Presentation Details
| Using AI to increase compliance with perioperative glucose testing at scale: A prospective observational study Cameron Holguin1, Robert Martinez1, Ido Zamberg2, Tomer Cramer2, Brian Masel1. 1Department of Anesthesiology, University of Texas Medical Branch, Galveston, TX, USA.2C8 Health, Inc., New York, NY, USA |
Abstract
BACKGROUND: Compliance with evidence-based best practices improves patient and financial outcomes, yet sustained adherence is challenging. Although leadership often have metric dashboards, individual clinicians frequently lack (i) timely visibility into their own performance (ii) actionable, point-of-care guidance; (iii) leadership also lacks tools to drive adherence at scale. Perioperative glucose-testing is a commonly tracked metric linked to lower surgical-site infections. Objective: Evaluate the impact of implementing an AI platform (C8 Health) on compliance with perioperative glucose-testing.
METHODS: Prospective observational study over 3 months at an academic health system. The intervention was an AI-enabled performance-feedback capability within C8 Health comprising: (i) an in-app dashboard with individual compliance and department aggregates; (ii) monthly department summaries highlighting trends/drivers; and (iii) twice-weekly individualized notifications tailored to clinician’s performance (improvement/decline vs prior month/week; year-to-date vs departmental goal) with links to “how-to-comply”. Clinicians were classified C8 users if they used the platform beyond initial sign-in and non-C8 users if they had no use or never signed in. Primary outcome: between-group difference in glucose-testing adherence, defined as pre-operative glucose recorded 6h before to 30min after anesthesia start and if duration ≥2h, post-operative glucose 30min before to 60min after anesthesia end (denominator: all cases with diabetes mellitus). Secondary outcome: change from baseline in overall glucose-testing compliance. Two-proportion z-tests were used.
RESULTS: Of 182 clinicians, 151 (83%) were C8 users and 31 (17%) non-C8 users. After 3 months, glucose-testing compliance was higher among C8 users (80.74% vs 67.4%; p=0.012), an absolute difference of 13.34 percentage points. Overall department compliance increased by ~2%, with estimated savings of ~$109k.
CONCLUSIONS: Use of an AI-Powered platform (C8 Health) was associated with significant increase in compliance to glucose-testing. These findings support the use of technology to automate targeted quality-interventions at scale, based on individual-performance, as a mechanism to improve compliance to quality metrics.
No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the author.
METHODS: Prospective observational study over 3 months at an academic health system. The intervention was an AI-enabled performance-feedback capability within C8 Health comprising: (i) an in-app dashboard with individual compliance and department aggregates; (ii) monthly department summaries highlighting trends/drivers; and (iii) twice-weekly individualized notifications tailored to clinician’s performance (improvement/decline vs prior month/week; year-to-date vs departmental goal) with links to “how-to-comply”. Clinicians were classified C8 users if they used the platform beyond initial sign-in and non-C8 users if they had no use or never signed in. Primary outcome: between-group difference in glucose-testing adherence, defined as pre-operative glucose recorded 6h before to 30min after anesthesia start and if duration ≥2h, post-operative glucose 30min before to 60min after anesthesia end (denominator: all cases with diabetes mellitus). Secondary outcome: change from baseline in overall glucose-testing compliance. Two-proportion z-tests were used.
RESULTS: Of 182 clinicians, 151 (83%) were C8 users and 31 (17%) non-C8 users. After 3 months, glucose-testing compliance was higher among C8 users (80.74% vs 67.4%; p=0.012), an absolute difference of 13.34 percentage points. Overall department compliance increased by ~2%, with estimated savings of ~$109k.
CONCLUSIONS: Use of an AI-Powered platform (C8 Health) was associated with significant increase in compliance to glucose-testing. These findings support the use of technology to automate targeted quality-interventions at scale, based on individual-performance, as a mechanism to improve compliance to quality metrics.
No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the author.
Content Locked. Log into a registered attendee account to access this presentation.