ABSTRACT
After deciding what to measure and how to measure it, with adjustments for casemix as necessary, we come to the business end of performance monitoring: deciding which units if any have performance levels that are worthy of further attention. What that attention is depends on the purpose of the monitoring, as has already been stressed. We generally want to know whether a unit’s performance is “unacceptable”, although P4P schemes, for instance, are also interested in “good” or “excellent” results. A quality improvement project may simply want to know whether care has improved at that unit – even if other units have improved more during the same time period. The drive for “high reliability systems”, using approaches borrowed from industry, includes a focus on reducing variation both within and between units. It’s therefore useful to be able to quantify the amount of inter-unit variation and monitor it over time. In this chapter, we’ll see how to do all of these. A couple of sections are necessarily more technical than the rest and are marked with an asterisk as usual.
