Friday, September 12, 2008

Keeping Things in Perspective

The New York Times has an interesting piece in their Business section today on implantable cardiac defibrillators (ICDs), but paints a remarkably pessimistic view of the future market for the devices. While I do not argue with the fact that ICD implants nationally have declined, I think it's relatively important to compare the cost of ICDs relative to other therapies we dispense in medicine relative to their life-saving value to our healthcare system.

Although admittedly assembled by Medtronic (who has a vested interest in promoting ICD therapy and yes, I suppose I do, too), there are data of the costs of ICDs relative to other preventative therapies doctors provide as "prevention." Here's a slide from them that adds perspective:

Click image to enlarge
Additionally, although the SCD-HeFT trial demonstrated that 15 patients were needed to be treated before a life was saved, other trials demonstrated a significantly better odds at saving a life:

Click image to enlarge

MUSTT @ 5 years from Kaplan Meier (KM) curve: 55%-24%, NNT=3 N Engl J Med 1999;341:1882-1890
MADIT @ average follow-up of 2.4 yrs, crude mortality rate: 39%-16%, NNT=4 N Engl J Med. 1996;335:1933-1940
MADIT-II @ 3 years from KM curve: 31%-22%, NNT=11 N Engl J Med. 2002;346:877-883
AVID @ 3 years from the KM curve: 36%-25%, NNT=9 N Engl J Med. 1997;337:1576-1583
SAVE (captopril, an ACE inhibitor) crude rate with average follow-up of 42 months: 25%-20%, NNT=20 N Engl J Med, 1992; 327:669-677.
Merit-HF (metoprolol, a BB in HF patients) @ 1 year from KM curve: 11%-7.2%, NNT 26 LANCET 1999; 353:2001-07.
4S (simvastatin) @ 6 years from KM curve: 12.3%-8.7%, NNT=28 LANCET 1994; 344: 1383-1389.
Amiodarone Meta-analysis of 15 trial @ average follow-up of 2 years: 19.2%-16.5%, NNT=37 Circulation, 1997; 96: 2823-2829.

Finally, in terms of cost of other preventative therapies that our society has deemed useful to save a life, I always appreciated this view:

Click image to enlarge
Just trying to keep it in perspective.



kyaw said...

It seems to me that if NNT is assumed to be 10, then cost per life saved is 500,000 (the procedure is said to cost 50,000.But how do you calculate cost per life year saved? Or Am I completely wrong?

Anonymous said...

But most pts with ICDs will have the ICD, be on a statin, an ARB or ACE-I, possibly a CCB and a PPI.
While we call it prevention for SCD, these pts already have serious health issues and use up a lot of resources before we even give then ICDs.