This research study is being conducted to evaluate digital software tools that collect health information through conversations. The study will test how acceptable, feasible, and usable these conversational software programs are for patients with chronic pain.
Purpose of the Study:
The goal is to understand whether digital conversational tools can effectively gather biopsychosocial information (biological, psychological, and social factors related to health) from patients experiencing chronic pain. Biopsychosocial information includes details about your physical symptoms, emotional well-being, social circumstances, and how these factors interact to affect your health. These digital tools are designed to have conversation-like interactions with patients to collect this comprehensive health data in a more natural and engaging way.
What Participation Involves:
If you join this study, you will interact with digital conversational software that will ask you questions about your health, pain experience, and related factors. The software is designed to communicate with you in a conversational manner, similar to talking with a healthcare provider, to gather information about your condition. You will be asked about various aspects of your chronic pain experience, including how it affects your daily life, your emotional state, and your social interactions.
Who Can Participate:
This study is looking for participants who have chronic pain conditions. Medical providers who treat patients with chronic pain are encouraged to refer eligible patients to this study. Healthcare providers interested in referring patients or learning more about the study can contact the study sponsor, AugMend Health Inc.
Why This Research Matters:
Chronic pain affects millions of people and understanding the full scope of how it impacts patients requires collecting detailed information about physical, psychological, and social factors. Traditional methods of collecting this information can be time-consuming and may not capture the complete picture. Digital conversational tools offer a potential new approach to gathering comprehensive health information in a more efficient and patient-friendly manner. This research will evaluate whether patients find these tools acceptable to use, whether they work practically in real-world clinical settings, and whether they are user-friendly and accessible for people with chronic pain.
Study Location:
The study is being conducted at AugMend Health in Cambridge, Massachusetts.
Study Sponsor:
This study is sponsored by AugMend Health Inc. Healthcare providers or potential participants who would like more information about the study or wish to make a referral can contact AugMend Health Inc.
Participation in this study is voluntary, and you can choose to stop participating at any time without affecting your regular medical care.
Adults aged 18 years or older
Diagnosis of chronic pain (pain lasting 3 months or longer)
Able to read and understand English
Able to provide informed consent
Under 18 years of age
Unable to provide informed consent
Non-English speaking
Acute pain only (pain lasting less than 3 months)
Parallel
Digital Biopsychosocial Conversational Data Collection Software
Other interventionThis intervention involves participants interacting with digital conversational software that uses artificial intelligence and natural language processing to collect biopsychosocial health information. The software engages participants in dialogue-based interactions, asking questions about their chronic pain experience, physical functioning, emotional well-being, and social circumstances. The conversational interface adapts to participant responses, can ask follow-up questions for clarification, and creates an interactive experience that mimics clinical interviews. Some implementations may utilize virtual reality (VR) environments to create immersive data collection experiences at the MIT.nano Immersion Lab. Participants interact with the software either at home using provided devices or at the MIT.nano Immersion Lab. The AI-driven platform standardizes data collection while maintaining conversational flexibility, collecting comprehensive biopsychosocial data through natural language
Standard Digital Questionnaire Data Collection
Other interventionThis intervention involves participants completing standard digital questionnaires or surveys to provide biopsychosocial health information. Participants respond to predetermined, fixed-format questions about their chronic pain experience, physical functioning, emotional well-being, and social circumstances. The data collection uses conventional electronic forms or surveys with static question sets that do not adapt based on participant responses. Questions are presented in a standard sequence without conversational elements or follow-up prompts. This represents traditional electronic data collection methods commonly used in clinical and research settings. Participants complete the questionnaires either at home using their own devices or provided equipment, or at designated study locations. This method provides a comparison baseline against which to evaluate newer digital data collection approaches, representing current standard practice for gathering patient-reported health informat
Acceptability
Acceptability will be measured using the Acceptability, Feasibility and Usability Survey (AFUS) - Acceptability Subscale, a researcher-developed instrument. This subscale includes items from Part 1 (Core Assessment Experience), Part 2 (culture), Part 3 (Future Use \& Preference), Part 4 (Privacy \& Trust), and Part 5 (Medical Disclosure Comfort). Items are rated on a 7-point Likert scale (1 = Strongly Disagree to 7 = Strongly Agree). Subscale scores will be calculated as the mean of all acceptability items. Higher scores indicate greater acceptability. Subscale breakdown will be accessible in supplemental figures of publications.
