ears said:
All basic science (possibly with clinical applications).[/b] As you probably know, there's a lot of basic science going on in medical centers. Big academic departments do a ton of work like this, and they have a lot of PhD-only PIs working for them. And this is a shorter path; like, ten years (five for school, three for post-doc training) before you could get a job. The only downside is, you can't practice medicine.
So really, only if you absolutely answer yes to both questions, "Do I want to practice medicine?" and "Do I want to do full-time research?" does an MD/PhD really make sense. If you're more into one or the other, you should just commit to one side, and do it.
I think about this part differently. The usual disclaimer applies: I am an MD/PhD student and this is my opinion... Just because I am moderator doesn't mean it's more right. This post gets back to that question of "What is the MD/PhD for?" It seems like everyone has their own views on that one--here's mine.
What is the goal of the MD/PhD program if you ask most program directors? Especially here at Penn (we hear it over and over again) it is: to train basic scientists who have their own labs as PIs. If there's any clinical duty in there at all, it's minimized down to 10-20% of work time. Most people I talk to say the 50/50 split is gone (with a few exceptions). The conclusion I've made then is that most MD/PhD programs are trying to train basic scientists who can translate questions to the bench and answers to the bed. Obviously, those who don't end up at that goal are not failures (though some would view the private practice MD/PhDs as such), however you should come into the MD/PhD program thinking that this is where they want you to go.
That being said, why shouldn't those students interested in being full-time researchers go into MD/PhD programs? Sure, if you can't stomach the idea of playing doctor or spending the extra years getting an MD, don't do it. But for those students who want to know and learn more for the purpose of better understanding the research they hope to do someday, I think it's a good idea. A PhD student has to face that in graduate school one gets very little education as to how your project fits into a bigger picture. I think that a broad based education, like that given in medical school, makes you better equipped to interpret data and ask questions of your own.
There was a time when graduate schools used to teach things like anatomy and histology for PhDs in related subjects. As such, alot of mature researchers got some of that broad-based education. Now, PhD education is more on specialized topics in their areas and the techniques that address them. At my undergrad, most Neuroscience students didn't even take Neuroanatomy (too expensive to get the brains and get someone to teach it!). Students in my lab doing pulse sequence development for MRI are not eligible to take human anatomy (one student tried recently), even though I think many would benefit greatly from it. In many labs, students must apply biochemistry and histology techniques, with little to no understanding of the fundamentals behind them. Undergraduate education is usually lacking in teaching one the fundamentals of biomedical science.
Still, I've never been a big advocate of PhD programs for another reason. I think graduate programs need more regulation for the protection of students. We all hear the stories of students who can't get a lab and get tossed out; students who get held in labs for bogus reasons; students floundering for years with little attention from their PIs. Sure, the PhD in a MD/PhD program is not guaranteed, but the funding comes with alot of oversight both from the institution (which most of your funding comes from, not the government in most cases) and from the NIGMS. This is a good thing. Some may look down on MD/PhDs as scientists because they have shorter times to PhDs and protections against bad labs... But, in my opinion, this is a GOOD THING for a training environment.
There's also that more practical issue of job options and security when you're done. You will be more attractive to research institutions with that MD even if you don't practice medicine. There's the one argument that it sets you apart from the droves of PhDs that are fighting for jobs and there's the other argument that if your research does tank you can switch over to being a clinician.
This advice can't be given to all PhD hopefuls. I've heard the reason from a few PhD students for getting their PhD is that they "Didn't know what else to do", "Wanted to drink some beer", etc... However, some PhD applicants are focused, and not always on academics but frequently on industry, law, or business. Though, if a student is focused on becoming an academic researcher, I would encourage them to consider MD/PhD. There's always that "I dunno" factor, and that's great. You don't know if you wanna do medicine or research, but you want to understand both is a good reason for choosing MD/PhD. You do still have those options for industry or business, but know that you'll probably end up in academic medicine or research. This is different from the truely clueless. They don't choose MD/PhD programs because they're "too long". They see (somewhat fallaciously) the PhD as a 5 year commitment and the MD/PhD as a 13 year commitment.
It's easy for PIs to say "PICK RESEARCH OR MEDICINE! YOU CAN'T DO BOTH SO WHY LEARN BOTH?". I heard it alot at my undergrad, which discourages students from persuing the MD/PhD. These PIs are well established in their careers and see what it takes to get to where they are. I had no idea I'd be in this lab for my PhD doing this kind of research when I started the MD/PhD. I don't know what I'm gonna be doing in another 10 years. But, I still love the idea of the 90/10 split... The idea of doing research all day and knowing how it can be applied to help people. When I end up being established, I'll see shorter roads to get to that point, but I still think I'll have benefitted from knowing what I know. Those who didn't take the curvy path don't know what they're missing, or maybe they just don't feel the same way as I do. After all, I'm one of those weirdos who wants to know just about everything just for the sake of knowing how the world works. I picked my lab based on "What do I know NOTHING about?" The PI being a nice guy and a history of getting students out quickly was a good touch.
They're actually getting paid, and for you, $42k a year is a big step up. Assume five years of post-grad training, and you don't see your first real paycheck until you're in your mid-thirties. 
. 
