Collective-Intelligence Recommender Systems: Advancing Computer Tailoring for Health Behavior Change Into the 21st Century

A comprehensive review of machine learning algorithms and their application in geriatric medicine: present and future

The increasing access to health data worldwide is driving a resurgence in machine learning research, including data-hungry deep learning algorithms. More computationally efficient algorithms now offer unique opportunities to enhance diagnosis, risk stratification, and individualised approaches to patient management. Such opportunities are particularly relevant for the management of older patients, a group that is characterised by complex multimorbidity patterns and significant interindividual variability in homeostatic capacity, organ function, and response to treatment. Clinical tools that utilise machine learning algorithms to determine the optimal choice of treatment are slowly gaining the necessary approval from governing bodies and being implemented into healthcare, with significant implications for virtually all medical disciplines during the next phase of digital medicine. Beyond obtaining regulatory approval, a crucial element in implementing these tools is the trust and support of the people that use them. In this context, an increased understanding by clinicians of artificial intelligence and machine learning algorithms provides an appreciation of the possible benefits, risks, and uncertainties, and improves the chances for successful adoption. This review provides a broad taxonomy of machine learning algorithms, followed by a more detailed description of each algorithm class, their purpose and capabilities, and examples of their applications, particularly in geriatric medicine. Additional focus is given on the clinical implications and challenges involved in relying on devices with reduced interpretability and the progress made in counteracting the latter via the development of explainable machine learning.

Therapist Feedback and Implications on Adoption of an Artificial Intelligence-Based Co-Facilitator for Online Cancer Support Groups: Mixed Methods Single-Arm Usability Study

BACKGROUND

The recent onset of the COVID-19 pandemic and the social distancing requirement have created an increased demand for virtual support programs. Advances in artificial intelligence (AI) may offer novel solutions to management challenges such as the lack of emotional connections within virtual group interventions. Using typed text from online support groups, AI can help identify the potential risk of mental health concerns, alert group facilitator(s), and automatically recommend tailored resources while monitoring patient outcomes.

OBJECTIVE

The aim of this mixed methods, single-arm study was to evaluate the feasibility, acceptability, validity, and reliability of an AI-based co-facilitator (AICF) among CancerChatCanada therapists and participants to monitor online support group participants' distress through a real-time analysis of texts posted during the support group sessions. Specifically, AICF (1) generated participant profiles with discussion topic summaries and emotion trajectories for each session, (2) identified participant(s) at risk for increased emotional distress and alerted the therapist for follow-up, and (3) automatically suggested tailored recommendations based on participant needs. Online support group participants consisted of patients with various types of cancer, and the therapists were clinically trained social workers.

METHODS

Our study reports on the mixed methods evaluation of AICF, including therapists' opinions as well as quantitative measures. AICF's ability to detect distress was evaluated by the patient's real-time emoji check-in, the Linguistic Inquiry and Word Count software, and the Impact of Event Scale-Revised.

RESULTS

Although quantitative results showed only some validity of AICF's ability in detecting distress, the qualitative results showed that AICF was able to detect real-time issues that are amenable to treatment, thus allowing therapists to be more proactive in supporting every group member on an individual basis. However, therapists are concerned about the ethical liability of AICF's distress detection function.

CONCLUSIONS

Future works will look into wearable sensors and facial cues by using videoconferencing to overcome the barriers associated with text-based online support groups.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

RR2-10.2196/21453.

Effect of a Machine Learning Recommender System and Viral Peer Marketing Intervention on Smoking Cessation: A Randomized Clinical Trial

IMPORTANCE

Novel data science and marketing methods of smoking-cessation intervention have not been adequately evaluated.

OBJECTIVE

To compare machine learning recommender (ML recommender) computer tailoring of motivational text messages vs a standard motivational text-based intervention (standard messaging) and a viral peer-recruitment tool kit (viral tool kit) for recruiting friends and family vs no tool kit in a smoking-cessation intervention.

DESIGN, SETTING, AND PARTICIPANTS

This 2 ×2 factorial randomized clinical trial with partial allocation, conducted between July 2017 and September 2019 within an online tobacco intervention, recruited current smokers aged 18 years and older who spoke English from the US via the internet and peer referral. Data were analyzed from March through May 2022.

