A snapshot of the biomedical research funding in Australia in 2015
Two weeks ago the National Health and Medical Research Council (NHMRC), the main biomedical funding agency in Australia released the 2015 funding outcome round for projects programs (regular project grants, new investigator grants, Development grants …) or people support for Early career researchers (Early career fellowships ECF), mid (Career development fellowships CDF) or late career researchers (Research fellowships RF) https://www.nhmrc.gov.au/grants-funding/outcomes-funding-rounds. Before going more into details, there are few comments I would like to make.
Firstly the results were anounced individually to all NHRMC applicants over a month ago and the embargo was just lifted 2 weeks after the announcement to all applicants by email. For the NHMRC special grant initiative for dementia, the embargo lasted for almost 3 months! While this is understandable the release of the results depends on the availability and the diary of the minister of health in Australia, I don’t believe that a lengthy embargo is acceptable. From previous years the emborgo lasted at most only 48 hours.
Secondly for non australian researchers, I would like to biefly summarise the peer-review process for NHMRC project grants. If you already know the process, please skip this section. All the details could be found on the NHRMC website (https://www.nhmrc.gov.au/print/book/export/html/51612). Once the grant is submitted, every proposal is assigned to external reviewers and to a Grant review Panel (GRP) (the equivalent of a study section in the NIH system) from the NHMRC assigners academy which is composed from leading experts in their fields (https://www.nhmrc.gov.au/grants-funding/peer-review/nhmrc-assigners-academy). Usually a proposal is assigned to 2 external reviewers, a secondary and a primary spokesperson. The role of the primary and secondary spokerspersons is to ensure the integrity of the review process, to score the application, to scrutinise the budget and to discuss the application at the grant review panel. Once the application is reviewed, the grant proposal is scored from the primary and secondary spokesperson and the reviewers’ comments (primary spokesperson as well as the 2 external reviewers) are sent to the applicants. The applicant is invited to rebut the comments. Once the rebuttal from the applicant is sent, the primary and secondary spokespersons correct or not their scoring and half of the applications of a GRP won’t go to the discussion panel and will be classified as not for further consideration (NFFC) in a same way as the NIH does. Once the proposal is sufficiently ranked to be assessed at the GRP, the proposal is discussed at the GRP. The primary spokesperson will lead the discussion, will summarise the strong points and the weaknesses of the grant and the secondary spokesperson will add further comments. Then the proposal will be discussed with other panel members and the proposal is ranked. A video of the all process is summarized on the NHMRC website (https://www.nhmrc.gov.au/grants-funding/peer-review/nhmrc-grant-review-panels-induction). Importanly the scoring system is as following: Scientific quality of the proposal contributes to 50% of the score while the significance of expected outcomes/innovation is 25% and the team quality and capability relative to opportunity is scored 25%.
The NHRMC projects grant results outcome
This year the success for NHMRC project grants was at its lowest (13.7 %, see Figure 1) compared to previous years (see previous years). This decline in funding is steady and not isolated to Australia. The same trend is observed for France, Canada or the USA. While worse was anticipated from the NHRMC (see http://www.theaustralian.com.au/higher-education/medical-research-grants-dry-up/story-e6frgcjx-1227603768010). The overall funding rate in Australia for biomedical research is dismal and clearly not sustainable for a high quality biomedical research in Australia. I will come back to this later. However, I must say many will be out of jobs at the end of the year. This is a real lost of talents in the research community and this is creating many difficult situations.
On the fellowship side, The results are as dismal as the NHMRC project grants with a steady decline in funding rate for senior fellowships (RF) as well as the mid career fellowships (CDF) while clinical fellowships or Early career fellowships are doing OK (see Figure 2)
In absolute numbers, the results are telling (Figure 3). This decline in funding is clearly associated with an increase in fellowship submission since 2012–2013 while the number of allocated fellowships remains relatively constant or slightly declining since 2010. This is particularly true for the CDF and RF and can be explained by the non retirement of “baby boomers” senior academics and the large cohort of finishing postdocs.
