For a couple years, I've worked for a company founded by an MD ~20 years ago, doing PACS (Radiology IT system; basically how X-Rays and CTs and MRIs get distributed within a hospital, viewed by doctors, archived, and sent between sites; my particular client is the US military in Iraq/Afghanistan/Kuwait/etc.). The founder is an MD (Ortho), but I don't think he's ever practiced medicine. I would probably say he's a better businessman than 99% of the MDs I've met, but probably would have been a worse doctor.<p>Generally, the perception of MDs in business, especially outside medical startups, is that they're horrible. First, they're usually convinced they're the smartest in the room, and even if they may be, that they're also the most knowledgeable about any given domain -- even if they have no idea about it. They often end up involved in mediocre small businesses, taken advantage of in scams, etc., due to being rich and bad at business.<p>They're also super-expensive; education and debt. The best are probably military doctors who leave after their first tour, but even then, they're mid/late 30s.<p>Within medicine, having an MD as the medical interface of your company (could be CEO, could be bizdev, could just be a chief medical officer) is essential. One of the factors limiting medical IT startups is that they're really just IT startups with some additional hassle, but the pool of suitable MDs willing to work in a startup is low.<p>I think there are a lot of great opportunities in medicine: the PACS/Telemedicine field is open (especially for consumer-facing systems, or for SaaS); various forms of automation (pharmacy, EHR/EMR, lab, etc.) exist, but generally have been developed by large companies and aren't as well engineered as one would like, or as innovative -- the long product cycle, conservative buyers, and FDA regulations slow this, but the effect is greater than this can explain. Probably the best opportunity for disruptive innovation is direct, consumer-facing health, especially in less or unregulated areas. Scott Adams (Dilbert)'s idea of passively monitoring one's daily life, then when you get sick, go back and look at the previous N months of data archive and identify causes or lifestyle issues; then do statistical analysis across people, is one idea. Providing better realtime advice and decision support about things with health impact (diet, exercise, various drugs, ...).<p>I guess the most relevant question is about you -- aside from liking HN/TechCrunch, and being an MD, are you someone who wants to code? Work on hardware? Business? Do you like quick/simple projects, well-funded low risk big projects, or true startups? Are you location-specific at all?<p>If it were me, if I didn't have 100% answers, I would probably try to intern or contract/consult with a variety of companies or projects, just to get a feel for how they work. Founding a startup is a 5-10 year project, so spending a year figuring out exactly what kind of startup, or if you want to do a startup at all, is probably worthwhile.<p>My next project is actually an IT infrastructure thing, which has healthcare and defense as primary markets.