INTERVENTIONS

Participants registering for the online intervention were randomly assigned to the ML recommender or standard messaging groups followed by partially random allocation to access to viral tool kit or no viral tool kit groups. The ML recommender provided ongoing refinement of message selection based on user feedback and comparison with a growing database of other users, while the standard system selected messages based on participant baseline readiness to quit.

MAIN OUTCOMES AND MEASURES

Our primary outcome was self-reported 7-day point prevalence smoking cessation at 6 months.

RESULTS

Of 1487 participants who smoked (444 aged 19-34 years [29.9%], 508 aged 35-54 years [34.1%], 535 aged ≥55 years [36.0%]; 1101 [74.0%] females; 189 Black [12.7%] and 1101 White [78.5%]; 106 Hispanic [7.1%]), 741 individuals were randomly assigned to the ML recommender group and 746 individuals to the standard messaging group; viral tool kit access was provided to 745 participants, and 742 participants received no such access. There was no significant difference in 6-month smoking cessation between ML recommender (146 of 412 participants [35.4%] with outcome data) and standard messaging (156 of 389 participants [40.1%] with outcome data) groups (adjusted odds ratio, 0.81; 95% CI, 0.61-1.08). Smoking cessation was significantly higher in viral tool kit (177 of 395 participants [44.8%] with outcome data) vs no viral tool kit (125 of 406 participants [30.8%] with outcome data) groups (adjusted odds ratio, 1.48; 95% CI, 1.11-1.98).

CONCLUSIONS AND RELEVANCE

In this study, machine learning-based selection did not improve performance compared with standard message selection, while viral marketing did improve cessation outcomes. These results suggest that in addition to increasing dissemination, viral recruitment may have important implications for improving effectiveness of smoking-cessation interventions.

TRIAL REGISTRATION

ClinicalTrials.gov Identifier: NCT03224520.

Human Decision-making in an Artificial Intelligence-Driven Future in Health: Protocol for Comparative Analysis and Simulation

BACKGROUND

Health care can broadly be divided into two domains: clinical health services and complex health services (ie, nonclinical health services, eg, health policy and health regulation). Artificial intelligence (AI) is transforming both of these areas. Currently, humans are leaders, managers, and decision makers in complex health services. However, with the rise of AI, the time has come to ask whether humans will continue to have meaningful decision-making roles in this domain. Further, rationality has long dominated this space. What role will intuition play?

OBJECTIVE

The aim is to establish a protocol of protocols to be used in the proposed research, which aims to explore whether humans will continue in meaningful decision-making roles in complex health services in an AI-driven future.

METHODS

This paper describes a set of protocols for the proposed research, which is designed as a 4-step project across two phases. This paper describes the protocols for each step. The first step is a scoping review to identify and map human attributes that influence decision-making in complex health services. The research question focuses on the attributes that influence human decision-making in this context as reported in the literature. The second step is a scoping review to identify and map AI attributes that influence decision-making in complex health services. The research question focuses on attributes that influence AI decision-making in this context as reported in the literature. The third step is a comparative analysis: a narrative comparison followed by a mathematical comparison of the two sets of attributes-human and AI. This analysis will investigate whether humans have one or more unique attributes that could influence decision-making for the better. The fourth step is a simulation of a nonclinical environment in health regulation and policy into which virtual human and AI decision makers (agents) are introduced. The virtual human and AI will be based on the human and AI attributes identified in the scoping reviews. The simulation will explore, observe, and document how humans interact with AI, and whether humans are likely to compete, cooperate, or converge with AI.

RESULTS

The results will be presented in tabular form, visually intuitive formats, and-in the case of the simulation-multimedia formats.

CONCLUSIONS

This paper provides a road map for the proposed research. It also provides an example of a protocol of protocols for methods used in complex health research. While there are established guidelines for a priori protocols for scoping reviews, there is a paucity of guidance on establishing a protocol of protocols. This paper takes the first step toward building a scaffolding for future guidelines in this regard.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

PRR1-10.2196/42353.

