Health Operations Research

Index - Major Sections
Home

_______________

Index - Same Level Subject

Data and Data Analyiss
Information Management
Computerized Patient Record
Health Operations Research
Data Mining
Data Dissemination
Management

Index - Child Subjects

Introduction

This section will in general paraphrase the great work performed by the PRICOR (Primary Health Care Operations Research) project.

PRICOR's objective was to promote operations research as a tool to help program managers and policymakers find solutions to problems they encounter in designing and operating primary health care (PHC) programs.

"Operations research provides a systematic approach to problem solving. In operations research, rather than relying on the costly process of trial-and-error, a well-defined plan of analysis is used to select the best of several possible alternatives. A specific operational problem is first defined and analyzed. Alternative solutions are developed and evaluated to identify those that are most appropriate and feasible. Recommendations are then made for testing, or in some cases directly implementing, the best solutions(s)" (PRICOR Methods Paper 1 1985)

Although, great attempts have been made in using operations research, in most cases it fail to meet the expectations. This is because of the mathematical sophistication required to implement its processes. Many school programs implemented what they called "operations research" programs but in a very diminutive form. These programs did little but teach students to "list out the problems" and to "list out the possible solutions." Very little attention was given to the mathematical methods of research and analysis; they were just too advance for the students taking these classes. Very few if any students could perform mathematical analysis using such methods as Markov Analysis, Monte Carlo Simulation, matrix analysis using linear algebra, decision trees, or even simple PERT problems.

Evaluation has often been confused with operations research and management. However, evaluation is aimed at a retrospective look back at how well a program has achieved its objectives.

Another type of activity that is sometimes confused with operations research is the development of basic Descriptive Indicators. These are important values which describe, for example, target population, aspects of health-related behavior , and causal factors.

These indicators often constitute a vital input to an operations research exercise, but their production does not by itself constitute an analysis that helps select courses of action aimed at accomplishing specific objectives.

(PRICOR Methods Paper 1 1985)

Operations research is characterized by its prospective orientation and focus on action-oriented decision making...management.

Defining the operation problem

Problem analysis can be defined as:

Define the operational problem
Analyze the problem and divide into smaller units
Establish research priorities

The major difficult in problem analysis is knowing that a problem exists!

Indicators do not indicate a problem unless they are compared with some standard. Budgets may or may not indicate a problem depending on wither the budget was correctly formulated. Administrators will rarely tell their supervisors that they have a problem if it may mean losing their job or their next promotion.

New ideas and new events may make the current operating process inefficient "compared" to the old.

Although it is important to get agreement on the objectives, not everyone will agree to all the objectives listed. What is important, both for a consensus and for analytical, is to to be able to "solve" the problems in several difference ways to reach different objectives. Then by using sensitivity analysis it is possible to have different "views."

It is possible that after each of the solutions are tested one or the other of the solutions then becomes more or less important. The first objective may really not be that important after all.

The objectives must be defined in terms of direct outputs or outcomes. Yet, very few organizations collect enough data to be able to estimate these values successfully.

PRICOR definition of operations research is:

"Operation research is the application of analytic methods designed to help the decision-maker choose between various courses of action available to accomplish specified objectives. The analysis is aimed at determining which course of action is most likely to achieve those objectives and to do so most effectively and efficiently when compared to other possibilities. Thus, operations research is a planning technology; it is future-oriented."

Operations Research is an literary process in that it continuously repeats its cycle.

Problem Analysis -> Design -> Implementation-> Evaluation ->

Approach

PRICOR's Summary of Operations research approach are:

Problem Analysis
- Define the operational problem
- Analyze the problem and divide into smaller units
- Establish research priorities
Solution Development
- Specify solution objectives
- Specify decision variables and stipulate constraints on the solution
- Identify or construct an appropriate model for solution development
- Determine and obtain required data
- Develop solution using analytic model.
- Perform sensitivity analysis
Solution Validation
- Design field test
- Implement field test
- Evaluate the proposed solution, modify if necessary
- Integrate the solution with the larger system.

Problems with Operations Research

(1) You must know that a problem exists. If the problem can not be defined, then there is not much hope in being able to create a solution. Many administrations do not know that they have a problem...they may think they are doing very well indeed!

(2) These "steps" have been developed from the traditional operations research model where the objectives were very clearly define: minimize cost, maximize profits, and subsystems and processes were clearly defined in commercial production systems, In health care and in the social sciences the objectives and processes may not be so clearly defined. Information is just not available on interactions, time sequences, or movement of resources.

(3) One must know the courses of action that are available in advance. Without sufficient data it is impossible to determine what these actions may be. InHCc will add to the PRICOR definition by stating that not only will our operations research must be able to "choose between various courses of action" but our operations research must collect data that will "suggest" possible solutions (through data mining techniques).

It is not comparing possible solutions to problems that make an organization efficient and effective, it is the creation of these solutions.

Will objectives may not be so clearly defined, outputs may be.

(4) Priorities are difficult to define. In health care "everything" is a priority.

(5) While money is always a constraints, how do you decide who is to get service and who is not.

(6) Processes are not static. While an analysis must be done using a "static model" it must be remembered that these processes are not static and that revision is always necessary. Too often the "proposed solutions" are developed and implemented and then never changed!

PRICOR gives an example of a courses of action by stating that primary health care often make it a criterion that new community health workers must be from the same village in which they would practice. This may or may not be a valid criterion, but it is one that is often made rather routinely and without direct analysis of the problem.

Without sufficient data, other chooses are hard to imagine

Another example the PRICOR gives is an example of price-setting. Price-setting for community financing of health services is often set by trial-an-error. Although trial and error may eventually lead to the desired outcome, it may also prove to be very wasteful of such resources as money, staff time and even good will.

Operations research is often used in the standard "waiting time" or quenching problem. If it is known how many clients arrive at a "counter" per unit of time and the average time required to server each client, then it can be easily calculated as to the number of staff needed to service these clients to have the client only waiting, say, 10 minutes.

Operations research is a good tool...but in the past there just wasn't enough data to be able to use this valuable tool. Now it is!

Today with the ability to collect greater amounts of data and having better and easier to use analytical tools, InHCc will attempt to integrate the great work from PRICOR with today's technologies.

Back to Top