Redesigning Surgical Care for Patients in Wisconsin

Summary

This study evaluates strategies to train surgeons to use Better Conversations, an evidence-based communication framework designed to improve informed consent by helping patients understand the goals of surgery, the downsides of treatment, and what to expect. Better Conversations supports deliberation, patient preparation, and alignment of decisions with patient goals, addressing known shortcomings in traditional informed consent.

The purpose of this study is to compare two methods of surgeon training: (1) training delivered by an education specialist using audit and feedback, and (2) training supported by computerized automation that identifies elements of Better Conversations in de-identified transcripts of surgical consultations. The central question is whether the automated training program is non-inferior to the specialist-delivered program.

Approximately 60 surgeons from two academic health systems will be randomized to one of these training approaches. Each surgeon will complete a didactic session, have outpatient surgical consultations audio-recorded for feedback, and complete assessment recordings after training. Patients of enrolled surgeons will complete surveys before and after their surgeon's training to evaluate patient-reported communication outcomes.

Findings from this study will assess the effectiveness, feasibility, and acceptability of automated training and support the development of a larger pragmatic study to evaluate the broader effects of Better Conversations on patient outcomes.

Conditions

• Communication
• Surgery
• Shared Decision Making

Interventions

BEHAVIORAL

Education Specialist Delivered Training

Training in the Better Conversations framework delivered by an education specialist, including a brief didactic session and audit-and-feedback based on 10 de-identified outpatient consultations, followed by assessment using five additional recordings scored with an adherence checklist.

BEHAVIORAL

Automated Training

Training in the Better Conversations framework supported by computerized automation. Procedures match the education-specialist approach (didactic session; 10 training recordings; 5-7 assessment recordings). For each training consultation, the de-identified transcript is processed using previously developed natural language processing with active/supervised machine learning to identify elements of Better Conversations that are present or absent and common errors. An education specialist reviews and edits the automated output and emails feedback and a score sheet within one week using the same standardized format. In one half of the automated-training arm, surgeons also receive intermittent disclosure messages indicating that some feedback is computer-generated.

Sponsors & Collaborators

• Wisconsin Partnership Program

  collaborator OTHER

• Medical College of Wisconsin

  collaborator OTHER

• University of Wisconsin, Madison

  lead OTHER

Principal Investigators

• Margaret (Gretchen) L Schwarze, MD, MPP · UW School of Medicine and Public Health

Study Design

Allocation

RANDOMIZED

Purpose

HEALTH_SERVICES_RESEARCH

Masking

TRIPLE

Model

PARALLEL

Eligibility

Min Age

18 Years

Sex

ALL

Healthy Volunteers

No

Timeline & Regulatory

Start

2026-02-27

Primary Completion

2028-03-31

Completion

2028-08-31

Countries

• United States

Study Locations