An Overview of Innovative Approaches to Support Timely and Agile Health Communication Research and Practice

Innovative approaches are needed to make health communication research and practice more timely, responsive, and effective in a rapidly changing information ecosystem. In this paper we provide an overview of strategies that can enhance the delivery and effectiveness of health communication campaigns and interventions, as well as research approaches that can generate useful data and insights for decisionmakers and campaign designers, thereby reducing the research-to-practice gap. The discussion focuses on the following approaches: digital segmentation and microtargeting, social media influencer campaigns, recommender systems, adaptive interventions, A/B testing, efficient message testing protocols, rapid cycle iterative message testing, megastudies, and agent-based modeling. For each method highlighted, we also outline important practical and ethical considerations for utilizing the approach in the context of health communication research and practice, including issues related to transparency, privacy, equity, and potential for harm.

Applying Collective Intelligence in Health Recommender Systems for Smoking Cessation: A Comparison Trial

Background: Health recommender systems (HRSs) are intelligent systems that can be used to tailor digital health interventions. We compared two HRSs to assess their impact providing smoking cessation support messages. Methods: Smokers who downloaded a mobile app to support smoking abstinence were randomly assigned to two interventions. They received personalized, ratable motivational messages on the app. The first intervention had a knowledge-based HRS (n = 181): it selected random messages from a subset matching the users’ demographics and smoking habits. The second intervention had a hybrid HRS using collective intelligence (n = 190): it selected messages applying the knowledge-based filter first, and then chose the ones with higher ratings provided by other similar users in the system. Both interventions were compared on: (a) message appreciation, (b) engagement with the system, and (c) one’s own self-reported smoking cessation status, as indicated by the last seven-day point prevalence report in different time intervals during a period of six months. Results: Both interventions had similar message appreciation, number of rated messages, and abstinence results. The knowledge-based HRS achieved a significantly higher number of active days, number of abstinence reports, and better abstinence results. The hybrid algorithm led to more quitting attempts in participants who completed their user profiles.

Evaluating the use of a recommender system for selecting optimal messages for smoking cessation: patterns and effects of user-system engagement

Background

Motivational messaging is a frequently used digital intervention to promote positive health behavior changes, including smoking cessation. Typically, motivational messaging systems have not actively sought feedback on each message, preventing a closer examination of the user-system engagement. This study assessed the granular user-system engagement around a recommender system (a new system that actively sought user feedback on each message to improve message selection) for promoting smoking cessation and the impact of engagement on cessation outcome.

Methods

We prospectively followed a cohort of current smokers enrolled to use the recommender system for 6 months. The system sent participants motivational messages to support smoking cessation every 3 days and used machine learning to incorporate user feedback (i.e., user’s rating on the perceived influence of each message, collected on a 5-point Likert scale with 1 indicating strong disagreement and 5 indicating strong agreement on perceiving the influence on quitting smoking) to improve the selection of the following message. We assessed user-system engagement by various metrics, including user response rate (i.e., the percent of times a user rated the messages) and the perceived influence of messages. We compared retention rates across different levels of user-system engagement and assessed the association between engagement and the 7-day point prevalence abstinence (missing outcome = smoking) by using multiple logistic regression.

Results

We analyzed data from 731 participants (13% Black; 73% women). The user response rate was 0.24 (SD = 0.34) and user-perceived influence was 3.76 (SD = 0.84). The retention rate positively increased with the user response rate (trend test P < 0.001). Compared with non-response, six-month cessation increased with the levels of response rates: low response rate (odds ratio [OR] = 1.86, 95% confidence interval [CI]: 1.07–3.23), moderate response rate (OR = 2.30, 95% CI: 1.36–3.88), high response rate (OR = 2.69, 95% CI: 1.58–4.58). The association between perceived message influence and the outcome showed a similar pattern.

Conclusions

High user-system engagement was positively associated with both high retention rate and smoking cessation, suggesting that investigation of methods to increase engagement may be crucial to increase the impact of the recommender system for smoking cessation.

Trial registration

Registration Identifier: NCT03224520. Registration date: July 21, 2017.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-021-11803-8.

Health Recommender Systems: Systematic Review

Background

Health recommender systems (HRSs) offer the potential to motivate and engage users to change their behavior by sharing better choices and actionable knowledge based on observed user behavior.

Objective

We aim to review HRSs targeting nonmedical professionals (laypersons) to better understand the current state of the art and identify both the main trends and the gaps with respect to current implementations.

Methods

We conducted a systematic literature review according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines and synthesized the results. A total of 73 published studies that reported both an implementation and evaluation of an HRS targeted to laypersons were included and analyzed in this review.

