0.09, intervention mean −0.52; P=.04), BMI (control mean 0.58, intervention mean −1.22; P=.01), and high-density lipoprotein cholesterol levels (control mean −4.86, intervention mean 3.56; P<.001). Significant improvements were seen in AM scores, SDSCA, and waist size for both groups and in diastolic and systolic blood pressure for the control group; the between-group differences were not statistically significant. The percentage of participants who were engaged with MHCE for ≥50% of days period was 60.7% (68/112; months 0-3), 57.4% (62/108; months 3-6), 49.5% (51/103; months 6-9), and 43% (42/98; months 9-12).Conclusions: Our study produced mixed results with improvement in PAM scores and outcomes in both the intervention and control groups. Structural design issues may have hampered the influence of tailored behavioral messaging within the intervention group.Trial Registration: ClinicalTrials.gov NCT02949037; https://clinicaltrials.gov/ct2/show/NCT02949037International Registered Report Identifier (IRRID): RR2-10.2196/resprot.6993
- 26 May 2020
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 22 (5) , e17968
Open AccessEnhancing Patient Activation and Self-Management Activities in Patients With Type 2 Diabetes Using the US Department of Defense Mobile Health Care Environment: Feasibility Study
Background: Past mobile health (mHealth) efforts to empower type 2 diabetes (T2D) self-management include portals, text messaging, collection of biometric data, electronic coaching, email, and collection of lifestyle information.Objective: The primary objective was to enhance patient activation and self-management of T2D using the US Department of Defense’s Mobile Health Care Environment (MHCE) in a patient-centered medical home setting.Methods: A multisite study, including a user-centered design and a controlled trial, was conducted within the US Military Health System. Phase I assessed preferences regarding the enhancement of the enabling technology. Phase II was a single-blinded 12-month feasibility study that randomly assigned 240 patients to either the intervention (n=123, received mHealth technology and behavioral messages tailored to Patient Activation Measure [PAM] level at baseline) or the control group (n=117, received equipment but not messaging. The primary outcome measure was PAM scores. Secondary outcome measures included Summary of Diabetes Self-Care Activities (SDSCA) scores and cardiometabolic outcomes. We used generalized estimating equations to estimate changes in outcomes.Results: The final sample consisted of 229 patients. Participants were 61.6% (141/229) male, had a mean age of 62.9 years, mean glycated hemoglobin (HbA1c) of 7.5%, mean BMI of 32.7, and a mean duration of T2D diagnosis of 9.8 years. At month 12, the control group showed significantly greater improvements compared with the intervention group in PAM scores (control mean 7.49, intervention mean 1.77; P=.007), HbA1c (control mean −0.53, intervention mean −0.11; P=.006), and low-density lipoprotein cholesterol (control mean −7.14, intervention mean 4.38; P=.01). Both groups showed significant improvement in SDSCA, BMI, waist size, and diastolic blood pressure; between-group differences were not statistically significant. Except for patients with the highest level of activation (PAM level 4), intervention group patients exhibited significant improvements in PAM scores. For patients with the lowest level of activation (PAM level 1), the intervention group showed significantly greater improvement compared with the control group in HbA1c (control mean −0.09, intervention mean −0.52; P=.04), BMI (control mean 0.58, intervention mean −1.22; P=.01), and high-density lipoprotein cholesterol levels (control mean −4.86, intervention mean 3.56; P<.001). Significant improvements were seen in AM scores, SDSCA, and waist size for both groups and in diastolic and systolic blood pressure for the control group; the between-group differences were not statistically significant. The percentage of participants who were engaged with MHCE for ≥50% of days period was 60.7% (68/112; months 0-3), 57.4% (62/108; months 3-6), 49.5% (51/103; months 6-9), and 43% (42/98; months 9-12).Conclusions: Our study produced mixed results with improvement in PAM scores and outcomes in both the intervention and control groups. Structural design issues may have hampered the influence of tailored behavioral messaging within the intervention group.Trial Registration: ClinicalTrials.gov NCT02949037; https://clinicaltrials.gov/ct2/show/NCT02949037International Registered Report Identifier (IRRID): RR2-10.2196/resprot.6993
- Open Access
- 1 April 2020
- Published by Elsevier in Diabetes Research and Clinical Practice
- Vol. 162, 108117
How patients make sense of a diabetes diagnosis: An application of Weick’s model of organizing
- 1 April 2020
- Published by Elsevier in Diabetes Research and Clinical Practice
- Vol. 162, 108117
Open AccessHow patients make sense of a diabetes diagnosis: An application of Weick’s model of organizing
- Open Access
- 16 March 2020
- Published by American Board of Family Medicine (ABFM) in The Journal of the American Board of Family Medicine
- Vol. 33 (2) , 211-219
Turning Points as Opportunities to Partner with Patients Living with type 2 Diabetes or Prediabetes
