Voice-Based Screening for SARS-CoV-2 Exposure in Cardiovascular Clinics (VOICE-COVID-19-II): Protocol for a Randomized Controlled Trial (Preprint)
Bibliographic record
Abstract
BACKGROUND The COVID-19 pandemic has disrupted the health care system, limiting health care resources such as the availability of health care professionals, patient monitoring, contact tracing, and continuous surveillance. As a result of this significant burden, digital tools have become an important asset in increasing the efficiency of patient care delivery. Digital tools can help support health care institutions by tracking transmission of the virus, aiding in the screening process, and providing telemedicine support. However, digital health tools face challenges associated with barriers to accessibility, efficiency, and privacy-related ethical issues. OBJECTIVE This paper describes the study design of an open-label, noninterventional, crossover, randomized controlled trial aimed at assessing whether interactive voice response systems can screen for SARS-CoV-2 in patients as accurately as standard screening done by people. The study aims to assess the concordance and interrater reliability of symptom screening done by Amazon Alexa compared to manual screening done by research coordinators. The perceived level of comfort of patients when interacting with voice response systems and their personal experience will also be evaluated. METHODS A total of 52 patients visiting the heart failure clinic at the Royal Victoria Hospital of the McGill University Health Center, in Montreal, Quebec, will be recruited. Patients will be randomly assigned to first be screened for symptoms of SARS-CoV-2 either digitally, by Amazon Alexa, or manually, by the research coordinator. Participants will subsequently be crossed over and screened either digitally or manually. The clinical setup includes an Amazon Echo Show, a tablet, and an uninterrupted power supply mounted on a mobile cart. The primary end point will be the interrater reliability on the accuracy of randomized screening data performed by Amazon Alexa versus research coordinators. The secondary end point will be the perceived level of comfort and app engagement of patients as assessed using 5-point Likert scales and binary mode responses. RESULTS Data collection started in May 2021 and is expected to be completed in fall 2022. Data analysis is expected to be completed in early 2023. CONCLUSIONS The use of voice-based assistants could improve the provision of health services and reduce the burden on health care personnel. Demonstrating a high interrater reliability between Amazon Alexa and health care coordinators may serve future digital tools to streamline the screening and delivery of care in the context of other conditions and clinical settings. The COVID-19 pandemic occurs during the first digital era using digital tools such as Amazon Alexa for disease screening, and it represents an opportunity to implement such technology in health care institutions in the long term. CLINICALTRIAL ClinicalTrials.gov NCT04508972; https://clinicaltrials.gov/ct2/show/NCT04508972 INTERNATIONAL REGISTERED REPORT DERR1-10.2196/41209
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Direct model labels (unvalidated)
Per-model category and study-design labels from the labeling rounds. They are machine output, unvalidated, and the disagreement between models ships as data. No study design here is MEDLINE-validated yet.
| Model arm | Categories | Study design | Confidence |
|---|---|---|---|
| gemma | no category Domain: not available · Genre: Protocol About the Canadian research system: no · About a Canadian topic: no | Randomized trial | low |
| gpt | no category Domain: not available · Genre: Protocol About the Canadian research system: no · About a Canadian topic: no | Randomized trial | low |
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.034 | 0.031 |
| Meta-epidemiology (narrow) | 0.006 | 0.003 |
| Meta-epidemiology (broad) | 0.010 | 0.005 |
| Bibliometrics | 0.003 | 0.003 |
| Science and technology studies | 0.003 | 0.003 |
| Scholarly communication | 0.004 | 0.004 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.008 | 0.007 |
| Insufficient payload (model declined to judge) | 0.089 | 0.013 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedLabeled directly by 2 models reading the full record.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".