Indeed, operations research (OR) and management science (MS) methods are not consistently used in practice for healthcare management decision-making. A report published by National Academy of Engineering and Institute of Medicine (Reid et al, 2005) states in an unusually blunt way, “In fact, relatively few health care professionals or administrators are equipped to think analytically about health care delivery as a system or to appreciate the relevance of engineering tools. Even fewer are equipped to work with engineers to apply these tools.”

Thus, it is often difficult for many administrators to appreciate the role of MS and OR methodology in the healthcare delivery process. A wide gap exists between the OR and MS publications that urge the use of this methodology in healthcare settings but provide few or no practical examples, and the publications with examples that are too specialized and complex for digesting by a typical hospital administrator. This gap is probably one of the reasons why too many administrators still have a vague idea of the practical value of healthcare OR and MS methodology. Many of them simply do not see ‘what’s in it for me’.

On the other hand, OR and MS professionals/engineers do not always have enough knowledge of healthcare or the role of physicians in making not only clinical but also management decisions. Healthcare has a culture of rigid division of labor. This functional division does not effectively support the methodology that crosses the functional areas, especially if it assumes significant change in traditional relationships.

Nonetheless, to address the challenge of transforming the system of care delivery in practice, some leading healthcare organizations have adopted this area as a strategic priority. For example, the Mayo Clinic, one of the largest integrated medical centers in the USA, has defined the Science of Healthcare Delivery as one of its four strategic directions. The others are Quality, Individualized Medicine, and Integration (Fowler et al, 2011). The Mayo Clinic has also created the Center for the Science of Healthcare Delivery, a new initiative that will focus on creating improved approaches to how healthcare is delivered (Mayo Clinic, 2011).

The bottom line: physicians and healthcare administrators are not supposed to have the knowledge of the OR/MS methods. They are too busy with other problems. Rather, they are supposed to understand why traditional management approaches and education guess are usually not accurate, short-lived or unsustainable; which quantitative technique is more appropriate for addressing a particular managerial problem; what can be expected from a particular technique and what are its strengths and limitations. For example, is queuing analytic theory (QAT) or discrete event simulation (DES) appropriate methodology for addressing a particular problem? What are the caveats in Linear Optimization for staffing and scheduling? What technique is the most appropriate for making a particular forecast type and why? What is the best approach to the fair cost (savings) allocation? And so on…Collaboration and trust between the healthcare/physicians leaders and OR/MS professionals is the key to progress in this area.

In complete agreement with Alexander Kolker

Physicians usually don't understand mathematical developments, and MCDM researchers are not aware of the complications of the medical practices that also involves strong participation of patients and their families

I think the research that is out there is too technical, only of academic interest to OR researchers. There should be more applied research, case studies, and papers written together with healthcare OR practitioners. Perhaps, a big problem is that there aren't too many healthcare OR practitioners. This is not the case in logistics and SCM, there are many practitioners with OR advanced degrees in those areas.

Cenk Çalışkan Please see, the book by Alexander Kolker that provides 'more applied research, case studies, and papers written together with healthcare OR practitioners':

"Healthcare Management Engineering: What Does This Fancy Term Really Mean? The Use of Operations Management Methodology for Quantitative Decision-Making in Healthcare Settings"

• DOI: 10.1007/978-1-4614-2068-2
• Publisher: Springer, 2012
• ISBN: 978-1-4614-2067-5

