Medical research can take on average 17 years to translate into practice. The time from discovering the link between smoking and ill health to implementing a TV ban on smoking advertising was 54 years. Cognitive behavioural therapy for schizophrenia has taken 48 years to gain a foothold in clinical practice, whereas a major drug advance, olanzapine, “only” took 20 years.
This is because in today’s business climate of ‘time is money’ the purchasing power of research funding is decreasing exponentially while the burden of proof is simultaneously increasing – after the mid-2000s National Institutes of Health research funding declined nearly 2% per year in real terms representing a 13% decrease in NIH purchasing power.
No one likes negative results, especially the funders of research.
When long-term funding decreases, so does the type of research that leads to new discoveries. Reliance on aggregating old results becomes more important, as we have seen from the central role the Cochrane collaboration has played in reviewing research
We’re just four months into the year and already the Cochrane database holds more than 2000 new articles highlighting “no difference” across a range of practices from widely diverse clinical settings such as laparoscopic cholecystectomy and pressure ulcers.
No one likes negative results (especially the funders of research). When the negative research juggernaut supports a “no difference” argument rather than a “better” and “worse” argument the burden of proof is higher, particularly if there is little difference between interventions. This kind of research will never eventuate in the current economic climate.
It is much quicker and good enough now to prove that something new is no worse – “non-inferiority” trials show that newer interventions “are not inferior to an acceptable extent” to any existing one. Over 1700 of these trials are mentioned in Google Scholar this year.
What can we do to move the medical research juggernaut more into discovery mode and speed up the process?
The IT industry may have the answer. We can learn a lot from information doctors about how to aggregate, validate and effect change through modern technology. For example, using content-mining software, as I have done with Google Scholar for this blog, we could unlock a rich source of valid data in case reports. Or we could predict outcomes of new advances by using the newer gaming technologies rather than normal statistics.
Medical research used to be a cornucopia of hope and success for clinical practice. Now the business model is turning it into a black hole of negative energy. We need medical research to continue to help us find answers, especially to the complex health issues that now occupy most of the time available in our clinical services.