Patient perspectives on the integration of an intensive online behavioral weight loss intervention into primary care

Leveraging Machine Learning to Develop Digital Engagement Phenotypes of Users in a Digital Diabetes Prevention Program: Evaluation Study

BACKGROUND

Digital diabetes prevention programs (dDPPs) are effective "digital prescriptions" but have high attrition rates and program noncompletion. To address this, we developed a personalized automatic messaging system (PAMS) that leverages SMS text messaging and data integration into clinical workflows to increase dDPP engagement via enhanced patient-provider communication. Preliminary data showed positive results. However, further investigation is needed to determine how to optimize the tailoring of support technology such as PAMS based on a user's preferences to boost their dDPP engagement.

OBJECTIVE

This study evaluates leveraging machine learning (ML) to develop digital engagement phenotypes of dDPP users and assess ML's accuracy in predicting engagement with dDPP activities. This research will be used in a PAMS optimization process to improve PAMS personalization by incorporating engagement prediction and digital phenotyping. This study aims (1) to prove the feasibility of using dDPP user-collected data to build an ML model that predicts engagement and contributes to identifying digital engagement phenotypes, (2) to describe methods for developing ML models with dDPP data sets and present preliminary results, and (3) to present preliminary data on user profiling based on ML model outputs.

METHODS

Using the gradient-boosted forest model, we predicted engagement in 4 dDPP individual activities (physical activity, lessons, social activity, and weigh-ins) and general activity (engagement in any activity) based on previous short- and long-term activity in the app. The area under the receiver operating characteristic curve, the area under the precision-recall curve, and the Brier score metrics determined the performance of the model. Shapley values reflected the feature importance of the models and determined what variables informed user profiling through latent profile analysis.

RESULTS

We developed 2 models using weekly and daily DPP data sets (328,821 and 704,242 records, respectively), which yielded predictive accuracies above 90%. Although both models were highly accurate, the daily model better fitted our research plan because it predicted daily changes in individual activities, which was crucial for creating the "digital phenotypes." To better understand the variables contributing to the model predictor, we calculated the Shapley values for both models to identify the features with the highest contribution to model fit; engagement with any activity in the dDPP in the last 7 days had the most predictive power. We profiled users with latent profile analysis after 2 weeks of engagement (Bayesian information criterion=-3222.46) with the dDPP and identified 6 profiles of users, including those with high engagement, minimal engagement, and attrition.

CONCLUSIONS

Preliminary results demonstrate that applying ML methods with predicting power is an acceptable mechanism to tailor and optimize messaging interventions to support patient engagement and adherence to digital prescriptions. The results enable future optimization of our existing messaging platform and expansion of this methodology to other clinical domains.

TRIAL REGISTRATION

ClinicalTrials.gov NCT04773834; https://www.clinicaltrials.gov/ct2/show/NCT04773834.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

RR2-10.2196/26750.

Overweight and obesity: effectiveness of interventions in adults

OBJECTIVE

To identify the most effective interventions in overweight and obese adults.

METHOD

A narrative review through a search of the literature in databases PubMed, Cochrane, Joanna Briggs Institute, EMBASE, Cuiden y Cinahl with free and controlled language (MeSH terms) using Boolean operators AND and NOT. The research was limited to articles published between 2007 and 2015. Eighteen articles were selected based on the established inclusion and exclusion criteria.

RESULTS

Different types of interventions were identified based on the modification of lifestyles, mainly diet, physical activity and behavior. Major differences were found in specific content, degree of intensity of interventions, time tracking and elements evaluated. Most of studies found statistically significant weight loss but this was limited in terms of weight and number of people. Web-based interventions have no uniform effect on weight loss but achieve similar levels to face-to-face interventions in maintaining weight loss.

CONCLUSIONS

The combination of personalised diet, exercise and cognitive behavioural therapy is the most effective form of intervention in overweight and obesity. There is insufficient data to indicate whether group or individual interventions are more effective. Online intervention allows greater accessibility and lower cost.

Using Health Information Technology to Foster Engagement: Patients' Experiences with an Active Patient Health Record

Personal health records (PHRs) typically employ "passive" communication strategies, such as non-personalized medical text, rather than direct patient engagement in care. Currently there is a call for more active PHRs that directly engage patients in an effort to improve their health by offering elements such as personalized medical information, health coaches, and secure messaging with primary care providers. As part of a randomized clinical trial comparing "passive" with "active" PHRs, we explore patients' experiences with using an "active" PHR known as HealthTrak. The "passive" elements of this PHR included problem lists, medication lists, information about patient allergies and immunizations, medical and surgical histories, lab test results, health reminders, and secure messaging. The active arm included all of these elements and added personalized alerts delivered through the secure messaging platform to patients for services coming due based on various demographic features (including age and sex) and chronic medical conditions. Our participants were part of the larger clinical trial and were eligible if they had been randomized to the active PHR arm, one that included regular personalized alerts. We conducted focus group discussions on the benefits of this active PHR for patients who are at risk for cardiovascular disease. Forty-one patients agreed to participate and were organized into five separate focus group sessions. Three main themes emerged from the qualitatively analyzed focus groups: participants reported that the active PHR promoted better communication with providers; enabled them to more effectively partner with their providers; and helped them become more proactive about tracking their health information. In conclusion, patients reported improved communication, partnership with their providers, and a sense of self-management, thus adding insights for PHR designers hoping to address low adoption rates and other patient barriers to the development and use of the technology.

