Jennie Gallimore (Committee Member), Pratik Parikh (Advisor), Xinhui Zhang (Committee Member)
Master of Science in Engineering (MSEgr)
The occurrence of one of the most common chronic conditions in the U.S., diabetes, is expected to rise 53% from 24 million cases in 2003 to 37 million cases in 2023. The U.S. Veterans Health Administration (VHA) is not immune to this. The VHA has experienced an $820 million increase spending on diabetes patients between 2000 and 2008. The VHA has tried to keep the growth of chronic care costs in check through improvements in patient access to care by expanding its network of community-based outpatient clinics. Other methods the VHA has used to curb chronic care spending costs are electronic health records (EHR), patient aligned care teams (PACT), telehealth, and e-consults. An e-consult is defined as an electronic communication between primary care physicians and specialists about general or patient-specific questions that may preclude the need for an in-person referral. The objective of this study was to evaluate the effects of increased e-consult demand on time-based outcomes, quantify the sensitivity of these outcomes to walk-in patient arrival rates, electronic view-alert notifications, and primary care physician (PCP) unavailability, and provide recommendations to alleviate the detrimental effects of factors that are determined to have a significant effect on these outcomes. We collected data from 5 different VHA outpatient clinics, which was used in a discrete event simulation (DES) model of a typical VA outpatient clinic. Factors analyzed in the model were e-consult demand, view-alert notification arrivals, walk-in patient arrivals, and PCP unavailability. After the model was validated with real data, a detailed experimental study was conducted to determine factors that have a significant effect on e-consult time-based outcomes, such as cycle time. A total of 495 experiments were run and statistical analysis of the results indicated that all four factors had a significant effect on e-consult cycle time (p<0.05). Results also showed that, generally, as e-consult demand increases, e-consult cycle time also increases. In a case where a PCP is always available, e-consult cycle time increases by only 5 days when demand is raised from 0.01 e-consults per day to 2.75 e-consults per day. However, the increase in cycle times is not linear. In the same case, as demand increases from 2.75 e-consults per day to 3.25 e-consults per day, cycle time increases by 17 days. To reduce the detrimental effect of PCP unavailability due to sickness and/or vacation on e-consult cycle time, we recommend splitting the additional notification and walk-in patient demand incurred by a PCP's unavailability over the remaining available PCPs. In doing so, the cycle time does not increase drastically with an increase in e-consult demand, compared to the current strategy where the team leader assumes all the responsibility for the additional workload. Further research in the areas, such as walk-in patient arrival rate reduction methods, notification arrival rate reduction methods, and notification prioritization strategies, is likely to improve the time-based outcomes and meet the VA-set goals for e-consult completion. For example, if notification arrival rates are reduced by 20%, a 75% decrease in e-consult cycle times can be expected.
Department or Program
Department of Biomedical, Industrial & Human Factors Engineering
Year Degree Awarded
Copyright 2013, all rights reserved. This open access ETD is published by Wright State University and OhioLINK.