|Has the incidence of deep vein thrombosis in patients undergoing total hip/knee arthroplasty changed over time? A systematic review of randomized controlled trials
|Xing KH, Morrison G, Lim W, Douketis J, Odueyungbo A, Crowther M
The authors concluded that the deep vein thrombosis incidence in patients undergoing elective total knee arthroplasty appeared to have decreased over time in patients receiving warfarin prophylaxis. Much of the review was well conducted and evidence appears to support the authors’ cautious conclusions, but presentation of the results of control groups and regression models would have helped its interpretation.
To evaluate the effects of warfarin thromboprophylaxis on rates of deep vein thrombosis (DVT) over time in patients undergoing total knee arthroplasty or total hip arthroplasty.
MEDLINE, EMBASE and the Cochrane Central Register of Controlled Trials (CENTRAL) were searched from inception to October 2006 for studies published in English. Search terms were reported. Reference lists of identified articles were screened and experts contacted.
Randomised controlled trials (RCTs) were eligible if they evaluated the effects of warfarin thromboprophylaxis on DVT in patients undergoing elective total knee or hip arthroplasty. Selected studies screened patients for DVT using venography and assessed proximal (popliteal, femoral, or iliac veins) and total (proximal and distal) DVT using venography. They reported international normalised ratio (INR) values.
In most studies, warfarin prophylaxis was started the evening before surgery. In some studies, concurrent methods of thromboprophylaxis were used including graduated pressure stockings, continuous passive motion devices, and non-steroidal anti-inflammatory drugs. Targets for INR ranged from 1.8 to 4.0 and most patients were undergoing unilateral joint replacement. Where reported, studies differed with respect to type of anaesthesia (types included general, region, and a mix of the two), tourniquet time, duration of surgery, and time to mobilisation after surgery. Hip studies rarely provided information about surgery or tourniquet time or mobilisation. The review also assessed symptomatic DVT. The authors stated that venograms were generally performed 7 to 14 days after surgery.
Two reviewers independently selected studies and disagreements were resolved by discussion with a third reviewer.
Assessment of study quality
Two reviewers independently assessed validity using Jadad criteria (randomisation, methods of double-blinding, and reporting of withdrawals). Studies that scored three or more out of the maximum five points were classified as high quality.
For each study, the proportion of patients experiencing any DVT, symptomatic DVT, and proximal DVT were extracted for warfarin groups only.
Two reviewers independently extracted data onto a standardised form. Discrepancies were resolved through discussion with a third reviewer.
Methods of synthesis
The studies were grouped by type of joint replacement (knee or hip). The weighted average proportion of DVT was calculated for each year. Weighted linear regression was used to assess the relationship between DVT incidence and year; the strength of the relationship was tested using the correlation coefficient (r).
Pooled odds ratio (OR) and 95% confidence interval (CI) data were calculated for total knee versus hip arthroplasty, using the fixed-effect Mantel-Haenszel method. Differences between studies were discussed with respect to changes in clinical practice over time including the duration of surgery and tourniquet time, age of patients, level of obesity, concurrent DVT prophylaxis methods, time of mobilisation, and type of implant.
Results of the review
Fourteen RCTs were included (n=4,423 patients). Nine trials were in patients undergoing total knee arthroplasty (n=2,973) and seven trials were in patients undergoing total hip arthroplasty (n=1,450); two trials included both types of procedure. Publication dates ranged from 1992 to 2005.
Eleven trials were classified as high quality. Nine reported adequate randomisation methods, seven reported adequate allocation concealment, eight were double-blinded, and all 14 described withdrawals. For patients receiving warfarin thromboprophylaxis the results were:
Total DVT: The risk of DVT was significantly greater after total knee compared with hip arthroplasty (36% versus 23%, OR 1.85, 95% CI 1.6 to 2.14). The risk of developing a DVT decreased significantly over time in patients undergoing knee surgery (r=-0.75, p=0.031), but there was no change in patients undergoing hip surgery (r=0.25, p=0.63).
Symptomatic DVT: The risk of symptomatic DVT was significantly greater after hip surgery compared with knee surgery (2.8% based on two hip trials versus 1.3% based on four knee trials; OR 2.18, 95% CI 1.11 to 4.27). There was insufficient data to evaluate changes in DVT incidence over time.
Proximal DVT: The proximal DVT rate was 5.9% after knee surgery and 4.3% after hip surgery. The risk of developing a proximal DVT decreased significantly over time in patients undergoing knee surgery (r=-0.86, p=0.007), but there was no change over time in patients undergoing hip surgery (r=0.13, p=0.80).
The incidence of DVT in patients undergoing elective total knee arthroplasty appears to have decreased over time in patients receiving warfarin prophylaxis, but this decline was not noted in patients undergoing total hip arthroplasty.
The review question was clearly stated and inclusion criteria were appropriately specified. Several relevant sources were searched, but no attempt was made to minimise language bias and it was not clear if any attempt was made to reduce publication bias. Appropriate methods were used to minimise reviewer error and bias during the review process. Validity was assessed and results were reported. Only RCTs were included, but since only data from warfarin treatment arms were used it is not clear why only RCTs were eligible. Trials were combined using meta-analysis, but it was unclear how the comparisons between knee and hip surgeries were made, or whether the results were valid, as neither the results for control groups, nor the pooled results for the knee or hip trials only, were reported.
Although some potential causes of differences between trials were discussed, there was no statistical testing of the significance of differences in DVT rates between trials. The authors used regression models to assess changes in DVT incidence over time, but as they highlighted, a limitation of this was that the publication year of the study might not reflect the years of recruitment. They also did not report full results of the models, only the correlation coefficients, and as there were only a few trials and one appears to have been a major outlier, it is unclear how valid the reported relationships are. There was no discussion of reasons for the apparent reduction in DVT incidence over time in knee and not hip surgery.
Much of the review was well-conducted and evidence appeared to support the authors’ cautious conclusions, but presentation of the results of the control groups, and the regression models would have helped its interpretation. Discussion of the reasons for the decline in DVT in total knee and not hip arthroplasty were needed.
Implications of the review for practice and research
Practice: The authors did not state any implications for practice.
Research: The authors stated that future studies evaluating warfarin for DVT prophylaxis should use contemporary rates of DVT after total knee arthroplasty. Non-pharmacological methods of DVT prophylaxis after orthopaedic surgery should also be evaluated.
No funding or financial support was received.
Xing KH, Morrison G, Lim W, Douketis J, Odueyungbo A, Crowther M. Has the incidence of deep vein thrombosis in patients undergoing total hip/knee arthroplasty changed over time? A systematic review of randomized controlled trials Thrombosis Research 2008; 123(1): 24-34
Subject indexing assigned by NLM
Adult; Age Factors; Aged; Arthroplasty, Replacement, Hip /adverse effects; Arthroplasty, Replacement, Knee /adverse effects; Humans; Incidence; Middle Aged; Postoperative Complications /epidemiology /prevention & control; Randomized Controlled Trials as Topic; Time Factors; Venous Thrombosis /epidemiology /prevention & control; Warfarin /therapeutic use
Date bibliographic record published
Date abstract record published
This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.