Operations Research in Healthcare
Hospitals and other healthcare providers have relied on methods developed using Operations Research (O.R.) for many of their processes since the 1940’s. O.R. has been used for allocating blood supplies, structuring queues of patients waiting for exams or treatment, providing systematic ways for diagnosing, treating and preventing disease, as well as developing many more processes for improving population health and quality of care and life.
The Mayo Clinic, a nonprofit medical practice and research center, was an early adopter of systems engineering principles, which were used in building and organizing the clinic. In 1947, under the leadership of Ernest Schlitigus the Mayo Clinic formed Procedures and Records, a team of industrial engineers who would “make systems engineering services available to support projects related to scheduling, workflows analysis and process improvement.” As these services evolved and encountered more complex problems, their use of O.R. evolved with them, becoming larger and more important.
According to Charles Flagle, who is sometimes known as the “father” of O.R. in healthcare, after these few smaller-scale applications were developed a few major events led to the expansion and mainstream application of this discipline. In Britain there was the post-World War II push to create a system of universal healthcare coverage and in United States to create a system to fund medical care facilities, which led to more interest in creating guidelines for design, construction and management. In the United States, the Hill-Burton legislation of 1945 created a research group within the Public Health Service, and provided an opportunity for O.R. professionals to receive funding for research in healthcare. In England, Norman Bailey published two articles on “Operational Research in Medicine,” and on appointment systems in outpatient departments explaining how O.R. principles could be applied to medicine. Bailey’s seminar at the School of Engineering at Johns Hopkins University helped make the field of O.R. more accessible and applicable. Lillian Gilbreth, Ruth Kuehn, and Harold Smalley conducted a two week workshop in Pittsburgh on hospital operations improvement in 1952. These developments came together to convince people to begin looking at the issues facing healthcare by using the principles and methods of O.R.
Spurred by a push at the Johns Hopkins Hospital for interdisciplinary study, Russell A. Nelson, MD, then director of the hospital, suggested the creation of a Division of Operations Research to provide “continuity from research on problem areas through recommendation, implementation, and evaluation of results.” Charles Flagle was asked to head this department, and developed a unit that worked closely with hospital staff, observing and beginning to solve problems that the staff dealt with daily. His work there caught the attention of the Hill-Burton program and eventually the Assistant Surgeon General. Flagle along with John Moss worked to implement similar practices to those being used at Johns Hopkins, and to develop a Public Health Service Guideline. The department continued to grow, applying O.R. to more areas of healthcare. With more government involvement in healthcare in the following years, including the creation of Medicare and Medicaid in the United States, the importance of O.R. in healthcare continued to increase.
An early example of the influence of O.R. in health care is the use of randomized controlled clinical trials. The approach of randomly assigning trial subjects to either treatment or control groups is often considered the “gold standard” for clinical trials. It was originally introduced in psychology and education, and then made its way to agriculture due to statisticians Jerzy Neyman (George Dantzig’s thesis advisor) and Ronald Fisher, and came to medicine in the 1940’s through the work of medical statistician Austin Bradford Hill.
Perhaps more directly attributable to O.R. is the work of Yale professor of industrial management Robert B. Fetter (later Chair of the Yale School of Organization and Management) in collaboration with John D. Thompson of the Yale School of Public Health. In the 1970’s they created a system to classify hospital cases into Diagnosis-related Groups (DRGs) as a way of categorizing each hospital ‘product’, for example providing care for an appendectomy, heart failure, pneumonia, etc. The DRG system is now the basis for Medicare reimbursements to hospitals in the United States, and has been called “the single most influential postwar innovation in medical financing.”
William P. Pierskalla is another pioneer of healthcare Operations Research. He built models for scheduling nurses and optimizing patient care in hospitals and developed a better system for scheduling, as well as including a behavioral analysis component – such as how many days off nurses typically wanted and how often to schedule them for weekends. This work was implemented in over a dozen hospitals in the first year of use.
Despite some initial resistance, more and more healthcare providers started to change and adopt these procedures. These initial applications spawned new professional organizations and journals. Harold Smalley, another pioneer of O.R. in healthcare, founded the society HIMSS. Other organizations and journals include Health Applications Society, and Healthcare Management Science, and Operations Research for Healthcare (ORHC). O.R. approaches expanded from logistical and managerial capacity problems to public health and medical decision-making problems regarding diagnosis and therapy, medical education, and disease prevention, detection, and treatment.
