Content area
Full text
It is essential to monitor the effectiveness of any service redesign on the process of care. Interpretation of process changes is complicated by natural or "common cause" variation. 1 Classic statistical tests can determine whether a process performance indicator is "significantly different" before and after service redesign. Statistical process control (SPC) charts, invented in the 1920s and used in industry for many years, provide a different approach that might be useful in healthcare. 2 - 9
An SPC chart is a chronological graph of process data with a centre line (usually the mean) and upper and lower control limits defined statistically. 3 4 If all process values fall between the upper and lower limits and exhibit no unusual patterns (eg, nine consecutive points on the same side of the centre line), the variation is considered to be "common cause". Processes exhibiting only common cause variation are in "statistical control". If process values fall outside the limits, or exhibit unusual patterns of variation, this provides evidence of a "special" cause of variation. 3 SPC charts of historical data can determine whether a process has been in "statistical control" (phase 1). SPC charts can then be used prospectively to detect process changes (phase 2). 3 When a chart indicates the establishment of a changed level of performance, calculation of a new centre line and limits is required.
Data on aspects of stroke care in three Lothian hospitals have been collected since July 2001, during which time stroke services have been developed in two of the hospitals. In April 2003, hospital A moved to a new site, increased the number of stroke beds from 22 to 24, acquired faster CT (7 min per brain scan rather than 10 min), and, in January 2004 acquired an additional CT session for stroke. In February 2004, hospital C appointed a staff grade stroke physician and in April 2004 increased stroke beds from 11 to 17. Hospital B's service remained structurally unchanged. In July 2002, the Medical Research Council Heart Protection Study (MRC HPS), indicating benefits of statins following stroke, was published. 10
We retrospectively control-charted process data to: determine whether the changes in stroke service structure and the MRC HPS were followed by "special cause variation" (indicating improvements in process...