Abstract

The book provides multiple concrete examples of applications management engineering/operations research methodology. The book illustrates in depth a concept of healthcare management engineering and its domain for hospital and clinic operations. Predictive and analytic decision-making power of management engineering methodology is systematically compared to traditional management reasoning by applying both side by side to analyze 26 concrete operational management problems adapted from hospital and clinic practice. The problem types include: clinic, bed and operating rooms capacity; patient flow; staffing and scheduling; resource allocation and optimization; forecasting of patient volumes and seasonal variability; business intelligence and data mining; and game theory application for allocating cost savings between cooperating providers. Detailed examples of applications are provided for quantitative methods such as discrete event simulation, queuing analytic theory, linear and probabilistic optimization, forecasting of a time series, principal component decomposition of a data set and cluster analysis, and the Shapley value for fair gain sharing between cooperating participants. A summary of some fundamental management engineering principles is provided. The goal of the book is to help to bridge the gap in mutual understanding and communication between management engineering professionals and hospital and clinic administrators. The book is intended primarily for hospital/clinic leadership who are in charge of making managerial decisions. This book can also serve as a compendium of introductory problems/projects for graduate students in Healthcare Management and Administration, as well as for MBA programs with an emphasis in Healthcare.

Thank you for the reference, Alexander Kolker .

Frankly, those working in health in situ, as it were, are often under pressures that don't permit the time to either understand or apply the mechanics. CfWI and its Australian counterpart both tried to complex systems thinking and to a degree they were successful but both units were in the end terminated for wider reasons. In the world of policy making the time scales and the luxury of time required by these processes are hamstrung by the politicization of process (not a criticism per se) and therefor policy making and decision making works to far shorter time scales than these processes can require, the capability within ministries and health systems is often not there but even if it is, timeliness, speed and the need to satisfice often take precedence. Its a real world issue.

I agree with Cenk

Many researchers are more inclined to develop analytical methods, using among others, complex fuzzy theories or tyring to justify aspects such as using different procedures to improve criteria weights, when they ignore if the problem on which their theories will apply, is or not feasible, or that they are not able to see the forest from the trees. They concentrate on a certain detail and not considering the whole scenario.

In my humble opinion, researchers have to have their feet on the ground, that is, to see the more practical and true aspects of a scenario. In other words, researchers should try to incur on real problems, try to solve them, and see where the MCDM models fail to represent a real scenario.

Nobody can solve a health or industrial or social problem if he/she does not spend some time in a hospital, an industry, or in community service; they can end up developing a technique that does not have a practical use, or that does not adjust to reality.

Nolberto Munier You are right that "Nobody can solve a health or industrial or social problem if he/she does not spend some time in a hospital, an industry, or in community service". I've spent almost 10 years working directly in the large hospitals in Wisconsin, such as Froedtert, Children's, as well as consulted some other healthcare organizations. I summarized my experience for solving some practical health problems in the above mentioned book in which I presented "26 concrete operational management problems adapted from hospital and clinic practice." The problem types included in the book were: clinic, bed and operating rooms capacity; patient flow; staffing and scheduling; resource allocation and optimization; forecasting of patient volumes and seasonal variability; business intelligence and data mining; and game theory application for allocating cost savings between cooperating providers.

Thank you for sharing your experiences, Alexander Kolker.

Thank you Alexander for confirming my ideas

Dear Javed

I think that you post this issue very clearly, but regarding MCDM applied to health care, or to social issues, or to industrial matters, is always a task reserved to specialists. That is, a doctor poses a problem, a farmer asks about the best crops to cultivate, and a railway expert asks about determining performance. Decision Making is like any other activity, you need knowledge about it. It is not like a kitchen recipe where you follow a written text and get the final product.

The DM needs to possess enough expertise to determine the criteria to apply, of course, constantly assessed by people in a specific business, or due that the DM own experience in that business. He has to request detailed information, considering the alternatives, in order for him to determine the criteria, albeit the performance values will be given by doctors, farmers, or social workers.

For this reason, I am against pair-wise comparisons because it allows people without the faintest idea of the project, to decide by intuition about something that is reserved for specialists. Do you think for instance that a DM without any knowledge of structural engineering may decide that a wooden beam is 2 much better than a steel beam? This borders the absurd, and however, it is done in compting weights.

Then, according to the problem and its characteristics, the DM may select the method, examine the results from the MCDM point of view, explain the result to the stakeholders, and consider what they have to say, and make corrections if necessary.

The DM needs to face normally different problems, in different fields and industries and with different conditions. No problem is identical to another.

