Rep Harris proposal shows why NIH should actually look at the data

Oct 10 2014 Published by under Uncategorized

A Rockey post that I originally missed last spring tells us that the median age of first R01 for Early Stage Investigator (ESI) applicants is 39. This is of comfort to my longstanding inability to match the 42 year old number with my experience- I, of course, mentally ignore a key part of the NI distribution from which that median derives.

Three year difference from the 42 yo median that includes the ESIs, so we can assume the difference from New Investigators (NIs) is larger...5-6 yrs?

We don't know the relative numbers but a birdy suggested maybe 60/40 ESI/NI back when they invented ESI. I bet it is still within 60/40 to 40/60.

Policies, like Rep Harris' proposal, that don't recognize the 42 yo numerical target includes substantial numbers of established investigators previously funded with major, non-NIH grants will have odd consequences. Remember, the NIH itself evinced astonishing ignorance over this when they issued the ESI designation. How many years of the NI checkbox mysteriously not working was it? Ignorance of the data led, apparently, to an undesired outcome. Not learning this lesson lets proposals like Rep Harris' become more dangerous than it has to be. IMO.

The picture is further complicated because the nonESI NI population also contains some people who have not previously had major funding. These are probably people who simply aged past the ESI limit, which is based on time since degree rather than time since first faculty appointment (which would have been better). For Rep Harris' stated goal, perhaps these should be further separated from the NI pool.

I'm sure we can disagree over which slice of the NI pool should be helped and how to do so. But I think we can agree that we should base decision making on comprehensive data and not smear together qualitatively different PI types.

29 responses so far

  • Anon says:

    This is actually complicated a bit by the fact that you only have 10 years post terminal degree in which you are an ESI. So for someone who went straight to grad school from college, then got a PhD in 6 years, you would stop being an ESI around 38. Obviously some take longer for grad school, others take off time after college, and MDs are much older, but it's safe to say a large portion of ESIs age out in the late 30s. So what the NIH numbers say is that the median ESI grant age is close to the oldest you can possibly be while being an ESI (for a PhD who didn't take much time off after college). That's not that much more informative than the NI age.

  • Former Technician says:

    A member of our lab, research track, just received her first NIH grant as an NI. She has long since been beyond ESI as she is approaching 60. Kinda messes up the medians.

  • Anon says:

    This is actually complicated a bit by the fact that you only have 10 years post terminal degree in which you are an ESI. So for someone who went straight to grad school from college, then got a PhD in 6 years, you would stop being an ESI around 38. Obviously some take longer for grad school, others take off time after college, and MDs are much older, but it's safe to say a large portion of ESIs age out in the late 30s. So what the NIH numbers say is that the median ESI grant age is close to the oldest you can possibly be while being an ESI (for a PhD who didn't take much time off after college). That's not that much more informative than the NI age. A median ESI age at the upper limit of ESI ages suggests that there are a probably whole lot of PIs getting their first grants after losing ESI status.

  • Philapodia says:

    In response to Representative Harris' Op-Ed in the New York Times, we get this letter:

    "To the Editor:

    Re “Young, Brilliant and Underfunded” (Op-Ed, Oct. 3):

    In general Representative Andy Harris is correct: The bulk of RO1 grants (the most common) from the National Institutes of Health go to researchers “esteemed in their fields.”

    But this does not necessarily starve talented young scientists. Established scientists tend to have large laboratories and programs that enlist and cultivate brilliant, highly trained scientists in the field, sometimes as co-principal investigators.

    In any event, these burgeoning investigators have great freedom to bring their ideas to fruition and to make valuable contributions.

    Dr. Harris urges Congress to take various steps to change how the N.I.H. awards its grants. This would certainly inject political influences into the scientific endeavor. That is a bad idea.

    Chapel Hill, N.C., Oct. 3, 2014

    The writer is a retired researcher and was the recipient of a number of RO1 grants."

    So the moral here is that you should give money to the old guys who know what to do with it and the golden boys/girls in the old guys lab will possibly get the honor of being a co-pi under the old guy (and maybe do an experiment of their own!!!).

    A fine example of a progressive scientist.

