Discuss Upsides and Downsides of K99-R00 Program

Nov 07 2014 Published by under Uncategorized

Given all the data that I have recently posted about the K99-Roo program (transitions to academic positions, transitions to R01 funding, distributions across institutes and centers and academic institutions, publication patterns, comparison to F32 program), I would be interested in any reactions to this program. Should it be expanded, contracted, modified? Why? Are their other data or analyses that would be informative?

43 responses so far

  • DrugMonkey says:

    Expanded. While the F32 should be significantly contracted. Your data on transition to grant funded independent positions supports this.

  • eeke says:

    Is there any advantage to restricting K99's to the first 48 months of post-doctoral training? I would be in favor of expanding this time frame, given the difficulty of obtaining funding, etc, but I can also see why the NIH would be interested in keeping post-docs short.

    Is the F32 award directed only toward individuals who want to pursue an academic career? I would not be in favor of reducing this program if part of the intent is to ease the burden on the mentor. There are so many other careers out there - I maintain that it is extremely narrow minded to bring trainees into your lab with the expectation that they are there only to go into an academic career. Post-docs who wish to pursue careers in the private sector should not be excluded from fellowships.

    • DrugMonkey says:

      Why should the NIH pay for this private sector training?

      • eeke says:

        The post-doc is presumably working in an NIH funded lab. Why would their work be any less valuable than that of someone who aspires to a faculty position?

        And the private-sector training, as you say, means that this person gets a job. Works her way up through the ranks of a company or starts a company of her own, hires more people. They all pay taxes. It all goes back to the NIH. If anything, it's a great investment.

      • SaG says:

        Why should NIH only pay for public sector training?

        • DrugMonkey says:

          Because it is a public entity using public finals for public good. That is why.

          • SaG says:

            Then I guess NIH should stop awarding grants to private universities. That will free up lots of money. Sorry Scrips, CSHL, Harvard Etc....

          • DrugMonkey says:

            Do you really not understand the difference between not-for-profit and for-profit non-public enterprise?

          • SaG says:

            I understand perfectly. I just get frustrated with folks using simplistic bumper sticker/twitter phrases to discuss complicated topics. What you say sounds nice but is so much more complex. I didn't even mention SBIR grants. NIH funds have never been meant as a special source of funds to support the salaries of highly paid academic profs, deans and presidents. It is to fund the best science irrespective of who is doing it. Sometimes that means private companies who are the better stewards of public money for the public good. The assumption that public nonprofits have a higher moral authority than private for profit companies is just not true. I suggest you compare the salaries of your senior execs to those of similar size companies. Someone is profiting off of the NIH dollars you are bringing in even if it is not you.

    • Dave says:

      The NIH should not be in the business of funding the training of those who are heading to industry, or any other 'alt-career' for that matter. Pharma makes plenty of dosh, let them train their own people.

      • SaG says:

        Then we better cut back the number of Post-docs NIH funds by about 80%.

        • Dave says:

          There is no doubt that the number of post-docs and grad students needs to be reduced, as well there is a need to limit soft-money faculty. Some even think we should be forcing retirements. If all we had were post-docs on training grants, they would probably have better job prospects.

      • eeke says:

        Right. So then the NIH should consider getting rid of the T32 program while we're at it. Shouldn't take a chance on someone getting a Ph.D. only to go out into industry.

        • SaG says:

          Might as well start with grad students. Have them sign a form pledging not to go into industry once they get their Ph.D. The horror of it all!!!

          But, what about academics who start companies? Are they good or evil?

  • Dave says:

    The value of the K99 would diminish linearly with an increase K99 awards. There wont be enough faculty jobs to go around, surely. I say keep them as they are. They obviously have value, but that value is tied to its 'exclusivity'.

  • DJMH says:

    Even aside from Dave's point, if the program were expanded, you essentially push the t-t faculty decision back to the early/mid postdoc stage, because if there are a lot of K99ers out there, only K99ers will be viable on the job market. And there are a LOT of reasons why people might have middling records by early/mid stage postdoc, and yet turn out to be phenomenal PIs. So I don't like the idea of the award conferment becoming in effect, the sine qua non for faculty positions.

    I'm a recipient, and happy enough, but have plenty of colleagues who weren't going to fit the K99 mold for one reason or another but are now awesome faculty.

