Sunday, 10 November 2013

Clot filters

It seems I will never run out of examples of treatments that sound obviously effective in theory, fall into common use based on the strength of the biological mechanism, and yet they fail to show a significant benefit when put to the test. The story of the IVC filter is one of these.

Deep venous thrombosis (DVTs, clots) in the leg can dislodge, traveling up through the main vein in the pelvis and abdomen (inferior vena cava, IVC), through the heart and then embolise in the lungs (pulmonary embolus, PE), sometimes causing rapid death. An IVC filter is a wire cage placed in the IVC that snares clots that have broken free from the leg veins, before they can travel to the lungs. The device has been used for decades, but without much evidence of benefit, as this recent report tells us.

The main randomised trial that was performed on IVC filters did show a reduction in the rate of PEs (extending over 8 years), but this did not translate into a mortality benefit, and the rate of DVTs was higher in the IVC group. The study has been criticised (here), but these criticisms do not mean that the filters work.

The article also notes that the regulatory approval of these devices is less than ideal. Basically, the first IVC filter was approved by the FDA in 1976, and all the subsequent devices were approved on the basis of being similar to the first one. None required evidence of effectiveness; after all, they are not drugs, they are only devices that are permanently implanted into the main vein in the body.

IVC filters are also associated with significant risks of harm, such as infection, dislodgement, migration, misplacement, blockage of the IVC, and need for removal.

When the evidence is not clear for a particular treatment, we often find that the use of that treatment varies a lot between hospitals or groups of doctors. Although tens of thousands of these devices are inserted each year in the US alone, there is significant geographic variation in the rate of usage (here). They may have some place in the rare patient who cannot take anticoagulants, but even in that case, the benefit is still theoretical: based on the same, admittedly attractive, biological mechanism.

The bottom line

The story of the IVC filter provides a good example of something that makes sense in theory, but fails the acid test, and for which there is an evidence-practice mismatch. It is another medical treatment whose benefit has been overestimated and harms underestimated. If you disagree with me and consider the device to be effective in some cases, please do a study to prove it.

Addit 11 Nov 2013: A recent systematic review on the subject reported similar findings: IVC filters lower the rate of PE, but not overall mortality, and that more studies are needed.


  1. Having my leg crushed when assaulted by a truck, the treatment was clot busters and rat poison, apparently more successful than the basketball net discussed here.

  2. The data on anticoagulants is nt all that robust either!

    1. Point me in the right direction and I will look it up