Robot assisted UKA a randomised trial

Robot assisted UKARobot assisted UKAa randomised triala randomised trial

Justin CobbJustin CobbChair of OrthopaedicsChair of Orthopaedics

Imperial CollegeImperial CollegeLondonLondon

CT based planningCT based planning Few 5mm slices through hips and anklesFew 5mm slices through hips and ankles 1mm contiguous slices at knee1mm contiguous slices at knee

Novel registration algorithmNovel registration algorithm Hip centreHip centre 4 initial points4 initial points 6 points in 5 zones6 points in 5 zones

On screen feedback more helpful On screen feedback more helpful than direct visionthan direct vision

Rest of operation is routineRest of operation is routine

ResultsResults

Prospective Randomised Controlled TrialProspective Randomised Controlled Trial Ethics committee approvedEthics committee approved Conventional mis uka by one of 4 surgeonsConventional mis uka by one of 4 surgeons All had ct based planningAll had ct based planning Surgeon showed what he was aiming atSurgeon showed what he was aiming at Patients not told until after 6 week checkPatients not told until after 6 week check Imaging reported by blind assessorImaging reported by blind assessor

Post-op CT vs pre-op planPost-op CT vs pre-op plan Whole bone images co-Whole bone images co-

registeredregistered White = planWhite = plan Yellow = achievedYellow = achieved

• Combination of errors:Combination of errors: registration registration bone prepbone prep implantationimplantation

Control examples cases 1,3Control examples cases 1,3

Control GroupControl Group

Acrobot example 1Acrobot example 1

Acrobot GroupAcrobot Group

Blind assessmentBlind assessment

Group Statistics

38 .071 4.3175 .7004

26 .104 2.0348 .3991

38 -2.918 5.1591 .8369

26 .796 1.5280 .2997

38 1.129 6.2594 1.0154

26 -.492 3.7476 .7350

38 -.016 2.6912 .4366

26 .315 .9967 .1955

38 .342 2.4720 .4010

26 .665 1.0202 .2001

38 -.095 1.7205 .2791

26 -.069 .5698 .1117

C0R10

1

0

1

0

1

0

1

0

1

0

1

VARVAL

FLEXEXT

AXIALROT

MEDLAT

ANTPOST

PROXDIST

N Mean Std. DeviationStd. Error

Mean

Difference Difference in stdevin stdev

2 times2 times

3 times3 times

1.5 times1.5 times

2.5 times2.5 times

2.5 times2.5 times

3 times3 times

Acrobot much more accurateAcrobot much more accurate

Tibio-Tibio-Femoral Femoral angleangle

Type of surgeryType of surgery

All kneesAll kneesAcrobot Acrobot SystemSystem

ConventionalConventional

2°2° 13 (100%)13 (100%) 6 (40%)6 (40%) 1919

> > 2° 2° 9 (60%)9 (60%) 99

TotalTotal 1313 1515 2828 P<0.001 (Fischer’s exact t test)P<0.001 (Fischer’s exact t test)

Correlate accuracy and AKS scoresCorrelate accuracy and AKS scores

change in knee score vs accuracy

ACCURACY

6050403020100

120

100

80

60

40

20

0

-20

Observed

Linear

Correlations

1.000 -.522**

. .002

32 30

-.522** 1.000

.002 .

30 30

Correlation Coefficient

Sig. (1-tailed)

N

Correlation Coefficient

Sig. (1-tailed)

N

SUMVAR

KSSDIF

Spearman's rhoSUMVAR KSSDIF

Correlation is significant at the .01 level (1-tailed).**.

Correlation is significant, p=0.002

Summary Summary

1MIS UKA will become the default operation for OA of the 1MIS UKA will become the default operation for OA of the knee, whether medial, lateral or patello-femoralknee, whether medial, lateral or patello-femoral

2 Meniscal bearing an excellent option2 Meniscal bearing an excellent option 3 Ct based modelling is an enabling technology3 Ct based modelling is an enabling technology