longest time taken to do an MD/PhD

[ganglia777]

A handful of colleagues I know have taken 10 or 11 years to do an MD/PhD.
Has anybody taken 12 or more years? This would mean 4 years for MD and 8 years for PhD.

It's not uncommon for normal PhD's to take 7 or 8 years just for a PhD, but for some reason MD/PhDs usually get a PhD in 4 or 5 years max. Likely due to being more motivated to 1. pick easy projects, 2. focus on the prize of becoming an MD and thus rush through the research.

---

Members don't see this ad.

 

[Lucca]

Are you an MD/PhD student? If so, this comment seems extremely bizarre. MD/PhD students have shorter graduation times on average due to eliminating course and teaching requirements (or those requirements overlap with things they must do for the MD as well). The requirements for graduating the PhD are the same as any other PhD at the programs I’ve interviewed at.

I don’t think there’s anybody out there picking “easy” projects.

 

[Neuronix]

I know a guy who took 13 years to do MD/PhD. To elaborate, this was one of those two separate institution MD/PhD programs in the same city. Several of these have a notorious reputation for 10+ year MD/PhD graduates happening with regularity.

Regardless,
1. You have to evaluate each person for themselves. You can't just make an assumption like this. There are impressive MD/PhDs who did well in a short period of time and PhDs who do lousy in a long period of time. Time served is a really poor judge of a trainee's potential.
2. Rushing isn't necessarily a bad thing if the quality is there. I worked very hard during my PhD and had a strong PhD at the end by every metric. I almost graduated the whole program in 7 years and regret that it took me 8.

 

[ClimbsRox]

The longest I have heard is 13 years. Did his PhD in chemistry and had a PI who set unrealistic goals and kept him for 9 years. The program was relatively new and not MSTP so the institution had little power to push him along against the PIs wishes.

 

[vaticrunner]

Are you an MD/PhD student? If so, this comment seems extremely bizarre. MD/PhD students have shorter graduation times on average due to eliminating course and teaching requirements (or those requirements overlap with things they must do for the MD as well). The requirements for graduating the PhD are the same as any other PhD at the programs I’ve interviewed at.

I don’t think there’s anybody out there picking “easy” projects.

I will say before I started interviewing I agreed with this sentiment. However, after travelling to a couple institutions where the average/desired PhD length was 3.5 years, I realized there are certainly “easier” projects for MSTP students to pick. The way it was explained to me by one faculty member who spearheaded the push for shorter time to degree was that there are certain projects, such as making a mouse model, which are extremely hit or miss. He said that type of project isn’t suitable for an MSTP student. Instead, he tries to place MSTP students on projects where the mouse model has already been generated and the student’s job is to ask peritent questions with the model already up and running. Obviously, this is a slightly niche example, but there is definitely a push to make sure that MSTP students pick “reasonable” projects having the right amount of return on investment. With the idea being, the research you conduct during fellowship is much more important for refining your interests. This is probably very institutionally dependent, but I have definitely seen this idea before just with different semantics.

 

[Lucca]

I will say before I started interviewing I agreed with this sentiment. However, after travelling to a couple institutions where the average/desired PhD length was 3.5 years, I realized there are certainly “easier” projects for MSTP students to pick. The way it was explained to me by one faculty member who spearheaded the push for shorter time to degree was that there are certain projects, such as making a mouse model, which are extremely hit or miss. He said that type of project isn’t suitable for an MSTP student. Instead, he tries to place MSTP students on projects where the mouse model has already been generated and the student’s job is to ask peritent questions with the model already up and running. Obviously, this is a slightly niche example, but there is definitely a push to make sure that MSTP students pick “reasonable” projects having the right amount of return on investment. With the idea being, the research you conduct during fellowship is much more important for refining your interests. This is probably very institutionally dependent, but I have definitely seen this idea before just with different semantics.

Definitely, the nature of a project changes the timeline of the PhD drastically. Deciding to do a mouse model or cryo-EM heavy project mivht net you some big rewards if it all works out but if there are major setbacks then that has obvious repercussions for finishing “on time”.

In my discipline, it’s not odd to see 7-8 yr PhDs from cryo-EM heavy projects and also not uncommon to see a straight PhD graduate in 4 yrs from a purely computational project. Does that mean the project is “easier”? In my admittedly naive opinion as someone who does both experimental and computational work, no.

 

[ClimbsRox]

Definitely, the nature of a project changes the timeline of the PhD drastically. Deciding to do a mouse model or cryo-EM heavy project mivht net you some big rewards if it all works out but if there are major setbacks then that has obvious repercussions for finishing “on time”.

