LLM-Generated Medical Summaries During Hospitalization to Improve Patient Activation in High-Risk Pregnancies

The Effect of Large Language Model-Generated Medical Summaries on Patient Activation During Antepartum Hospitalization: A Randomized Controlled Trial

The goal of this clinical trial is to learn if AI-generated medical summaries help hospitalized pregnant participants better understand and actively participate in their own care. The main questions it aims to answer are:

Do AI-generated summaries increase participant activation (knowledge, skills, and confidence in managing their health)? How accurate are the AI-generated summaries, as assessed by physicians? Do AI-generated summaries affect depressive symptoms during hospitalization?

Researchers will compare participants who receive periodic AI-generated summaries of their medical care with participants who receive standard care to determine whether the summaries improve patient activation during hospitalization.

Participants will:

Be randomly assigned to receive AI-generated medical summaries or standard care.

Complete the Patient Activation Measure (PAM-13) at baseline before the intervention and again after the intervention.

Complete the Edinburgh Postnatal Depression Scale (EPDS) at baseline and again after the intervention.

Receive up to 3 summaries during their hospitalization (intervention group only).

Study Overview

• Status: Not yet recruiting
• Conditions: Hospitalization; Patient Activation; Artificial Intelligence (AI); High-risk Pregnancy
• Intervention / Treatment: Other: LLM-Generated Medical Summaries

Detailed Description

Hospitalized pregnant participants in high-risk antepartum units may face complex medical conditions and treatment plans and may have difficulty understanding all of the medical information provided during hospitalization. This may reduce their ability to actively participate in their own care.

In this study, participants will be randomly assigned in a 1:1 ratio to one of two groups. Randomization will be stratified by gestational age at enrollment (<32+0 weeks vs. ≥32+0 weeks) using permuted blocks with randomly varying block sizes.

Participants in the intervention group will receive periodic medical summaries generated by a large language model (LLM) using a secure, privacy-protected AI environment at Sheba Medical Center. Each summary will be reviewed by a physician for medical accuracy before being provided to the participant. If a summary is considered inadequate, the physician may correct it before providing it to the participant or reject it, in which case it will not be provided.

The first summary will be generated on days 3-5 of hospitalization. Updated summaries will subsequently be generated every 3-5 days, with a maximum of 3 summaries per participant during hospitalization.

Participants in the control group will receive standard care without LLM-generated medical summaries.

The researchers will investigate:

Whether LLM-generated summaries improve patient activation, defined as the knowledge, skills, and confidence participants have in managing their own health, as measured by the Patient Activation Measure (PAM-13).

Whether LLM-generated summaries affect depressive symptoms during hospitalization, as measured by the Edinburgh Postnatal Depression Scale (EPDS).

The accuracy of LLM-generated medical summaries as assessed by physicians, including whether summaries require modification or rejection before being provided to participants.

Obstetric, neonatal, and postpartum clinical outcomes collected from the medical record.

• Study Type: Interventional
• Enrollment (Estimated): 200
• Phase: Not Applicable

Contacts and Locations

• Study Contact: Name: Avihu Krieger, MD; Phone Number: +972547421909; Email: avihuey@gmail.com
• Study Contact Backup: Name: Abraham Tsur, MD; Phone Number: +972525418828; Email: Avi.Tsur@sheba.health.gov.il

Study Locations

• Israel
  • Ramat Gan, Israel, 5262000
    • Sheba medical center
    • Contact: Avihu Krieger, MD; Phone Number: 0547421909; Email: avihuey@gmail.com
    • Contact: Email: avihuey@gmail.com
    • Principal Investigator: Avihu Krieger, MD
    • Sub-Investigator: Abraham Tsur, MD

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Child; Adult; Older Adult
• Accepts Healthy Volunteers: No

Description

Inclusion Criteria:

• Admitted to the antepartum high-risk department at Sheba Medical Center
• Able to provide written informed consent
• Able to read and understand Hebrew, English, or Arabic

Exclusion Criteria:

• Expected hospitalization of less than 3 days

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Health Services Research
• Allocation: Randomized
• Interventional Model: Parallel Assignment
• Masking: None (Open Label)

Number of Arms

2

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
Experimental: LLM-Generated Medical Summaries; Participants receive periodic patient-facing written medical summaries generated by a large language model (LLM) using a secure, privacy-protected AI environment at Sheba Medical Center. This is an informational intervention and does not involve a drug, medical device, or medical procedure. Each summary is reviewed by a physician for medical accuracy before delivery. Summaries are generated every 3-5 days, beginning on days 3-5 of hospitalization, with a maximum of 3 summaries per participant.	Other: LLM-Generated Medical Summaries; Participants receive patient-facing written medical summaries generated by a large language model (LLM) using a secure, privacy-protected AI environment at Sheba Medical Center. The intervention is informational and does not involve administration of a drug, use of a medical device, or performance of a medical procedure. Each summary is reviewed by a physician for medical accuracy before being provided to the participant. If a summary is deemed inadequate, it may be corrected by the reviewing physician before delivery or rejected and not provided. The first summary is generated on days 3-5 of hospitalization, with updated summaries every 3-5 days and a maximum of 3 summaries per participant.
No Intervention: Standard Care; Participants receive standard care, which does not include LLM-generated medical summaries.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Change in Patient Activation Measure Score	Change in Patient Activation Measure (PAM-13) score from baseline (pre-intervention) to post-intervention, compared between the intervention and control groups. The PAM-13 is a 13-item instrument assessing patient knowledge, skills, and confidence in managing their own health. Scores range from 0 to 100, with higher scores indicating greater patient activation.	From baseline (pre-intervention) to post-intervention during hospitalization, approximately 1-8 weeks.
Number of LLM-Generated Medical Summaries Requiring Physician Modification or Rejection	Number of LLM-generated medical summaries that require physician modification or are rejected following physician review and prior to delivery to the participant.	Throughout hospitalization, approximately 1-8 weeks

