This is a multicenter prospective clinical study using a modified Knee
Society scoring system which evaluated the effect of age, gender, wei
ght, preoperative range of motion and knee score, previous surgery, an
d modification of the posterior femoral condyle geometry on postoperat
ive range of motion. The primary outcome variable was change in flexio
n. The data were collected from 5 surgeons using a single total knee s
ystem. The current study has 621 patients enrolled, of which 282 total
knee replacements have followup of 12 months and 86 have followup of
24 months. Multivariate analysis was used to evaluate the data. The va
riables listed were examined as to their relationship to changes in fl
exion. Patients were divided into 3 groups: preoperative flexion less
than 90 degrees, 91 degrees to 105 degrees, and greater than 105 degre
es. When comparing the patients with preoperative motion less than 90
degrees to those with motion greater than 105 degrees, the first group
improved 26 degrees more than the latter. They also improved 12 degre
es more than the midrange group. The midrange group improved 14 degree
s more than the upper range group. These values are all adjusted to el
iminate differences due to the other variables. None of the other vari
ables showed a significant correlation with the flexion outcome. To an
alyze the knee score, the group was also divided into 3 groups: preope
rative score less than 27, 28 to 40, and greater than 40. The preopera
tive knee score was the best predictor of the postoperative knee score
. The patients with preoperative knee scores below 27 improved 16 poin
ts more than those in the 27 to 40 range and 33 points greater than th
e greater than 40 group. To analyze functional evaluation, the patient
s were divided into 3 groups based on preoperative score: less than 40
, 41 to 50, and greater than 50. Those in the less than 40 group impro
ved 14 points more than the midrange group and 35 points more than the
greater than 50 group. Analysis of delta range of motion and delta pa
in showed similar results. Age, weight, previous open surgical procedu
re, and altered femoral component contour, did not seem significantly
correlated with changes in postoperative flexion. The best predictors
of postoperative clinical results are the preoperative scores.