Are you contemplating a knee replacement and want to know the raw facts? Like "what are the chances of actually dying from this procedure?".


A medical paper published in August 2001 (ref 1) examines this question in more than 22,000 consecutive knee replacement patients from the same organisation and the results suggest that:

  • there is about a 0.21% chance of dying in the first 30 days after a total knee replacement.
  • cemented implants carry a far greater risk than uncemented implants. In this study 47 people died out of the 22,000, and in all of these the cemented variety was used.
  • dying is more likely during the first knee replacement than in subsequent procedures which may become necessary to replace the first knee replacement - that probably means that if you have the risk factors, then they are likely to take you out in the first operation and anyone going on to subsequent revision surgery are likely to have lower risk factors in the first place.
  • pre-existing lung and heart problems really do matter and must be taken seriously. 43 of the 47 patients who died had pre-existing lung or heart problems. Although not stated, it seems that one might safely conclude that if you do have these problems it would be best to consider an un-cemented replacement rather than a cemented one (the cement is pretty toxic when it is setting and does nasty things to the blood pressure for the first few minutes when it goes in).
  • having both knees done at the same time is also riskier than one at a time. Sometimes however, eg. with severe bow legs, it is not practicable to do the two knees at separate times. A bow-legged person given one straight knee during surgery would have trouble rehabilitating as the other leg would be much shorter.
  • age seems to be an important factor. 39 of the 47 patients who died were over 70.
  • dying during the actual operation is less common than dying after it, so one needs to understand that the first 30 days are all critical.
  • women and men seem to fare equally well and gender does not seem to be a real issue.


    1. Parvizi, J et al (2001) Thirty-day mortality after total knee arthroplasty. Journal of Bone and Joint Surgery 83-A, 8, 1157 - 1161