The sequential course of the knee score and functional score of the Kn
ee Society rating system of 276 press fit condylar modular unconstrain
ed total knee arthroplasties performed for osteoarthritis between June
1988 and December 1992 was documented prospectively. The knee score i
ncreased significantly and stayed on a constant level from 2 years on,
whereas the function score reached a maximum at 2 years and declined
subsequently. Multiple regression analysis was performed testing the s
tatistical significance and correlation of preoperative predictors wit
h criteria at followup to determine their influence on outcome, Preope
rative predictors were knee score and function score, body mass index,
age, gender, patient category, and implant factors. Criteria studied
were pain, knee score, and function score at 2 years followup. The fun
ction score is influenced significantly by the walking distance, age,
body mass index, and patient category correlating moderately. The knee
score is not affected by any of these factors. Pain was found to corr
elate low with the walking distance. Rating systems are influenced by
numerous factors linked to the patient's general health and condition.
Their impact on the overall result can be controlled by separate rati
ng of the knee score and function score as the dual Knee Society ratin
g system does. Scoring systems adding up knee and functional rating to
an overall result should not be used. There is a need for additional
improvement of total knee arthroplasty rating such as patient based ev
aluation and establishing reliability and validity.