Development and implementation of a clinical decision support-based initiative to drive intravenous fluid prescribing

Michelle C Spiegel; Annie N Simpson; Achsah Philip; Carolyn M Bell; Nandita R Nadig; Dee W Ford; Andrew J Goodwin

doi:10.1016/j.ijmedinf.2021.104619

Development and implementation of a clinical decision support-based initiative to drive intravenous fluid prescribing

Int J Med Inform. 2021 Dec:156:104619. doi: 10.1016/j.ijmedinf.2021.104619. Epub 2021 Oct 15.

Authors

Michelle C Spiegel¹, Annie N Simpson², Achsah Philip³, Carolyn M Bell⁴, Nandita R Nadig⁵, Dee W Ford⁵, Andrew J Goodwin⁵

Affiliations

¹ Division of Pulmonary, Critical Care, Allergy, and Sleep Medicine, Department of Medicine, Medical University of South Carolina, Charleston, SC, United States. Electronic address: Spiegel@MUSC.edu.
² Department of Health Care Leadership and Management, Medical University of South Carolina, Charleston, SC, United States.
³ Department of Information Solutions, Medical University of South Carolina, Charleston, SC, United States.
⁴ Department of Pharmacy, Medical University of South Carolina, Charleston, SC, United States.
⁵ Division of Pulmonary, Critical Care, Allergy, and Sleep Medicine, Department of Medicine, Medical University of South Carolina, Charleston, SC, United States.

PMID: 34673308
DOI: 10.1016/j.ijmedinf.2021.104619

Abstract

Objective: Studies suggest superior outcomes with use of intravenous (IV) balanced fluids compared to normal saline (NS). However, significant fluid prescribing variability persists, highlighting the knowledge-to-practice gap. We sought to identify contributors to prescribing variation and utilize a clinical decision support system (CDSS) to increase institutional balanced fluid prescribing.

Materials and methods: This single-center informatics-enabled quality improvement initiative for patients hospitalized or treated in the emergency department included stepwise interventions of 1) identification of design factors within the computerized provider order entry (CPOE) of our electronic health record (EHR) that contribute to preferential NS ordering, 2) clinician education, 3) fluid stocking modifications, 4) re-design and implementation of a CDSS-integrated IV fluid ordering panel, and 5) comparison of fluid prescribing before and after the intervention. EHR-derived prescribing data was analyzed via single interrupted time series.

Results: Pre-intervention (3/2019-9/2019), balanced fluids comprised 33% of isotonic fluid orders, with gradual uptake (1.4%/month) of balanced fluid prescribing. Clinician education (10/2019-2/2020) yielded a modest (4.4%/month, 95% CI 1.6-7.2, p = 0.01) proportional increase in balanced fluid prescribing, while CPOE redesign (3/2020) yielded an immediate (20.7%, 95% CI 17.7-23.6, p < 0.0001) and sustained increase (72% of fluid orders in 12/2020). The intervention proved most effective among those with lower baseline balanced fluids utilization, including emergency medicine (57% increase, 95% CI 0.7-1.8, p < 0.0001) and internal medicine/subspecialties (18% increase, 95% CI 14.4-21.3, p < 0.0001) clinicians and substantially reduced institutional prescribing variation.

Conclusion: Integration of CDSS into an EHR yielded a robust and sustained increase in balanced fluid prescribing. This impact far exceeded that of clinician education highlighting the importance of CDSS.

Keywords: Decision support systems, clinical; Implementation science; Medical informatics; Quality improvement.

MeSH terms

Decision Support Systems, Clinical*
Electronic Health Records
Humans
Medical Order Entry Systems*
Quality Improvement