Healthcare Analytics for Quality and Performance Improvement

CHAPTER 9

Basic Statistical Methods and Control Chart Principles

If your experiment needs a statistician, you need a better experiment.

—Ernest Rutherford

There are complementary methods to measure the impact of a change or innovation on quality and performance—statistically and using control charts. Statistical methods to determine changes in performance rely on the performance of statistical tests to determine if changes in quality, performance, or other metrics are “statistically significant.” Graphical approaches, on the other hand, use specialized charts known as statistical process control (SPC) charts (and specific rules to aid the interpretation of those graphs) to determine if a change in quality or performance is in fact occurring. This chapter discusses how both of these methods can be employed for quality and performance improvement.

Statistical Methods for Detecting Changes in Quality or Performance

I chose the epigraph at the start of this chapter rather tongue-in-cheek. My intent with the quotation isn’t to say that statistics (and statisticians) should be avoided, but rather that the job of analytics professionals (including statisticians) is to make statistics more accessible and easily understood to all users of information through the use of the right tools in addressing the right problems (those of the quality and performance issues of the organization).

Statistics offers a wide range of methods with which to analyze quality and performance data. In this section, ...