Interquartile Range
The Interquartile Range (IQR) Method is used to identify healthcare providers (hospitals, nursing homes, home health agencies, etc.) that stand out from other healthcare providers in their performance on quality, safety, patient experience, or cost measures. The interquartile range is calculated by first arranging the data values for a given performance measure in ascending order and then dividing it into 4 equal parts or 'quartiles'. The first quartile value (Q1) splits off the lowest 25% of the data values from the highest 75% of the data values. The third quartile value (Q3) splits off the highest 25% of the data values from the lowest 75%. The interquartile range (IQR) is the difference between Q3 and Q1 and shows how spread out the data values are. It can be used to identify data values that stand out from the rest of the data.
The respected statistician, John Tukey, picked cut-offs of 1.5 x IQR below Q1 and 1.5 x IQR above Q3 as the demarcation lines for box-and-whisker plots. Tukey chose the 1.5 x IQR cutoff because it worked well to identify outliers - that is, data points that are well outside the expected range of. Our goal is to identify a broader range of “stand out” performance values, not just the very outlier values. So, we chose a cut-off of 0.5 x IQR, a third of that used by Tukey to identify outliers. Hospitals with low values are defined as those with a value less than Q1 – (0.5 x IQR). Hospitals with high value are defined as those with a value greater than Q3 + (0.5 x IQR).
IQR is most useful when only values are present in a data set and no sample size is given.
Data sets that use the IQR method include:
- Emergency Department Timeliness
- HCAHPS Patient Satisfaction Survey
- Costs and Charges Data
Identifying Outliers:
It is important to note that IPRO uses 0.5IQR method to determine its High and Poor performers