What IQR/Median Stand For
IQR stands for Interquartile Range, and Median represents the 50th percentile value of a dataset—these are robust descriptive statistics used to express central tendency and data spread, particularly for skewed distributions common in medical research. 1
Median
- The median is the middle value when data points are arranged in order, representing the 50th percentile of the distribution 2, 1
- It is not influenced by extreme values or outliers, making it ideal for skewed distributions like incubation periods, survival times, and other clinical measurements 1
- The median provides a more accurate representation of "typical" values compared to the mean when data are not normally distributed 1
Interquartile Range (IQR)
- The IQR represents the distance between the 25th percentile and the 75th percentile of the data distribution 2
- It captures the middle 50% of all data points, providing information about the spread of the central portion of the dataset 1
- A smaller IQR indicates more consensus or less variability in the data, while a larger IQR indicates greater spread 2
- The IQR is resistant to outliers and provides a more accurate representation of typical variation than standard deviation when data are skewed 1
Clinical Application
- In medical literature, data are commonly reported as "Median (IQR)" format, such as "8.0 (3.0–15.0)" where 8.0 is the median and the values in parentheses represent the 25th and 75th percentiles 2
- These statistics are particularly useful for expressing variables like hospital length of stay, laboratory values, survival times, and other clinical measurements that often have skewed distributions 2
- The American Thoracic Society specifically recommends using median and IQR for small sample sizes and non-normally distributed data 1
Why Not Use Mean and Standard Deviation?
- Unlike mean and standard deviation, median and IQR are not distorted by extreme outliers that can occur in clinical datasets 1
- The Pearson coefficient and t-tests are designed for normally distributed data and may not be appropriate for skewed distributions 1
- When data are skewed (as is common with incubation periods, treatment durations, and many clinical outcomes), median/IQR provides more clinically meaningful information 1