Statistical Process Control (SPC) is a methodology that makes up the backbone of quality control. Simply, SPC is the ruler of industrial engineering.
It is based on a simple principle, all processes have a degree of variability. When throwing darts at a board, a player will not hit the bulls-eye all of the time. A very skilled player may hit it most of the time, but they will eventually miss. If we could only know where the darts landed and nothing else, how would we be able to differentiate between natural hit/miss of the player or whether something significant has changed?
Statistical process control allows statisticians to detect when the player has just missed or if there is an actual change in performance like a differently skilled player, throwing darts with broken tails, or the player stepping closer.
A Little History
SPC was pioneered in the early 1920s by Walter A. Shewhart at Bell Laboratories. This methodology became a heavily utilized technique to measure when a process was in control or when a process was changing or deviating. Manufacturing was the primary beneficiary of this development. With this tool, changes in intentional processes could be measured and unintentional processes detected. “In statistical control,” has become an important term in manufacturing that defines capability of any process.
Healthcare has slowly begun to reap the benefits of this methodology. By incorporating SPC, it has become possible to benchmark processes or entire hospitals.
With statistical process control, events become quantitative values that can be used to discover the True State of a system. Outliers can be worked around and true values can be compared. Likelihood of perioperative infection, rate of cases taken in each day, number of accidental needle pokes, average wait time for tests. These values can be quantified by statistical process control and compared. Changes in processes can be measured to be improved.
Statistical process control provides a tool for healthcare providers to better analyze their own processes and find the best ones. Better processes mean faster diagnostics, faster patient turnover for lack of a better statement. Better processes mean less drug use, reduced fraud, and decreased medical mistakes. The use of statistical process control in healthcare provides an understanding that will improve patient care and decrease costs. This is a benefit reaped by the healthcare provider, the patient, and society.
The Big Question: How?
How should a hospital ranked? By the number of “healed?” Many illnesses cannot be cured. Customer satisfaction? I would not be surprised to find people diagnosed with cancer to be unsatisfied, but I don’t believe many people would be willing to fill out additional forms.
In addition, the application of statistical control charts is complicated. Industry professionals have misapplied these concepts before. Healthcare responsibility for people’s lives means the risks are too high to allow for misapplied and poorly understood quality control concepts to influence decision-makers. Healthcare processes are uniquely different from the assembly line structures of manufacturing. A basic understanding of statistical process control is not enough. Nothing less than a strong understanding of statistical process control within healthcare is needed. Healthcare systems engineers must be able to correctly distinguish the correct chart and measure for any situation
While research has been conducted to discover which techniques and charts within statistical process control is most applicable for each situation, the decision methodology is still in its infancy. I believe a comprehensive manual for how, when, and where healthcare system engineers should apply statistical process control must be created.This document may and should change as new standards and techniques are discovered, but the beginnings of a framework must be created to establish a standard.
Simply put: How does a healthcare systems engineer approach the problem of analyzing a process with statistical process control?