Results

Recommended items were classified into four major categories: lifestyle, nutrition, general health care information, and specific health conditions. The majority of HRSs use hybrid recommendation algorithms. Evaluations of HRSs vary greatly; half of the studies only evaluated the algorithm with various metrics, whereas others performed full-scale randomized controlled trials or conducted in-the-wild studies to evaluate the impact of HRSs, thereby showing that the field is slowly maturing. On the basis of our review, we derived five reporting guidelines that can serve as a reference frame for future HRS studies. HRS studies should clarify who the target user is and to whom the recommendations apply, what is recommended and how the recommendations are presented to the user, where the data set can be found, what algorithms were used to calculate the recommendations, and what evaluation protocol was used.

Conclusions

There is significant opportunity for an HRS to inform and guide health actions. Through this review, we promote the discussion of ways to augment HRS research by recommending a reference frame with five design guidelines.

Natural Language Processing-Based Virtual Cofacilitator for Online Cancer Support Groups: Protocol for an Algorithm Development and Validation Study

Background

Cancer and its treatment can significantly impact the short- and long-term psychological well-being of patients and families. Emotional distress and depressive symptomatology are often associated with poor treatment adherence, reduced quality of life, and higher mortality. Cancer support groups, especially those led by health care professionals, provide a safe place for participants to discuss fear, normalize stress reactions, share solidarity, and learn about effective strategies to build resilience and enhance coping. However, in-person support groups may not always be accessible to individuals; geographic distance is one of the barriers for access, and compromised physical condition (eg, fatigue, pain) is another. Emerging evidence supports the effectiveness of online support groups in reducing access barriers. Text-based and professional-led online support groups have been offered by Cancer Chat Canada. Participants join the group discussion using text in real time. However, therapist leaders report some challenges leading text-based online support groups in the absence of visual cues, particularly in tracking participant distress. With multiple participants typing at the same time, the nuances of the text messages or red flags for distress can sometimes be missed. Recent advances in artificial intelligence such as deep learning–based natural language processing offer potential solutions. This technology can be used to analyze online support group text data to track participants’ expressed emotional distress, including fear, sadness, and hopelessness. Artificial intelligence allows session activities to be monitored in real time and alerts the therapist to participant disengagement.

Objective

We aim to develop and evaluate an artificial intelligence–based cofacilitator prototype to track and monitor online support group participants’ distress through real-time analysis of text-based messages posted during synchronous sessions.

Methods

An artificial intelligence–based cofacilitator will be developed to identify participants who are at-risk for increased emotional distress and track participant engagement and in-session group cohesion levels, providing real-time alerts for therapist to follow-up; generate postsession participant profiles that contain discussion content keywords and emotion profiles for each session; and automatically suggest tailored resources to participants according to their needs. The study is designed to be conducted in 4 phases consisting of (1) development based on a subset of data and an existing natural language processing framework, (2) performance evaluation using human scoring, (3) beta testing, and (4) user experience evaluation.

Results

This study received ethics approval in August 2019. Phase 1, development of an artificial intelligence–based cofacilitator, was completed in January 2020. As of December 2020, phase 2 is underway. The study is expected to be completed by September 2021.

Conclusions

An artificial intelligence–based cofacilitator offers a promising new mode of delivery of person-centered online support groups tailored to individual needs.

International Registered Report Identifier (IRRID)

DERR1-10.2196/21453

Comparison of a Collective Intelligence Tailored Messaging System on Smoking Cessation Between African American and White People Who Smoke: Quasi-Experimental Design

BACKGROUND

The Patient Experience Recommender System for Persuasive Communication Tailoring (PERSPeCT) is a machine learning recommender system with a database of messages to motivate smoking cessation. PERSPeCT uses the collective intelligence of users (ie, preferences and feedback) and demographic and smoking profiles to select motivating messages. PERSPeCT may be more beneficial for tailoring content to minority groups influenced by complex, personally relevant factors.

OBJECTIVE

The objective of this study was to describe and evaluate the use of PERSPeCT in African American people who smoke compared with white people who smoke.

METHODS

Using a quasi-experimental design, we compared African American people who smoke with a historical cohort of white people who smoke, who both received up to 30 emailed tailored messages over 65 days. People who smoke rated the daily message in terms of perceived influence on quitting smoking for 30 days. Our primary analysis compared daily message ratings between the two groups using a t test. We used a logistic model to compare 30-day cessation between the two groups and adjusted for covariates.