In my discipline, it’s not odd to see 7-8 yr PhDs from cryo-EM heavy projects and also not uncommon to see a straight PhD graduate in 4 yrs from a purely computational project. Does that mean the project is “easier”? In my admittedly naive opinion as someone who does both experimental and computational work, no.

Thats interesting that you say CryoEM projects are long. I haven't done any structure work yet, but the PIs I talked to seemed to be solving CryoEM structures quite rapidly. One even said he gives every rotation student a CryoEM structure to do and they finish it in about 6-8 weeks.

 

[Lucca]

Thats interesting that you say CryoEM projects are long. I haven't done any structure work yet, but the PIs I talked to seemed to be solving CryoEM structures quite rapidly. One even said he gives every rotation student a CryoEM structure to do and they finish it in about 6-8 weeks.

Depends a lot on the system and the question you’re trying to answer. If you are looking at an abundant, easily isolated protein and want to ask “What does it look like” then ya it’s nbd. If you’re entire PhD is about cryo-EM and you’re trying to develop new methods to figure out the precise mechanism of how a protein mechanically does what it does then you’re going to be there for a while. Cryo EM is all about statistics, which means how easily you can obtain and reproduce the system you want to image will tune how challenging the problem is (suppose you want to image a catalytic process with 1 vs multiple conformational changes) on top of the facilities available for actually getting images and how much access your PI has. I just used it as an example because a friend of mine is doing a cryo EM based project at Berkeley looking at motor proteins and he’s finally graduating after 8 yrs. This is my impression as someone who does super-Res but not CryoEM so happy to be corrected.

As for mouse models, if someone offered me 2X PhD stipend to create a mouse model I’d tell them to hire a tech with that money and email me if they want to collaborate lol.

 

[ClimbsRox]

Depends a lot on the system and the question you’re trying to answer. If you are looking at an abundant, easily isolated protein and want to ask “What does it look like” then ya it’s nbd. If you’re entire PhD is about cryo-EM and you’re trying to develop new methods to figure out the precise mechanism of how a protein mechanically does what it does then you’re going to be there for a while. Cryo EM is all about statistics, which means how easily you can obtain and reproduce the system you want to image will tune how challenging the problem is (suppose you want to image a catalytic process with 1 vs multiple conformational changes) on top of the facilities available for actually getting images and how much access your PI has. I just used it as an example because a friend of mine is doing a cryo EM based project at Berkeley looking at motor proteins and he’s finally graduating after 8 yrs. This is my impression as someone who does super-Res but not CryoEM so happy to be corrected.

As for mouse models, if someone offered me 2X PhD stipend to create a mouse model I’d tell them to hire a tech with that money and email me if they want to collaborate lol.

Makes sense. I am interested in CryoEM for my graduate work, so something to consider. Super res is no walk in the park either.

 

[Lucca]

Off-topic, but when you say super-res do you mean optical single molecule imaging?

Yup. PALM/STORM/PAINT/STED that sort of thing

 

[Lucca]

Wacky. Didn't expect to see someone else on SDN in my field of research.

I chose my avatar pic because it looks like the diffraction limited STORM images I take!

 

[Lucca]

Lol yep, I totally see it.

Interestingly, somewhat back on topic, two biophysics labs I interviewed with this cycle abandoned single molecule studies in the past in favor of cryo-EM. Obviously these are two entirely different imaging paradigms, but I almost wonder if single molecule/super res techniques are going out of vogue. Cryo-EM is everywhere now.

Live cell super res is alive and well. Pm me if u want to talk more about it

 

[gbwillner]

I've seen 14 as the longest.... 3 lab changes were involved. Obviously, a lot of things did not go right.

10 sounds really long, but I know lots of people who've done that.

 

[nonamesleft]

Depends a lot on the system and the question you’re trying to answer. If you are looking at an abundant, easily isolated protein and want to ask “What does it look like” then ya it’s nbd. If you’re entire PhD is about cryo-EM and you’re trying to develop new methods to figure out the precise mechanism of how a protein mechanically does what it does then you’re going to be there for a while. Cryo EM is all about statistics, which means how easily you can obtain and reproduce the system you want to image will tune how challenging the problem is (suppose you want to image a catalytic process with 1 vs multiple conformational changes) on top of the facilities available for actually getting images and how much access your PI has. I just used it as an example because a friend of mine is doing a cryo EM based project at Berkeley looking at motor proteins and he’s finally graduating after 8 yrs. This is my impression as someone who does super-Res but not CryoEM so happy to be corrected.

As for mouse models, if someone offered me 2X PhD stipend to create a mouse model I’d tell them to hire a tech with that money and email me if they want to collaborate lol.

1. I think I know your friend! (if recent graduate)
2. I used crystallography. Solving was the "easy" part. The tray setup and getting good diffraction definitely added time to my graduate work.