NHMRC project grants: Demographic and gender distribution
Let’s go back to the project grants now. I have plotted the distribution of the funded grants for principal investigator (CIA) (Figure 4) from their academic qualifications as well as the gender. Sadly I don’t have the demographic and gender distribution for the total number of submitted grants as it requires a Freedom of Information request to be able to have access to these data. Therefore there are limitations to this.
As I’ve noticed before for the Australian Research Council (https://medium.com/@GaetanBurgio/a-note-on-the-australian-research-council-arc-discovery-program-e459151cd570#.alosm7y50) a large proportion of the funded grants (48%) went to senior academics (Professors; academic Level E) while the other funded grants went equaly to junior mid-career scientists (Level academic B/C) and mid-career to senior scientists (Associate Professor; Level academic D). Interestingly I found many senior academics getting multiple grant funded with up to 3 funded projects (14 occurrences) as well as program grants holder funded with additional grants projects (over 10 projects). While the track record only counts for 25% in the grant assessment, the result shows that more weight is given to the track record as it should be in the grant review process. I will come back to this later on.
Regarding the gender, the results are dismal for females academic with only 25% of the funded projects were leaded from CIA females. The gender repartition is as following for each academic level: 30% for Level B/C, 35% for Level D and 21% for Level E. Sadly only 6 out of 33 (18%) new investigator grants were funded to female academics. Interestingly the NHRMC has established gender equity policies since 2013–14 at institutional level (see https://www.nhmrc.gov.au/research/women-health-science/institutional-gender-equity-policies) as well for the assessment of project grant (see the career disruption section https://www.nhmrc.gov.au/grants-funding/apply-funding/project-grants). However the results are fairly disappointing (see for more https://www.nhmrc.gov.au/2015-funding-outcomes-gender-summary-findings) which means that the females academics are clearly disadvantaged to compete for NHMRC project grants.
Repartition of the funded NHMRC project grants per broad research areas
One interesting question from this decline in project grant funding rate is which broad area of research amongst basic science, clinical science, health services and public health lost the most grants and funding allocation. I’ve plotted the total funding as well the average funding per project for each broad area of research (Figure 5)
Interestingly while the total funding is declining for basic science and not for the other areas of research, the average money allocation per broad area of research is increasing for all areas except the health services. Additionally the increase allocation per grant is fairly modest for basic science compared to public health or clinical science while salary cost is increasing. If we look at the proportion of funded grants per broad research areas, there is almost no changes since 2010. Although, looking at the absolute numbers, the fall is fairly substantial for basic science (Figure 6).
Together this means that the number of funded grant per broad research area remains the same in proportion, but not in absolute numbers. This might be explained by the fact that is easier to justify for 5 years funding to conduct longitudinal studies as well as clinical studies than a basic research project where it is more challenging to predict and justify research experiments 5 years in advance. Therefore basic science projects are the biggest losers of the 5 years implementation of the project grants as well as the increase in research cost in average per grant.
Are researchers in top institutes in Australia have a better to get funded?
The biomedical research in Australia is performed by universities as well as the CSIRO and medical research institutes such as the Walter and Eliza Hall Institute for medical research at Melbourne (WEHI http://www.wehi.edu.au) as well as the Garvan institute at Sydney (http://www.garvan.org.au) or the Queensland Institute for Medical research Berghofer (http://www.qimrberghofer.edu.au). These independant research institutes usually highly successful for NHRMC project grants as well for fellowships. In a context of project grants, I’ve looked at the performance of these 3 top institutes in Australia (Figure 7).
What is absolutely remarkable is while the number of grants per institution has considerably increased (almost doubled) since 2012 following the general trend, the number of allocated grants remains the same. As a result, a substantial fall per institution is obsverved at a greater rate than the general trend (except for QIMR at Brisbane), which is worrying given that these institutions relies heavily on NHMRC grants and fellowships to function.
So, What to make of this?