  • MorganPhD says:

    My concern for K99s in general is the focus on an idealized career trajectory. At this point, any hiccup can take you out of the running for an award, especially with the recent change in eligibility from 5 to 4 years post-PhD.

    A personal example, the first 2 years in my postdoc lab was unproductive (demonstrated that a previous SCN-level pub from a previous post doc was likely BS, although if true, had great potential). I regrouped and just had 2 papers accepted in "high impact" journals that would look great on a K99 app. Accepted 51 months after my PhD, 3 months later than the 48 month deadline for K99 applications under new rules. So I and many others who didn't take a straight path are out of luck, despite potentially being great candidates otherwise.

  • Morgan Price says:

    "The NIH should not be in the business of funding the training of those who are heading to industry, or any other 'alt-career' for that matter." -- If pharma and biotechs are too short-sighted to fund this training then NIH surely should, to accelerate the development of treatments and diagnostics. Also, I'm guessing that since postdocs are generally the most productive people in academic labs, F32s are a better deal for the NIH than K99s (although this would be very difficult to quantify).

    • SaG says:

      NIH funds people with fellowships not industry sectors. If a person goes on to an industry job great! An academic job cool! A government job sweet! Why should NIH care which sector they enter? how much of a real diff is there between a big biotech and a major research university anyway? We all know it's all about the benjis in both cases.

  • lurker says:

    Upsides: makes a postdoc candidate an easy no-brainer for schools to hire into Ass Prof, immediate indirects and hiring committees outsources the"risk" of hiring a dud because blame can be directed at study sections.

    Downsides: independent phase way too short and thus not enough balanced money, since noob Ass prof feels he has to spend it all right away. It also works against the Ass Prof because it gives ammo to study sections to claim insufficient productivity since this "ESI" has R00 but not enough pubs. R00 also disqualifies Ass Prof from new investigator bump at NSF. Some Ass Profs also get nickled/dimed on startup even though the stipulations of R00 is that institution can't do this. This so called "Pathway" only just shortens the period before the Ass Prof veers of the funding cliff like the rest of the cohort of Going-Up-for-tenure ass profs not getting their first R01 renewed. As you pointed out, JB, >80% of K99's transition to R00 (no-brainer for schools), but only half get some other grant after R00, just like the cohort trying to renew their 1st R01's. Crapshoot.

    Those without K99's and are getting hired as Ass Prof are able to get first R01 as true ESI and 5 years modular, but the R00 only has 3 years modular, and then half the cohort is seeing their funding evaporate, stuck in the rut sooner than those who had a 5 year modular shot. Only the lucky and the golden child R00's are getting their R01's, I bet within just the year the R00 started, so productivity excuse can't be applied. "Pathway to Independence" should be renamed "Delay of the Inevitable", where the finite (or shrinking) NIH trough can't feed so many piglets and the big old boars already there continue to feed and crowd everyone else out.

    It's not science anymore, it's thunderdome and rolling the dice (a trio, actually, not just a pair, since it takes only one SS member to "spread their score" and you're done). No wonder every state now is building more and more casinos. Bethesda has already turned into one, but no Hat Wearer (Rockey) would ever admit that....

  • datahound says:

    Thanks for all of the comments and keep 'em coming.

    The comments raise many important issues:

    (1) The initial impetus for the K99/R00 program was the desire (from the scientific community and the NIH) to shorten the time to independent funding by decreasing the length of postdoc time as well as the time in faculty positions without funding. This is the driver for shortening the eligibility time since some were concerned (me included) that the program could have the opposite effect on postdoc time, increasing the time of being "mentored" rather than shortening it and allowing younger scientists to go out on their own.

    (2) Design of these programs is always a balancing act between policy goals such as (1) above and making the programs sufficiently inclusive to avoid indirectly disadvantaging other groups. In the present highly constrained system, any program that provides an advantage to one group (e.g. those with K99 awards) automatically disadvantages another (those without).

    (3) NIH never intended that K99s would become a de facto pseudo-requirement for faculty positions. It would be interesting to know really how much this true. I know it certainly is a factor for many searches at many institutions, but I also know that other individuals are being hired.

    (4) The K99-R00 requires considerable program officer time at NIH. This is both a good thing and a challenge. In my experience, this helps the program officers understand between what is going on in the trenches for young investigators as they face the challenge of finding suitable positions that work for them professionally and personally. At the same time, this does limit the scale to some degree.

