Bench research as an MD only?

This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.

reese07

Full Member
15+ Year Member
Joined
Feb 4, 2007
Messages
516
Reaction score
18
Let's say that I am set to just pursue just my MD. What pathways would I take if I ultimately want to do bench research with some clinical work on the side, similar to the 70/30 that MD/PhDs do? I have come to this decision because, while I love research, I cannot see myself living on a 30k salary as a postdoc for years of getting my PhD.

Members don't see this ad.
 
Let's say that I am set to just pursue just my MD. What pathways would I take if I ultimately want to do bench research with some clinical work on the side, similar to the 70/30 that MD/PhDs do? I have come to this decision because, while I love research, I cannot see myself living on a 30k salary as a postdoc for years of getting my PhD.

:laugh: So you can see yourself living on a 60k-70k salary as a fellow for years while getting PhD equivalent bench research experience, while trying to pay off massive loans and seeing your friends make 200k or more? That's your alternative.
 
Let's say that I am set to just pursue just my MD. What pathways would I take if I ultimately want to do bench research with some clinical work on the side, similar to the 70/30 that MD/PhDs do? I have come to this decision because, while I love research, I cannot see myself living on a 30k salary as a postdoc for years of getting my PhD.

You could do a residency/specialty that favors those who do research, such as IM and pathology. Many fellowships in IM have a built-in research component.

Unfortunately, you would still need to "put your time in" at some point. You could extend your residency research into a post-doc. You will need to show evidence of being able to produce papers and grants before anyone would give you significant protected research time, let alone space for a lab or start-up funds. On the bright side, many departments will pay you PGY salary to do the post-doc, meaning you're getting 60K instead of 30K over this time. But you will see that fewer people will be willing to take a chance on you than they would a MD/PhD- that 60K they are giving you is taking away from the funds they will pay someone to provide clinilal services.

Good luck.
 
Members don't see this ad :)
:laugh: So you can see yourself living on a 60k-70k salary as a fellow for years while getting PhD equivalent bench research experience, while trying to pay off massive loans and seeing your friends make 200k or more? That's your alternative.

what's so funny about that? 60k-70k/yr is a much better deal than 30k and one can start paying off loans on that salary, not to mention supplement that income with moonlighting, which many fellowship programs let you do. MD/Ph.D's still need to do a subspecialty fellowship if they want to stay in research.

As for my colleagues earning 100k-200k in private practice straight out of residency, i couldnt care less, because I love having a research component and would go bonkers having to spend countless hours in the primary care clinic.
 
what's so funny about that? 60k-70k/yr is a much better deal than 30k and one can start paying off loans on that salary, not to mention supplement that income with moonlighting, which many fellowship programs let you do.

Your post-tax income after those loans is going to be roughly equivalent, except in the MD-only pathway you have loans to keep paying off as an attending as well.

Moonlighting is going to be a situation specific thing. But if you're serious about getting the bench research you need to launch a bench research career, how do you have time to moonlight? I certainly didn't when I was in graduate school. My residency has a significant chunk of time built in for research, but moonlighting is forbidden. The attitude is that if you have spare time it should be spent on academic pursuits.

MD/Ph.D's still need to do a subspecialty fellowship if they want to stay in research.

Usually. But the research portion of this fellowship should be shorter than a MD who is coming in without or with minimal basic science research.

As for my colleagues earning 100k-200k in private practice straight out of residency, i couldnt care less, because I love having a research component and would go bonkers having to spend countless hours in the primary care clinic.

That's fine for someone who is truly dedicated. I think the MD->research pathway is a fine choice for the right person. However, if someone has difficulties with the low pay associated with research training right off the bad, I was pointing out that these issues don't go away with time.
 
You could do a residency/specialty that favors those who do research, such as IM and pathology. Many fellowships in IM have a built-in research component.

