Randomized, controlled trials (RCTs) are generally accepted as the “gold standard” for the provision of the most unbiased measures of the efficacy of interventions but are often criticized for the lack of external validity. We assessed the external validity of a RCT examining the efficacy of local infiltration analgesia (LIA) compared with continuous epidural infusion after total knee arthroplasty (TKA) During a one-year period, all patients consecutively admitted for elective, unilateral, primary TKA were identified as potential participants. All underwent eligibility screening to determine who were eligible for participation in a randomized controlled trial. We investigated the distribution of preoperative characteristics and postoperative variables among excluded patients, non-consenters, and enrolled and randomized participants.Background
Methods
At follow-up weight-bearing AP pelvis radiographs were obtained. OA was present when the minimum joint space width was <
2.0 mm. Radiographs of sex- and age-matched controls were obtained from The Copenhagen City Heart Study. The following criteria for exclusion were applied:
insufficient or missing radiographs patients who refused to participate, emigrated persons, persons lost to follow-up, patients with previous surgery to pelvis or lower limbs and dead persons. 52 patients (55 hips) were enrolled in the study and 115 patients (136 hips) were excluded. Mean age for men at follow-up was 53 years and for women 55 years.
Radiographs of sex- and age-matched controls for the follow-up group were obtained from The Copenhagen City Heart Study. The following criteria for exclusion were applied
emigrated persons, persons lost to follow-up and patients with previous surgery to pelvis or lower limbs. 135 patients (156 hips) were enrolled in this study and 32 patients (35 hips) were excluded.
The intra-osseous pressure, PO2, and PCO2 were measured in 32 hips (21 patients) which were painful but showed no severe degenerative changes. Pre-operative scintigraphy and radiography was performed in all patients. Thirteen hips showed early osteoarthritis, eight had early osteonecrosis, and 11 had no changes. Core biopsies were performed and the bone was examined histologically and graded for necrosis. Histologically, necrosis was present in 27 specimens. Scintigraphic findings did not correlate with the histological results but were more closely related to the radiographic findings. The intra-osseous pressure in hips with histological necrosis (mean 47 mmHg) was significantly higher than in hips without necrosis (mean 26 mmHg). The PO2 was lower in bone with histological necrosis (mean 44 mmHg) than in bone without (mean 71 mmHg). PO2 increased and intra-osseous pressure decreased after decompression. The results confirm that ischaemia plays a central role in the development of necrotic changes in bone. Histological necrosis was found in hips with radiographic signs of osteonecrosis and in those with osteoarthritis. Radiography, and scintigraphy are shown to be insensitive methods for differentiating between those disorders.
We studied the safety of combining the postoperative use of a non-steroidal anti-inflammatory drug with low-dose heparin. In a double-blind, placebo-controlled clinical trial we reviewed the complications in 235 patients after total hip replacement, all treated with low-dose heparin and either indomethacin or a placebo. The incidence and type of complications in the two groups were nearly equal; indomethacin-treated patients had no increase in complications related to bleeding. Postoperative bleeding into drains was marginally greater in the indomethacin group, although the difference was not statistically significant. We conclude that treatment with indomethacin and low-dose heparin after hip replacement does not significantly increase the bleeding or other complications. We also found that patients receiving indomethacin were mobilised an average of one day before those on placebo.