RESULTS

The study included 119 people who smoke (African Americans, 55/119; whites, 64/119). At baseline, African American people who smoke were significantly more likely to report allowing smoking in the home (P=.002); all other characteristics were not significantly different between groups. Daily mean ratings were higher for African American than white people who smoke on 26 of the 30 days (P<.001). Odds of quitting as measured by 30-day cessation were significantly higher for African Americans (odds ratio 2.3, 95% CI 1.04-5.53; P=.03) and did not change after adjusting for allowing smoking at home.

CONCLUSIONS

Our study highlighted the potential of using a recommender system to personalize for African American people who smoke.

TRIAL REGISTRATION

ClinicalTrials.gov NCT02200432; https://clinicaltrials.gov/ct2/show/NCT02200432.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

RR2-10.2196/jmir.6465.

Dissemination and Effectiveness of the Peer Marketing and Messaging of a Web-Assisted Tobacco Intervention: Protocol for a Hybrid Effectiveness Trial

BACKGROUND

Smoking continues to be the leading preventable cause of death. Digital Interventions for Smoking Cessation (DISCs) are health communication programs accessible via the internet and smartphones and allow for greater reach and effectiveness of tobacco cessation programs. DISCs have led to increased 6-month cessation rates while also reaching vulnerable populations. Despite this, the impact of DISCs has been limited and new ways to increase access and effectiveness are needed.

OBJECTIVE

We are conducting a hybrid effectiveness-dissemination study. We aim to evaluate the effectiveness of a machine learning-based approach (recommender system) for computer-tailored health communication (CTHC) over a standard CTHC system based on quit rates and risk reduction. In addition, this study will assess the dissemination of providing access to a peer recruitment toolset on recruitment rate and variability of the sample.

METHODS

The Smoker-to-Smoker (S2S) study is a 6-month hybrid effectiveness dissemination trial conducted nationally among English-speaking, current smokers aged ≥18 years. All eligible participants will register for the DISC (Decide2quit) and be randomized to the recommender system CTHC or the standard CTHC, followed by allocation to a peer recruitment toolset group or control group. Primary outcomes will be 7-day point prevalence and risk reduction at the 6-month follow-up. Secondary outcomes include recruitment rate, website engagement, and patient-reported outcomes collected via the 6-month follow-up questionnaire. All primary analyses will be conducted on an intent-to-treat basis.

RESULTS

The project is funded from 2017 to 2020 by the Patient Centered Outcomes Research Institute. Enrollment was completed in early 2019, and 6-month follow-ups will be completed by late 2019. Preliminary data analysis is currently underway.

CONCLUSIONS

Conducting a hybrid study with both effectiveness and dissemination hypotheses raises some unique challenges in the study design and analysis. Our study addresses these challenges to test new innovations and increase the effectiveness and reach of DISCs.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

DERR1-10.2196/14814.

How recommender systems could support and enhance computer-tailored digital health programs: A scoping review

Objective

Tailored digital health programs can promote positive health-related lifestyle changes and have been shown to be (cost) effective in trials. However, such programs are used suboptimally. New approaches are needed to optimise the use of these programs. This paper illustrates the potential of recommender systems to support and enhance computer-tailored digital health interventions. The aim is threefold, to explore: (1) how recommender systems provide health recommendations, (2) to what extent recommender systems incorporate theoretical models and (3) how the use of recommender systems may enhance the usage of computer-tailored interventions.

Methods

A scoping review was conducted, using MEDLINE and ScienceDirect, to identify health recommender systems reported in studies between January 2007 and December 2017. Information was subsequently extracted to understand the potential benefits of recommender systems for computer-tailored digital health programs. Titles and abstracts of 1184 studies were screened for the full-text screening, in which two reviewers independently selected articles and systematically extracted data using a predefined extraction form.

Results

A total of 26 articles were included for data extraction. General characteristics were reported, with eight studies reporting hybrid filtering. A description of how each recommender system provides a recommendation is described; the majority of recommender systems used messages as recommendation. We identified the potential effects of recommender systems on efficiency, effectiveness, trustworthiness and enjoyment of the digital health program.

