Dashboard Terminology. December 2017

Transcription

1 Dashboard Terminology December 2017

2 Contents Green Boundary O/E Ratio Equity Ratio Ideal Level Standard Deviation Z-Score Statistical process control SPC methods Run chart Control chart Cumulative sum (CUSUM) chart with V Mask

3 Green Boundary The green boundary is defined in different ways for different indicators. National rate (per xx population) National numerator/national denominator per xx population National average Average of all the DHB scores Equity Ratio Difference between national value and 1 O/E Ratio Difference between national value and 1

4 O/E Ratio The O/E (observed/expected) ratio is a ratio of the number of observed cases to the number of expected cases The green boundary of the O/E ratio is the difference between the national value and 1 The ideal level of the O/E ratio is 0 Any DHB that is better than (inside) the green boundary has an O/E ratio that is less than the national O/E ratio.

5 Equity Ratio The Equity ratio compares the rates of two populations. The base population is either European & Other or non-māori & non-pacific. Māori, Pacific and Asian ethnicities are compared against these base rates. The green boundary of the equity ratio is the difference between the national equity ratio and 1. The ideal level of the equity ratio is 1. Any DHB equity ratio that is better than (inside) the green boundary is closer to 1 than the national equity ratio.

6 Ideal Level There are different ideal levels between indicators depending on their measures. Ideal = 0 for: Indicators with an O/E Ratio Some Safety, Access and Effectiveness indicators Ideal = 1 for: Indicators with an Equity Ratio Ideal = 10 for: Some Patient centred indicators Ideal = 100% for: Some Patient centred and Safety indicators

7 Standard Deviation Standard Deviation indicates how much the DHBs vary. The bigger the standard deviation, the more the DHBs vary. The standard deviation for each indicator is calculated based on the DHB values as the population using the formula: σ = XX ii XX 2 nn σσ is the standard deviation XX ii is each DHB value XX is the mean of the DHB values nn is the number of the DHBs (20)

8 Z-Score All the DHB values are represented on the dartboard using z-scores from a standard normal distribution. A z-score indicates how many standard deviations a value is away from the green boundary. These scores are the results of standardising the DHB values distribution.

9 Statistical process control Statistical process control (SPC) allows the observer to distinguish between changes which are likely to be random variation (common cause variation) and those that indicate something significant has changed (special cause variation).

10 SPC methods The following SPC methods were used in the Health Quality & Safety Commission s health quality and safety dashboards: run charts control charts cumulative sum (CUSUM) charts.

11 Run chart Run charts are line graphs where a measure is plotted over time, often with a median (the middle value of those plotted so that half are above and half are below) also shown. Two most commonly used rules of identifying special causes: Shift rule: Six or more consecutive data points either all above or all below the median (see chart opposite). Trend rule: Five or more consecutive data points either all going up or all going down. Shift down For more rules and application details see IHI virtual training.

12 Control chart A control chart plots a rate over time, with a central line for mean, the upper and lower control limits are 1, 2 and 3 sigma above and below the mean, see an example on the right. It is more precise than a run chart for identifying special causes. Three most commonly used rules are: Outlier rule: a single point outside the control limits (see top chart). Shift rule: Eight or more consecutive points above or below the mean (see bottom chart). Trend rule: Six or more consecutive points increasing (trend up) or decreasing (trend down). Shift up For more rules and application details see IHI virtual training.

13 Cumulative sum (CUSUM) chart with V Mask Cusum is a similar technique to SPC but it measures the cumulative sum of deviations (higher or lower) from the mean value for a given variable over a set period. By the end of this period these deviations will have cancelled each other out meaning that the Cusum figure will always end at 0. The V-MASK fulfils a similar function to that of the control limit in an SPC chart. It identifies when changes in recorded results are likely to reflect a genuine change in conditions or process. However, as change rather than raw data is being plotted, the meaning of direction is reversed on the chart. If the Cusum line crosses the higher V line, this indicates that rates have fallen significantly, and vice versa. Continued...

14 Cumulative sum (CUSUM) chart with V Mask The Cusum is generally regarded as being more effective when changes or numbers are small. Consequently, for DHB-level data we have tended to use Cusum more than SPC. Further information about this approach, including how it was used to track methicillin-resistant Staphylococcus aureus infections in Scotland can be found at: entley.edu.centers/files/csbigs/henderson. pdf. By crossing the Vhi line at this point, from above it to below it, this indicates that the number of falls with fractured neck of femur was significantly reduced in recent months.