Accuracy and precision of joint space width measurements in standard and macroradiographs of osteoarthritic knees.

OBJECTIVES To improve the reproducibility and accuracy of joint space width (JSW) measurement as an assessment of cartilage loss in patients with osteoarthritis (OA) of the knee by determining how precision and accuracy of JSW measurement were altered by a computerised method of measurement, correction for radiographic magnification, radiography of the knee in the standing semiflexed view, and high definition macroradiography of the knee in the semiflexed view--taking JSW measurements from standard radiographs of OA knees in the extended view as the standard for comparison. METHODS Twenty five OA and 10 nonarthritic knees were radiographed in the extended view and minimum JSW was measured manually. Conventional and x5 macroradiographs were taken in the semiflexed view. All radiographs were taken twice on the same day and repeated two weeks later. Automated computerised measurement of minimum JSW was obtained from digitally stored images of all radiographs. RESULTS For medial compartment JSW measurements, computerised was more accurate than manual, correction for radiographic magnification improved precision and accuracy, measurements in the semiflexed view were more precise and accurate, and macroradiography increased measurement precision. For the lateral compartment JSW measurements, correction for radiographic magnification improved precision and accuracy, and the semiflexed view improved precision only. CONCLUSIONS Protocols defining radiographic and mensural procedures are essential for quality control of knee radiography in the semiflexed view to permit accurate and reproducible measurement of JSW. Macroradiography provides greater precision of JSW measurement.