There are numbers of lessons to learn for this and previous NHRMC grant project funding round as well as the fellowships:
The funding is undoubtely tight for medical researchers in Australia as well as elsewhere except Switzerland, Germany or Japan. As a results, the number of applications for fellowships as well as NHMRC project grants has considerably increased from the last 2–3 years, roughly from the end of the stimulus package post GFC from the Rudd government. This 13.7% success rate is in fact a little bit more than anticipated due to the fact that many grants were not funded to 5 years, especially for basic science and this has artificially increased the funding rate. Worryingly with such a low funding rate at an unsustainable rate for biomedical researchers in Australia, the allocation of the grants becomes more a lottery rather than a ranking based on the real innovative aspect of the proposal.
Secondly the funded project grants are in majority allocated to Level E academics with only a proportion to junior scientists and females academics got funded. This means to me that firstly the NHRMC grant review process is conservative and risk adverse allocating funds preferentially to those with good publications records in Cell, Nature or Science journals or other high impact factor journals. While officially Impact Factors as well as H index metrics are not part of the grant assessment, this is practice not the case at all. Secondly as more funds are allocated towards early career researchers (fellowships) as well as senior researchers (project grants and fellowships), mid-career researchers are right in a funding hole and it becomes very challenging for a mid career researcher to gain research independence in Australia. Thirdly, given the weight on the track record of the chief investigators, especially the CIA, it encourages to nominate as a CIA senior researchers that has not designed and perform the experiments. The actual system favors ghostwriting applications from junior academics.
Thirdly the basic science research in Australia is the bigest loser of the decline in funds these last few years. This is a direct consequence of the extension from 3 to 5 years grants as per the McKeon review recommendation page 151 (http://www.mckeonreview.org.au/downloads/Strategic_Review_of_Health_and_Medical_Research_Feb_2013-Final_Report.pdf) without increasing the total funding allocation for NHRMC dedicated to standard project grants or fellowships. This had a catastrophic consequence to Australian researchers, especially in the basic science area of research and many early-mid career researchers will leave Australia or simply science. Therefore there will be a massive gap in the research workforce in the next 5 years as many of Level E professors will retire.
Fourthly, the decline in funding from institute is at a rapid pace and there is nowadays no advantage to be a researchers in a leading institution to attract NHMRC project grants. This might have some consequences on the quality of the research conducted in these institutions.
Where to go from there?
There is no doubt the funding lansdcape in Australia is fairly tight and many ways to address the situation have been proposed
The Medical Future research fund (http://www.health.gov.au/internet/main/publishing.nsf/Content/mrff) is a way to increase funds for medical research in Australia as well as a way to offer more stability in this tight funding landscape. Basically the funds will be originated from health savings. While I won’t go into details here on this initiative, there are many unknowns on how the money will be spent, who will receive the money and how this initiative will be complementary to the NHRMC.
There are voices in Australia advocating a change in the grant allocation attribution on researchers rather than projects (see: https://www.mja.com.au/insight/2015/32/are-we-killing-innovation or https://www.mja.com.au/insight/2015/41/john-mattick-be-brave). In fact as the Australian system of the evaluation of the grants (ARC and NHMRC) is already more person rather than project centric. We are already funding people rather than projects and this in anyway fosters innovation and favour well established investigators. Additionally, this idea has been implemented in Canada from the Canadian Institute for Health Research (CIHR) and had dramatic consequences on early-mid career researchers in Canada (see this post from Michael Hendricks from McGill on the foundation grant scheme https://medium.com/@MHendr1cks/the-impact-of-cihr-reforms-on-early-career-biomedical-scientists-in-canada-6aef2f14bb12#.a9mrttyel). In short this will in its implementation foster more conservatism, will create a researcher elite and ultimately will kill any innovation in Australia biomedical research.