Conclusions

Incorporating a collaborative method with demographic filtering as a second step to knowledge-based filtering could potentially add value to traditional tailoring with regard to enhancing the user experience. This study illustrates how recommender systems, especially hybrid programs, may have the potential to bring tailored digital health forward.

Mining User-Generated Content in an Online Smoking Cessation Community to Identify Smoking Status: A Machine Learning Approach

Online smoking cessation communities help hundreds of thousands of smokers quit smoking and stay abstinent each year. Content shared by users of such communities may contain important information that could enable more effective and personally tailored cessation treatment recommendations. This study demonstrates a novel approach to determine individuals' smoking status by applying machine learning techniques to classify user-generated content in an online cessation community. Study data were from BecomeAnEX.org, a large, online smoking cessation community. We extracted three types of novel features from a post: domain-specific features, author-based features, and thread-based features. These features helped to improve the smoking status identification (quit vs. not) performance by 9.7% compared to using only text features of a post's content. In other words, knowledge from domain experts, data regarding the post author's patterns of online engagement, and other community member reactions to the post can help to determine the focal post author's smoking status, over and above the actual content of a focal post. We demonstrated that machine learning methods can be applied to user-generated data from online cessation communities to validly and reliably discern important user characteristics, which could aid decision support on intervention tailoring.

Transforming health policy through machine learning

In their Perspective, Ara Darzi and Hutan Ashrafian give us a tour of the future policymaker's machine learning toolkit.

Intuitive and interpretable visual communication of a complex statistical model of disease progression and risk

Computer science and machine learning in particular are increasingly lauded for their potential to aid medical practice. However, the highly technical nature of the state of the art techniques can be a major obstacle in their usability by health care professionals and thus, their adoption and actual practical benefit. In this paper we describe a software tool which focuses on the visualization of predictions made by a recently developed method which leverages data in the form of large scale electronic records for making diagnostic predictions. Guided by risk predictions, our tool allows the user to explore interactively different diagnostic trajectories, or display cumulative long term prognostics, in an intuitive and easily interpretable manner.

Analyzing recommender systems for health promotion using a multidisciplinary taxonomy: A scoping review

BACKGROUND

Recommender systems are information retrieval systems that provide users with relevant items (e.g., through messages). Despite their extensive use in the e-commerce and leisure domains, their application in healthcare is still in its infancy. These systems may be used to create tailored health interventions, thus reducing the cost of healthcare and fostering a healthier lifestyle in the population.

OBJECTIVE

This paper identifies, categorizes, and analyzes the existing knowledge in terms of the literature published over the past 10 years on the use of health recommender systems for patient interventions. The aim of this study is to understand the scientific evidence generated about health recommender systems, to identify any gaps in this field to achieve the United Nations Sustainable Development Goal 3 (SDG3) (namely, "Ensure healthy lives and promote well-being for all at all ages"), and to suggest possible reasons for these gaps as well as to propose some solutions.

METHODS

We conducted a scoping review, which consisted of a keyword search of the literature related to health recommender systems for patients in the following databases: ScienceDirect, PsycInfo, Association for Computing Machinery, IEEExplore, and Pubmed. Further, we limited our search to consider only English-language journal articles published in the last 10 years. The reviewing process comprised three researchers who filtered the results simultaneously. The quantitative synthesis was conducted in parallel by two researchers, who classified each paper in terms of four aspects-the domain, the methodological and procedural aspects, the health promotion theoretical factors and behavior change theories, and the technical aspects-using a new multidisciplinary taxonomy.

RESULTS

Nineteen papers met the inclusion criteria and were included in the data analysis, for which thirty-three features were assessed. The nine features associated with the health promotion theoretical factors and behavior change theories were not observed in any of the selected studies, did not use principles of tailoring, and did not assess (cost)-effectiveness.

DISCUSSION

Health recommender systems may be further improved by using relevant behavior change strategies and by implementing essential characteristics of tailored interventions. In addition, many of the features required to assess each of the domain aspects, the methodological and procedural aspects, and technical aspects were not reported in the studies.

CONCLUSIONS

The studies analyzed presented few evidence in support of the positive effects of using health recommender systems in terms of cost-effectiveness and patient health outcomes. This is why future studies should ensure that all the proposed features are covered in our multidisciplinary taxonomy, including integration with electronic health records and the incorporation of health promotion theoretical factors and behavior change theories. This will render those studies more useful for policymakers since they will cover all aspects needed to determine their impact toward meeting SDG3.

