Association Between Left Ventricular Assist Device Infections and Number of Shared Patients Among Care Providers: A Network Analysis

Abstract

Background: For patients on durable left ventricular assist devices, prior work has documented that more shared patients among clinicians (connectedness) is associated with improved long-term outcomes. This study examines the association between preadmission clinician connectedness, patient acuity, and postimplant outcomes including infections.

Methods: The Society of Thoracic Surgeons Intermacs National Database was linked to Medicare claims to identify patients undergoing durable left ventricular assist device implantation between 2009 and 2017. Patients were stratified by terciles (low, moderate, high) of provider connectedness. The association between provider connectedness and major infections within 180 days of implant was evaluated by regression analysis, with an interaction term to assess effect modification by patient acuity via Intermacs Patient Profile.

Results: Of the 4985 patients across 127 hospitals, those within networks of low vs high provider connectedness were more commonly Intermacs Profile 1 (12.6% vs 6.7%, P < .001), and designated for destination therapy (3.4% vs 0.7%, P = .002). Patients cared for in centers with high connectedness had a lower risk of infection (β = -0.100, P < .026) and fewer total infections (β = -0.215, P = .014). The risk of infection was most pronounced in higher acuity Intermacs Profile 1 patients (P = .006).

Conclusions: Patients in networks with lower clinician connectedness (fewer shared patients) presented with higher acuity. After adjusting for preimplantation risk, lower clinician connectedness was associated with an increased risk of major infections, particularly among high-acuity patients. Measuring and improving network connectedness may be an important quality metric for timely access to advanced heart failure treatment and reducing complications.