Effectiveness of AI-Assisted Antibiotic Prescribing

NCT07833293 | Not Yet Recruiting

• 1 other identifier: LUMS-IRB-0492
• Study Type: interventional
• Target: 50
• Locations: 1 country
• Sites: 1

Brief Summary

This randomized controlled trial aims to evaluate the impact of providing LLM access to physicians on antibiotic prescribing appropriateness, using a composite outcome instrument, the Antibiotic Prescribing Appropriateness Score (APAS), that simultaneously evaluates antibiotic selection, dosing, duration, clinical reasoning, and management planning. Participants will be randomly assigned to one of two groups: the intervention group will have access to an LLM alongside conventional resources, while the control group will use conventional resources only (e.g. UpToDate, PubMed and Google Search with AI features disabled). Both groups will respond to a set of clinical vignettes covering common infectious disease scenarios requiring antibiotic prescribing decisions, with responses evaluated using an expert-validated grading rubric and independently scored by blinded raters.

Conditions

Healthy Adult

Keywords

antibiotic prescribing; antimicrobial resistance; clinical reasoning; large language model

Outcome Measures

Primary Outcomes (1)

• Antibiotic Prescribing Appropriateness Score (APAS)

  The primary outcome will be the composite score (expressed as a percentage) based on an expert-validated rubric that evaluates physician responses across four domains for each clinical vignette: Antibiotic Selection \& Spectrum (0-6 points), Dosing, Route \& Duration (0-3 points), Clinical Reasoning (0-2 points), and Management Plan (0-2 points). The total score for each vignette is the sum of scores across the four domains (maximum 13 points). APAS is expressed as a percentage, calculated as the total score divided by 13, multiplied by 100. The primary endpoint is vignette-level APAS. Responses will be independently evaluated by three licenses physicians blinded to participant identity and treatment assignment. The arithmetic mean of the three raters' total scores will constitute the vignette-level APAS used in the primary analysis.

  Assessed at a single time point for each case, during the scheduled diagnostic reasoning evaluation session, which takes place between 0-6 days after participant enrollment.

Secondary Outcomes (2)

• Antibiotic Selection, Spectrum & Dosing Subscore

  Assessed at a single time point for each case, during the scheduled diagnostic reasoning evaluation session, which takes place between 0-6 days after participant enrollment.

• Time per Vignette

  Assessed at a single time point for each case, during the scheduled diagnostic reasoning evaluation session, which takes place between 0-6 days after participant enrollment.

Study Arms (2)

Intervention Arm

ACTIVE COMPARATOR

Participants will have access to ChatGPT in addition to conventional diagnostic and reference resources (UpToDate, PubMed, Google search with AI features disabled). They will evaluate the same set of clinical vignettes covering common infectious disease scenarios requiring antibiotic prescribing decisions, with vignette order randomized independently for each participant.

Other: ChatGPT

Control Arm

NO INTERVENTION

Participants will use conventional resources only (UpToDate, PubMed, Google Search with AI features disabled). They will evaluate the same set of clinical vignettes presented in a randomized order.

Interventions

ChatGPT OTHER

Participants in the intervention arm will have access to ChatGPT in addition to conventional diagnostic and reference resources.

Intervention Arm

Eligibility Criteria

• Sex: all
• Healthy Volunteers: Yes
• Age Groups: Child (0-17), Adult (18-64), Older Adult (65+)

You may qualify if:

• Completed Bachelor of Medicine, Bachelor of Surgery (MBBS) or equivalent degree. The equivalent degree of MBBS in the US and Canada is Doctor of Medicine (MD).
• Full or Provisionally Registered Medical Practitioners with the Pakistan Medical and Dental Council (PMDC).

You may not qualify if:

• Any other Registered Medical Practitioner (Full or Provisional) with PMDC not holding an MBBS or equivalent degree (e.g., practitioners with a Bachelor of Dental Surgery (BDS)).

Sponsors & Collaborators

• Lahore University of Management Sciences: lead
• King Abdullah University of Science and Technology: collaborator

Study Sites (1)

Lahore University of Management Sciences

Lahore, Punjab Province, 54792, Pakistan

Location

Related Publications (1)

• Qazi, I.A., Ali, A., Khawaja, A.U. et al. Large language model diagnostic assistance for physicians in a lower-middle-income country: a randomized controlled trial. Nat. Health 1, 198-205 (2026). https://doi.org/10.1038/s44360-025-00007-8

  BACKGROUND

Central Study Contacts

Ihsan Ayyub Qazi

CONTACT

+92 42 35608368; ihsan.qazi@lums.edu.pk

Study Design

• Study Type: interventional
• Phase: not applicable
• Allocation: RANDOMIZED
• Masking: SINGLE
• Who Masked: OUTCOMES ASSESSOR
• Purpose: TREATMENT
• Intervention Model: PARALLEL
• Sponsor Type: OTHER
• Responsible Party: PRINCIPAL INVESTIGATOR
• PI Title: Professor of Computer Science

Model Details: This study will be an assessor-blind, randomized controlled trial.

Study Record Dates

• First Submitted: September 16, 2026
• First Posted: September 22, 2026
• Study Start: September 1, 2026
• Primary Completion (Estimated): December 1, 2026
• Study Completion (Estimated): December 1, 2026
• Last Updated: September 22, 2026

Record last verified: 2026-09

Locations

PA (1)