A recent posting on Derek Lowe’s “In the Pipeline” presents a dismal view of the returns on R&D in Pharma. BCG had included a graph (from Bernstein Research) of the trend of new molecular entities (“nme”) per $B spent over the past six decades. Supposedly adjusted for inflation, the graph suggests that fewer than one nme is being approved per $B spent on R&D in the industry overall. This backward focus may give a huge misimpression of what is happening now, including some cause for optimism (?)
There are several problems with this analysis, not the least of which infers that we were creating around 50 “nme’s” per year, per $B back in 1950. But the biggest problem is using “new molecular entities” as the key metric for our industry. I have been looking at the effects of improvements on how we are doing drug R&D for several years, and strongly recommend that firms focus on “New MEDICAL Entities” (“NME’s“) instead of nme’s. Approved NME’s in theory offer greater value to society and are a better signal of R&D productivity, and future returns. More on the nme versus NME arguments at the bottom. However the Lowe posting caused me to look at my data from that aggregate perspective to surprising results!
What could be the reasons for this reversal? Is it financially significant?
This chart reflects the yearly NME’s approved by the FDA, versus PhRMA-cited spending. It shockingly shows an inflection point in 2006, and recent information (like the projected drop of 3% in overall spending in 2010), suggests the industry is actually improving its R&D returns and innovation!
First, some explanations on the analysis then the implications:
The denominator is PhRMA spending cited in the 2006 Congressional Budget Office report on the industry. However since the results of NME approvals depend on actions and spending that occurs mostly five years earlier, we apply a lag to better connect investment to results. For instance, the five NME’s approved in 2005 are divided by the industry R&D spend of $21.7B from the year 2000. Scale and slope change slightly if one does not lag the data, but the trends and the inflection point remain the same.
The NME data is from the FDA’s public database, culled of very similar follow-ons, salts and obvious use approvals (versus the initial first in class therapeutic form.) Details on this approach and its rationale are here as part of a larger analysis to reveal what R&D improvement initiatives are doing to R&D returns on investment (initially impact of mergers, later on outsourcing, six-sigma and IT tools.)
Implications: I would like readers to think and comment on what may have caused the above reversal. Are we getting smarter – with a five-year lag that prevented us from seeing this? Has the change been hidden underneath “me-too” drug approvals – and has the FDA been tightening the screws on them so a greater proportion of true NME innovation is “flourishing” (maybe)? Are the investments in outsourcing, shifts from small molecules to biotech, six-sigma, translational medicine, … at work finally? What are your thoughts?
Thinking forward: As I noted the Bernstein graph and so much of our gloom is based on retrospective analysis. Ours is a complex industry that tries to serve society, is regulated, takes a heck of a long time and lots of money. Resulting in so much historic data that it affects our thinking. We often fail to look at the bottom line, or to really look forward. This uptick in NME productivity MAY NOT reflect a large increase in the value (and resulting revenues and profits) that society will be happy or willing to pay. That connection or proof is still to be made. However, this could be one of the first signals of a turnaround that will have a positive effect on BioPharma. Again, comments? Ideas supporting or challenging that notion?
You can search your own and I encourage others to read the actual articles, but some highlights year to year: 
