The goal of this randomized crossover trial is to examine whether the AI-empowered chatbot will effectively collect comprehensive migraine feature data and demonstrate higher compliance, accuracy, utilization, acceptability, feasibility, and validity in migraine feature tracking compared to traditional paper-based migraine diaries among Hong Kong women with migraine. Participants will use different tools (AI-empowered voice-interactive chatbot or paper migraine diary) for migraine tracking. Participants will be randomly allocated to one of two intervention sequences: (1) AI-empowered voice-interactive chatbot followed by paper migraine diary, or (2) paper migraine diary followed by AI-empowered voice-interactive chatbot. The study will consist of two intervention periods, each lasting one month, with a 1:1 allocation ratio.
Study Type
INTERVENTIONAL
Allocation
RANDOMIZED
Purpose
OTHER
Masking
NONE
Enrollment
49
Participants will interact with the chatbot on a daily basis to record their migraine experiences. They will be asked to initiate a conversation with the chatbot whenever they experience a migraine attack and provide the required information, including the date and time of the attack, presence of aura, migraine intensity, accompanying symptoms, triggers, and medication use. The chatbot will engage participants in a conversation to collect these data points and store them securely within the chatbot system.
Participants will manually document their migraine experiences on the paper migraine diary, including the date and time of the attack, presence of aura, migraine intensity, accompanying symptoms, triggers, and medication use.
Participants will manually document their migraine experiences on the paper migraine diary, including the date and time of the attack, presence of aura, migraine intensity, accompanying symptoms, triggers, and medication use.
Participants will interact with the chatbot on a daily basis to record their migraine experiences. They will be asked to initiate a conversation with the chatbot whenever they experience a migraine attack and provide the required information, including the date and time of the attack, presence of aura, migraine intensity, accompanying symptoms, triggers, and medication use. The chatbot will engage participants in a conversation to collect these data points and store them securely within the chatbot system.
The Hong Kong Polytechnic University
Hong Kong, HONG KONG, Hong Kong
timeliness
Timeliness will be defined as the proportion of migraine episodes recorded within 24 hours of onset. This measure will be derived from the data collected through the chatbot and paper diary.
Time frame: At the end of each 1-month intervention period
Completeness
Completeness will be defined as the proportion of diary entries with all required migraine-related information fully recorded. This measure will be derived from the data collected through the chatbot and paper diary.
Time frame: At the end of each 1-month intervention period
Missed episodes
Missed episodes will be defined as the proportion of migraine episodes not recorded; the number of missed episodes in each period will be obtained at the end-of-period questionnaire
Time frame: At the end of each 1-month intervention period
Accuracy
Accuracy will be defined as the proportion of diary entries in which all required migraine-related fields were recorded correctly. Accuracy will be adjudicated by assessing the consistency between migraine features recorded using the chatbot or paper diary and those reported during end-of-period online interviews with study staff. Discrepancies will be documented; records will be not altered for primary analyses. During these interviews, researchers will summarise the participant's recorded migraine information for that period, regardless of modality, and ask the participant to identify any errors.
Time frame: At the end of each 1-month intervention period
Utilization
Utilization will be evaluated using the Chinese version of the System Usability Scale (SUS), a widely used and validated 10-item questionnaire that assesses the usability of a system on a 5-point Likert scale. The SUS yields a single score ranging from 0 to 100, with higher scores indicating better usability. Specifically, chatbot utilization will be objectively assessed using system-generated data, such as the number of logins, user-initiated conversations, and user responses.
Time frame: At the end of each 1-month intervention period
Acceptability
Acceptability will be measured using the validated Chinese version of Technology Acceptance Model (TAM) scale, an 11-item questionnaire on a 7-point Likert scale with higher scores indicating greater acceptance.
Time frame: At the end of each 1-month intervention period
Usage
Usage will be assessed by the number of uses (entries) per tool.
Time frame: At the end of each 1-month intervention period
Time per recording
Time required to record each migraine episode, using system-logged durations for the chatbot and self-recorded start and finish times for the paper diary.
Time frame: At the end of each 1-month intervention period
Validity
The validity of the migraine data collected through the chatbot and paper diary will be assessed using convergent validity and divergent validity. Convergent validity will be evaluated by examining the correlation between migraine frequency and severity recorded using the chatbot or paper diary and the Chinese version of the Migraine Disability Assessment (MIDAS) scores (0 to 270, higher scores indicating greater migraine-related disability). Divergent validity will be assessed by analyzing the correlation between migraine frequency and severity and the Chinese version of the Brief Self-Control Scale (7 to 35, with a higher score indicating greater self-control).
Time frame: At the end of each 1-month intervention period
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.