I have been thinking quite a bit about health management, quality improvement and some of the tools we used back in my big corporate, manufacturing days. This resulted in a rather long article on SPC, which I have split into 4 blog posts. Enjoy part 1.
SPC, or Statistical Process Control to use its Sunday name, is a technique that has been widely used in industry for many decades. Walter A. Shewhart developed the control chart in 1924 whilst working at Bell Laboratories (the source of many innovations that reverberate into the 21st century).
A critical insight Shewhart developed was that, while all processes exhibit variation, some variation is inherent to the specific process (‘normal’ or ‘chance’ sources of variation) and others are ‘special’ or ‘assignable’ sources of variation due to things that might be controlled or eliminated. Essentially, ‘normal’ sources of variation are those that consistently acts on a process and collectively produce a statistically-stable and repeatable distribution over time, while ‘special’ causes are t hose things that cause intermittent and unpredictable variation, often only affecting some of the process.
Shewhart described process control in terms of natural, chance-cause vs. special assignable-cause variation and developed the control chart as a useful tool for distinguishing between the two. He stressed the importance of bringing a production process into a state of statistical control, where there is only chance-cause variation; keeping it in control is necessary to predict future output and to manage a process economically.
Figure 1 SPC describes multiple control charts
Value of SPC
Understanding this allows organisations to methodically identify and eliminate the ‘special’ sources of variation in order to improve the process, and avoid wasted effort attempting to control the ‘normal’ sources that are an inherent feature of the process.
Using control charts to differentiate between the ‘special’ sources and the ‘normal’ sources allow organisations to modify their processes so that all variation remains within the ‘normal’ bounds (i.e. the process is stable). Eliminating all variation is not possible, but unexpected variation triggers an alert that the system or process has moved out of the normal range allowing the cause to be investigated. There are different variants of control charts appropriate to different situations, for example, the number of measurements being made, whether you are controlling the number of defects rather than the length or weight of something, etc. There is a useful summary here.
In part 2 I’ll explore successes and concerns of applying SPC to the Healthcare settings.
Part 3 will take a sidewise look at SPC in Healthcare and ask it it really is the next big thing, or just another bandwagon
When part 4 comes along it will be about how to use SPC with your current data tools, namely Excel and Power BI. I’ll reach some kind of a conclusion too.
If you are interested in #SPC, #QualityImprovement and the #NHS then you might find this multipart article interestingTweet