You can fill your tax return or give it to an accountant, who follows strict rules and procedures established by the government. It is like an algorithm that can be applied to different actions, and it also tells you what is related to what.

But that is inexistent in MCDM. Not only do even similar projects have differences, but also there are no norms or regulations, and for certain, no mathematical formula to reach a result, and no known result you can compare to.

Consequently, in my opinion, I don't think that a doctor must be proficient in MCDM models, this is not his/her function, as is not ours to diagnose ourselves. We should be happy if they are aware that methods exist to help him in a certain endeavour, for instance in selecting a treatment for a disease

It is difficult to impose MCDM methods in real life applications. First of all, a practical, simple and capable MCDM method should be produced. It should also be easy to calculate manually (or even mentally). And it should be the best MCDM method.

The alternative to this solution is to integrate complex and good methods such as TOPSIS, PROMETHEE or VIKOR into business analytics, artificial intelligence and machine learning applications. Thus, MCDM enters all areas of life as a decision support element.

But before that, there is something important to do. Discovering and proving the best or most appropriate MCDM method is essential. Forgive me, but what good is PROMETHEE if you can't prove that PROMETHEE is superior to simple weighted sum? If the simple weighted sum is sufficient then why are there more than 100 MCDMs?

Dear Mahmut

I don’t believe that it is difficult. I have modelled and solved about 300 /400 MCDM problems, and the only problem that I couldn't solve using MCDM is Supply Chain, albeit I believe that it is possible. In addition, you know that there are thousands of problems solved by different practitioners. Sorry, I don’t share your theory.

Regarding solving manually, yes for trivial problems; have you tried to solve by hand a problem with 15 alternatives and 33 criteria? Of course, it can be done, and I did, but on top of being very prone to errors, you get tired, and then, all attempt to be precise is normally lost. Did you say 'mentally' ?. I guess it is an error......

You propose an option to be added to the more than 100 methods, but I ask you. On what grounds do you say that TOPSIS, PROMETHEE, or VIKOR are the best? I share your opinion, but I don’t dare to say that they are the best

As far as my understanding you can’t incorporate AI in MCDM because you don’t have the necessary database for that task.

PROMETHEE is certainly better than SAW, because the first is based on rationality, and since it admits resources, as well as maximums and minimums

You answer your own question; there are more than 100 methods because SAW is very elemental and simple, to address complex scenarios.

There are some simple computational methods in the literature that are easy to use and you can do mental calculations in small scenarios and manual calculations in complex scenarios. It is not difficult . For example, the FUCA method proposed by Mendoza is an example of convenience and simplicity. It is even simpler than SAW.

Dear Mahmut

The FUCA method is used to find in a Pareto Front which is the best point, which normally happens to correspond to a change of the curve slope, for instance from concave to convex

However, I don't think that the drawing of the Pareto Front is easy, and in my opinion, it is not an MCDM method

FUCA is the French acronym for Faire Un Choix Adéquat (Make An Adequate Choice) . There are a limited number of scientists using the FUCA method. And they use terminological descriptors like "pareto front" or "pareto optimal" about it. Also do the same with TOPSIS, MOORA, VIKOR etc. they do it for other methods as well. Frankly, these are not a problem for me. Using the word "optimal" for MCDM methods doesn't make sense, and "pareto optimal" sounds logical. And the absence of that word "pareto" certainly wouldn't have affected my decisions or analysis because FUCA just does its job like a classic MCDM method.

Do you know to what Pareto refers? It is not a word.

Nolberto Munier Mahmut Baydas I am sorry to interrupt your exchange but it is a vivid illustration of the original question why operations research rarely used in practice. Can you imagine anyone from healthcare really involved or interested in this type of your discussion ? It is completely irrelevant to healthcare professionals both doctors and administrators. Could you both please stay in the original question track?

You are right Alexander; we are far from the original question, and I will follow your recommendation.

Perhaps with Mahmut you should open a new channel

Thank you for your contribution

Hello

This technique has not been well introduced among researchers and they are not aware of its benefits, so we must introduce this technique to other scientific groups.