Name
Abstract | ||
|---|---|---|
Statistical process control is increasingly used by single hospitals or centres to monitor their performance, but national monitoring across multiple centres, measures and groups incurs higher false alarm rates unless the method is modified. We consider setting the threshold for cumulative sum charts to produce the desired false alarm rate, taking into account the centre volume and expected outcome rate. We used simulation to estimate the false alarm and successful detection rates for a variety of chart thresholds. We thereby calculated the 'cost' of a higher threshold compared with one set to give a false alarm rate of 5% for three clinical groups of common interest. The false alarm rate often showed non-linear relations with the threshold, volume and expected mortality rate but an equation was found with good approximation to the simulated values. The relation between these factors and the 'cost' of a higher threshold was not straightforward. The 'cost' (difference in number of deaths) incurred by raising the chart threshold provides an intuitive measure and is applicable to other settings. Journal of the Operational Research Society (2011) 62, 1711-1718. doi:10.1057/jors.2010.121 Published online 25 August 2010 |
Year | DOI | Venue |
|---|---|---|
2011 | 10.1057/jors.2010.121 | JORS |
Keywords | Field | DocType |
computer science,reliability,statistical process control,logistics,location,investment,forecasting,operational research,project management,production,information technology,marketing,cumulant,operations research,scheduling,management science,information systems,communications technology,false alarm rate,mortality rate,inventory | False coverage rate,False alarm,Computer science,Statistical process control,Chart,Health services,Constant false alarm rate,Statistics,Operations management,Mortality rate | Journal |
Volume | Issue | ISSN |
62 | 9 | 0160-5682 |
Citations | PageRank | References |
0 | 0.34 | 2 |
Authors | ||
2 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Alex Bottle | 1 | 1 | 1.36 |
| Paul Aylin | 2 | 1 | 1.02 |