Unfortunately, you would still need to "put your time in" at some point. You could extend your residency research into a post-doc. You will need to show evidence of being able to produce papers and grants before anyone would give you significant protected research time, let alone space for a lab or start-up funds. On the bright side, many departments will pay you PGY salary to do the post-doc, meaning you're getting 60K instead of 30K over this time. But you will see that fewer people will be willing to take a chance on you than they would a MD/PhD- that 60K they are giving you is taking away from the funds they will pay someone to provide clinilal services.

Good luck.

This is my current set-up actually. I had thought about doing a Ph.D. program after finishing residency but ultimately decided i'd rather not take the 50% pay cut from my residency salary. I do have loans to pay but on the flip side, i could have deferred them as a grad student. I figured since I have a doctoral degree already, I could just go straight to a post-doc and not worry about tests and coursework requirements. My PI is an MD (no Ph.D), and does great clinical & translational research so it definitely can be done and done well, and happens often actually.
 
My PI is an MD (no Ph.D), and does great clinical & translational research so it definitely can be done and done well, and happens often actually.

This is the key to this discussion. Clinical/translational research generally does not require the level of training and commitment that bench research requires.
 
Your post-tax income after those loans is going to be roughly equivalent, except in the MD-only pathway you have loans to keep paying off as an attending as well.

I pay about $500/month in student loans x 12 months comes out to about $6000. My postdoc salary is about $50k after taxes. I mean yeah it comes out just slightly higher but grad student salary of $30k is a pre-tax figure as well.


Moonlighting is going to be a situation specific thing. But if you're serious about getting the bench research you need to launch a bench research career, how do you have time to moonlight? I certainly didn't when I was in graduate school.
It really depends on the project(s) you're working on, your other commitments obviously. As a grad student i'm sure you also had coursework and exams to worry about. maybe your boss was very demanding also. I hope all that time commitment paid off for you though...

Also, the way i see it, an MD doesn't really launch a research career from a post-doc (or grad school for that matter). An MD launchs a research career from their subspecialty fellowship. Any lab training that happens previous to that is just to gain extra lab experience and insights (and maybe a few extra publications that may or may not be relevant to the field you will ultimately be working with--if not, doesn't really help all that much tbh). Your relevent pubs are going to come from your subspecialty fellowship work, which is what you're then going to use for grants to become and independent researcher. Of course the more prior research experience you have, the more likely you will be productive in your subspecialty fellowship research. (this is, of course, referring to IM subspecialties... I cant comment on other fields of medicine).

As for moonlighting, I personally strive to keep a balance between my research and clinical work. I MAKE time for it. I only moonlight like 2 day shifts per month, but I know colleagues (in subspecialty fellowships) who do more, and work night shifts. Moonlighting the weekend day shifts does the number of full weekends i have to 2 per month, and I do have a pretty busy life, but hey it's still an easier schedule than residency and I like the flexibility.


Usually. But the research portion of this fellowship should be shorter than a MD who is coming in without or with minimal basic science research.
True, and more productive i imagine. But that is why we are discussing the utility for an MD, of a going back to grad school to get a Ph.D, versus just doing a post-doc for double the salary to gain the research experience before or after doing subspecialty clinical fellowship.


That's fine for someone who is truly dedicated. I think the MD->research pathway is a fine choice for the right person. However, if someone has difficulties with the low pay associated with research training right off the bad, I was pointing out that these issues don't go away with time.
oh, most certainly... I do agree with you there. And there are CERTAINLY pro's to getting a Ph.D. particularly if one wants to do very molecular, biochem research. The niche for MD's though is in the translational to clinical research spectrum, and if one doesnt have a Ph.D. one certainly has to recognize that there will be a niche in the research world where you wont be able to compete with the Ph.D.s. But there's the other niche where the Ph.D.s cant compete with you because of your clinical insights and experience with patients. And well, the MD/Ph.D's certainly have the best of both worlds, but that doesn't mean an MD without a Ph.D can't have a research career if that's what you ultimately want.
 
My postdoc salary is about $50k after taxes.

The post-docs in my lab get paid $35K before taxes. $50K after taxes is very high. How much clinical service do you do besides the 2 days/month of moon-lighting?
 