There are many discussions on the NHMRC review process and how writing grants is a waste of time (see http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3664356/ or http://bmjopen.bmj.com/content/5/1/e006912.full). Recently a trial is being conducted on the development grant scheme consisting in the increase of external reviewers from 2 to 5 to improve the quality of the review and the reproducibility of the scoring (see http://www.theaustralian.com.au/higher-education/nhmrc-to-test-peer-review-of-research-project-grants/story-e6frgcjx-1227612838745) and ditching grant rebuttal (see http://www.theaustralian.com.au/higher-education/nhmrc-trials-ditching-grant-rebuttal/story-e6frgcjx-1227615815470) based on the success of the Canadian system. While this is a good idea on a paper, the reality is very different. The CIHR experience shows that 1) adding more reviewers doesn’t improve the reproducibility of the scoring. In Australia a discovery projects for the Australian Research Council coud get reviewed from up to 5 referees and the scoring is still highly inconsistent 2) The pool of reviewers is relatively small and the referees will be overloaded with grants to review and this will exacerbate the “reviewer fatigue” 3) This fascination for the CIHR in Canada is beyond belief. The reform of the CIHR is an absolute catastrophe and has frustrated many canadian researchers (see Jim Woodgett post https://medium.com/@jwoodgett/a-decade-of-mishandling-science-in-canada-d0b8f73e432e#.qua4xojpk)
How to improve the NHMRC grant review process?
The NHMRC grant review process put too much enphasis on the track record of the Chief investigators, especially the CIA and not enough on the project. Additionally the power of the primary and secondary spokesperson in the review process is too strong. The scoring should take put more weight on the project ~60% and the innovation ~30% and less on the track record of the investigators ~10%. I don’t believe that increasing the number of reviewers would improve the grant assessment, it will instead worsen the situation as the inconsistancy of the referees will remain. The GRPs should get renewed more often than every 3 years and should include a reasonable proportion of females and junior academics (gender equity and at least 30–40% of junior academics). The simplification of the grant application is to me a way to divert from the real issues. Compared for instance to the NIH, CIHR or the ARC, the NHMRC grant application system is relatively simple and I don’t believe this is such an issue.
From this round and previous rounds of funding. I’ve seen too many senior scientists got multiple funding such as 4 to 6 project grants as CIA or more frequently a NHMRC program + project grants. While there is no doubt that senior scientists are an important part of the mentoring process for early-mid career researchers, one way to foster talents and allows the researchers to gain their independence, the number of project grants should be capped to 4 or program grant holders won’t be able to apply for a project. There are too many researchers, generally at senior level, monopolizing research funds. I would also propose to cap the age to apply for a project grant as well as a fellowship. This will allow senior researchers to mentor junior scientists, to avoid ghostwriting applications and break any conservatism. As many senior researchers are males, this will allow for more early-mid career researchers females to apply as a CIA as the pool of female at academic level E is relatively small. Finally this will favour early-mid career researchers to establish their own research program.
The shift from 3 to 5 years while the funding allocation for project grants is stagnant has been an absolute catastrophe for basic science. This should be ablolished and we should go back to 3 years funding as long as the budget is flat.
Finally the universities and research institute have to take their responsibilities. The NHMRC cannot assure tenure as well as grant funding to all biomedical researchers in Australia. The NHMRC cannot do everything for biomedical research in our country. Institutions must 1) offer a sustainable way to fund researchers in Australia with institutional funding 2) cap the number of grants and fellowships to submit to the NHRMC and establish a internal peer-review system to select the proposals that are potentially fundable. Many grants were submitted and these were no near to get funded. This is everyone’s waste of time! If we keep the trends in sending more and more applications as long as the NHMRC budget remains the same (which will be the case in 2016), the success rate will fall again and this will exhaust everyone who is part of the writing + review process. The data on the independant institutes shows that an increase in grant submission is definitively not correlated to a better success for NHRMC. In fact this is the exact opposite.
In conclusion:
The funding is undoubtedly tight for all biomedical researchers in Australia and this will persist for several years. The Australian research community needs to break the existing conservatism and favour early-mid career researchers to establish their research programs and to keep funding basic science. This is a key to innovation in biomedical research.
