When Good Isn’t Wells Enough: A Resident‐Driven Ehr Decision Support Tool for Appropriate Ct‐Pa Ordering

1University of Colorado —
Anschutz Medical Campus, Aurora, CO
2University of Colorado School of Medicine, Aurora, CO

Meeting: Hospital Medicine 2014, March 24-27, Las Vegas, Nev.

Abstract number: 250


As hospitals face increasing pressure to provide safe and high‐value care, quality improvement initiatives that address both evidence‐based care and cost mitigation will produce significant impact. Trainee engagement in such initiatives can address ACGME QI educational requirements as well as augment trainee understanding of value in healthcare. Moreover, as immersed experts in hospital workflow, resident physicians are uniquely positioned to contribute to high‐value care delivery solutions.


We established an interdisciplinary team of internal medicine residents and faculty representatives from hospital medicine, emergency medicine, and radiology to develop a novel EHR decision support tool for evidence‐based CT‐PA ordering.


Internal medicine residents training at our large academic medical center can elect to participate in a focused Hospitalist Training Track (HTT) where resident teams devise a longitudinal faculty‐mentored QI project. Beginning in July 2013, our HTT team formed a coalition with radiology and emergency medicine physicians at our hospital to examine adherence to evidence‐based CT‐PA ordering to diagnose PE. A standardized medical record review protocol based on established PE diagnostic algorithms had previously demonstrated high rates (39%) of avoidable CT‐PA ordering within our emergency department. These rates persisted even after providers gained access to an electronic PE diagnostic algorithm. We utilized the same chart review protocol on 100 instances of CT‐PA ordering on internal medicine ward services and found that 49% of CT‐PAs ordered to rule out PE were potentially avoidable. Beginning in October 2013, we launched an emergency department pilot of a “soft stop” within the EHR CT‐PA order set that asked providers to voluntarily calculate a Wells score and enter the D‐dimer for patients with a Wells score less than or equal to 4 prior to ordering a CT‐PA. 35 studies were ordered during the initial pilot. Only 15% of ordering providers voluntarily entered Wells data and 48% of ordered studies were avoidable. With this pilot data, we were able to initiate a “hard‐stop” pilot within the EHR that forced entry of Wells score (and D‐dimer for low‐risk cases) or to actively opt out of the algorithm. In the 14 days after initiation of the “hard stop,” 53 studies were ordered with 84% of providers entering data, and 85% of orders following evidence‐based guidelines. Given its initial success, a long term pilot is ongoing with plans to extend utilization of this decision‐support tool throughout the hospital system.


Resident‐driven process improvement initiatives can promote high‐value healthcare including the appropriate utilization of CT‐PA studies to diagnose PE. Such initiatives can align ACGME and program QI requirements for trainees with increased hospital attention to quality, safety, and value.

To cite this abstract:

Lamb K, Nitsch A, Handoyo K, Locke B, Suby‐Long T, Kneeland P, Nordenholz K. When Good Isn’t Wells Enough: A Resident‐Driven Ehr Decision Support Tool for Appropriate Ct‐Pa Ordering. Abstract published at Hospital Medicine 2014, March 24-27, Las Vegas, Nev. Abstract 250. Journal of Hospital Medicine. 2014; 9 (suppl 2). https://www.shmabstracts.com/abstract/when-good-isnt-wells-enough-a-residentdriven-ehr-decision-support-tool-for-appropriate-ctpa-ordering/. Accessed April 1, 2020.

« Back to Hospital Medicine 2014, March 24-27, Las Vegas, Nev.