Our RSA data do not support the view that tourniquet use will improve fixation. We found no statistically significant effect on prosthesis migration. The confidence interval excludes an increase in median value of 0.13 mm if a tourniquet is not used. Such a small difference is not likely to have any clinical importance. However, the distribution of the migration values was far from normal, and there were a number of outliers. These outliers are probably patients who may be at risk of future loosening (and one of these has already undergone revision). Indeed, RSA data may often show discontinuous distributions, reflecting distinct subcategories of the postoperative course (some are well fixed and others not) (Aspenberg et al. 2008
). In this study, there were similar numbers of outliers in both groups.
Apart from prosthesis fixation, bleeding is generally considered to be a good reason for tourniquet use. However, the possible advantage of less bleeding should be balanced against the disadvantage of increased postoperative pain and a smaller final range of motion. We found a reduction in overt bleeding in the tourniquet group, as has been described in a recent meta-analysis (Alcelik et al. 2011
). However, there was no substantial difference in total bleeding as estimated by the hemoglobin dilution method, and any increase exceeding 22% from lack of tourniquet use could be excluded with 95% confidence. It is conceivable that the hidden, postsurgical blood loss is less in patients in whom the surgeons have been able to identify and cauterize bleeding vessels.
The main reasons for refraining from tourniquet use, as judged by our results, are postoperative pain and range of motion. Before the study, we had the impression of a clear reduction in the patients’ complaints about thigh pain after we stopped using a tourniquet. Although patients reported more pain in the tourniquet group, there was no statistically significant difference in the use of morphine-analog analgesics. This may reflect that the difference between the groups was only minor, so that most patients’ pain was controlled by standard postoperative analgesics given to all patients. The range of motion may be more important: the difference between 113° and 124° may mean the difference between being able to ride a bicycle or not. There have been very few previous randomized trials on tourniquet use and range of motion. Short-term better flexion has been shown in 2 studies (Abdel-Salam and Eyres 1995
, Wakankar et al. 1999
), but neither publication reported a difference between treatment groups in the longer term. Total range of motion was not reported.
The mechanism behind the reduced range of motion is unclear. However, the postoperative thigh pain might reflect muscle injury due to physical damage, as well as reperfusion injury, which might both cause a degree of muscle fibrosis. The increased pain might also reduce the patient’s ability to perform postoperative training, which has been shown to negatively affect the final range of motion (Newman 1984
, Silver et al. 1986
, Clarke et al. 2001
The main weaknesses of our study lie in the limitations of the RSA method to predict clinical outcomes, in spite of the fact that RSA studies are recommended before new joint prosthesis are commercially deployed (Derbyshire et al. 2009
). The study size is sufficient for construction of relevant confidence intervals, but not for analysis of the number of outliers in the respective groups. This must be considered. The findings regarding range of motion and pain—although highly statistically significant—should be interpreted with caution, as they were not primary variables. Estimates of bleeding have a high degree of uncertainty, as the number of patients may be too small.
In conclusion, our RSA data do not support the use of a tourniquet to improve fixation. On the contrary, tourniquets appear to cause more postoperative pain and less range of motion.