Skip navigation
Home
News & Events
About WHAIP
Public Health Wales Main Site
Contact us
About Healthcare Associated Infections (HCAI)
About HCAI Surveillance
Infection Surveillance Reports
Point Prevalence Survey 2011
Education and Training
National Infection Control Policies for Wales
1000 Lives Plus
Links
Site Map
Feedback
 
  • Advanced search

Explanation of Statistical Process Control (SPC) Charts

Changes to the presentation of information in the Staphylococcus aureus bacteraemia quarterly reports

 
The format of presenting trend information in the 14th Staph. aureus bacteraemia Quarterly Report has changed.
 
Previously, trend data for each Health Board was presented as a series of moving annual rates with accompanying 95% confidence intervals. 
 
Individual quarterly rates are now presented in the form of a Statistical Process Control (SPC) P-chart. SPC charts assume that rates within a Health Board will be largely similar over time. They present the occurrence of Staph. aureus bacteraemias in a Health Board in relation to what would be expected, based on the mean rate for the Health Board and calculated statistical process control limits for the time period. 
 
These control limits, derived from plus or minus 3 standard deviations from the mean, represent the range of variation in rates of Staph. aureus bacteraemias that might be expected to occur in a Health Board due to chance alone. The control limits for a Health Board may change from quarter to quarter, depending on changes in the sample size ie the number of bed days.
 
 
SPC charts allow the distinction to be made between natural variation (rates that fall within the limits) and special cause variation, where something unusual is occurring in a Health Board (rates that fall outside the limits).
 
If a Health Board's rate in any quarter falls outside its control limits, it means that the rate differs from its mean rate by more than is likely to be due to chance, and special circumstances are more likely to operate. Such a result should prompt an investigation into the possible cause, which could be a true high or low rate of bacteraemia or may have arisen as a result of reporting biases from incomplete or over reporting.
 
As well as investigating when rates fall outside the control limits, investigations should be prompted when certain trends in the data are observed. These trends are listed below and should be read in conjunction with Figure 1.
 
SPC charting
Figure 1. Example Statistical Process Control Chart
 

Trends in an SPC Chart to Prompt an Investigation

  1. Eight consecutive values on the same side of the mean i.e. in zones A, B or C or in zones D, E or F
  2. Any 12 of 14 consecutive values on the same side of the mean i.e. in zones A, B or C or in zones D, E or F
  3. Three consecutive values in either the top third or bottom third of the expected range i.e. in zone A or F
  4. Five consecutive values in the top two thirds or bottom two thirds of the expected range, i.e. in zones A or B or in zones E or F
  5. Thirteen consecutive values in the middle thirds of the expected range ie in zones C and D.
  6. Eight consecutive values either increasing or decreasing.
  7. Cyclic or periodic behaviour.
The occurrence of the above trends should prompt an investigation into their possible causes since:
  • Trends 1 to 4 may indicate an improvement or deterioration in infection control.
  • Trend 5 may indicate an improvement in the consistency of application of infection control measures.
  • Trend 6 may indicate improving or deteriorating application of control measures.
  • Trend 7 may indicate an underlying pattern to changes in the application of control measures that repeats itself at regular intervals, such as seasonality.
The overall mean rate for all Welsh Health Boards will also be plotted on each Board's SPC chart, for comparison purposes.
 


Last updated: 09/12/2010