I pay about $500/month in student loans x 12 months comes out to about $6000. My postdoc salary is about $50k after taxes. I mean yeah it comes out just slightly higher but grad student salary of $30k is a pre-tax figure as well.

That would correlate roughly to a total loan burden of around $50,000 unless you are using IBR. That's about one year of loans for many medical students. The average debt nationally is $140,000 as of 2007 (http://www.amsa.org/AMSA/Homepage/About/Committees/StudentLife/DebtFactSheet.aspx).

There is a graph that I currently cannot find showing the exponential growth in medical student debt versus a relatively flat residency salary. I'm not sure when you graduated, but for current students starting medical school, student loan debt is a much bigger problem. Also note that Stafford loans issued when you took them were at about 2.25% interest. Most student loans are now well over 6.5% interest. There was also once student loan deferment during residency, but this is now long gone as well.

In total, IMO the increased salary as a fellow does not justify the MD->postdoc route for someone seriously considering a bench research career. The lack of student loan debt during the MD/PhD program still makes the MD/PhD pathway the best financial plan given the tremendous current loan burden of medical school graduates.

It really depends on the project(s) you're working on, your other commitments obviously. As a grad student i'm sure you also had coursework and exams to worry about. maybe your boss was very demanding also. I hope all that time commitment paid off for you though...

Running experiments alone was a 60-80 hour a week job for me. My boss was not particularly demanding until it was time for me to actually finish, hah. As for the time commitment paying off, we will see. I don't think it helped me obtain a residency. At this point, I am strongly leaning towards private practice when I finish residency.

An MD launchs a research career from their subspecialty fellowship.

For bench research, that only happens if the subspecialty fellowship includes years of protected bench research time. Whether you call that protected research time "post-doc" or "fellowship" is a matter of semantics.

True, and more productive i imagine. But that is why we are discussing the utility for an MD, of a going back to grad school to get a Ph.D, versus just doing a post-doc for double the salary to gain the research experience before or after doing subspecialty clinical fellowship.

I completely agree. I don't think an MD should go back for a PhD, and I have stated so in this forums several times (see: http://forums.studentdoctor.net/showpost.php?p=11854704&postcount=5). This is a pre-medical student deciding whether to do a PhD now as part of an MD/PhD program or to do MD to post-doc later. This is a different situation.

The niche for MD's though is in the translational to clinical research spectrum

That is not what the op is posting about. MD/PhD programs are meant to generate bench researchers. The op in this thread asked about alternate pathways to bench research. if the op had asked about clinical research, I would have recommended they not complete a PhD.

but that doesn't mean an MD without a Ph.D can't have a research career if that's what you ultimately want.

I agree, but I think you're underestimating the time and effort required to launch a bench research career in the current funding environment as an MD or MD/PhD. Launching a career comprised of a large amount of clinical research is much easier and does not require a PhD.

echod said:
The post-docs in my lab get paid $35K before taxes. $50K after taxes is very high

A good rule of thumb for expected post-doc salaries is the NIH post-doc minimum. See: http://grants.nih.gov/grants/guide/notice-files/NOT-OD-11-067.html. A good rule of thumb for expected resident/fellow salary is take the equivalent number of years on that post-doc chart and add about $10,000/year. Depending on your debt burden, this may make up the difference between MDs and PhDs in student loan debt, but most likely will not.

It is true however that some (IMO unscrupulous) PIs do find ways to pay PhD post-docs less than NIH minimum.
 
excellent thread here! reese07 asked a question that I was planning to ask too because I got offers for both MD-only and MD-PhD programs. I'm having trouble choosing between the 2 because the MD-only program is very research oriented, with 1 out of 4 years in med school spent on research. I also have the option of doing a clinician-scientist residency afterwards, where up to 1 year can be spent doing full-time research.

I was just wondering if there are any repercussions if I choose to go back to fully clinic work after spending 5 years doing 80/20? perhaps someone who has gone through this can explain?
 
I was just wondering if there are any repercussions if I choose to go back to fully clinic work after spending 5 years doing 80/20? perhaps someone who has gone through this can explain?

