Welcome to DrPHealth

Please leave comments and stimulate dialogue. For those wanting a bit more privacy or information, email drphealth@gmail.com. Comments will be posted unless they promote specific products or services, or contain inappropriate material or wording. Twitter @drphealth.

Monday 18 July 2011

Evidence - putting numbers into action

Don't be shy, your comments are welcomed.

There is a small collection of individuals who are coming back to the site, I thank you.  Pass the word along. Email the address to a friend and ask them to read. 

The issue today is applying evidence.  If there was no evidence of activity on this blog, I'd be foolish to continue it. 

The gold standard for evidence has rapidly become the double blinded randomized control trial.  Neither the observer or the recipient are aware of who is getting the intervention and the results are only analyzed when the study is completed.  It is the closest thing to laboratory based research that we can get with humans.

The problem, is that not every intervention can be neatly wrapped up in a pill.  When you start intervening at the level of communities or populations, the techniques to amass and measure "evidence" of the effectiveness become increasingly challenging.  Those that review such evidence (check out Health Evidence Canada http://health-evidence.ca/  (within Canada)  or the Cochrance databases http://www.cochrane.org/ ) often dismiss such studies are methodologically flaws or weak design. 

When we look at the things that have improved the population's health, the quality of evidence is not good.   Who would question that reducing tobacco use rates had not been instrumental in saving millions, that water chlorination doesn't work, that safer vehicle and road design is not worth it?  

However, if you were to take these issues and do a science review, the conclusion may well be that there is insufficient evidence to demonstrate effectiveness. 

Insufficient evidence should not be interpreted as the evidence supports the contrary.  It is better stated that the research undertaken on the topic does not have the gold standard type of research that is used when making individual level decisions.  Ultimately, there is a bias against population level health interventions even when they can be more efficient, less costly and more effective.

In the meantime, we are channelling limited health resource away from population level activity because of "insufficient evidence".   We all suffer.  

No comments:

Post a Comment