Total knee arthroplasty had been reported to present similar amount of blood loss in external and hidden form. We studied whether lateral patellar release made any differences on both forms of blood loss.

Material and methods: We studied 91 patients (83 women an 8 men) undergoing primary unilateral total knee arthroplasty cemented posterior stabilised with patellar substitution, in 40 lateral patellar release were performed. Surgery is performed under tourniquet. The blood drain was recorded. We assumed that blood volume on the fourth postoperative day was the same as before surgery. Blood volume was estimated taking sex, body mass and height into account. Haemoglobin was recorded on preoperative and four days postoperative.

Results: We found more external blood loss in patients without lateral patellar release (p< 0,05) but no statistical differences in total and hidden blood loss. The amount of hidden blood in both groups were more than twice (975 mL) the external blood loss (443 mL). No relationship was found between body mass and any type of blood loss.

Conclusions: Lateral patellar release made no difference in the amount of total blood loss. In total knee arthroplasty hidden blood loss doubled external blood loss.

Aims: In order to reduce the waiting list, the Spanish National Institute of Health sent a large number of patients from Badajoz to other private hospitals, from October 1996 to December 2000, to undergo knee replacement. No medium or long-term follow-up was arranged. Our aims were to compare revision operations in those patients with the ones performed locally.

Material and methods: In that period of time 791 knee arthroplasties were sent to distant centres and 620 were performed in our institution. All revision surgery was performed in our hospital after two months of the index operation in the distant hospitals. We stopped entry of patients in this study in December 2003.

Results: 82 (10.3%) knees have so far required revision surgery in the distant centres group. Of these, 45 have been for deep infection, 26 for aseptic loosening, 5 failed unicompartmental, 3 stiffness, 2 painful non-replaced patellas, 1 non-union of the tibial tuberosity

In the local group 17 (2.6%) knees have so far been revised in the same period of time. Of these 10 have been for deep infection, 3 for aseptic loosening, 3 for instability, and 1 for soft tissue impingement.

Conclusion: The causes for such a difference in revision rates were analysed and include implant selection, surgical technique and absence of follow-up. A constant and angry complaint of all patients sent to other hospitals and subsequently revised was the lack of follow-up.