Between November 1994 and June 1999, 35 patients referred to our Problem Fracture Service with chronic diaphyseal osteomyelitis were treated using a closed double-lumen suction irrigation system after reaming and arthroscopic debridement of the intramedullary canal. This is a modified system based on that of Lautenbach.

Between June and July 2007 the patients were reviewed by postal questionnaire and telephone and from the case notes. At a mean follow-up of 101 months (2 to 150), 26 had no evidence of recurrence and four had died from unrelated causes with no evidence of recurrent infection. One had been lost to follow-up at two months and was therefore excluded. Four had persisting problems with sinus discharge and one had his limb amputated for recurrent metaplastic change.

Our results represent a clearance of infection of 85.3% (29 of 34), with recurrence in 11.8% (4 of 34). They are comparable to the results of the Papineau and Belfast techniques, but with considerably less surgical insult to the patient.

We have studied the progression of healing in 103 unstable fractures of the tibia. In 76 patients we removed the external fixator once the stiffness had reached 15 Nm/° in the sagittal plane. Deformity at the site of the fracture subsequently occurred in four patients. In a further 27, we measured stiffness in several planes and removed the fixator only when the stiffness reached 15 Nm/° in each. We found that stiffness in two orthogonal planes may differ widely (maximum difference 9.0 Nm/°, mean 4.1 Nm/°). There were no failures in the second group. We advocate that fracture stiffness be measured in two orthogonal planes when assessing tibial healing and suggest that values above 15 Nm/° in two planes give an indication that it is safe to remove the fixator.

Nomograms derived from mathematical analysis indicate that the level of malunion is the most important determinant of changes in the moment arm of the knee, the plane of the ankle and alterations in limb length. Testing in five patients undergoing reconstruction showed a mean error of postoperative limb length of 2.2 mm (sd 0.8 mm), knee moment arm of 4.7 mm (sd 3.3 mm) and ankle angle of 2.6° (sd 2.3°). These nomograms provide the information required when assessing whether a particular degree of angulation may be accepted.