How to Learn and Memorise as Fast as Possible at Medical School

articles Med School

I used to think there was a fairly direct correlation between the number of hours you spent studying and the results you eventually got in the exam. If somebody revised for four hours and somebody else revised for ten, surely the person doing ten hours would learn more and therefore do better…..right?

How wrong I was.

During my pre-clinical years at Cambridge, there were periods before exams where I was revising for around 10–12 hours a day. I stopped going to the gym, cut out most of my social life and tried to turn essentially every available hour into revision because I thought that was what working hard looked like. It affected me negatively, and despite spending an enormous amount of time studying, my results were not improving in proportion to the amount of work I was putting in.

Fourth year was almost the opposite. I kept going to the gym throughout exam term, kept some separation between medicine and the rest of my life and became much more deliberate about what I actually did when I sat down to revise. It ended up being my best year at university academically, finishing roughly within the top 15% of the year.

The lesson is not that studying less somehow produces better marks, because obviously you still need to put the work in. The lesson is that time spent studying and learning are not the same thing, and once you get to medical school, where the amount of information you could theoretically learn is almost endless, the efficiency with which you convert an hour of work into something you can actually retrieve becomes far more important.

“Work smarter, not harder” sounds like one of those phrases that belongs on a slightly questionable productivity Instagram page, but unfortunately it is true.

The first mistake: confusing familiarity with learning

One of the biggest changes in how I think about studying is distinguishing between becoming familiar with information and actually being able to produce it from memory.

You can read a lecture, understand every sentence and recognise every diagram, then close the document and discover that you can reproduce surprisingly little of it. The problem is that while the information is sitting in front of you, your brain is constantly being given cues, so recognising something as familiar can very easily create the impression that you know it.

Reading still matters, particularly when you first encounter a topic, because you obviously cannot retrieve something that you have never understood in the first place. What I think is inefficient is repeatedly returning to the same notes and treating every reread as though it is producing the same amount of learning.

The research on retrieval practice supports this distinction. In a classic experiment by Roediger and Karpicke, students who repeatedly studied material performed better after a very short delay, but students who repeatedly retrieved it from memory retained substantially more after longer delays. The important point is not that reading is useless, but that deliberately trying to produce information from memory changes what happens to it afterwards.

This is also one of the central ideas in Stanislas Dehaene’s How We Learn. He describes four pillars that support learning: attention, active engagement, error feedback and consolidation. Once I came across this framework, a lot of the study techniques that had worked for me started to make much more sense, because an efficient revision system should really be trying to exploit all four.

How I wasted entire evenings in pre-clinical medicine

In pre-clinical medicine, our learning was mostly structured around lectures, and my usual approach was to go to the lecture, use the lecture notes to understand the topic and then turn the important information into Anki cards.

The problem was making the cards.

There were premade Anki decks available, but I never completely trusted them because information could be missing and the cards were often not written in the way that I personally found easiest to learn. So I made my own, which solved one problem but created another: there were evenings where making the Anki for the next set of lectures could take essentially the entire evening.

That is a real problem because creating revision material can very easily masquerade as revision. I might have spent hours reading lecture slides, deciding what information mattered, copying things across and turning everything into neat questions, but much of that time was still fundamentally spent with the information sitting directly in front of me. By the end of the evening I had built the tool I was going to use for active recall without necessarily having done very much active recall.

The same problem occurs when the cards themselves are badly designed. I personally find Anki cards with enormous answers extremely difficult to memorise, particularly when one card expects you to reproduce a paragraph almost word for word, so I would much rather split the same information across several smaller cards. As a rough rule for myself, I try to keep the back of an ordinary card to a maximum of around two lines wherever possible.

This is not a scientific law about the optimum number of words on a flashcard, and different people will prefer different styles. It is simply what works much better for me because each card asks one relatively precise question, which makes it much easier to tell whether I genuinely knew the answer or not.