  • Established PI says:

    Saw that letter in this morning's paper - groan. As for the statistics on ESIs, they are a little less grim, but let's not airbrush the problem. Yes, late-entering PIs who were previously richly supported or were working in funded research institutes or NIH intramural before getting their first R01 skew the figures, but my guess is that they represent a pretty small fraction of first-time R01s. What I am witnessing at my own institution are more and more junior PIs struggling longer and longer to win their first R01 and running low on startup money before they are able to get significant outside funding. This is happening at a place that is very highly ranked in overall NIH funding (single digits) . I would love to see statistics on time to first R01 among junior faculty (at universities and in non-clinical med school departments). This information is not be so easy to get at but would be highly informative.

  • drugmonkey says:

    How can the median move by 3 years from the ESI-included pool just by removing the ESIs if there are not large numbers of nonESI NIs? How does that work?

  • Philapodia says:

    Since ESI status only lasts 10 years after your terminal degree and post-docs are lasting longer than ever before (let's say 6 years total), ESI status tends to run out half-way through the tenure clock. and you're left as a NI. Perhaps the superstars that have truncated post-docs (3 years) prior to getting a TT position are skewing the ESI age because they can keep ESI and NI through tenure, whereas the rest tend to get funded as NI but not ESI? Since an R01 is essential for tenure at a lot of institutions, perhaps Asst Profs tend to work harder close to the end of their tenure clock to get an R01 (or finally figure out how to write a grant), whereas there is a lag time at the beginning of the clock to get the lab set up, figure out all the crap that goes into running the lab, teaching commitments, etc.

  • Established PI says:

    The most likely explanation is that they are largely NIs that aged out of ESI status before getting their first R01. Someone who did a 5-year postdoc and then took 6 years to get their first R01 is an NI, not an ESI. Given that many people are doing longer postdocs and second postdocs, this moves them just beyond ESI status. I haven't followed all the postings close enough to know whether there are data on how "new" the NIs are. It would be great to get numbers on this if they aren't already out there. I just find it hard to imagine that there are that many 50-ish former NIH intramural, HHMI and DOE grantees to significantly impact the average age (although I do not have any hard information on this so I can't swear I am right). And why would there be more of them now?

  • Philapodia says:

    @ Established PI

    We haven’t done any quantitative research on this trend, but I think we're probably right

  • Anonymous says:

    If you go straight to grad school from college, and take 6 years to get a PhD, your ESI status ends around age 38. Of course some don't go straight to grad school, some take longer to graduate, and MDs are much older ESIs. But the 39 year old figure says to me that an ESI who gets a grant is at the tail end of the ESI status. This also fits with my anecdotal experience. So a lot of people are probably getting their first grants as NIs and making the distinction for this analysis doesn't mean that much.

  • drugmonkey says:

    There are not available data on how "new" the remaining parts of the NI distribution are or are not.

    It is very clear from the old "surprise" about ESIs that there were at one point a LOT of highly experienced investigators getting the "NI" grants. This, btw, was readily apparent to anyone who served 1-2 rounds of study section back in the pre-ESI, NI-checkbox days. Given that the 42 years old number has only flatlined following the ESI affirmative action policy, I think it very likely that established NIs are still a big part of the first-timer distribution.

  • E rook says:

    I'm curious what the group believes a target age/distribution for first R01 should be. Should it be different for women vs men because women are more likely to take maternity leave? (and pause the tenure clock per whatever policy their institute allows). What is a good age at first R01 and why? I am asking because I genuinely don't have the answer and I want to read some arguments. Seems people agree it should be lower or at least not be moving up. I would say MY Current age, but obviously I'm biased.

  • DrugMonkey says:

    36. IQR 33-37

  • DrugMonkey says:

    Absolutely no difference in male/female ages either.

  • rxnm says:

    Age is irrelevant...we should be looking at time from eligibility/independence.

  • DrugMonkey says:

    For some goals, absolutely. But age is needed to combat certain bad trends in the business of science.

  • K99er says:

    Totally hypothetical question here, but let's say you were a very young PI who is trying hard to get a first R01. Is there a way to use age to your advantage? Like maybe tactfully dropping hints to your program officer about just how much being chosen for select pay would help that institute's age statistics?

  • dsks says:

    K99er, what you're suggesting would be illegal on the basis of age-discrimination. This is why you often see funding application guidelines along the lines of, "only folk x number of years from graduation/first faculty position/etc may apply", but never, "Only folk aged x to y may apply".

