Academic journal article International Journal of Child and Adolescent Health

Evidence-Based Medicine at a Crossroad: Is the Therapeutic Value (TV) of Complementary and Alternative Medicine (CAM) Really Larger Than the Therapeutic Value of Pharmaceutical Drugs?

Academic journal article International Journal of Child and Adolescent Health

Evidence-Based Medicine at a Crossroad: Is the Therapeutic Value (TV) of Complementary and Alternative Medicine (CAM) Really Larger Than the Therapeutic Value of Pharmaceutical Drugs?

Article excerpt

More and more often pharmaceutical drugs are found to be inefficient and harmful (1,2). In a recent article in BMJ we have suggested that the ratio of benefit/harm should be expressed as the therapeutic value (TV=NNHtotal/NNT), where NNT is the Number Needed to Treat and NNHtotal is the total Number Needed to treat to Harm, i.e., the likelihood for experiencing one of the drug's adverse effects (3).

One of the arguments for the clinical relevance of this formula is that as a rule both positive and negative effects are counted only when they are clinically relevant. Another argument is that in reality we do not have any other measure, so we need to use what we have. The measure TV is still only a rough indicator of benefit vs. harm.

What genuinely surprised us when we looked at the many Cochrane Reviews of single drugs was that NNT often is 20-50, while NNHtotal often is 1-3 (3,4). Therapeutic value of pharmaceutical drugs is thus normally between 3/20 and 1/50 (TV= NNHtotal/NNT= 0.02 - 0.67). When we looked at Cochrane reviews for whole drug groups (like antidepressant groups (outcome: less depressed), antipsychotic drugs (outcome: mental state improved) and anticancer chemotherapy (outcomes: survival, quality of life)), we found that the therapeutic effect often completely disappeared (4,5), while NNTtotal raised to about 1. If these studies are to be believed, drugs from these groups are - on average - harmful to most patients and helpful to very few at best, with the studies not even able to tell us which of the drugs might be useful. Evidence-based drug medicine seems to be in severe trouble; BMJ deputy editor Trish Groves has called this a sad state of affairs for "evidence-debased medicine" (1).

When it comes to alternative and holistic medicine (CAM) the development has been the opposite. From very little evidence, dozens of different CAM treatments for almost any clinical condition have now been documented in hundreds of Cochrane reviews (6). More than 38% of the Cochrane reviews showed a positive or possibly positive (12.4%) effect. A typical effect level was NNT=2-30, while harm is extremely rare in CAM (NNHtotal>1,000) (7). An NCCAM report on therapeutic massage indicated a NNTtotal [asymptotically =] 1,000,000 (8).

We thus found the TV for CAM in general to be TV=NNHtotal/NNT=1,000/30 - 1,000/2 = 33 - 500. For some of the most effective CAM types (mindbody medicine) the TV was calculated to over 9000. (9). As NNT for therapeutic massage often is 5-20 in the Cochrane reviews, TVmassage=50,000-200,000! No wonder Hippocrates said "first do no harm," using mainly therapeutic talk and therapeutic massage.

The real surprise comes when we compare the TV for pharmaceutical drugs (TV=0.02-0.67) with the TV of evidence-based CAM (TV=33-500). We found in the two BMJ papers (3,7) that the TV of CAM is 50-25.000 times larger than the TV of pharmaceutical drugs. If we take the lowest numbers of TV (NNHtotal/NNT) from Cochrane studies of whole drug groups (4,5) and compare these with the similar numbers from the best types of CAM (for the same clinical conditions) (6,7), we find this number going up to TVCAM/TVdrugs=0.01:9,000=900,000!

The question we want to pose to the medical community is this: Is the best and most efficient type of CAM really a little million times better than drugs for clinical conditions like depression and schizophrenia, cancer and coronary heart disease (10)? …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.