    (5) NIH-supported training does help those who go into many career paths and those in many career paths do contribute to the NIH mission of improving the health of the nation. Furthermore, as supported by considerable data, many scientists supported by NIH do go on to non-academic careers, either by choice or necessity. I think there is a substantial difference between having training programs that are agnostic with regard to career path versus have programs tailored specifically to prepare folks for non-academic careers.

  • Physician Scientist says:

    I am running a asst prof search for my top15-30 R1 med school department. We are looking for success in two places (grad school and post-doc) - meaning success follows the applicant and we need to see some evidence of granting ability (F32 minimum, CCFA, leukemia/lymphoma, ACS, K22, K01, K08, K99). Of our pool of about 100 this year, we have 3 K99ers. Only one is getting an interview. Other interviews (so far) are going to 2 K22s and a K01.

    My first post-doc has a K99 award now. This is beneficial as he has 3 first author papers in solid (but not CNS) journals and it allows him to spread his wings with his own money and follow some of his own ideas. It'll be sink or swim based on how well these play out.

    In short, at least in our search committee, we want to see evidence of being able to grant, but a K99 isn't sufficient or necessary for an interview.

    • SaG says:

      What else are you looking for in addition to a k99? Or, What was it about the two k99 applicants that you didn't call in for an interview that you didn't like?

      • Physician Scientist says:

        We want to see solid science in both grad school and post-doc. No one-hit wonders - no Kajagoogoo or Dexy's Midnight runners - and most (if not all) papers in society level journals or higher without 60 authors on them. They also have to prove they can write a grant (at least an F32 or fellowship with more senior career development being preferred).

        Some of the K99ers have a single CNS paper from a lab that should have a CNS paper. Some have multiple sub-society level journals and this makes us concerned as we're not a top 10 school with the attendant excellent post-doc population - we don't want someone who is satisfied shooting for PlosOne for instance.

        With this being said, this is our third year of hiring (we are looking to hire 1/2 a year x 4 years) and the applicant pool is 50% stronger this year.

        • SaG says:

          Hmm. I would have expected that every k99 awarded would be at the top of the list. What do you think the study section reviewers are missing? Or, is it that you have an extra year or two of data on which to base your Decisions?

          • Dave says:

            I met a current K99 awardee who was funded with an impact score of 30, and only because his PO pulled for him in council after his one and only first author paper was accepted in Nature Neuro before council. He has struggled since the award, even though he got a TT at a big school. I cannot see him getting an R-level grant and he seems out of his depth.

            There are a few one-hit wonders with K99s. There are even more from big labs that always publish at the top. It must be difficult to filter the K99 pool.

          • Dave says:

            .....without 60 authors on them

            Now, now. Did you see Datahounds analyses on this?

          • SaG says:

            I did. But, it didn't answer my question. What are the employers looking for that the study sections are not? Or is it, like most of science, just not practical to expect a group of reviewers to see into the future when all the can evaluate is the document in front of them?

          • K99er says:

            I was told a few things about why I was hired. First, my research fit amazingly well with my current department without overlapping with anyone else. Second, I had a follow-up paper accepted by the time I had my interview, i.e., I had been productive during the K99 phase and wasn't a one-hit wonder. Third, I was productive in grad school and my postdoc while working in small, non-famous labs. Therefore I was given more credit for my success than comparable applicants who came from high profile labs.

  • K99er says:

    My vote would be to get rid of K99s completely and increase the number of ESI R01s that are awarded. My university doesn't interview anyone without a K, and that eliminates far too much of the applicant pool. Second, schools shortchange K99 recipients with regard to startup funds (despite the NIH's best intentions to prevent this, there are just too many ways for it to happen, e.g., being required to immediately pay high salary %, or being required to buy into a departmental shared equipment plan, or not being given standard TA slots for students). Thus, the K99 is in a lot of cases not really giving the recipient a "leg up." Third, R01 study sections expect miracles from a K99 recipient and don't realize that R00s are a "total cost" award (meaning they are NOT equivalent to R01 funding).

    My experience has been that the R00 has been a detriment to me getting an R01. Despite excellent critiques, my last two R01 submissions have both missed the payline by 3%. The only advice my program officer gives is that, due to the funding climate, study section members are unlikely to give me another grant on the same topic until the R00 money is completely gone (even though I only have two months left and it's actually overspent). Meanwhile, this is exactly the time that my university's expectations for my salary contribution goes way up (year 4). Most others in my IC's K99 cohort have not fared any better. Only one has received an R01. She got this immediately upon starting her faculty job, and curiously her department chair was also the study section chair. Based on what's on NIH reporter, she appears to have relinquished the R00 funding as a stipulation for the R01.

