More details on the Pfizer accident have been released. The initial reports of a fire in the research labs were only correct insofar as all of a pharmaceutical company's facilities are used for research. This was a large-scale affair in a chemical storage area next to a pilot plant, not something that happened in someone's fume hood. I've been unable to find out what solvent or reagent set off the blaze, but it sounds like something that's intrinsically reactive (like a solution of some organometallic reagent.)
Chemical research is a hazardous job. Not as dangerous as, say, coal mining, but there are some real risks. Everyone who's worked in the field for a few years will have stories to tell of accidents that happened in their lab, or down the hall. Most of these stories have reasonably happy endings, but some, inevitably, don't.
You can minimize the risks with attention and intelligence. Almost all the bad accidents I've been around in research labs have grown out of poor decisions: how to set up a reaction, how large to run it, what reagents to use and how. It lets other chemists think "that wouldn't have happened to me," and often, that's true. The two worst accidents I can think of at my current company certainly would not have happened to me, for example.
But that's not always the case. Some bad things just happen, and when you're surrounded by flammable solvents and air-reactive reagents, a small bad thing has the potential to become a large one very quickly. The key is to be ready for what could happen, know how to deal with it, and keep it from getting worse.
You can have your training courses, lab inspections, safety meetings, your standard operating procedures. None of these will protect you, though, from every possible contingency, from random chance or invincible human error. I don't know what the root cause of the Pfizer explosion was, but I assume that it will be determined, and steps will be taken to keep it from happening again. That's a good thing, don't get me wrong - one more accident that's less likely to happen is always a good thing. But there are others out there, and it's one of our jobs as chemists to realize that and be ready.