AUTHORS
Anna L. Parks, MD, Christopher Boyer, PhD2, Simon Lynn, MS3, Oleg Lisheba, MS4, Aaron Hamilton, MD, MBA5, Matthew Pappas, MD, MPH6, Michael Rothberg, MD, MPH7
AFFILIATIONS
1 University of Utah Health Care, Salt Lake City, Utah, United States
2 Cleveland Clinic Foundation, Cleveland, Ohio, United States
3 Cleveland Clinic Foundation, Cleveland, Ohio, United States
4 Cleveland Clinic Foundation, Cleveland, Ohio, United States
5 Cleveland Clinic Foundation, Cleveland, Ohio, United States
6 Cleveland Clinic Foundation, Cleveland, Ohio, United States
7 Cleveland Clinic Foundation, Cleveland, Ohio, United States
ABSTRACT
Background
Electronic venous thromboembolism (VTE) risk assessment models (RAMs) can improve appropriate targeting of pharmacologic thromboprophylaxis among hospitalized medical patients. But clinicians often override interruptive alerts regarding prophylaxis. Whether making RAM use mandatory prior to VTE thromboprophylaxis order entry leads to sustained improvement in clinician VTE risk classification and pharmacologic thromboprophylaxis is unknown. We evaluated the long-term impact of mandatory RAM use following a previously completed pragmatic stepped-wedge trial.
Methods
We conducted a pragmatic stepped-wedge trial across 10 Cleveland Clinic hospitals followed by approximately 6 years of mandatory health system-wide use of an electronic VTE RAM. Adult medical inpatients admitted between January 2015 and December 2025 were included. During the trial (2017-2019), hospitals sequentially transitioned from usual care to voluntary RAM use. After the trial was completed, RAM completion became mandatory through an electronic health record hard stop as part of admission orders (2019-2025). Outcomes included RAM utilization, physician classification of VTE risk, and pharmacologic thromboprophylaxis use.
Results
Among 492,151 eligible admissions, RAM utilization increased from 0% before implementation to 31% during voluntary use and 85% immediately following mandatory implementation. RAM utilization continued to increase to >95% over the subsequent 6 years of routine clinical practice. During the same period, the proportion of patients classified as high risk declined from 90% before implementation to 67% during voluntary use and 38% following mandatory implementation. Pharmacologic thromboprophylaxis use decreased from 77% to 64% to 47%, a 39% relative reduction from baseline (RR = 0.614; 95% CI: 0.611 to 0.617). These changes in physician VTE risk classification and thromboprophylaxis use were maintained throughout 6-year follow-up.
Conclusions
Making the RAM mandatory increased RAM utilization and produced sustained changes in physician classification of VTE risk and pharmacologic thromboprophylaxis use. These findings support mandatory RAM use as a strategy to improve appropriate targeting of pharmacologic thromboprophylaxis for hospitalized medical patients and suggest that mandatory implementation may overcome the limited clinician uptake observed with voluntary RAM use.
DISCLOSURES
Anna L. Parks, MD: No relevant conflicts of interest to declare.