NMSU Engineers Work To Improve Hospital Care
Visits to the doctor are something people never look forward to, but thanks to some research conducted by a New Mexico State University professor and his student, those visits might just become a little less burdensome.
Yu-Li Huang, NMSU industrial engineering instructor, and graduate student Justin Marcak spent three months collecting data at Mimbres Memorial Hospital in Deming, N.M. During that time, the two studied the patient scheduling system and the factors that affect the length of patients' appointments. Results from that research have aided Huang and Marcak in developing what seems to be a more efficient scheduling system.
"Basically, the research is to develop a scheduling system that takes characteristics from a patient into account," explained Marcak. "The characteristics we looked into include a patient's gender, the time of day that a procedure was performed, mobility of a patient, what type of procedure and what part of the body the procedure was performed on."
These characteristics were first tested using analysis of variance (ANOVA) to determine if they were significant in determining the amount of time a patient will need to have a procedure performed.
According to a report by Huang and Marcak, the factors defined from the data collection for analysis also included where the patient was sent from (walk-in, emergency room, in- or out-patient unit) and appointment time. These factors were significant in determining the amount of time a patient required, and in redesigning the length of procedure time slots.
"The data was entered into a computer software program that allowed us to perform a decision tree method," Marcak said. "The software gave us a number of groups based on the procedures. We then grouped together the patients that had similar statistics into 15-minute slots."
Five groups were placed into 15-, 30-, and 45-minute time slots and tested (moved around) to determine the coefficient of variance.
"The setup that gives the lowest CV is the best solution," added Marcak. "Using the setup with the best CV and the amount of workload for each of the time slots, a patient schedule can be developed."
Huang and Marcak studied a variety of procedures including X-rays, CAT scans, ultrasounds, mammograms and MRIs. Because of low patient volume, nuclear medicine is one procedure they were unable to study, said Huang.
They concluded that using their new scheduling procedure, patient wait time could be reduced by up to 71 percent for X-rays - from 13 to four minutes.
"Patient access to care is increased by 125 percent for both CAT scans and ultrasounds," they reported. "The proposed scheduling scheme does allow for some flexibility in needing to meet different patient needs that may arise on any given day."
"For example, if more 15-minute slots are needed, a 30-minute slot can be split up," said Marcak. "For future research, it is proposed to come up with a grid system that would help schedulers based on available resources."
Huang explained that the county's population - approximately 25,000 residents - was one of the reasons the hospital was chosen for the experiment.
"Mimbres Memorial Hospital is the only facility serving the population of Luna County with radiological services," he said. "It is critical to improve the patient access to care for this facility."
The research was conducted May through August 2011, and the analysis was done during the fall. Results were provided to the hospital January 2012. The algorithm is generalized enough to be used at other facilities, but the data collection is specific to the facility and is required to be completed again.
Marcak said he hopes to continue working on the project, and hopefully conduct research at other hospitals in the future.