Re: NICE’s recommendations for thromboembolism are not evidence based
Dr Mark Welfare, in his personal view piece “NICE’s recommendations for thromboembolism are not evidence based”, makes several valid observations, particularly about the lack of good quality evidence in some areas and the need for more research to address this. The prophylaxis of venous thrombo-embolism is not alone in lacking top grade evidence; it is alarming how many apparently tried and tested therapies fail that test when a proper search is made for proof. We agree there is a lack of randomised trial data in the field of VTE prophylaxis for general medical in-patients, despite the fact that expert systematic reviewers provided all the findable evidence and the Guideline Development Group made every effort to examine the evidence that was available. It is matter of record that when the CLOTS Trial Collaboration report findings on DVT prophylaxis in stroke patients  we stalled the publication of our guidance, met to reconsider, and changed our guidance in the light of this most recent evidence.
Dr Welfare takes the guideline developers to task on the use of surrogate endpoints and, in principle, he is right to do so. If there were ample randomised evidence with the important endpoints of fatal bleeds versus fatal embolism they would have used that alone. As he says, there is insufficient evidence. The Guideline Development Group paid particular attention to whether asymptomatic DVT was a surrogate or whether there was in fact a continuum and revisited the question several times during the development of the guidance. The GDG studied the point specifically and found that where both the major and lesser outcomes were available there was a proportionality which ran through individual studies and their meta-analyses. There are underpinning biological reasons to believe in a common cause hypothesis for DVT and pulmonary embolism and there are empirical reasons to believe that Heinrich’s triangle applies in the relationship between venous thrombosis and death from pulmonary embolism. We therefore decided to take the line that there was a continuum between fatal PE and asymptomatic DVT and if lower limb thrombosis was reduced that would translate into lives saved from PE. The NICE GDG is not alone in taking this approach. This is imperfect, but better to use the evidence available than waste it. Some unbiased evidence is better than none. As and when new evidence becomes available, NICE will assess it.
Unfortunately Dr Welfare then departs from cogently reasoned argument by implying that the guideline developers were in the pocket of pharma and that, by implication, NICE was acquiescent in putting its name to a deliberately flawed guideline. The gentlest interpretation is that Dr Welfare has an incomplete understanding of how NICE guidelines are developed. Be that as it may, the Guideline Development Group members, unfairly accused by Dr Welfare, might be owed an apology. Dr Welfare’s use of the fact that Boehringer Ingelheim issued a press release on behalf of Lifeblood without its consent as proof of complicity between the two is similarly unconsidered as evidenced by his omission of the fact that Lifeblood reported Boehringer to the ABPI over the matter.
We have to defend our reputations against the implication that an independent Guideline Development Group could be so easily suborned by self-interest and that due vigilance on behalf of each other was not exercised. Every meeting started with a rigorous update on conflicts of interest and an interrogation by the chair where there was any doubt. It is quite natural that leaders in the field also are involved in funded meetings and help with clinical trials of pharmaceutical products but during the work of the GDG we ensured that they distanced themselves from these involvements.
Competing interests: Memebr of NICE CG92 group