Impact of a structured referral algorithm on the ability to monitor adherence to appropriate use criteria for transthoracic echocardiography

Steven Promislow; Joseph G Abunassar; Behnam Banihashemi; Benjamin J Chow; Girish Dwivedi; Kasra Maftoon; Ian G Burwash

doi:10.1186/s12947-016-0075-2

Impact of a structured referral algorithm on the ability to monitor adherence to appropriate use criteria for transthoracic echocardiography

Cardiovasc Ultrasound. 2016 Aug 15;14(1):31. doi: 10.1186/s12947-016-0075-2.

Authors

Steven Promislow¹, Joseph G Abunassar¹, Behnam Banihashemi¹, Benjamin J Chow¹, Girish Dwivedi¹, Kasra Maftoon¹, Ian G Burwash²

Affiliations

¹ Division of Cardiology, Department of Medicine, University of Ottawa Heart Institute, University of Ottawa, 40 Ruskin Street, Ottawa, ON, K1Y 4W7, Canada.
² Division of Cardiology, Department of Medicine, University of Ottawa Heart Institute, University of Ottawa, 40 Ruskin Street, Ottawa, ON, K1Y 4W7, Canada. iburwash@ottawaheart.ca.

Abstract

Background: Many free-form-text referral requisitions for transthoracic echocardiography (TTE) provide insufficient information to adequately evaluate their adherence to Appropriate Use Criteria (AUC). We developed a structured referral requisition algorithm based on requisition deficiencies identified retrospectively in a derivation cohort of 1303 TTE referrals and evaluated the performance of the algorithm in a consecutive series of cardiology outpatient referrals.

Methods: The validation cohort comprised 286 consecutive TTE outpatient cardiology referrals over a 2-week period. The relevant AUC indication was identified from information extracted from the free-form-text requisition. The structured referral algorithm was applied prospectively to the same cohort using information from the free-form-text requisition, electronic medical record and ordering clinicians. Referrals were classified as appropriate, uncertain, non-adherent (inappropriate) or unclassifiable based on the American College of Cardiology Foundation 2011 AUC.

Results: Only 28.7 % of free-form-text requisitions provided adequate information to identify the relevant AUC indication, as compared to 94.4 % of referrals using the structured referral algorithm (p < 0.001). The structured algorithm improved identification in the AUC categories of general evaluation of cardiac structure/function (100 % vs. 43.0 %, p < 0.001); valvular function (100 % vs. 23.0 %, p < 0.001); hypertension, heart failure or cardiomyopathy (100 % vs. 20.3 %, p < 0.001); and adult congenital heart disease (100 % vs. 0 %, p < 0.001). By applying the algorithm, the number of identifiable non-adherent studies increased from 2.6 to 10.4 % (p <0.001).

Conclusions: Use of a structured TTE referral algorithm, as opposed to a free-form-text requisition, allowed the vast majority of referrals to be monitored for AUC adherence and facilitated the identification of potentially inappropriate referrals.

Keywords: Appropriate use criteria; Diagnostic requisitions; Quality improvement; Transthoracic echocardiography.

Publication types

Validation Study

MeSH terms

Aged
Algorithms*
Cardiovascular Diseases / diagnosis*
Echocardiography / standards*
Female
Guideline Adherence*
Humans
Male
Middle Aged
Monitoring, Physiologic / methods*
Patient Selection
Referral and Consultation*
Reproducibility of Results
Retrospective Studies