The only reprocussions are that you're probably rusty and not as good as a clinician as you would otherwise have been (but you will catch up), and that you will (likely) be poorer than you otherwise would have been.
 
The only reprocussions are that you're probably rusty and not as good as a clinician as you would otherwise have been (but you will catch up), and that you will (likely) be poorer than you otherwise would have been.

I've never understood this argument. I mean, if you're a rusty not so good clinician because you're doing so much research, why should you be practicing clinically 20% of the time at all?

I was just wondering if there are any repercussions if I choose to go back to fully clinic work after spending 5 years doing 80/20? perhaps someone who has gone through this can explain?

The repercussion is that you've spent 5 years of your life plus training for that 5 years of your life (probably ~5 years), and then you basically have nothing to show for it. You go back to square one on the clinical ladder, with a lot of training time, much decreased assets, and not much else. I've seen a few of these attendings, and they are not a happy bunch.
 
Members don't see this ad :)
One PI we regularly collaborate with "only" has an MD his lab does great bench and clinical work. Also, I don't think Peter Agre is sitting there with his Nobel thinking "if only I'd gotten a PhD too." I think it's a very personal decision, but the PhD isn't always necessary/worth it. That being said I'm doing a DVM/PhD program, since I'm going DVM route I've resigned myself to never making money.
 
Thanks for the replies, gbwillner and Neuronix!!! They were very helpful.
 
I don't often see a thread that lays out the difference in pay between those with the medical degree only and the PhD only who work in labs. Its a nice thing to see, including the fact of physician only debt being factored in. Thanks Neuronix and everyone else. Great thread.
 
I've never understood this argument. I mean, if you're a rusty not so good clinician because you're doing so much research, why should you be practicing clinically 20% of the time at all?

A. You need clinical revenue to supplement your salary to "pay for you"

B. You want to do some clinical work.

If you are working 20% of your time on something, it stands to reason you will not be as proficient as someone who devotes 100% of their time to it. It does not mean you are incompetent.
It may mean you necessarily sub-specialize to a small niche field in your specialty. Working full time may mean you need to broaden your practice because you cannot find enough volume for your previous work. This happens often in pathology. For example, you sign out renal pathology- a niche field with relatively few specialists. You can justify your salary with grants and this niche field. If you lose your lab, it is likely you will not have sufficient volume and your director/chairman may ask you to sign-out general surgical pathology. If you haven't done that in 5 years, you are going to suck at it and will need to "catch up".
 
That would correlate roughly to a total loan burden of around $50,000 unless you are using IBR. That's about one year of loans for many medical students. The average debt nationally is $140,000 as of 2007 (http://www.amsa.org/AMSA/Homepage/About/Committees/StudentLife/DebtFactSheet.aspx).

good point... I am about average as far as loan burden, but I am on a graduated repayment plan for the bulk of my loans...I am locked into it, but it is actually not available anymore. I think the closest thing is the income contingent plan maybe?

I also was fortunate enough to have locked in a pretty low fixed rate interest by consolidating with the government back in 2006. I was able to consolidate with the government because of a couple tiny Direct loans i'd taken out in college. So I was lucky that way.

In total, IMO the increased salary as a fellow does not justify the MD->postdoc route for someone seriously considering a bench research career. The lack of student loan debt during the MD/PhD program still makes the MD/PhD pathway the best financial plan given the tremendous current loan burden of medical school graduates.
You're talking about MSTP here and you are absolutely correct in this. Not arguing what you said here in the least. But for someone who is not MSTP, what is the best way to become a physician scientist? Is investing the time/pay cut to get a Ph.D necessary? I say no.

Running experiments alone was a 60-80 hour a week job for me. My boss was not particularly demanding until it was time for me to actually finish, hah. As for the time commitment paying off, we will see. I don't think it helped me obtain a residency. At this point, I am strongly leaning towards private practice when I finish residency.
I mean it really depends on what the project is, how efficient one is, how motivated one is, how much energy one has, a bit of luck as well. Bench research is the kind of thing that you can easily spend your every waking hour doing assays, or you can take your sweet time getting things done, or somewhere in between. It's so self-directed that the onus is on you to define a work-life balance.

