Random Walks on Multiplex: A Simple Approach for Integrative Functional Module Identification (Lifan) and Reducing patient wait time: modeling the impact of an intervention at an outpatient clinic in Malawi (Menna)
Lifan's Abstract: Integrating multiple sources of biological data has great potential to provide a better understanding of how genes function together. This study adopted the multiplex framework to integrate different high-throughput data for functional module identification. Our results showed that, with appropriate clustering algorithms, the multiplex formulation achieved a better accuracy for functional module identification. In addition, topic modeling on biomedical literature is a promising data source to complement high-throughput experimental data.
Menna's Abstract: Long patient wait times are a significant problem in clinics and hospitals throughout Africa. Located in Lilongwe, the capital city of Malawi, Daeyang Luke outpatient clinic sees an average of 150 patients per day. The current workflow and patient flow cause congestion at the outpatient clinic. Specifically, the reception area is one of the most congested areas because patients have to pre-pay for every service. This pay-as-you-go system requires patients to queue for payment up to four times during a clinic visit. Our research aims to model and measure the impact of an intervention that would alleviate the congestion problem at the reception desk. We hypothesize that a reconfiguration of the patient reception area will result in reduced patient wait time in the outpatient clinic. We developed discrete event simulation (DES) models to predict the impact of an intervention on patient waiting time. The intervention consisted of a transition from the current configuration at the reception (two cashiers and one data clerk) to having three universal servers (a single universal server can process registrations and payments). The simulation models were built using Rockwell’s Arena software. We aimed to determine whether an intervention will ultimately affect the total wait time of patients from their arrival to the clinic to their exit. After validation of our models, we made a comparison of patient wait times between the baseline model and the universal server model using statistical methods. First, we compared the reception desk models; then we measured time in the system for a patient visit. The models showed that the intervention would result in a decrease in patient wait time at the reception desk. The DES tool is useful in making high stake decisions in low resource hospitals.