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Change in Edinburgh Postnatal Depression Scale Score	Change in EPDS score from baseline (admission) to end of hospitalization (prior to discharge), compared between the intervention and control groups. The EPDS is a 10-item self-report instrument with scores ranging from 0 to 30. A score of 10 or above indicates clinically significant depressive symptoms.	From admission to discharge, approximately 1-8 weeks
Gestational Age at Delivery	Gestational age at delivery, measured in weeks and days, collected from the medical record and compared between the intervention and control groups.	At delivery, up to approximately 6 months after enrollment.
Mode of Delivery	Mode of delivery, categorized as spontaneous vaginal delivery, operative vaginal delivery, or cesarean delivery, collected from the medical record and compared between the intervention and control groups.	At delivery, up to approximately 6 months after enrollment.
Indication for Cesarean Delivery	Indication for cesarean delivery among participants undergoing cesarean delivery, collected from the medical record.	At delivery, up to approximately 6 months after enrollment.
Maternal Fever or Chorioamnionitis	Proportion of participants diagnosed with maternal fever or chorioamnionitis during pregnancy or delivery, collected from the medical record and compared between the intervention and control groups.	From enrollment through delivery, up to approximately 6 months.
Length of Antepartum Hospitalization	Duration of antepartum hospitalization in days, collected from the medical record and compared between the intervention and control groups.	From admission to discharge from the antepartum hospitalization, approximately 1-8 weeks.
Antepartum Readmission or Emergency Department Visit	Proportion of participants with hospital readmission or an emergency department visit following discharge from the index antepartum hospitalization and prior to delivery.	From discharge from the index hospitalization through delivery, up to approximately 6 months after enrollment.
Birth Weight	Neonatal birth weight in grams, collected from the medical record and compared between the intervention and control groups.	At delivery, up to approximately 6 months after enrollment.
Apgar Score at 1 Minute	Apgar score at 1 minute after birth, ranging from 0 to 10, with higher scores indicating better neonatal condition.	At delivery.
Apgar Score at 5 Minutes	Apgar score at 5 minutes after birth, ranging from 0 to 10, with higher scores indicating better neonatal condition.	At delivery.
Umbilical Arterial Cord Blood pH	Umbilical arterial cord blood pH measured at delivery and collected from the medical record.	At delivery.
Umbilical Venous Cord Blood pH	Umbilical venous cord blood pH measured at delivery and collected from the medical record.	At delivery.
Umbilical Arterial and Venous Cord Blood Base Excess	Umbilical arterial and Venous cord blood base excess measured at delivery and collected from the medical record.	At delivery.
Neonatal Intensive Care Unit Admission	Proportion of newborns admitted to the neonatal intensive care unit (NICU), collected from the medical record and compared between the intervention and control groups.	From birth through neonatal hospital discharge, assessed up to 6 months after birth.
Postpartum Endometritis	Proportion of participants diagnosed with postpartum endometritis, collected from the medical record and compared between the intervention and control groups.	From delivery through 6 weeks postpartum.
Postpartum Hemorrhage	Proportion of participants experiencing postpartum hemorrhage, defined as estimated blood loss greater than 1000 mL, collected from the medical record and compared between the intervention and control groups.	From delivery through postpartum hospital discharge, assessed up to 6 weeks postpartum.
Length of Postpartum Hospitalization	Duration of postpartum hospitalization in days, collected from the medical record and compared between the intervention and control groups.	From delivery through postpartum hospital discharge, assessed up to 6 weeks postpartum.
Postpartum Readmission or Emergency Department Visit	Proportion of participants with hospital readmission or an emergency department visit during the postpartum period.	From postpartum hospital discharge through 6 weeks postpartum.

Collaborators and Investigators

• Sponsor: Sheba Medical Center

Study record dates

Study Major Dates

• Study Start (Estimated): October 1, 2026
• Primary Completion (Estimated): January 1, 2028
• Study Completion (Estimated): July 23, 2028

Study Registration Dates

• First Submitted: May 15, 2026
• First Submitted That Met QC Criteria: September 12, 2026
• First Posted (Actual): September 16, 2026

Study Record Updates

• Last Update Posted (Actual): September 16, 2026
• Last Update Submitted That Met QC Criteria: September 12, 2026
• Last Verified: September 1, 2026

More Information

Terms related to this study

• Keywords: Artificial Intelligence; Postnatal Depression; Patient Activation; PAM-13; High-Risk Pregnancy; Large Language Model; EPDS; Antepartum Hospitalization; Medical Summaries
• Additional Relevant MeSH Terms: Urogenital Diseases; Mental Disorders; Female Urogenital Diseases and Pregnancy Complications; Pregnancy Complications; Mood Disorders; Puerperal Disorders; Depressive Disorder; Behavior; Treatment Adherence and Compliance; Health Behavior; Patient Acceptance of Health Care; Depression, Postpartum; Patient Participation
• Other Study ID Numbers: SMC-25-2019

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO
• IPD Plan Description: This study is planned as a single-center trial, and at this stage, there are no plans or agreements in place for sharing individual participant data (IPD). Future considerations for data-sharing agreements may be evaluated based on future research collaborations and institutional policies.

Drug and device information, study documents

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