Abstract 287: The Utilization pattern of Stress Myocardial Perfusion Imaging in a rural health care system

Abstract

Background: Appropriate use criteria (AUC) for Single-Photon Emission Computed Tomographic Myocardial Perfusion Imaging (MPI) were revised in 2009 to include 15 new clinical scenarios. Prior studies showed that ∼15 % of MPI studies were requested for inappropriate indications, mostly by non-cardiology providers. Although awareness of the AUC has improved, few studies have specifically evaluated the impact of 2009 AUC in an electronic, integrated, rural health care system. Methods: All MPI studies done between April - Sep 2011 were reviewed in this single centre study at a rural, electronic, integrated health care system. Using 67 scenarios in AUC guidelines, these studies were classified into four categories: Appropriate (A), Inappropriate (I), Uncertain (U) and Unclassifiable. To estimate the independent impact of ordering provider specialty on level of appropriateness, multivariable analysis was performed using backward stepwise variable selection. Results: During 6 month study period, 328 patients underwent MPI. Overall, 287 (87.5%) studies were classified as (A), 18 (5.5%) as (I), 23 (7%) as (U) and none were deemed as unclassifiable out of these 328 studies. Preoperative testing accounted for 8 (44%) of the total 18 (I) studies. Of the 23 studies classified as (U), 16 (70%) were performed for patients with new or worsening symptoms and prior normal coronary angiography or prior normal stress imaging study. The ordering physician specialty (cardiologists vs. non-cardiologists) did not show a multivariable correlation with appropriateness of the test (p=0.46). Results are summarized in Fig. 1 Conclusion: In a rural, integrated, electronic, health care system; majority of providers, regardless of their specialty utilized MPI studies for (A) indications. Only 5.5% of MPI studies were ordered for (I) indications, suggesting a significant decrease in (I) tests compared to prior reports, which may reflect an increase in awareness of the revised 2009 criteria. However, certain common scenarios still account for a majority of small proportion of (I) studies. These findings may suggest a continuing need for provider education and possibly focusing the preauthorization triage process only for (I) clinical scenarios.