Designing more engaging computer-tailored physical activity behaviour change interventions for breast cancer survivors: lessons from the iMove More for Life study

BACKGROUND

Participating in regular physical activity is a recommended cancer recovery strategy for breast cancer survivors. However, tailored support services are not widely available and most survivors are insufficiently active to obtain health benefits. Delivering tailored programs via the Internet offers one promising approach. However, recent evaluations of such programs suggest that major improvements are needed to ensure programs meet the needs of users and are delivered in an engaging way. Understanding participants' experiences with current programs can help to inform the next generation of systems.

PURPOSE

The purposes of this study are to explore breast cancer survivor's perspectives of and experiences using a novel computer-tailored intervention and to describe recommendations for future iterations.

METHODS

Qualitative data from a sub-sample of iMove More for Life study participants were analysed thematically to identify key themes. Participants long-term goals for participating in the program were explored by analysing open-ended data extracted from action plans completed during the intervention (n = 370). Participants negative and positive perceptions of the website and recommendations for improvement were explored using data extracted from open-ended survey items collected at the immediate intervention follow-up (n = 156).

RESULTS

The majority of participants reported multi-faceted goals, consisting of two or more outcomes they hoped to achieve within a year. While clear themes were identified (e.g. 'being satisfied with body weight'), there was considerable variability in the scope of the goal (e.g. desired weight loss ranged from 2 to 30 kg). Participants' perceptions of the website were mixed, but clear indications were provided of how intervention content and structure could be improved.

CONCLUSIONS

This study provides insight into how to better accommodate breast cancer survivors in the future and ultimately design more engaging computer-tailored interventions.

HealthRecSys: A semantic content-based recommender system to complement health videos

BACKGROUND

The Internet, and its popularity, continues to grow at an unprecedented pace. Watching videos online is very popular; it is estimated that 500 h of video are uploaded onto YouTube, a video-sharing service, every minute and that, by 2019, video formats will comprise more than 80% of Internet traffic. Health-related videos are very popular on YouTube, but their quality is always a matter of concern. One approach to enhancing the quality of online videos is to provide additional educational health content, such as websites, to support health consumers. This study investigates the feasibility of building a content-based recommender system that links health consumers to reputable health educational websites from MedlinePlus for a given health video from YouTube.

METHODS

The dataset for this study includes a collection of health-related videos and their available metadata. Semantic technologies (such as SNOMED-CT and Bio-ontology) were used to recommend health websites from MedlinePlus. A total of 26 healths professionals participated in evaluating 253 recommended links for a total of 53 videos about general health, hypertension, or diabetes. The relevance of the recommended health websites from MedlinePlus to the videos was measured using information retrieval metrics such as the normalized discounted cumulative gain and precision at K.

RESULTS

The majority of websites recommended by our system for health videos were relevant, based on ratings by health professionals. The normalized discounted cumulative gain was between 46% and 90% for the different topics.

CONCLUSIONS

Our study demonstrates the feasibility of using a semantic content-based recommender system to enrich YouTube health videos. Evaluation with end-users, in addition to healthcare professionals, will be required to identify the acceptance of these recommendations in a nonsimulated information-seeking context.

Impact of a Collective Intelligence Tailored Messaging System on Smoking Cessation: The Perspect Randomized Experiment

Background

Outside health care, content tailoring is driven algorithmically using machine learning compared to the rule-based approach used in current implementations of computer-tailored health communication (CTHC) systems. A special class of machine learning systems (“recommender systems”) are used to select messages by combining the collective intelligence of their users (ie, the observed and inferred preferences of users as they interact with the system) and their user profiles. However, this approach has not been adequately tested for CTHC.

Objective

Our aim was to compare, in a randomized experiment, a standard, evidence-based, rule-based CTHC (standard CTHC) to a novel machine learning CTHC: Patient Experience Recommender System for Persuasive Communication Tailoring (PERSPeCT). We hypothesized that PERSPeCT will select messages of higher influence than our standard CTHC system. This standard CTHC was proven effective in motivating smoking cessation in a prior randomized trial of 900 smokers (OR 1.70, 95% CI 1.03-2.81).

