Re: NICE’s recommendations for thromboembolism are not evidence based
In his Personal View published in 2011 , Dr Welfare highlights important points surrounding the use of anticoagulants in the prevention of VTEs and their inclusion in the NICEs guidelines for VTE prevention . We would like to add to his comments our personal review of the evidence supporting the use of mechanical methods of prophylaxis, most commonly translated as graduated compression stockings (GCS).
The 2010 NICE guidelines recommend that all surgical patients start mechanical prophylaxis on admission, continue during their procedure and post-operatively until fully mobile. For medical patients, mechanical prophylaxis should be considered in patients who have a contraindication to pharmacological prophylaxis. NICE makes an exception for stroke patients, who should not be offered mechanical prophylaxis in any circumstances. Mechanical prophylaxis is usually translated in the use of Graduated Compression Stockings (GCS).
NICE largely base their recommendations on evidence from a 2010 Cochrane review (3) and a separate 2009 trial by Dennis et al (4). After review of these publications, we would like to question the certainty with which the Cochrane authors, and subsequently NICE, made their recommendations.
The Cochrane authors included 18 trials encompassing both medical and surgical patients with and without additional background pharmacological prophylaxis against VTE. These trials included on average 140 patients (range of 18 to 252 patients). The Cochrane authors excluded two large trials, one in 2518 stroke patients (4) and one in 874 orthopaedic patients (5). Reasons for exclusion were that the population was too speciifc for the former and the study too pragmatic for the later. Both of these trials did not support the use of GCS. In 6 of the 18 included trials, CGS were applied only to one leg, with the patient’s other leg acting as a control, a fact known to impede blood flow in both legs (6). All of the trials included in the Cochrane review had detection of VTE as their primary outcome, which were detected radiologically by scanning patients on set days.It is unclear whether or not the patients with the VTEs were symptomatic or asymptomatic, an important fact when the significance of asymptomatic VTEs is uncertain (7). The authors identified 10 of the 14 trials (out of a total of 18) that had received some form of support from stocking manufacturers, ranging from simply supplying the stockings to outright sponsoring the study. Finally, the Cochrane authors decided to part from the Cochrane guidelines when considering publication bias, and provided 2 funnel plots to conclude that no bias was present. We repeated this simple statistical exercise using the Cochrane guidelines and our findings are different (Figures a, b) indicating a statistical bias may well have misled their conclusion that “GCS are effective in diminishing the risk of DVT in hospitalised patients”.
NICE chose to consider both the Cochrane review and one of the study rejected by the Cochrane authors. Including this trial formed the basis that GCS should not be used in stroke patients. NICE does not comment if this conclusion might also apply (or why it would not) to the medical patients included in the studies reviewed by the Cochrane authors, neither does it comment if the inclusion of that study in the statistical analysis would have modified their conclusion.
Both NICE and Cochrane acknowledge that none of the individual trials, or the Cochrane analysis, are statistically powerful enough to see any differences in more clinically important outcomes such as all-cause mortality and pulmonary embolus.
In conclusion, we believe that NICE’s recommendation on GCS should be reviewed. Given the low confidence in any potential benefit it is important that any recommendation does not prevent research required to establish the validity of GCS in routine clinical practice for both medical and surgical patients,. The cost associated with these measures is considerable, and the benefits are uncertain.
Figures a, b. Funnel plots are a way to consider publication bias (8). In figure a, the Cochrane authors created two separate funnel plots with 8 and 10 studies respectively. They concluded on that basis that there was no publication bias. In figure b, we plotted all the studies in one funnel plot, and the plot indicates a possibility of publication bias. Our method is in line with the Cochrane review guidance (8), whilst the one used by the authors is not.
Dr Laurence Whittaker
Foundation Year 1 Doctor
County Durham and Darlington NHS Foundation Trust, Hollyhurst Road, Darlington DL3 6HX
Dr Trevor Baglin
Divisional Director of Pathology & Consultant Haematologist
Box Number: 234, Addenbrooke's Hospital, Cambridge University Hospitals NHS Foundation Trust
Hills Rd, Cambridge CB2 0QQ
Dr Alain Vuylsteke
Consultant in Anaesthesia and Intensive Care
Department of Anaesthesia and Intensive Care, Papworth Hospital NHS Foundation Trust, Cambridge CB23 3RE
1. Welfare M. NICE’s recommendations for thromboembolism are not evidence based. BMJ 2011;343:d6452
2. National Institute for Health and Care Excellence. Venous thromboembolism: reducing the risk of venous thromboembolism (deep vein thrombosis and pulmonary embolism) in patients admitted to hospital. CG92. London: National Institute for Health and Clinical Excellence;2010
3. Sachdeva A, Dalton M, Amaragiri SV, Lees T. Elastic compression stockings for prevention of deep vein thrombosis. Cochrane Database Syst Rev. [online] 2010(7). Available from DOI 10.1002/14651858.CD001484.pub2 [date of access 24/04/13]
4. Dennis M, Sandercock PA, Reid J, Graham C, Murray G, Venables G, et al. Effectiveness of thigh-length graduated compression stockings to reduce the risk of deep vein thrombosis after stroke (CLOTS trial 1): a multicentre, randomised controlled trial. Lancet. 2009 Jun 6;373(9679):1958-65
5. Cohen AT, Skinner JA, Warwick D, Brenkel I. The use of graduated compression stockings in association with fondaparinux in surgery of the hip. A multicentre, multinational, randomised, open-label, parallel-group comparative study. J Bone Joint Surg Br. 2007 Jul;89(7):887-92
6. Spiro M, Roberts VC, Richards JB. Effect of externally applied pressure on femoral vein blood flow. Br Med J. 1970 Mar 21;1(5698):719-23
7. Kim YH, Oh SH, Kim JS. Incidence and natural history of deep-vein thrombosis after total hip arthroplasty. A prospective and randomised clinical study. J Bone Joint Surg Br. 2003 Jul;85(5):661-5
8. Sterne JAC EM, Moher D (ed.). Chapter 10: Addressing reporting biases. In: Higgins JPT, Green S (ed.) Cochrane Handbook for Systematic Reviews of Intervention. Version 5.1.0 (updated March 2011). The Cochrane Collaboration. 2011; Available from www.cochrane-handbook.org [date of access 24/04/13
Competing interests: No competing interests