Evaluating the Effectiveness of an AI-powered Physician Assistant in Improving Patients' and Physician's Satisfaction in an Outpatient Setting of a Tertiary Care Hospital.

Patients' satisfaction depends on several factors, including health care costs, access to care, and the waiting time to see a healthcare professional. In Pakistan, hospitals face overcrowding, which in turn results in long waiting times, particularly in outpatient departments. Longer waiting times not only hurt patients' experience and hospitals' performance but also increase stress on the physicians.

These challenges can be addressed with the effective use of Artificial Intelligence (AI) and related technologies. By leveraging machine learning algorithms and advanced data prediction models, AI can augment healthcare providers in clinical decision-making and streamline their work processes. However, these applications are largely studied and implemented in high-income countries, creating a lack of evidence from low- and middle-income countries.

Hence, a randomized controlled trial will be conducted to assess the effectiveness of an AI physician assistant in improving patient and physician satisfaction within outpateint clincis of a resource constrained setting.

Study Overview

• Status: Not yet recruiting
• Conditions: Patient Centered Care; Integration in Clinical Workflows
• Intervention / Treatment: Other: AI Physician Assistant

• Study Type: Interventional
• Enrollment (Estimated): 367
• Phase: Not Applicable

Contacts and Locations

• Study Contact: Name: Saqib Bakhshi; Phone Number: +923062750710; Email: saqib.dow@gmail.com
• Study Contact Backup: Name: Shifa Habib; Phone Number: +923018222783; Email: shifa.habib@aku.edu

Study Locations

• Pakistan
  • Karachi, Pakistan
    • Aga Khan University Hospital
    • Contact: Saqib Bakhshi; Phone Number: +923062750710; Email: saqib.dow@gmail.com
    • Contact: Hamdan Pasha; Phone Number: +923333129014; Email: hamdan.pasha@aku.edu
    • Principal Investigator: Saqib Bakhshi

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: Yes

Description

Inclusion Criteria (Patients):

• Informed consent before enrolment.
• Adults aged 18 years and above.
• Initial patients registering at the clinic during the entire trial duration.
• Possession of a digital device for an OTP (one-time password)
• Can read and write Urdu and/or English

Inclusion Criteria (Physicians):

• Informed Consent
• Agree to include AI physician assistant in their workflows

Exclusion Criteria (Patients):

• Patients requiring emergency care
• Patients who refuse to complete the history process with the AI physician assistant.

Exclusion Criteria (Physicians):

- Physicians from non-surgical specialties

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Other
• Allocation: Randomized
• Interventional Model: Parallel Assignment
• Masking: Single

Number of Arms

2

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
Experimental: AI Physician Assistant; The intervention group will comprise participants enrolled in the application (AI physician assistant) in addition to the standard of care The study participant allocated to the intervention will interact with the AI-physician assistant application "Hami" before they consult with the physician. The application will collect the medical history of the patient. This will then be followed by an AI-generated clinical summary, which their physicians will receive before the consultation begins. Physicians will review this summary and ask further questions of patients if required, and update the patient's record through an inbuilt scribe feature in the application.	Other: AI Physician Assistant; The intervention evaluated here is an AI Physician Assistant. The assistant takes the patient's history using a specialty-specific line of questioning. Once the interaction ends, the application converts the information into an AI-generated clinical summary for physicians to review. The physician reviews the summary and asks the patient additional questions, if required. Any additions or changes to the patient's history are recorded in the application. The physician then conducts a physical examination and can view AI-generated and guideline-based recommendations for assessment and treatment within the application. These recommendations may be selected, modified, or disregarded according to the physician's clinical expertise. All additions to the patient's record can be entered manually or dictated verbally and automatically added through the application's ambient scribe feature. Once the treatment plan has been documented, the application generates a SOAP note.
No Intervention: Standard of Care; The arm will comprise participants who receive standard care. In surgical clinics, standard care involves residents seeing the patients before the physicians. However, as part of the study, we will include physicians who agree to see patients without residents taking the history first. Hence, the trial uses the term 'physician' as part of the control group or standard care terminology.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Patient's satisfaction	Patient satisfaction is conceptualized through the lens of perceived quality of care, which is influenced by the effective utilization of waiting time and the provision of patient-centred care. Effective utilization of waiting time refers to patients' perceptions regarding whether their waiting time was used meaningfully during the visit. The domains of patient-centred care have been adapted from the Institute of Medicine (IOM) framework and include respect for patients' values and preferences, coordinated and integrated care, adequacy of information and communication, emotional support, involvement of family and friends, and physical comfort. These questions have been adapted based on the study objectives. The questionnaire will include demographic questions and five-point Likert-scale items (Strongly Agree to Strongly Disagree) and one open-ended question to obtain additional feedback regarding patients' experiences and satisfaction.	Every day from each patient for a period of 2 months

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Physician Satisfaction	It will be assessed with regards to integration of an AI-powered physician assistant, focusing on usability, impact on workflow efficiency, evidence based treatment recommendations and improved patient-physician interaction. Physician's satisfaction will be calculated utilizing mean scoring system, where each question will be scored on a 5 point Likert scale (Strongly Agree to Strongly Disagree). Additionally, we will ask one open-ended question at the end of the survey as part of physician satisfaction. This tool will be made exclusively for this study and will undergo content validation.	From each physician at the end of each day for two months.
Mean consultation time	Consultation time (calculated in minutes) refers to the time taken by the physician while the patient is in the physician's room and the time taken by the physician for each of the following: to inquire about symptoms, conduct an examination, prescribe treatment, and provide counselling. It will be measured using timestamps from a stopwatch from the time the patient enters the consultation room till the time they leave.	Every day for each patient consultation for a period of 2 months

Other Outcome Measures

Outcome Measure	Measure Description	Time Frame
Process flow evaluation outcome - Mean queuing time	Mean queuing time for each patient before the consultation process begins. Queuing time (calculated in minutes) refers to the time spent by a patient in the waiting area after their registration has been completed till the start of their consultation. It will be recorded using timestamps in two steps: one starting from the registration till the vitals are taken, secondly after vitals have been recorded till the patient visit the physician. These timings will be combined into a single aggregated time and will be calculated once for each patient.	Every day for each patient visit for a period of 2 months

Collaborators and Investigators

• Sponsor: Aga Khan University

Study record dates

Study Major Dates

• Study Start (Estimated): September 1, 2026
• Primary Completion (Estimated): November 1, 2026
• Study Completion (Estimated): November 1, 2026

Study Registration Dates

• First Submitted: August 7, 2026
• First Submitted That Met QC Criteria: August 7, 2026
• First Posted (Actual): August 10, 2026

Study Record Updates

• Last Update Posted (Actual): August 10, 2026
• Last Update Submitted That Met QC Criteria: August 7, 2026
• Last Verified: January 1, 2026

More Information

Terms related to this study

• Keywords: Quality of Care; Patient Satisfaction; Physician Satisfaction; Workflows
• Additional Relevant MeSH Terms: Behavior; Treatment Adherence and Compliance; Health Behavior; Patient Satisfaction
• Other Study ID Numbers: 2026-12176-39541

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No