Use of Integrated Health Record Viewer Reduces Duplicate Imaging

Wednesday, June 26, 2019: 8:00 AM
Madison A (Marriott Wardman Park Hotel)

Presenter: Yingzhe Yuan

Co-Authors: Megan E. Price; David F. Schmidt, MD; Merry Ward, PhD; Jonathan R. Nebeker, MD; Steven Pizer

Discussant: Jeffrey S. McCullough

We estimate the impact of using an integrated health record viewer on duplicate imaging for service members transitioning from the Department of Defense (DoD) Military Healthcare System (MHS) to the Veterans Health Administration (VHA). Joint Legacy Viewer (JLV) is an integrated health information viewer that allows healthcare providers in VHA and DoD to read patients’ medical records from multiple sources. JLV assists VHA healthcare providers to gain a better understanding of their patients’ medical histories and helps providers to make better medical decisions, such as ordering fewer duplicate images. Health information exchange between VHA and DoD through JLV will support improvements in efficiency and quality of VHA care, induce cost savings and generate health benefits to Veterans, especially recently transitioned service members.

This study uses encounter data from the VA’s Corporate Data Warehouse (CDW), DoD claims data provided by the VA Informatics and Computing Infrastructure (VINCI) and JLV audit data obtained from the VA’s JLV database. Recently separated service members who had a primary care visit at VHA in Fiscal Year (FY) 2018 within 90 days of an imaging study conducted at DoD are included in the sample. CPT codes indicating imaging studies are categorized by mode and body part to compare VHA and DOD imaging records and determine potential duplicate images. The subset of patients who received at least one imaging study at VHA within 90 days of a DoD imaging study from the same imaging mode and on same body part are considered to have received potentially duplicative imaging studies.

The independent variable “JLV encounter” refers to whether the provider had a JLV audit linked to the patient on the visit date. The proxy “JLV provider” indicates whether the provider was a JLV user, determined by if the provider had 10 or more JLV audits in the prior 6 months. Potential confounders include the provider’s image ordering rate in prior 6 months, patient age, gender, and health status indicated by Elixhauser comorbidity scores. OLS and logistic regression models are adopted to estimate the effect of JLV encounter or JLV provider on duplicate imaging rates. The proxy of JLV provider is also applied as an instrumental variable (IV) to deal with potential endogeneity of JLV use.

We find that provider use of JLV is significantly associated with lower rates of ordering duplicate imaging studies. The duplicate imaging rate is 4.41% for JLV encounters compared to 6.96% for non-JLV encounters. In the logistic model with time effects, JLV provider has an odds ratio of 0.56 (p = 0.025) for ordering duplicate images. This effect is consistent across OLS and logistic models using IV or proxy. The results also show that the provider’s practice pattern of ordering imaging studies and the patient’s health status are powerful confounders. This evidence that JLV use reduces duplicate imaging rates supports efforts to promote health information exchange between DoD and VHA to improve healthcare quality and reduce costs.