DVT, uncertainty and hospitals

You and Yours today had a feature about thrombosis. They suggest on their website that “hospitals are failing to risk assess patients for DVT resulting in thousands dying needlessly”.

The mortality issue around preventing DVT – blood clots – in hospitals is a very interesting one. NICE guidelines last year said that ” A note of caution must remain however. We generally lack evidence for reduction in all-cause mortality which would require very large trials.” (page 3)

So we don’t have good evidence that we can prevent deaths from blood clots in hospitals.  Many thrombosis are without symptoms and consequence. Others can maim or kill. While mechanical means of getting blood moving – getting up as soon as possible after an operation, or compression stockings – don’t involve drugs, the use of anticoagulants is now common, and mainly protocol driven. These potentially have side effects. Even if these are rare effects, they stack up when the drugs are prescribed to bigger groups of people.

NICE are still using proxy outcomes for mortality, which are less reliable, by nature, than real life outcomes. This is important, especially when there is ample opportunity to study this better; this is a big patient group.

Anticoagulation has side effects; take this exchange from the BMJ about the previous guidelines in 2007, for example:

We are concerned that the authors do not appreciate the catastrophic sequelae of an epidural haematoma following spinal surgery. In surgical operations when spinal osteotomy or vertebrectomy is undertaken, the blanket guidelines seen in the BMJ will lead to an increase in neurological complications, especially as epidural bleeding may be very difficult to control during these operations. It is our policy not to use chemical thromboprophylaxis routinely in spinal surgery but to use mechanical methods and to mobilise patients early with the in patient use of compression stockings. We also encourage patients to perform active foot and ankle movements whilst in bed. Even with this policy the numbers of epidural haematomas seen in our unit is greater than the rate of symptomatic venous thromboembolic episodes.”

So these doctors were concerned about side effects from anticoagulants causing major problems to the site of surgery and being a bigger hazard than blood clots.

The authors of the guidelines responded (scroll down on the BMJ site)

“In some patients as highlighted by Gardner and colleagues(9) even a very small bleed, insignificant in other contexts, might be devastating. Thus they have interpreted the guidelines as being too aggressive, a view shared by many orthopaedic surgeons. This issue was debated extensively by the GDG. Guidelines are however what they say they are: guidelines – not directives and on every NICE Guideline is a statement that healthcare professionals are expected to take the guideline fully into account when exercising their clinical judgement. The guidance does not, however, override the individual responsibility of healthcare professionals to make decisions appropriate to the circumstances of the individual patient, in consultation with the patient and/or guardian or carer. Clinical policy coloured by a recent tragedy (in this instance a bleed) which leads to subsequently not using anticoagulation (and ignoring the entirety of the evidence base) can disadvantage subsequent patients who suffer thrombosis”

I think it’s incredibly hard to follow ‘guidelines’ when they are based on inadequate evidence. It’s too simplistic to think that blood clots are simple a failure of treatment. If the debate is to be balanced, we should also hear from people who have been harmed through overtreatment of risk factors for blood clots. Even then, we need better data to know how to place these, and it’s rather sad that, two sets of guidelines in, we still don’ t have decent mortality data in.

One Response to “DVT, uncertainty and hospitals”

  1. Dr Grumble July 17, 2011 at 4:40 pm #

    Where does the figure of 25,000 deaths come from? Could it be from a lobby group or is it from reliable data?