The website-based eaTracker®‘My Goals’ feature: a qualitative evaluation

Objective

In 2011, Dietitians of Canada added ‘My Goals’ to its website-based nutrition/activity tracking program (eaTracker®,http://www.eaTracker.ca/); this feature allows users to choose ‘ready-made’ or ‘write-your-own’ goals and to self-report progress. The purpose of the present study was to document experiences and perceptions of goal setting and My Goals, and report users’ feedback on what is needed in future website-based goal setting/tracking tools.

Design

One-on-one semi-structured interviews were conducted with (i) My Goals users and (ii) dietitians providing a public information support service, EatRight Ontario (ERO).

Setting

My Goals users from Ontario and Alberta, Canada were recruited via an eaTracker website pop-up box; ERO dietitians working in Ontario, Canada were recruited via ERO.

Subjects

My Goals users (n23; age 19–70 years; 91 % female;n5 from Alberta/n18 from Ontario) and ERO dietitians (n5).

Results

Dietitians and users felt goal setting for nutrition (and activity) behaviour change was both a beneficial and a challenging process. Dietitians were concerned about users setting poor-quality goals and users felt it was difficult to stick to their goals. Both users and dietitians were enthusiastic about the My Goals concept, but felt the current feature had limitations that affected use. Dietitians and users provided suggestions to improve My Goals (e.g. more prominent presence of My Goals in eaTracker; assistance with goal setting; automated personalized feedback).

Conclusions

Dietitians and users shared similar perspectives on the My Goals feature and both felt goal use was challenging. Several suggestions were provided to enhance My Goals that are relevant to website-based goal setting/tracking tool design in general.

Use of the internet in the treatment of obesity and prevention of type 2 diabetes in primary care

Cardiovascular prevention, including emphasising healthy diet and physical activity patterns for weight loss and diabetes prevention, is an important focus of primary care medicine, yet provision of intensive lifestyle counselling in the primary care setting remains uncommon. Online approaches for providing lifestyle counselling are emerging as a new avenue for bringing support to outpatient patient populations. Features of successful programmes include the use of a structured behavioural programme, electronic counselling support and feedback, and uncomplicated user interfaces. Online tools may be used independently or to complement in-person approaches. Limited data also suggest that the internet can be used to support the maintenance of weight loss. In addition, the internet may help overcome various clinical barriers to lifestyle support, including significant time limitations, a need to prioritise acute care and maintain clinical workflow, and the high cost of counselling. Furthermore, the continuity of the primary care patient-provider(s) relationship provides an established source of long-term support which has been difficult to create in other community settings. As the field of online lifestyle counselling matures, nutrition and physical activity experts will face new challenges in providing asynchronous counselling without the assistance of traditional non-verbal communication cues. However, the potential for reaching a wider population in a convenient and accessible manner also creates unique opportunities for providing lifestyle support.

Promoting patient phronesis: communication patterns in an online lifestyle program coordinated with primary care

Phronesis, or practical wisdom developed through experience, is an Aristotelian concept that can shed light on the capacities of patients to make health-related decisions and engage in healthy behaviors. In this article, the authors develop a conceptual framework for understanding the role of phronesis in lifestyle change as well as its relationship to patient activation, which is considered to be a critical component of the Chronic Care Model and patient education in general. The authors develop the concept of phronesis by analyzing qualitatively the comments made by 35 participants working to manage chronic health issues in a weight-loss study. The authors iteratively coded transcribed passages of exit interviews for phronesis and patient activation. These passages provide experientially grounded content for evaluating the use of phronesis and its development among individuals engaging in lifestyle change. Phronesis is expressed in 31% of participant responses to questions regarding the relationship between the online lifestyle intervention, participant health, and participant readiness to engage in productive clinical encounters with health care practitioners. Of those responses, 73% express some level of patient activation. The authors conclude that phronesis may be an important new tool for understanding successful self-management support, with potential usefulness in the creation of tailored lifestyle interventions, the development of patient activation, and the ability of participants to enact health-related behaviors.

Primary care providers' perspectives on online weight-loss programs: a big wish list

Background

Integrating online weight-loss programs into the primary care setting could yield substantial public health benefit. Little is known about primary care providers’ perspectives on online weight-loss programs.

Objective

To assess primary care providers’ perspectives on online weight-loss programs.

Methods

We conducted focus group discussions with providers in family medicine, internal medicine, and combined internal medicine/pediatrics in Texas and Pennsylvania, USA. Open-ended questions addressed their experience with and attitudes toward online weight-loss programs; useful characteristics of existing online weight-loss programs; barriers to referring patients to online weight-loss programs; and preferred characteristics of an ideal online weight-loss program. Transcripts were analyzed with the grounded theory approach to identify major themes.

Results

A total of 44 primary care providers participated in 9 focus groups. The mean age was 45 (SD 9) years. Providers had limited experience with structured online weight-loss programs and were uncertain about their safety and efficacy. They thought motivated, younger patients would be more likely than others to respond to an online weight-loss program. According to primary care providers, an ideal online weight-loss program would provide—at no cost to the patient—a structured curriculum addressing motivation, psychological issues, and problem solving; tools for tracking diet, exercise, and weight loss; and peer support monitored by experts. Primary care providers were interested in receiving reports about patients from the online weight-loss programs, but were concerned about the time required to review and act on the reports.

Conclusions

Primary care providers have high expectations for how online weight-loss programs should deliver services to patients and fit into the clinical workflow. Efforts to integrate online weight-loss programs into the primary care setting should address efficacy and safety of online weight-loss programs in clinic-based populations; acceptable methods of sending reports to primary care providers about their patients’ progress; and elimination or reduction of costs to patients.