Perhaps you could try new techniques in tackling questions. For example, for long questions, you could always try to look at the last sentence or two. Get an idea what the question wants.
"What's the diagnosis? What's the best next step?"
Sometimes you can even answer the question entirely by reading the last two lines. Once you know exactly what the question wants, you could easily skim through the question looking for the things you need. For example, if it was a "what's the diagnosis" question, I always find myself skimming for some red flags, and buzz words that could strongly hint for a diagnosis. If it's the best next step, you'd probably need to pay closer attention to the supplied exam / investigations. All in all, you need to see how you could formulate a good idea about the question in less than a minute.
Additionally, put a limit on each question and keep your eye on the timer. This limit should not be exceeded under any circumstance. For example, 44 questions, you should give a maximum of 1 minute and 30 seconds for each question. If you get close to that limit, simple mark it, make a best guess and move on. Don't dwell on it. You also should not be spending 1 minute 30 second on every question, that should be your maximum, because 1:30 means 66 minutes for 44 questions , and obviously you're not allowed to spend that much. But the reason such a limit could be allowed is that you are expected to finish plenty of other questions in much less than a minute, and thus make up for the difference.