Anki itself is not magic either. A recent systematic review of its use in medical education found some positive associations between consistent use and examination performance, particularly for standardised examinations, but the evidence across university examinations was mixed and largely observational. Deck quality, consistency and how the cards are actually used probably matter at least as much as having Anki installed on your laptop.

What actually works

The study session I now think makes the most sense has four stages. You first test what you already know without looking at the material, then compare what you produced with the source and correct what was missing, then apply the knowledge in whatever format your exam is actually going to test, and finally use your performance to decide what needs to happen in your next revision session.

The order matters. If I open the notes immediately before testing myself, I am partly testing whether I remember something I looked at five minutes ago rather than whether I can retrieve it from long-term memory.

In pre-clinical medicine, particularly for essays, I would therefore try to produce an essay plan from scratch before going back to my notes. Once I had written everything I could remember, I could compare the plan with the lecture material, identify what I had missed and fill the gaps, which gave me far more useful information than simply reading another model answer.

For factual material, Anki achieves something similar by forcing you to answer before revealing the card. In clinical medicine, question banks became even more useful because they combined retrieval with application, forcing me not just to remember information but to discriminate between several plausible answers and decide which fact actually mattered.

This is also why I think application should begin much earlier than most students expect. If your exam contains essays, practise producing essay plans and essays throughout the year; if it contains SBAs, start doing questions; and if it contains OSCEs, you eventually need to practise examinations and histories rather than simply reading about them. Understanding the material and performing the thing that will eventually be assessed are related skills, but they are not identical.

For pre-clinical essays, I found that writing plans from memory and then using my notes and supervisor feedback to identify what I had missed was far more useful than endlessly rereading the same information. There were fewer useful questions for some of the pre-clinical multiple-choice papers, so the exact system necessarily changed depending on what resources were available, but the principle stayed the same: produce first, check afterwards.

Mistakes are probably the most useful part

The biggest improvement I made in clinical school came when I changed how I treated mistakes.

When you use a question bank, it is very easy to reduce the session to a percentage. You get something right and feel good, get something wrong and feel slightly annoyed, read the explanation and move on. The problem is that the wrong answers are often the most useful questions in the entire session because they tell you exactly where your knowledge or reasoning failed.

In How We Learn, error feedback is one of Dehaene’s four pillars because learning depends on comparing an internal prediction with what actually happened and then adjusting the model. That describes question-bank learning remarkably well: every incorrect answer is evidence that something inside your current understanding needs to change.

So whenever I got something wrong on PassMed, Quesmed or another question bank in fourth year, I started asking why I had got it wrong. Sometimes I had forgotten the investigation of choice, sometimes I had confused two similar presentations and sometimes there was simply one fact or contraindication that separated the correct answer from the option I chose.

Whatever that missing concept was, it deserved an Anki card.

The important distinction is that I did not simply copy the original question into Anki. I had tried repeating incorrect questions before, but sometimes I would get them right later because I recognised the wording rather than because I had repaired the underlying gap, so instead I made a new card testing the general concept that had caused the mistake. That meant the same piece of knowledge could then help me answer a completely different question in the future.

This significantly improved my marks because the question bank stopped being something that merely measured how good I currently was and became something that actively decided what I should learn next.

Spacing is what makes the memory last

Retrieving something successfully once does not mean it will still be there in three months, which is where spaced repetition becomes useful.

The basic idea is that instead of repeatedly reviewing the same information in one sitting, you revisit it over progressively separated periods so that each retrieval requires some genuine effort. The spacing effect is one of the better-established findings in learning science; a large review covering hundreds of experiments found that distributed learning generally improves later retention compared with massing the same learning episodes together, although the best interval depends on how long you ultimately need to retain the information.

This is the part Anki handles particularly well for me. Rather than deciding manually whether I should see a card tomorrow, next week or in a month, I use the scheduling system and concentrate on actually answering the cards.

