Methods to monitor quality of care

This programme aims to understand better control charts, a graphical method that can be used to display hospital safety and quality performance measures. These charts differ from other display methods , such as tables and line charts, in that they distinguish between special cause and chance variation. This is valuable, because investigating data that lies within the play of chance consumes resources better devoted to other quality improvement efforts. Control charts constrain such investigations of chance variations and direct decision-makers’ attention to the data that lay outside of chance variation, resulting in more efficient quality improvement efforts.

The plan was to appreciate control charts from two different perspectives:

• The normative use of control charts, specifically how should control charts be constructed and interpreted. This will be used to provide guidance and advice to NHS trusts about what, why and how they should use control charts. 
• To describe the current use of control charts by NHS trusts. This would allow us to understand if NHS trusts were already using control charts. If they were using control charts, we looked at how often and what types of control charts they used. 

By achieving these aims, recommendations can be formulated to support the increased use of control charts in NHS trusts across the West Midlands in order to improve the analysis of information and appropriate decision-making. 

Programme Outcomes

• The results of the investigation, which have been published in the British Medical Journal (BMJ) Quality and Safety, showed that only 88 (6%) of charts in the 30 trust board papers examined depicted the role of chance, and only 17 of the papers included any charts at all depicting the role of chance.
• The understanding of the use of control charts normative use is much clearer now. Control charts are actually a category of charts which can display safety and quality information across time and space. Control charts advantage over simpler charts is that control charts make atypical data immediately apparent. In so doing direct quality improvement initiatives in one of two directions: investigate the special cause data or chance the process underlying all the data. The reason control charts readily enable this better data-driven decision making can be explained by dual process theory, a psychological theory of judgement and decision-making the behavioural science group at Warwick Business School is well suited to investigate. 
• The understanding of control charts current use is also better. Control charts are being used in approximately half of NHS board papers, but not often in most. Funnel charts displaying mortality information were the most common control chart used. Often these control charts appeared to be copied from external sources, indicating a bottom-up supply chain problem. 

Programme Lead

Kelly Schmidtke
e: kelly.schmidtke@wbs.ac.uk
t: 0121 371 8061