Early work in identifying opportunities for O.R. to play a role in healthcare have been transformed over time into approaches with significant societal impact. For example, a 1969 article in the journal Radiology, titled “The Method of Linear Programming Applied to Radiation Treatment Planning,” by Bahr et al. has led to wide spread adoption of mathematical programming into radiation treatment planning systems. Organ transplantation policy provides another important example of the use of O.R. to guide healthcare decision making. In the 1990s the United Network of Organ Sharing (UNOS) in the United States began the process of developing simulation models to guide public policy decisions about how to allocate the limited supply of organs from donors to recipients (Harper et al. 2000).
O.R. methods that are now routinely used in healthcare include statistics and machine learning, stochastic models, computer simulation, mathematical optimization, and decision analysis. O.R. professionals also continue to develop new methods to apply to healthcare, to solve new or previously unsolved healthcare problems, and to improve different healthcare systems. The discipline that started with small and simple applications has become an integral part of our healthcare systems which in turn is driving new methods and applications of O.R. to improve human health and safety.
Edited by Brian Denton, Linus Schrage, Pinar Keskinocak.
Links and References
Flagle C. D. (2002) Some Origins of Operations Research in the Health Service. Operations Research, 50(1): 52-60. (link)
Mohan Lal T. & Kuchera D. (2013) Roundtable profile: Mayo Clinic. OR/MS Today, 40(2). (link)
Bahr, G. K., Kereiakes, J. G., Horwitz, H., Finney, R., Galvin, J., & Goode, K. (1968). The method of linear programming applied to radiation treatment planning. Radiology, 91(4), 686-693.
Flagle, Charles D. (1997) Interview, October 21, 1997, (video) National Library of Medicine
Flagle, Charles D. (2002) "Some Origins of Operations Research in the Health Services." Operations Research Vol 50 No 1, Jan-Feb 2002, pp. 52-60 (link)
Fries, B (1976) Bibliography of Operations Research in Health-Care Systems, Operations Research Vol 24 No2, Sept-Oct 1976, pp. 801-814
Harper, A.M., Taranto, S.E., Edwards, E.B., Daily, O.P., An Update on a Successful Simulation Project: The UNOS Liver Allocation Model, Proceedings of the Winter Simulation Conference, 2000.
Lal, Tarun Mohan and D. Kuchera (2013). "Roundtable Profile - The Mayo Clinic." ORMS Today, April 2013, Vol. 40, No. 2, pp. 18-20. (link)
Means, Chip. "Slideshow: HIMSS 50 in 50 recipients - 1960s" January 25, 2011. (link)
Ozcan, Yasar A., . "Health Care Management" in Encyclopedia of Operations Research and Management Science. Gass, Saul and Fu, Michael, editors, New York: Springer, 2013. pp 679-84. (link)
Pierskalla, William P. (2015) Interview by Matt Sobel, November 2, 2015 (link)
Pierskalla, William P. (2010) “Operations Research: A Valuable Resource for Improving Quality, Costs, Access and Satisfaction in Health Care Delivery” in Engineering the system of healthcare delivery, Rouse WB and Cortese DA eds. IOS Press, Amsterdam ; Washington, DC, 2010.
Reid PP, Compton WD, Grossman JH, Fanjiang G, eds. (2005) ”Building A Better Delivery System: A New Engineering/Health Care Partnership, The National Academies Press, Washington, DC
Zhang J, Mason J, Denton BT and Pierskalla WP, “Disease Prevention, Detection, and Treatment” in Encyclopedia of Operations Research and Management Science. Gass, Saul and Fu, Michael, editors, New York: Springer, 2013. pp 437-447.
Associated Historic Individuals
Arrow, Kenneth J.Balintfy, Joseph L.
Bass, Frank M.
Bellman, Richard E.
Caywood, Thomas E.
Edie, Leslie C.
Esogbue, Augustine O.
Flagle, Charles D.
Green, Paul E.
Hausman, Warren H.
Horvath, William J.
Johnson, Ellis A.
Johnson, Ellis L.
Kolesar, Peter J.
Larson, Richard C.
Liebman, Judith
Neyman, Jerzy
Pierskalla, William P.
Puterman, Martin
ReVelle, Charles S.
Schwarz, Leroy B.
Settles, F. Stan
Sobel, Matthew J.
Staelin, Richard
Starr, Martin K.
Tien, James M.
Urban, Glen L.
Winkler, Robert L.
Wolman, Eric