Usability
Usability will be measured using the Acceptability, Feasibility and Usability Survey (AFUS) - Usability Subscale (Part 6), a researcher-developed instrument adapted from established usability frameworks. The subscale consists of 10 items rated on a 7-point Likert scale (1 = Strongly Disagree to 7 = Strongly Agree). Negatively-worded items are reverse-scored. The total score is calculated as the mean of all items. Higher scores indicate better usability.
Technical Feasibility
Technical feasibility will be measured by the proportion of data collection sessions completed without significant technical failures. A significant technical failure is defined as any system error, software malfunction, connectivity interruption, hardware issue, or other technical problem that results in early termination of the session. Technical success rate will be calculated as the number of sessions without significant technical failures divided by total sessions attempted, expressed as a percentage. Higher percentages indicate greater technical feasibility.
Implementation Feasibility
Implementation feasibility will be measured using the Acceptability, Feasibility and Usability Survey (AFUS) - Future Use \& Preference Subscale (Part 3). Items assess participant willingness to use the data collection method again, likelihood of recommending it to others, and trust in the method for ongoing health evaluation. Items are rated on a 7-point Likert scale (1 = Strongly Disagree to 7 = Strongly Agree). The subscale score is calculated as the mean of all items. Higher scores indicate greater implementation feasibility.
Participant Experience & Satisfaction
Patient experience will be measured using the Acceptability, Feasibility and Usability Survey (AFUS) - Overall Experience Subscale (Part 7). Items are rated on a 7-point Likert scale (1 = Strongly Disagree to 7 = Strongly Agree). The subscale score is calculated as the mean of all items. Higher scores indicate more positive patient experience.
Patient Engagement - Session Duration
Patient engagement will be measured by session duration, defined as the total time in minutes participants actively spend on data collection tasks, recorded via system logs. Longer durations indicate greater time engagement with the data collection process.
Patient Engagement - Response Elaboration
Patient engagement will be measured by mean response length, defined as the average number of words per participant response across the conversation transcript. Higher word counts indicate greater verbal engagement with the data collection process. (experimental group only)
Perceived Therapeutic Alliance
Perceived therapeutic alliance with the data collection interface will be measured using the Acceptability, Feasibility and Usability Survey (AFUS) - Therapeutic Alliance Subscale (Module A). Items are rated on a 7-point Likert scale (1 = Strongly Disagree to 7 = Strongly Agree). The subscale score is calculated as the mean of all items. Higher scores indicate stronger perceived therapeutic alliance.
Correlation Between Physiological Measures and Engagement
The association between physiological arousal and participant engagement will be measured using correlation coefficients between Empatica-derived physiological measures (electrodermal activity, pulse rate) and engagement metrics (time spent answering questions, response word count). Correlation coefficients range from -1.0 to 1.0, with values further from zero indicating stronger associations.
Correlation Between Physiological Measures and Self-Reported Affect
The association between physiological measures and self-reported emotional states will be measured using correlation coefficients between Empatica-derived data (electrodermal activity, pulse rate, skin temperature) and participant-reported affective states from the transcripts and standardized questionnaires. Correlation coefficients range from -1.0 to 1.0, with values further from zero indicating stronger concordance between physiological and self-reported measures.
Emotive & Linguistic Analysis
Four linguistic metrics will be derived from transcripts: (1) Response Elaboration-mean words per response, with higher counts indicating greater engagement; (2) Linguistic Complexity-Flesch-Kincaid Grade Level score (0-18+) representing text sophistication; (3) Vocabulary Diversity-Type-Token Ratio (unique words/total words, 0-1.0); and (4) Self-Referential Language-percentage of first-person singular pronouns (I, me, my, mine) calculated via LIWC-22, with higher percentages associated with depression and pain catastrophizing.
Sentiment Valence of Participant Responses
Overall sentiment will be measured using automated sentiment analysis of transcripts. Sentiment scores will be calculated using validated natural language processing methods, yielding a score ranging from -1.0 (entirely negative) to +1.0 (entirely positive), with 0 representing neutral sentiment. Higher scores indicate more positive sentiment.
Click on the lead sponsor to view all of their trials.