And what you said about not knowing whether a Ph.D. paid off yet... this is exactly my point. You invested all this time and effort getting a Ph.D. and it may not have really mattered in the end. I personally dont think extra education is ever a waste, but if one hasn't done it yet and is trying to decide how to reach their goal, I personally dont think a Ph.D. is necessary to be successful in the research world. It might end up being a different kind of research than Ph.D.s might have more expertise with, but all of these kinds of research are needed.

For bench research, that only happens if the subspecialty fellowship includes years of protected bench research time. Whether you call that protected research time "post-doc" or "fellowship" is a matter of semantics.
very true and very good point, thank you for mentioning this Neuronix. When i was interviewing for fellowship, there were definitely some programs who emphasized research and some who were excellent but purely clinical training programs. If one is thinking research career, it is REALLY important to make sure you end up in a fellowship that does have that research emphasis and does have ample research opportunities for your topic of interest (if you have one).

I completely agree. I don't think an MD should go back for a PhD, and I have stated so in this forums several times (see: http://forums.studentdoctor.net/showpost.php?p=11854704&postcount=5). This is a pre-medical student deciding whether to do a PhD now as part of an MD/PhD program or to do MD to post-doc later. This is a different situation.
Oh, if so then I apologize i missed that detail. Well my advice is, if you are deciding MSTP then do MSTP. It's a free ride through med school, additional research experience, usually gives you a leg up for residency match (not always though), and if you have a chance to get a free couple of degrees you shouldn't miss out on that opportunity. If it's an MD/Ph.D program where you will still have to pay for med school, then depending on the kind of research you want to do, i recommend opting for the MD only, doing some sort of additional research exposure, whether its a year-long research program during med school (can also help BIG time for residency match and give you valuable connections for later on) or a post-doc after residency, or just going straight into a research-oriented subspecialty fellowship after residency, or all of the above.


That is not what the op is posting about. MD/PhD programs are meant to generate bench researchers. The op in this thread asked about alternate pathways to bench research. if the op had asked about clinical research, I would have recommended they not complete a PhD.
Translational research involves bench research as well as clinical research and is usually much more interesting and relevant to an MD-trained individual, and can usually be successfully handled by pure MD's with additional research experience. But yes, the OP's decision will really have to rest on what KIND of research he/she wants to do. Very basic science will likely require a Ph.D. background to have the expertise needed to compete with Ph.D. trained scientists, though I argue that the Ph.D. course for an MD/PhD is usually more rushed and abbreviated than the typical pure Ph.D course, so the niche may still remain distinct.

I agree, but I think you're underestimating the time and effort required to launch a bench research career in the current funding environment as an MD or MD/PhD. Launching a career comprised of a large amount of clinical research is much easier and does not require a PhD.
Yes. Translational research included in the latter. IME, MD-only researchers tend to have a lab doing translational bench research as well as clinical research protocols going on at the same time. Some MDs do only do clinical non-bench research, it really depends.

A good rule of thumb for expected post-doc salaries is the NIH post-doc minimum. See: http://grants.nih.gov/grants/guide/notice-files/NOT-OD-11-067.html. A good rule of thumb for expected resident/fellow salary is take the equivalent number of years on that post-doc chart and add about $10,000/year. Depending on your debt burden, this may make up the difference between MDs and PhDs in student loan debt, but most likely will not.
OK i am going to address my higher post-doc salary here. I am being paid higher than the average post doc. Yes. Why is this? Two things. My year doing research in med school counted as prior experience, as did my years of residency training. The extra years of experience boosted me on the pay scale. BOOM. Great reason to do MD only, then residency, then postdoc.

MSTP is a much sweeter deal though, so I am not comparing this to MSTP... it is in a class by itself.
 
Last edited:
The post-docs in my lab get paid $35K before taxes. $50K after taxes is very high. How much clinical service do you do besides the 2 days/month of moon-lighting?


