Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Inpatient general medicine attendings will be randomized to have an LLM feature turned on to provide a draft of an off-service handoff within Carelign (an EHR-adjacent provider communication tool). Providers who have access to this feature will be clearly instructed that if they use the LLM-generated draft, they must review and edit it as necessary before finalizing. The study will assess measures of documentation burden (as it relates to writing handoff) - including time spent writing handoff - and work exhaustion in both intervention and control groups.
The Structured Handoff Using Intelligent Framework for Transitions (SHIFT) Trial is a pragmatic, parallel-group randomized controlled trial designed to evaluate whether an artificial intelligence (AI) tool integrated into the Carelign handoff platform can reduce documentation time and burden among inpatient hospitalists. The AI feature uses a large language model (LLM) tuned for clinical summarization to generate an editable draft of the end-of-rotation handoff. Hospitalists randomized to the intervention arm will have access to a "Draft Handoff" button within Carelign, which automatically generates a structured, editable handoff draft using data from recent notes in the electronic health record. Clinicians are required to review and edit all draft content before finalizing and sharing the handoff with colleagues. Providers in the control arm will continue using standard handoff workflows within Carelign without AI assistance.
The study will be conducted on general medicine services at the Hospital of the University of Pennsylvania (HUP) and Penn Presbyterian Medical Center (PPMC). Eligible participants include attending hospitalists scheduled for at least five consecutive days on service. Providers will be randomized 1:1 to the intervention or control arm, stratified by site and team. The study will enroll approximately 90 hospitalists contributing about 120 eligible rotations over a 12-week period. The primary outcome is total time (in minutes) spent drafting the end-of-rotation handoff, measured using automated Carelign audit logs. Secondary outcomes include documentation burden (modified NASA Task Load Index), work exhaustion (Stanford Professional Fulfillment Index), and self-reported drafting time. Exploratory measures include tool usability (Net Promoter Score), receiving provider ratings of handoff quality, and electronic health record (EHR) usage metrics (total and after-hours "pajama time").
The trial is unblinded and will be conducted under real-world conditions to maximize generalizability. Analyses will follow an intention-to-treat approach, using linear mixed-effects models with random intercepts for provider to account for repeated rotations. All data will be stored and analyzed in secure, HIPAA-compliant institutional environments. The study has received approval from the University of Pennsylvania Institutional Review Board and Penn Medicine's AI Governance Council.
Not provided
Not provided
Not provided
Not provided
Not provided
| Label | Type | Description | Intervention Names |
|---|---|---|---|
| Standard of care | No Intervention | The control arm will continue usual care without access to the AI drafting feature. Attendings will prepare handoffs themselves using current workflows within Carelign. | |
| LLM Tool | Experimental | The intervention arm will have access to an LLM-assisted draft generation feature within Carelign. This will generate a draft of the handoff, which the provider will then edit as necessary before finalizing. |
|
| Name | Type | Description | Arm Group Labels | Other Names |
|---|---|---|---|---|
| LLM tool to draft handoff | Other | The intervention arm will have access to an LLM-assisted draft generation feature within Carelign. The feature will be accessed via a 'Draft Handoff' button in the attending handoff tab. The LLM is hosted in Penn's HIPAA-compliant environment and prompt engineering was performed through a series of handoff-specific iterative prompts with continuous quality assessments by the study team. In addition to the structured prompt, it will receive the most recent progress note from the primary team (or admission note, when no progress note is available), and the most recent specialty consult notes (within 72 hours of date of service). Generated drafts are editable; clinicians must review and finalize all content prior to sharing with their colleagues. |
| Measure | Description | Time Frame |
|---|---|---|
| Time spent drafting end-of-rotation handoff | Time (minutes) spent drafting the entire end-of-rotation handoff, calculated from Carelign audit logs as the duration between first and last edit per patient (capped at 30 minutes per patient), summed per rotation. | Over the last 3 days of a given rotation |
| Measure | Description | Time Frame |
|---|---|---|
| Documentation burden | Measured via the modified NASA-Task Load Index | Beginning (baseline) and end of study (3 months) |
| Work exhaustion | Measured via the Stanford professional fulfillment index |
| Measure | Description | Time Frame |
|---|---|---|
| Incoming provider ratings of handoff quality | Survey measures of clarity, accuracy, and usability of handoff (measured on a 5-point Likert scale, with higher being better) | Within 3 days of rotation start |
| Overall time spent on EHR |
Inclusion Criteria:
Exclusion Criteria:
- Jeopardy attendings and moonlighters
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
| Facility | Status | City | State | ZIP | Country | Contacts |
|---|---|---|---|---|---|---|
| Hospital of the University of Pennsylvania and Penn Presbyterian Medical Center | Philadelphia | Pennsylvania | 19104 | United States |
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
|
| Beginning (baseline) and end of the study (3 months) |
| Total time spent drafting handoff | Measured via first and last edits (overall) in Carelign on last day of rotation | Last day of rotation (varies by schedule, typically day 5, 7, or 14 on service) |
Total and after-hours time in the EHR on the last day of service
| Last day of rotation (varies by schedule, typically day 5, 7, or 14 on service) |