  • Susan says:

    The fact that the median is 39 for ESI reiterates the problem: even with the 'handicap', the average ESIs is only finally getting an R01 at their last eligibility .... it's really pretty stunning that the median hits the end of the eligibility range.

  • Ola says:

    Worst example of abuse I saw was an NIH intramural scientist, early 50s with a big lab, moved out to academia as a full professor with an endowed chair and promptly submitted a shit-ton of R01 apps claiming newbie status. Technically speaking, he'd never had an R grant before, but it left many of us on study section with a bad aftertaste. I've also seen one example of a grey-beard forced to retire at 65 in his home country in Europe, got a very sweet recruitment deal to come the US, then claimed NI status on his R01 submissions.

    I'm guessing these are not common, and so don't affect the overall balance tremendously. The question is how to write the rules to stop abuse by big dogs, while still allowing the mid-50s "lifer" R.A.P. to get that long-coveted first grant?

  • Ola says:

    Oops - I guess I should read DM's post in more detail first. He's essentially saying that the oldie non ESI but still NI crowd DOES affect the numbers.

  • jojo says:

    I'm currently trying to get my >60 year old PI to write the lab's first NIH grant (he's always been NSF funded, but the direction of the research has been moving into NIGMS territory in the last decade). I can't imagine he would submit intentionally with New Investigator status, but technically it would be his first R grant, so maybe he'd be lumped in?

  • Eli Rabett says:

    NIH grants require an entirely different mindset than writing NSF grants, and the grunt work (biosketch, budget, etc) is a lot different to. In that sense notifying the study section that this is someone coming in from the cold makes sense, explaining small lacuna in the bureaucratic aspects of a proposal, which, without NI designation, could be notified in the text by a sentence such as "this proposal has developed from a series of NSF/EU projects which have moved in the direction of . . . "

  • drugmonkey says:

    To be clear, I'm okay with there being a tiny bit of extra incentive or skids-greasing for established-NIs. I'm okay with that.

    The problem is that the NIH was trying, as they perennially claim, to help the newly-transitioned scientists. And the NI apps that did well were predominantly the established-but-other-funded PI. They ignored the outcome of their policy change and surprise, surprise got an outcome inconsistent with the stated* goal.

    *of course they may have actually had a different real goal under the stated goal. I get this.

  • Comradde PhysioProffe says:

    This concept of "abuse" of NI status is a total red herring. First, the idea that it garners gentler treatment by study sections is totally false, both anecdotally and statistically. Second, many (most?) ICs apply a relaxed payline only to ESIs and not to NIs. Third, program staff *always* has the discretion to fund or not fund any grants they choose.

  • drugmonkey says:

    dude, we're talking about the interval of the NI checkbox prior to the invention of ESI status.

    The point is not whether "NI"s were treated better/worse than established investigators but rather how different qualitative types of NI-qualified were treated relative to each other.

  • drugmonkey says:

    And both "anecdotally and statistically" the newly-transitioned were not receiving the boost that the NI checkbox (and associated review policies) was intended to produce. That is why ESI was invented and how it was justified.

  • girlparts says:

    If an investigator gets a K99 at the end of the (old) eligibility window- 5 years postgrad, and does the 2 years K99 + 3 years R00, s/he would no longer be an ESI at the end of it. Ditto someone who postdoc'd for 6 or 7 years, works off a foundation grant, or an R21 for a few years then finally lands an R01. There are tons of non-ESI, NI out there who are truly newbies.

  • drugmonkey says:

    I reiterate. NIH invented ESI as a way to get funding preferentially to [some subset of] "truly newbies". They did this because they realized a lot of "NI" awards were going to not-newbies, not I would argue, because they were so keen to screw over "truly newbies" that happened to be older than desired.

    They were driven in this by the 42-years-to-first-R01 soundbite/graphbite.

    They thought, I surmise, that if they pushed the funding to the "truly newbies" for a few years it would put a big counterweight on the years-to-first-R01.

    Obviously the aggregate effect of the ESI push on the total NI population has been only to stabilize the 42 year number.

    Looking only at the ESIs part of the NI distribution moves the age by only 3 years, down to 39.

    So while I agree with you that there are plenty of "truly newbies" in the remaining non-ESI part of the NI pool, the numbers suggest that there are a whole lot of non-newbies in there as well.

Leave a Reply