Yes it is, thank you for noting this. This is my salary for just my post-doc (just research). I also moonlight on the side, which is also a perk to being a fully trained board-certified physician at the same time. I made about $20k moonlighting last year, about two 10-12 hour hospitalist shifts per month.

As i wrote in my post above to Neuronix, my extra years of experience (residency, year off for research in med school) boosted me to a higher post-doc pay category. That was a big reason I opted against getting a Ph.D. Looking at $50k for being a postdoc vs $30k for being a grad student... I really had to consider whether that extra degree was really necessary and I decided that for my purposes it really wasnt especially since I had loans to start paying off.
 
Last edited:
This is the key to this discussion. Clinical/translational research generally does not require the level of training and commitment that bench research requires.

I have to correct you on this.

Translational research most certainly does involve bench research.

PI's with an MD only who do translational research, my current boss among them, do frequently supervise Ph.D. scientists in their labs who might do the more molecular work.
 
A. You need clinical revenue to supplement your salary to "pay for you"

B. You want to do some clinical work.

If you are working 20% of your time on something, it stands to reason you will not be as proficient as someone who devotes 100% of their time to it. It does not mean you are incompetent.
It may mean you necessarily sub-specialize to a small niche field in your specialty. Working full time may mean you need to broaden your practice because you cannot find enough volume for your previous work. This happens often in pathology. For example, you sign out renal pathology- a niche field with relatively few specialists. You can justify your salary with grants and this niche field. If you lose your lab, it is likely you will not have sufficient volume and your director/chairman may ask you to sign-out general surgical pathology. If you haven't done that in 5 years, you are going to suck at it and will need to "catch up".


Interestingly, I've found that many MD researchers (including MD/Ph.Ds) are SUPERB clinicians even though they dont spend their entire work day seeing patients. That scientific thought process and perspective can bring a unique facet to the practice of clinical medicine.
 
MSTP is a much sweeter deal though, so I am not comparing this to MSTP... it is in a class by itse

If you can get through the MD without loans (a rarity but still), I think it's very close, even when your interest is in a basic/translational career. The MD/PhD might make you more marketable for research residencies and research fellowships, but the PhD won't count much, if at all, for the grant process. Most likely you'll be working on a different subject, maybe with different methods, where's embarking on research after the MD means you'll be heading off exactly on the path for grants.
 
If you can get through the MD without loans (a rarity but still), I think it's very close, even when your interest is in a basic/translational career. The MD/PhD might make you more marketable for research residencies and research fellowships, but the PhD won't count much, if at all, for the grant process. Most likely you'll be working on a different subject, maybe with different methods, where's embarking on research after the MD means you'll be heading off exactly on the path for grants.

4-5 year postdoc as an MD only, 3-5 year PhD... what is the advantage? Do most MD-PhDs anticipate doing a postdoc in addition to fellowship/residency?
 
4-5 year postdoc as an MD only, 3-5 year PhD... what is the advantage? Do most MD-PhDs anticipate doing a postdoc in addition to fellowship/residency?

Yes, most MD/PhDs who want to run a lab should ALSO anticipate doing a post-doc. The advantage for the MD/PhD is the formal training and mastery of the scientific process. You are (supposedly) more apt for basic science, and should have more experience writing grants/obtaining funding. If you want to do clinical research, or even most translational research, the advantage of an MD/PhD is minimal.

Again, one issue is that you have to have the department allow you to do a postdoc without the PhD, which is not by any means a guarantee. You may be free to your lab (or not), but someone is paying your salary and insurance, and NOT receiving any clinical work. Someone with a PhD is a safer bet for the department.

Cheers
 
I see. I hadn't talked with enough young MD-PhD graduates to realize how common the postdoc was in addition to training and residency. Good to know, thank you! Do most people take a full 5 years before applying for assistant professor positions?
 
If you mean a full 5 years post fellowship, probably not. I would venture to guess (based on my peers) that most do ~2-5 years research, including built-in full-time research in residency/fellowship. If you do a 3 year fellowship with 2 years built-in research, you may get a startup and space as an asst. prof., but you may have a difficult time securing your first grant. Also, you can definitely get these types of jobs at lower-tiered institutions with minimal full-time research, but your chance of success may be limited.
 
