The Reliability of Orthopaedic Measurements

In making clinical decisions, orthopaedic surgeons rely heavily on measurements obtained from the patient’s history, physical signs and radiological examination. A measurement is regarded as reliable, or reproducible, if the same result is obtained when the same entity is measured again. The importance of measurement reliability can be illustrated with an example of a child with scoliosis. An incremental difference of 5#{176} between two measurements of the Cobb angle is generally believed to represent a deterioration that may require spinal fusion, but two recent studies have shown that different observers may vary up to 10#{176} in the measurement of the Cobb angle (Carman, Browne and Birch 1990; Morrissy et a! 1990). Therefore, true progression of a scoliotic curve is proven only when the difference between two measurements exceeds this limit. Any change less than 10#{176} may simply arise from measurement variability and may possibly lead to an unnecessary operation. The problem creates a scientific challenge for orthopaedic surgeons in documenting, and if necessary, reducing the variability of orthopaedic measurements. Measurement variability is present when different results are obtained on repeated measurement of the same entity. Although laboratory measurements are generally regarded as trustworthy because of wellestablished methods and standards for quality control,