Methods

PERSPeCT is an innovative hybrid machine learning recommender system that selects and sends motivational messages using algorithms that learn from message ratings from 846 previous participants (explicit feedback), and the prior explicit ratings of each individual participant. Current smokers (N=120) aged 18 years or older, English speaking, with Internet access were eligible to participate. These smokers were randomized to receive either PERSPeCT (intervention, n=74) or standard CTHC tailored messages (n=46). The study was conducted between October 2014 and January 2015. By randomization, we compared daily message ratings (mean of smoker ratings each day). At 30 days, we assessed the intervention’s perceived influence, 30-day cessation, and changes in readiness to quit from baseline.

Results

The proportion of days when smokers agreed/strongly agreed (daily rating ≥4) that the messages influenced them to quit was significantly higher for PERSPeCT (73%, 23/30) than standard CTHC (44%, 14/30, P=.02). Among less educated smokers (n=49), this difference was even more pronounced for days strongly agree (intervention: 77%, 23/30; comparison: 23%, 7/30, P<.001). There was no significant difference in the frequency which PERSPeCT randomized smokers agreed or strongly agreed that the intervention influenced them to quit smoking (P=.07) and use nicotine replacement therapy (P=.09). Among those who completed follow-up, 36% (20/55) of PERSPeCT smokers and 32% (11/34) of the standard CTHC group stopped smoking for one day or longer (P=.70).

Conclusions

Compared to standard CTHC with proven effectiveness, PERSPeCT outperformed in terms of influence ratings and resulted in similar cessation rates.

ClinicalTrial

Clinicaltrials.gov NCT02200432; https://clinicaltrials.gov/ct2/show/NCT02200432 (Archived by WebCite at http://www.webcitation.org/6lEJY1KEd)

Prioritizing the mHealth Design Space: A Mixed-Methods Analysis of Smokers' Perspectives

Background

Smoking remains the leading cause of preventable disease and death in the United States. Therefore, researchers are constantly exploring new ways to promote smoking cessation. Mobile health (mHealth) technologies could be effective cessation tools. Despite the availability of commercial quit-smoking apps, little research to date has examined smokers’ preferred treatment intervention components (ie, design features). Honoring these preferences is important for designing programs that are appealing to smokers and may be more likely to be adopted and used.

Objective

The aim of this study was to understand smokers’ preferred design features of mHealth quit-smoking tools.

Methods

We used a mixed-methods approach consisting of focus groups and written surveys to understand the design preferences of adult smokers who were interested in quitting smoking (N=40). Focus groups were stratified by age to allow differing perspectives to emerge between older (>40 years) and younger (<40 years) participants. Focus group discussion included a “blue-sky” brainstorming exercise followed by participant reactions to contrasting design options for communicating with smokers, providing social support, and incentivizing program use. Participants rated the importance of preselected design features on an exit survey. Qualitative analyses examined emergent discussion themes and quantitative analyses compared feature ratings to determine which were perceived as most important.

Results

Participants preferred a highly personalized and adaptive mHealth experience. Their ideal mHealth quit-smoking tool would allow personalized tracking of their progress, adaptively tailored feedback, and real-time peer support to help manage smoking cravings. Based on qualitative analysis of focus group discussion, participants preferred pull messages (ie, delivered upon request) over push messages (ie, sent automatically) and preferred interaction with other smokers through closed social networks. Preferences for entertaining games or other rewarding incentives to encourage program use differed by age group. Based on quantitative analysis of surveys, participants rated the importance of select design features significantly differently (P<.001). Design features rated as most important included personalized content, the ability to track one’s progress, and features designed to help manage nicotine withdrawal and medication side effects. Design features rated least important were quit-smoking videos and posting on social media. Communicating with stop-smoking experts was rated more important than communicating with family and friends about quitting (P=.03). Perceived importance of various design features varied by age, experience with technology, and frequency of smoking.

Conclusions

Future mHealth cessation aids should be designed with an understanding of smokers’ needs and preferences for these tools. Doing so does not guarantee treatment effectiveness, but balancing user preferences with best-practice treatment considerations could enhance program adoption and improve treatment outcomes. Grounded in the perspectives of smokers, we identify several design considerations, which should be prioritized when designing future mHealth cessation tools and which warrant additional empirical validation.