What are some of the goals for this time period? The same as for a PhD postdoc? Developing direction of research, do preliminary work to see which projects are doable, secure grant funding, publish? Which do you see as most important to long and short-term success?
(Assuming a basic research career)
 
What are some of the goals for this time period? The same as for a PhD postdoc? Developing direction of research, do preliminary work to see which projects are doable, secure grant funding, publish? Which do you see as most important to long and short-term success?
(Assuming a basic research career)

in short: maybe, yes, yes, yes, yes.


IMHO, the most important factors for short term success are getting as many papers out as possible in the shortest amount of time possible, and writing grants. Papers + grants= job (as asst. prof.). It's that simple. my chairman wants to see at least 2 first author papers during the post-doc and having at least applied for a K08 to be considered for faculty. And that's on top of him already liking you enough to bring you in as a research track applicant (i.e., you are already a stud and have proven youself in grad school). I don't know if that would be sufficient for any other outside applicant (I doubt it). At this stage, it's more about quantity than quality (of course the more quality the better). Save the 5 year Cell paper for after you have a job (unless you are so lucky it comes quickly). Why? When you apply for grants, they are going to look at your publication record in the field you are applying to. My first grant submission was not heavily criticized for experimental design. It was criticized because "applicant has not published in the relevant field, no 1st author papers since 2007" (before residency).

For long term success, I would say building solid relationships with top scientists in your (and other related) field is the key. Part of this would be to be at a top-tier institution. Having very good and famous scientists as mentors (and accessory PIs on your grants) means much higher chance of success. Success breeds success. There is nepotism in science. Grant reviewers feel much more confident about your prospects if famous guy X is you mentor- they know you will be steered in the right direction. They will also note your institution in the grant. That may seem unfair, but that's just the way it is. Furthermore, being involved in a $15-20M grant, even if you are not the primary PI, means you will always be important and have money for science, even of things aren't working well for a while.
 
I thought I would bump this thread with this pair of articles from this week's Science Translational Med:
http://stm.sciencemag.org/content/4/135/135fs14.full
http://stm.sciencemag.org/content/4/135/135cm6.full

Some choice excerpts:

"A: I think that risk aversion is one of the greatest risks to young clinician-investigators. ... One problem is that M.D.-Ph.D. candidates and early-career clinician-investigators are typically people who have been very successful in their young lives and who worry about failing to live up to that early success. Taking the kinds of risks that are necessary to make truly new discoveries is a gut check for many people. In my view, in order to try something completely new, a person must be willing to fail; that doesn’t come easily to accomplished young people who—having devoted so many years of their lives to training and to establishing independent research careers—feel the need to diminish their risk of failure...

Q.:
But when trainees assess successful senior scientists, how can we distinguish between the possibility that junior clinician-scientists who took big risks failed miserably and left academic science versus the possibility that successful researchers generally meet the risk-taking phenotype—that they asked the questions other scientists weren’t asking or asked the questions in a new way?"

"In 1980, average ages at time of appointment to assistant professorship in the biomedical sciences were 33.5, 33.5, and 34.8 years for Ph.D.s, M.D.s, and M.D.-Ph.D.s, respectively, rising in 27 years to 38.5, 37, and 39.9 years, respectively. Similarly, in 1980 the average ages of Ph.D.s, M.D.s, and M.D.-Ph.D.s in the biomedical sciences when they received their first U.S. National Institutes of Health (NIH) RO1 grants were 35.7, 37.7, and 36.1 years, respectively, rising to 41.7, 44.2, and 42.3 years, respectively, in 2007 (3). All took substantially longer to gain faculty slots and the R01 status of independence; M.D.s became faculty members sooner than those with Ph.D.s but became independent later. Thus, it appears that postdoctoral training nearly doubled in duration over this time period but was virtually required before a clinician-investigator could achieve some level of professional security."
 
Top