Behavioral nudges to reduce doctors' overreliance on AI diagnostic suggestions
Mitigating Automation Bias in Physician-LLM Diagnostic Reasoning Using Behavioral Nudges
This test tries to see if short behavioral nudges help doctors who use ChatGPT-5.1 notice and reject incorrect diagnostic suggestions.
Quick facts
| Phase | Not applicable |
|---|---|
| Study type | Interventional |
| Enrollment | 50 (estimated) |
| Sex | All |
| Sponsor | Lahore University of Management Sciences Academic / other |
| Locations | 1 site (Lahore, Punjab Province) |
| Trial ID | NCT07328815 on ClinicalTrials.gov |
What this trial studies
In this randomized controlled trial PMDC-registered physicians who have completed at least 10 hours of structured LLM training are randomized to view ChatGPT-5.1 recommendations with or without a dual-mechanism behavioral nudge. Each participant reviews six clinical vignettes paired with LLM-generated recommendations, three of which contain deliberate clinically significant errors, and then records diagnostic judgments. The behavioral nudge combines a baseline accuracy anchor with case-specific color-coded confidence signals designed to prompt critical review of the AI output. The primary outcome is the rate of uncritical acceptance of incorrect LLM recommendations, compared between nudge and control groups at a single site in Lahore, Pakistan.
Who should consider this trial
Good fit: Ideal participants are PMDC-registered physicians who have completed a minimum of 10 hours of structured training on ChatGPT or a comparable LLM including hands-on prompt engineering and content evaluation.
Not a fit: Patients whose care is delivered without LLM support or by clinicians who are not PMDC-registered or who lack the required LLM training are unlikely to benefit directly from this intervention.
Why it matters
Potential benefit: If successful, the intervention could reduce harmful overreliance on AI suggestions and improve diagnostic safety when physicians use LLMs.
How similar studies have performed: Prior studies of decision-support nudges and accuracy-anchoring have shown mixed but promising reductions in automation bias in simulated and clinical contexts, while direct evidence with large language models remains limited and novel.
Eligibility criteria
Show full inclusion / exclusion criteria
Inclusion Criteria: * Full or Provisionally Registered Medical Practitioners with the Pakistan Medical and Dental Council (PMDC). * Completed Bachelor of Medicine, Bachelor of Surgery (MBBS) Exam. The equivalent degree of MBBS in US and Canada is the Doctor of Medicine (MD). * Participants must have completed a structured training program on the use of ChatGPT (or a comparable large language model), totaling at least 10 hours of instruction. The program must include hands-on practice related to LLM's key aspects, specifically prompt engineering and content evaluation. Exclusion Criteria: * Any other Registered Medical Practitioners (Full or Provisional) with PMDC (e.g., professionals with Bachelor of Dental Surgery or BDS).
Where this trial is running
Lahore, Punjab Province
- Lahore University of Management Sciences — Lahore, Punjab Province, Pakistan (Recruiting)
Study contacts
- Principal investigator: Ihsan Ayyub Qazi, PhD — Lahore University of Management Sciences (LUMS)
- Study coordinator: Ihsan Ayyub Qazi, PhD
- Email: ihsan.qazi@lums.edu.pk
- Phone: +923233333766
How to participate
- Review the eligibility criteria above with your treating physician.
- Visit the official trial page on ClinicalTrials.gov for the most current contact information and recruitment status.
- Contact the listed study coordinator or principal investigator to request pre-screening. Pre-screening is free and never obligates you to enroll.