Tibiofemoral alignment has a direct correlation with the survival of total knee arthroplasty. Traditionally, it has been measured using a goniometer on radiographs. We describe new software which measures this alignment on scanned radiographs by automatically detecting bones in the image.

Two surgeons used conventional methods and two clerical officers used the computerised routine to assess 58 radiographs of the knee on two occasions. There were no significant differences between any of the paired comparisons. The largest mean difference detected was 1.19°. Across all comparisons, the mean correlation was 0.755. A standardised routine for measuring tibiofemoral alignment was the greatest factor in reducing error in our study.

These results show that non-medical staff can reliably use the software to measure tibiofemoral alignment. It has the potential to measure all the parameters recommended by the Knee Society.

We describe a method of audit of a type of total knee replacement, including some details of the organisational difficulties of administering multicentre studies, and draw attention to how this can be done using industrial funding without prejudicing the study. This is a prospective record of 1439 patients who had an Insall-Burstein II (IBII) prosthesis implanted between 1990 and 1994. The data were collected using the American Knee Society scoring system. A method of storing radiographs digitally at low cost is also described.

The results emphasise the need for the long-term collection of data on commonly used devices implanted by a cross-section of surgeons. We conclude that for most patients the IBII cemented, posteriorly stabilised, cruciate-substituting prosthesis will relieve pain and give excellent functional results throughout the patients’ remaining years with a very small incidence of revision, except in cases of infection.