Introduction: Understanding patients’ perspectives about their diabetes and what causes those perspectives to shift is critical to building a treatment strategy with the patient and facilitating patient self-management behavior. Key “turning points” can provide crucial opportunities to enact a change in perspective. The goal of this study is to identify “turning points” that have significance to diabetes-related health. Methods: Research coordinators interviewed 33 patients aged 25 to 65 diagnosed with type 2 diabetes mellitus or prediabetes at medical centers in Augusta, Georgia, and Las Vegas, Nevada. Retrospective interview technique and turning point analysis was employed to plot health or diabetes management changes from diagnosis up to the present day. The constant comparative method was used to conduct a thematic analysis. Axial coding identified properties characterizing each turning point. Results: Patients reported 5 interrelated turning points occurring at various times after diagnosis: 1) gaining knowledge, either through patients own research and/or a health care class; 2) making lifestyle changes, including exercising and healthier eating; 3) encountering a life-changing event/transition, including events that derailed healthy behavior, motivated health behavior, and removed barriers to enacting healthy behavior; 4) receiving social support, either through holding patients accountable or encouraging them to enact healthy behavior; and 5) interacting with clinicians, such as medication changes or behavior changes critical to disease management. Discussion: These turning points provide specific moments throughout diabetes care in which family physicians can effectively partner with patients. By prompting, facilitating, or attending to these turning points, family physicians can partner with patients throughout diabetes care.
- 16 March 2020
- Published by American Board of Family Medicine (ABFM) in The Journal of the American Board of Family Medicine
- Vol. 33 (2) , 211-219
Open AccessTurning Points as Opportunities to Partner with Patients Living with type 2 Diabetes or Prediabetes
Introduction: Understanding patients’ perspectives about their diabetes and what causes those perspectives to shift is critical to building a treatment strategy with the patient and facilitating patient self-management behavior. Key “turning points” can provide crucial opportunities to enact a change in perspective. The goal of this study is to identify “turning points” that have significance to diabetes-related health. Methods: Research coordinators interviewed 33 patients aged 25 to 65 diagnosed with type 2 diabetes mellitus or prediabetes at medical centers in Augusta, Georgia, and Las Vegas, Nevada. Retrospective interview technique and turning point analysis was employed to plot health or diabetes management changes from diagnosis up to the present day. The constant comparative method was used to conduct a thematic analysis. Axial coding identified properties characterizing each turning point. Results: Patients reported 5 interrelated turning points occurring at various times after diagnosis: 1) gaining knowledge, either through patients own research and/or a health care class; 2) making lifestyle changes, including exercising and healthier eating; 3) encountering a life-changing event/transition, including events that derailed healthy behavior, motivated health behavior, and removed barriers to enacting healthy behavior; 4) receiving social support, either through holding patients accountable or encouraging them to enact healthy behavior; and 5) interacting with clinicians, such as medication changes or behavior changes critical to disease management. Discussion: These turning points provide specific moments throughout diabetes care in which family physicians can effectively partner with patients. By prompting, facilitating, or attending to these turning points, family physicians can partner with patients throughout diabetes care.
- Open Access
- 24 January 2020
- Published by JMIR Publications Inc. in JMIR Preprints
Enhancing Patient Activation and Self-Management Activities in Patients With Type 2 Diabetes Using the US Department of Defense Mobile Health Care Environment: Feasibility Study (Preprint)
BACKGROUND Past mobile health (mHealth) efforts to empower type 2 diabetes (T2D) self-management include portals, text messaging, collection of biometric data, electronic coaching, email, and collection of lifestyle information. OBJECTIVE The primary objective was to enhance patient activation and self-management of T2D using the US Department of Defense’s Mobile Health Care Environment (MHCE) in a patient-centered medical home setting. METHODS A multisite study, including a user-centered design and a controlled trial, was conducted within the US Military Health System. Phase I assessed preferences regarding the enhancement of the enabling technology. Phase II was a single-blinded 12-month feasibility study that randomly assigned 240 patients to either the intervention (n=123, received mHealth technology and behavioral messages tailored to Patient Activation Measure [PAM] level at baseline) or the control group (n=117, received equipment but not messaging. The primary outcome measure was PAM scores. Secondary outcome measures included Summary of Diabetes Self-Care Activities (SDSCA) scores and cardiometabolic outcomes. We used generalized estimating equations to estimate changes in outcomes. RESULTS The final sample consisted of 229 patients. Participants were 61.6% (141/229) male, had a mean age of 62.9 years, mean glycated hemoglobin (HbA1c) of 