It also means that revision does not need to begin again from zero every time exam term arrives. If important information has been repeatedly retrieved over the year, the period immediately before exams can increasingly be used for application, questions and identifying the remaining weak areas rather than desperately trying to relearn an entire course.

I would also review new material relatively soon after first encountering it rather than leaving the lecture untouched for several weeks, although I would not obsess over claims that everyone forgets a fixed percentage after exactly 24 hours. Real forgetting curves vary enormously depending on the person, material and how well it was learnt initially; the useful principle is simply that one exposure is usually not enough, and well-timed retrieval is more valuable than repeatedly rereading something while it still feels completely familiar.

The most important thing is still not more revision

The other mistake I made in pre-clinical medicine was assuming that every hour spent away from my desk represented an hour in which I could have been getting ahead.

During exam season I would revise for ten or twelve hours, stop going to the gym and reduce most of the things I enjoyed doing because I thought this was the price of taking the exams seriously. In reality, at least for me, the longer I tried to sustain that schedule, the less useful many of those hours became.

In fourth year, my gym session remained part of my routine even during exam term, and I became much less guilty about protecting time away from medicine. I cannot prove that continuing to exercise caused my results to improve, because plenty of other things about the way I studied had also changed, but I was certainly able to work more effectively while maintaining a much more normal life.

Sleep is particularly difficult to treat as expendable if the goal is learning. Modern reviews of memory research describe sleep as an active period of memory consolidation, during which recently encoded information is reprocessed and stabilised, while experimental evidence also suggests that restricting sleep impairs memory formation. Staying awake for another hour to squeeze in more revision therefore does not automatically give you an extra hour of useful learning.

Working smarter does not mean trying to find the minimum amount of work required to scrape through medical school. It means recognising that attention, retrieval quality, feedback, spacing and recovery all affect how much useful learning comes out of the time you put in, so simply counting hours is a fairly poor measure of whether a day has gone well.

There is one major limitation to all of this

There is a danger in becoming too good at optimising for medical school examinations, because an examination and clinical medicine are ultimately testing overlapping but different things.

Anki is excellent for making facts retrievable. Question banks are excellent for learning common presentations, finding gaps and distinguishing between similar answers. If you become very efficient at both, you can become very good at retrieving exactly the information an exam wants.

The uncomfortable part is that this can sometimes come at the expense of the bigger picture. When I am revising purely for an exam, I can find myself reading something with the aim of extracting the facts I need to memorise rather than sitting with the material long enough to understand the physiology, mechanisms and uncertainties underneath it.

That is probably a very efficient way of maximising marks, particularly in a clinical written paper that rewards broad knowledge and pattern recognition. It is not necessarily the same thing as learning medicine as deeply as possible, and real patients are obviously not single-best-answer questions with five conveniently separated options.

I therefore would not use any of this as an argument for abandoning lectures, textbooks, placements or deeper understanding. The initial big-picture model still matters because isolated facts become much more useful when you understand how they connect, while patients and clinical experience teach variation, judgement and uncertainty in a way that an Anki card never can.

Final thoughts

If I could go back to the start of medical school, I would spend far less time trying to make revision feel difficult and far more time thinking about what each hour was actually accomplishing.

I would still go to the lecture and use the notes to understand the topic, but I would move into retrieval as quickly as possible rather than reading the same material repeatedly. I would test myself before reopening the notes, compare what I could produce with what I should have produced, practise the format the examination actually rewards and treat every important mistake as information about what needs to be learnt next.

I would also be much more careful about the amount of time spent constructing the revision system itself. Spending an entire evening manually turning tomorrow’s lectures into flashcards might eventually create a very nice deck, but the deck is only useful because of what happens when you start retrieving from it.

The question I used to ask myself was how many hours I had revised that day. I now think a much more useful question is how much I can retrieve today that I could not retrieve yesterday, and which mistakes have shown me what I need to learn next.

There will always be more medicine you could read and another hour you could theoretically spend studying. The aim is not to win that competition.

It is to make the hours you do spend actually count.