Changing Prescribing Behavior of Primary Care Providers (PCPs) to Reduce Risk of Vascular Events (VEs) During Testosterone Replacement Therapy (TRT) By Improving Access Through Electronic Consultation and Incorporating Education on Ordering and Interpreting Testosterone (T) Levels in a Lab Order Set

Introduction: Endocrinologists at this institution have adhered since 2008 to a policy governing who and when to prescribe TRT, akin to Endocrine Society guidelines. The policy, which does not apply to PCPs, excludes patients with a history of VEs (MI/CAD, CVA, VTE, PVD) <1y prior (absolute contraindication [CI] to TRT), or 1-3y prior (relative CI), and recommends strict diagnostic criteria, based on ≥2 early AM T levels by LC/MS/MS, with Total T <200ng/dl, or calculated bioavailable T <100ng/dl; or free T by Equilibrium Dialysis <5ng/dl. Data showed that 6 of 7 patients prescribed TRT by PCPs prior to 2014 (812/945 [85.9%]) did not meet criteria, and 3 of 10 had a prior VE (283/945[30.1%]). To change PCP prescribing behavior, two initiatives were implemented. One, in 7/2014, offered E-consultation to increase access to endocrinology input (EC ACCESS), and the other, in 5/2018, installed a Lab Order (LO) set with Education on how to order and interpret T levels (LO EDU). Objective: To determine the impact of the initiatives on TRT prescribing behavior and the risk of VEs. Methods: Retrospective cohort study of TRT prescribing behavior (adhering to diagnostic criteria and abiding by contraindications) before (2008-2014) and after implementation of EC ACCESS (2015-5/2018) and LO EDU (6/2018-6/2020) initiatives, and the impact on VE incidence. Results: TRT prescriptions decreased from 945 Pre-ACCESS (~135/y) to 121 after EC ACCESS (~31/y; p<0.001), and 61 (~31/y) after LO EDU. Endocrine input into TRT decisions increased from 164/945 (17.4%] Pre-ACCESS to 67/121 (55.4%) with EC ACCESS, and even further to 51/61 (83.6%; p<0.001) with LO EDU. The initiatives changed TRT prescribing behavior in 3 significant ways. First, PCPs were more likely to use ≥2 early AM T levels by LC/MS/MS when considering TRT (Pre-ACCESS: 196/945 [20.7%]; EC ACCESS: 62/121 [51.2%]; LO EDU: 47/61 [77%]; p<0.001). Second, strict diagnostic criteria were more likely to be met in those prescribed TRT (Pre-ACCESS:133/945 [14.1%]; EC ACCESS: 43/121 [35.5%]; LO EDU: 41/61 [67.2%]; p<0.001). Third, TRT was much less likely to be prescribed in those with prior VEs (Pre-ACCESS: 283/945 [30.1%]; EC ACCESS: 19/121 [15.7%]; LO EDU: 8/61 [13.1%]; p<0.001). The changes in TRT prescribing behavior effected by the EC ACCESS and LO EDU initiatives were associated with a significantly lower incidence of VEs on TRT (Pre-ACCESS: 142/945 [15%]; Post-ACCESS: 17/182 [9.3%]; p=0.043), despite a significantly longer mean (±SE) TRT duration (Pre-ACCESS: 22±0.7mo; Post- ACCESS: 26±1mo; p=0.0158) Conclusion: Changes in TRT prescribing behavior after EC ACCESS and amplified by LO EDU resulted in a 75% reduction in total TRT prescriptions, a nearly 5-fold increase in appropriate TRT (meeting strict criteria), and a 2.5-fold decrease in contraindicated TRT (with prior VEs). These changes were associated with a significant decrease in the incidence of VEs during TRT.