7.5%, mean BMI of 32.7, and a mean duration of T2D diagnosis of 9.8 years. At month 12, the control group showed significantly greater improvements compared with the intervention group in PAM scores (control mean 7.49, intervention mean 1.77; P=.007), HbA1c (control mean −0.53, intervention mean −0.11; P=.006), and low-density lipoprotein cholesterol (control mean −7.14, intervention mean 4.38; P=.01). Both groups showed significant improvement in SDSCA, BMI, waist size, and diastolic blood pressure; between-group differences were not statistically significant. Except for patients with the highest level of activation (PAM level 4), intervention group patients exhibited significant improvements in PAM scores. For patients with the lowest level of activation (PAM level 1), the intervention group showed significantly greater improvement compared with the control group in HbA1c (control mean −0.09, intervention mean −0.52; P=.04), BMI (control mean 0.58, intervention mean −1.22; P=.01), and high-density lipoprotein cholesterol levels (control mean −4.86, intervention mean 3.56; P<.001). Significant improvements were seen in AM scores, SDSCA, and waist size for both groups and in diastolic and systolic blood pressure for the control group; the between-group differences were not statistically significant. The percentage of participants who were engaged with MHCE for ≥50% of days period was 60.7% (68/112; months 0-3), 57.4% (62/108; months 3-6), 49.5% (51/103; months 6-9), and 43% (42/98; months 9-12). CONCLUSIONS Our study produced mixed results with improvement in PAM scores and outcomes in both the intervention and control groups. Structural design issues may have hampered the influence of tailored behavioral messaging within the intervention group. CLINICALTRIAL ClinicalTrials.gov NCT02949037; https://clinicaltrials.gov/ct2/show/NCT02949037 INTERNATIONAL REGISTERED REPORT RR2-10.2196/resprot.6993
- 24 January 2020
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessEnhancing Patient Activation and Self-Management Activities in Patients With Type 2 Diabetes Using the US Department of Defense Mobile Health Care Environment: Feasibility Study (Preprint)
BACKGROUND Past mobile health (mHealth) efforts to empower type 2 diabetes (T2D) self-management include portals, text messaging, collection of biometric data, electronic coaching, email, and collection of lifestyle information. OBJECTIVE The primary objective was to enhance patient activation and self-management of T2D using the US Department of Defense’s Mobile Health Care Environment (MHCE) in a patient-centered medical home setting. METHODS A multisite study, including a user-centered design and a controlled trial, was conducted within the US Military Health System. Phase I assessed preferences regarding the enhancement of the enabling technology. Phase II was a single-blinded 12-month feasibility study that randomly assigned 240 patients to either the intervention (n=123, received mHealth technology and behavioral messages tailored to Patient Activation Measure [PAM] level at baseline) or the control group (n=117, received equipment but not messaging. The primary outcome measure was PAM scores. Secondary outcome measures included Summary of Diabetes Self-Care Activities (SDSCA) scores and cardiometabolic outcomes. We used generalized estimating equations to estimate changes in outcomes. RESULTS The final sample consisted of 229 patients. Participants were 61.6% (141/229) male, had a mean age of 62.9 years, mean glycated hemoglobin (HbA1c) of 7.5%, mean BMI of 32.7, and a mean duration of T2D diagnosis of 9.8 years. At month 12, the control group showed significantly greater improvements compared with the intervention group in PAM scores (control mean 7.49, intervention mean 1.77; P=.007), HbA1c (control mean −0.53, intervention mean −0.11; P=.006), and low-density lipoprotein cholesterol (control mean −7.14, intervention mean 4.38; P=.01). Both groups showed significant improvement in SDSCA, BMI, waist size, and diastolic blood pressure; between-group differences were not statistically significant. Except for patients with the highest level of activation (PAM level 4), intervention group patients exhibited significant improvements in PAM scores. For patients with the lowest level of activation (PAM level 1), the intervention group showed significantly greater improvement compared with the control group in HbA1c (control mean −0.09, intervention mean −0.52; P=.04), BMI (control mean 0.58, intervention mean −1.22; P=.01), and high-density lipoprotein cholesterol levels (control mean −4.86, intervention mean 3.56; P<.001). Significant improvements were seen in AM scores, SDSCA, and waist size for both groups and in diastolic and systolic blood pressure for the control group; the between-group differences were not statistically significant. The percentage of participants who were engaged with MHCE for ≥50% of days period was 60.7% (68/112; months 0-3), 57.4% (62/108; months 3-6), 49.5% (51/103; months 6-9), and 43% (42/98; months 9-12). CONCLUSIONS Our study produced mixed results with improvement in PAM scores and outcomes in both the intervention and control groups. Structural design issues may have hampered the influence of tailored behavioral messaging within the intervention group. CLINICALTRIAL ClinicalTrials.gov NCT02949037; https://clinicaltrials.gov/ct2/show/NCT02949037 INTERNATIONAL REGISTERED REPORT RR2-10.2196/resprot.6993