    • drugmonkey says:

      So you think that people w/o an R00 never miss the payline by low single digit percentile points?

    • SaG says:

      " due to the funding climate, study section members are unlikely to give me another grant on the same topic"

      Umm, I believe it is the job of the Program Officer to "give" you a grant. The study section only gives you a score. The suggestion that reviewers can collude to so finely calibrate your score to get you close to but not over the funding line (which was probably unknown at the time the study section met) sounds unreasonable. That sounds like bad advice from your PO to me.

  • Susan says:

    What are the employers looking for that the study sections are not?

    Evidence of ability and creativity to think beyond the current set of Aims*. We, a low-ish R1, have declined to interview K99ers when their research statement did not indicate independent thought or direction or vision beyond the K99. An award based on mentor strength can be pretty obvious. Well polished, but obvious.

    I will relate two specific-ish examples from interviewees. One: 'My k99 was for the role of protein Y1a in little thing X. My next plan is to look at the role of Y1b in little thing X'. The other, we needed to explicitly ask during the job talk -- where do you see this in 5 or 10 years? The answer was something like, I expect these (current) aims to take a while, so that might be ongoing.

    *there are other issues of course, like research overlap and fit, but these are not addressed here.

    So my answer to Datahound is: decrease the role and importance of the mentor in originating the grant. Easier said than done, I understand, but important to do.

  • jmz4 says:

    As someone who just submitted one, I agree with Susan, and have a specific suggestion. There needs to be extra space, preferably an entire separate section, reserved for the more speculative, big picture ideas about where you think the science is going as it relates to the larger field.
    Right now that info is split between two sections. First, the career development goals sections, which also includes a lot of training info (who's lab you'll get expertise/equipment from, which grad students you might mentor, classes you'll take, etc). Second is in the actual research proposal, but there's not really a lot of room for that, since you're expected to fit 2-3 aims and there's a high bar for feasibility (hard to get big ticket stuff done in 2-4 years). Technically you could incorporate it into the significance, but that also takes away from your background, which you need to set-up the proposal.

    I'd like a one page section where you're specifically allowed to speculate on the direction of the field, how your current small project fits within it, and how it will act as a doorway to a substantive area of inquiry. Having a section like this will allow evaluation of the candidate's ability to read trends and synthesize data. Hopefully this will also give them space to separate their work from their mentor's work. If nothing else, if they're blindly parroting what a major name in the field is saying, then you know they've got nothing original.

  • Curiosity says:

    A couple of points:
    1. When starting a lab with more funding for salaries (say 3+PI for R00-funded and 1+PI for basic start-up w/o R00) the PI is taken out of the lab earlier with the R00 than without, since management is time consuming. This SLOWS them down to an unfortunate degree when the initial elbow grease of skilled PI-elbows is useful for kickstarting a research program. In other words, new PIs should be in the lab working to get things off the ground. With more money, they are hiring, managing, and doing other later-stage PIy things that are detrimental to starting things up quickly.
    2. As has been mentioned by other commenters, study sections expect R01-like renewals after R00 phase -- papers, etc. When starting a lab, there is a bottleneck of time and skill that is not ameliorated linearly with more money. At the beginning, money does not buy time. See comment 1.
    3. Datahound, any hint on how T&P committees are dealing with weirdo timelines for R00-R01 funding? It is my understanding that in the past, a PI would receive an R01 early on, then after renewal, the PI goes up for promotion. Is a single R01 good enough for promotion with an R00 but not without? This kind of information could change my opinion that K99/R00 mech is more harm than good at this point.... depending.

    • datahound says:

      I think T&P committees vary considerably about how they deal with this and other issues. At institutions with fixed tenure clocks (say 6 years), many investigators have not really had a chance to renew an R01 prior to coming up for promotion and this is made more likely, perhaps, by having an R00. The lack of renewal is handled different ways; at some institutions, the lack of renewal appears to be a strong negative whereas at others the institution appears to have adjusted expectations with the times. Others should weigh in with their experiences. In any case, it is best to ask questions at your institution to figure out what expectations (and therefore best strategies) are.

Leave a Reply