OCR: IBM, Rambam Hospital and the Technion       

links

OCR: IBM, Rambam Hospital and the Technion - overview


Healthcare collaboration takes a page from industrial engineering playbook

An IBM Open Collaborative Research project with Rambam Hospital and the Technion in Israel is utilizing massive quantities of data to optimize work processes that often begin in the emergency room and then fan out to other hospital departments, ensuring fair and efficient policies and procedures for staff and correct treatment regimens for patients.

hospital conference If this is a story about creating the hospital of the future where physicians and other medical staff utilize massive quantities of data to ensure optimal patient care, why should we start by talking about engineering paradigms?

The answer has little to do with "processing" people like so many microchips and everything with understanding that a hospital is a system of processes aimed at achieving a desired outcome. To study the processes that define the maddeningly complex universe of emergency rooms, surgery theaters, premie wards and intensive care units at Rambam Hospital in Haifa (Israel), researchers at IBM are working with a range of industrial engineers at the Technion in Israel Institute of Technology to, well, re-engineer the hospital setting. The goal: Optimize hospital processes to improve efficiency, reduce costs, eliminate errors and improve patient care.

The IBM-sponsored collaboration is a perfect fit. In the industry research corner — IBM, with decades-long expertise in data analysis and an affinity for systems coded deep in its DNA. In the academic corner — Rambam and the Technion, the former a teaching hospital that ponders everything from diabetic nephropathy to human evolution; the latter an institution that studies a range of scientific issues, from the role of micro-robotic helicopters in intelligence operations to determining the genetic risk factors for kidney disease.

How the collaboration began

The three-way collaboration came about when Boaz Golani, dean of industrial engineering and management at the Technion, and Oded Cohn, director of the IBM Research Lab in Haifa, explored the possibility of pooling their intellectual capital to work on issues of importance to each institution.

It was a natural partnership. Avishai Mandelbaum, one of the Technion professors involved in the project, was doing work in operations research and statistics and hoped to apply what he had learned about call center queueing, a previous area of expertise, to the business of running a medical center. Rambam was looking to apply statistical and data analysis to improve its healthcare services while reducing cost. And IBM, whose cross-disciplinary approach to service science, sought to expand a relationship with the two Israeli organizations that began in 2007 with an Open Collaborative Research (OCR) project to improve healthcare outcomes and processes.

The expanded OCR project

The three partners put together a steering committee that began the data and statistical analysis project by targeting four departments at Rambam:

  • Emergency department. The ED is frequently the gateway to other hospital services.
  • Operating room. In terms of income and cost, the OR is the most capital intensive unit in the hospital. Moreover, the OR is where medical advances are put into practice.
  • Premie ward. The hospitalization period here is long and expensive: Each premature baby can cost hundreds of thousands of dollars and demands the attention of a relatively large number of staff.
  • Trauma unit. Hospitalization here is relatively short and requires individual case management: A team of physicians and nurses has about 45 minutes to save a life.

In the early stages of the project, the OCR participants looked primarily at the ED, largely because the ED had generated and stored a lot of data. The Technion's late Professor David Zinreich, along with then student and current post-doc Yariv Marmor, had built a database of healthcare data and simulated a new ED that Rambam wanted to build. The simulation proved to be a useful tactic in helping the hospital play out what if scenarios before committing time and money to building the ER.

Rambam as the Technion's engineering laboratory

sleeping baby All industrial engineering students at the Technion must complete a project before graduation. Building on their relationship with Rambam, the Technion's professors of industrial engineering asked the students to craft projects that identified critical work process issues at Rambam and to recommend solutions.

One set of projects, for example, looked at the flow of patients from the ED to internal wards at Rambam. If patients stayed too long, they would block the arrival of new patients to the ED. Blocking the ED also undermined patient care, with patients running the risk of exposure to nosocomial, or hospital–based, infections or not receiving the correct treatment.

Rambam historically dealt with ED "blockage" by relying on a "justice chart." Members of the ED staff would decide where the patient should go next. A problem arose when the justice chart experienced software glitches. Some students worked on an algorithm that would determine where and when the patient would be sent.

Although the algorithm was successful, another problem arose when the justice chart "pushed" patients to the more efficient departments. Now these departments were carrying a bigger load than the slower departments — and their employees resented it. A student team finessed the algorithm to ensure that the faster departments would have faster turnover and a manageable number of filled beds. Voila, hospital staff and patients were well served — and hospital administrators were happy.

As with most systems, one solution uncovers yet another problem. Rambam had a case, for example, with two similar departments, where people in one department felt they were working harder than people in the other. The OCR collaborators brought in a psychologist who worked with staff in both departments to discuss issues related to the perception of stress and overwork. The psychologist's attention to psychological and cognitive issues contributed to creating an overall picture of hospital processes — the goal of every healthcare service engineering project.

It is no coincidence that IBM has upped its hiring of industrial engineering graduates.

Measuring the results and next steps

The OCR project has given IBM, Rambam and the Technion deeper insight into the workings of a hospital — and deeper knowledge of the healthcare industry. The project was hampered somewhat only because the Rambam Emergency Department was transitioning to a new ED from two other EDs — the original ED and a temporary one. Once the transition is complete, the OCR will have a steady state environment from which to build a baseline.

The hospital is also working on implementing a radio frequency identification (RFID) system so that it can track occupied and unoccupied beds as well as the movement of patients from one department to another.

Despite the dearth of measurable results to date, the OCR project has generated many publications and a prize from the Service Research Innovation Institute (SRII).

IBM has demonstrated its commitment to the project by granting two IBM Faculty Awards to the Technion. It also has donated equipment related to the OCR's business intelligence and information retrieval projects.

As with all IBM Open Collaborative Research projects, the hospital data analysis and industrial engineering project with Rambam and the Technion is already generating many publications to promote the advances of this research with developers and researchers around the world.

Research for this story came from Chani Sacharen.

Related link

Collaborative Research Initiatives




Members of the IBM-Rambam-Technion collaboration

Pnina Vortman, Smart Solutions Leader at IBM Research — Haifa and initiator of the OCR and many of its related activities

Segev Wasserkrug, optimization researcher at IBM Research — Haifa

Boaz Carmeli, healthcare researchers at IBM Research — Haifa

Avishai Mandelbaum, Professor of Operations Research at The Technion

Avi Shub, Professor of Management and Learning at The Technion

Danny Glopher, Professor of Human Factor Engineering at The Technion

Eitan Naveh, Professor of Quality Research at the Technion

Anat Rafaeli, Professor of Psychology at the Technion