Measuring functional abilities of patients with knee problems: rationale and construction of the DynaPort knee test.

We present the rationale and design of the DynaPort knee test. The test aims at measuring knee patients' functional abilities in an unobtrusive, user-friendly way. Test persons wear several belts around their trunk and legs. The belts contain accelerometers, the signals of which are stored in a recorder, embedded in one of the belts. The knee test consists of a set of 29 tasks related to activities of daily life ("test items"). Accelerometer signals are analyzed in terms of 30 "movement features" (accelerations, angles, durations, frequencies, and some dimensionless numbers). In data analysis, the beginning and end of each test item is marked by hand; otherwise, analysis is automatic. We compared 140 knee patients with 32 healthy controls and found 541 of the 29 x 30=870 test item x movement feature combinations differed significantly between the two groups. From these 541 combinations the DynaPort knee score is calculated by the weighted averages of movement features per item, then weighted averages of items per cluster (locomotion, rising and descending, transfers, lifting and moving objects), and finally the average of the clusters. In an initial study the test-retest reliability of the knee test proved high, and the test turned out to be sufficiently responsive (0.7 patients' standard deviations improvement after 24 months). However, it remains difficult to interpret the scores in more meaningful terms than merely "better" or "worse." Extensive reliability studies in the future will further assess the validity of the test and provide more insight into the meaning of the scores. The DynaPort knee test may thus become an important instrument for evaluating patients'functional abilities in knee-related clinical practice and research.