Taking a Different View Using Spatial Data and Simulations

Naoru Koizumi loves solving a mystery. She enjoys discovering hidden facts and uses computer modeling and simulations, as well as geographic information systems (GIS), to tease out answers to important questions.

“[These methods] often reveal information that wouldn’t necessarily be found through statistics alone,” says Koizumi, an assistant professor in the School of Public Policy. “Sometimes if you visualize [the data], the answer is so obvious.”

After completing an undergraduate degree in business administration from Aoyama Gakuin University in Japan, Koizumi traveled to the United States to study regional science at the University of Pennsylvania. But when she finished her PhD, she realized that she was interested in health care topics and returned to Japan to pursue another PhD, this time in environmental and preventive medicine at Hyogo College of Medicine.

By following this particular path, Koizumi has carved herself a niche and many of her research projects have involved applying GIS or modeling to health care problems or issues. Over the course of her career, she has used spatial techniques and simulations to examine such health issues as kidney transplants in the United Kingdom, cadmium concentrations in rice and renal failure mortality in Japan, and drug-resistant pneumococcal rates in U.S. inner cities.

One area of ongoing analysis for Koizumi is mental health system congestion. Her interest in this area began with her dissertation. In that project, Koizumi examined the causes of patient flow congestion in a psychiatric care system, using Philadelphia as a case study. The results of her study were surprising and contrary to popular perceptions.

By creating a model of the system, Koizumi found that systemwide congestion was due to shortages in supported housing and not due to a shortage of beds throughout the system. By increasing the number of supported-housing placements, the final stop in the system where patients begin to get on with their lives while still receiving some services, the City of Philadelphia could alleviate congestion throughout the psychiatric care system.

“It was the bottom of the system that was working as a bottleneck,” Koizumi says, “and it was not something health care administrators would have been able to see. They can only look at the part of the system they are involved in. Perhaps the most important policy implication of this analysis is that removal of such facility-specific bottlenecks may often be the most cost-efficient way to reduce congestion in the system as a whole.”

All of Koizumi’s analyses have had policy implications. With her kidney transplant research, she discovered that geography makes a difference. Despite the fact that kidney transplants account for almost 60 percent of all transplants worldwide, Koizumi found that, particularly in the United States, the patient’s proximity to a treatment facility was one of the greatest factors affecting whether he or she was even offered the opportunity for a transplant.

“By using geography and statistical analysis, we can tease out reasons why these patients aren’t treated as well as they could be,” she says. “Everyone should have an equal opportunity of getting treatment, and that brings up policy. What kind of policy could we formulate to ensure these patients get to hear what their treatments options are, regardless of where they live? I think that is an interesting question.”

Working with data on patients in London has proved more difficult. Koizumi is still working on this analysis. “Europe has a completely different [transplant] system and a more complicated allocation protocol.”

With funding from the National Institute of Mental Health, she also is working on an extension of her doctoral project on mental health systems that captures the differences in patient diagnoses and a cost-optimization model. “The goal of this project is to create a decision support tool for mental health policy makers,” she says.

In addition, Koizumi has begun working with Faye Taxman, a professor in Mason’s Department of Administration of Justice and an expert in correctional systems and recidivism. “A lot of mental health issues are being treated in prison and correctional systems,” she says. “In a way, prisons are becoming mental hospitals, and that’s also creating congestion in the correctional system.”

“Recidivism is a big problem for inmates with mental health issues. [Research has shown that] if you have a mental health episode, you are much more likely to return to prison,” Koizumi continues. “How can the mental health system be revised so we can treat those people outside of prisons? What kind of treatment is most effective in the correctional system, and what’s most effective outside? Those are some of the questions we hope to find answers to by using dynamic modeling.”

This story by Colleen Kearney Rich originally appeared in Mason Research 2009 magazine.