What Should Medical School Admissions Look Like? (Part II)

In Part I, I wrote about the some more conceptual changes in how I believe medical school admissions could be improved. Here I'll give an outline for an admissions process I'd like to see. It won't be perfect - there are certainly important aspects I may have overlooked or addressed poorly - but it's an attempt. I welcome any criticisms or suggestions.

Introduction
I would still want to look at many of the same things current admissions committees look at, namely:

- Undergraduate marks

- MCAT

- Extra-curriculars

- Letters of Reference

- Essays

- Interviews

What differs is how I would use these factors. All these factorss are proxies for the attributes we actually want in physicians - intelligence, work ethic, empathy, communication skills, etc. They're imperfect proxies, but they have some overlap. I think admissions processes should capitalize on that overlap to better get at the underlying attributes rather than treating the proxies as sufficient measures alone.

Step 1: Is this person academically intelligent enough?

An applicant passes this stage if:

- They have alright marks and a very strong MCAT (~75% and 12/11/12 minimum)

- They have very high marks and a mediocre MCAT (~90%  and 8/8/8 minimum)

- They have good marks and a good MCAT (~80% and 10/10/10 minimum)

Rationale: The idea here is to capture a lot of people with the intellect necessary to be physicians. The numbers can be played with a bit depending on the applicant pool, Not trying to get the smartest people here, just a good number who are all likely to be smart enough.

Step 2: Does this person have a strong enough work ethic?

An applicant passes this stage if:

- They have stellar marks (~90% minimum) and ECs that show some productivity outside of their curriculum

- They have good marks (~80% minimum) and ECs that show a plenty of productivity outside of their curriculum
- They have alright marks (~75% minimum) and ECs that show extensive productivity outside of their curriculum

Rationale: We want hard-working physicians, whether they're working hard at school or hard outside of school. In evaluating ECs at this stage, the focus should be on the quantity of hours worked, not necessarily the quality (though effort involved should play a factor). Anything and everything should count here - jobs of any sort (including fast food or retail), child-rearing, taking care of family members - literally anything. Medical school admissions don't give enough credit to people who are struggling just to keep life going and thereby give preference to those who are fortunate enough not to have that struggle - this step could help to level the playing field, even slightly.


Step 3: Has this person explored the medical field?
An applicant passes this stage if:
- They have any schooling or ECs with direct exposure to medicine, health care, or individuals with medical conditions
- They have any personal experiences with the medical system, either from their own medical conditions or those of loved ones

Rationale: Very easy to pass this step - just a quick check to make sure applicants have actually investigated medicine as a career. Being a physician should be an active choice, not a default position.

Step 4: Does this person work well with others?
An applicant passes this stage if:
- Their LOR has no red flags

Rationale: Again, an easy step to pass. Anything above a dismal or indifferent LOR will be fine. The main challenge with LORs is having people who know the applicant well enough to vouch for their competency. That's an important skill in itself, especially in the relationship-dependent field of medicine.


Step 5: Does this person bring something to the field of medicine?
An applicant passes this stage if:
- Their ECs and essays demonstrate a good average degree of achievement and skill-development in the 7 CanMEDS roles, with achievement above a certain level in at least 5 of the 7 roles.

Rationale: Here's where the quality and diversity of ECs factors in. Ideally we want physicians who are strong in all the CanMEDS competencies, but that's unrealistic for individuals just starting their career. Being good - but not great - at all aspects is a reasonable place to start. So is being incredible in a select few, as long as there is at least some development in most of the other competencies. The main goal here is to cut out the one-trick ponies who might be amazing in one way, but sorely lacking in the other aspects. Not everyone who would make a good physician is going to be a great scholar, or a great manager, or a great collaborator. Yet, individuals with no experience in scholarly activity, no experience managing others, and no experience collaborating are going to be in trouble, even if they happen to be superb health advocates and experienced communicators. This is intended to be the main EC cut-off.


Step 6: Can this person communicate?
An applicant passes this stage if:
- Their interview scores (MMI scores ideally) are in the top three quaters of those invited to interview

Rationale: Interviews suck. They're high stress and often fairly subjective. Unfortunately, so is medicine, so interviews have value. This is especially true in the margins, either in the very high end or very low end. The idea here is to eliminate the low end. I prefer the MMI to allow for increased, independent sampling of applicants' interview ability (though some MMI questions are just plain weird).


Step 7: Putting it all together
This is the final step. An applicant passes this stage if:
- Their combined score from all aspects of their application is high enough to be granted an invitation for admissions

Rationale: A final rating is necessary in just about any large-scale application scheme, so here it is. I picture the combined score comprising 50% of the interview score and 10% for each of the following: essays, ECs (evaluated for work ethic as was done in Step 2), ECs again (evaluated for achievement as was done in Step 5), undergraduate marks, and combined MCAT score. It's all a bit arbitrary at this point, but the idea is that the truly deficient are already out by this stage, so the specific metric shouldn't matter all that much. I put the additional emphasis on interview scores because intelligence is almost never lacking in medical students, but the same can't always be said about communication skills.


Conclusion
Again, I won't claim that the system I've presented is perfect. In many ways, it's quite similar to what many schools do now, maybe just a bit more complex in some respects. The main differences I'd like to emphasize are flexibility in academic requirements as well as an overall increase in the number of applicants passing the initial cutoffs and receiving interview invites. This would undoubtedly be an expensive, labour-intensive system. That's by design, but it's not a small consideration.

Would this be a better approach? I have absolutely no idea! If anyone has an opinion on this, please comment. I've treated this mostly as a mental exercise. What do I care about when looking at future colleagues and what would I like medical school to look at? Ultimately, my process above reflects my priorities: I want to see a greater emphasis on work ethic, communication skills and diversity of experiences than currently seems to exist, with perhaps a somewhat lower emphasis on raw intellect. Individuals with different priorities would undoubtedly design a system with an emphasis on different qualities. Nevertheless, I enjoy discussions about what it means to be a good physician, and I can think of no better way to focus that discussion than on the topic of medical school admissions, where the physicians of the future are chosen.

Back Up Plans

Whenever I advise someone considering medical school on how they should be preparing, I almost always tell them to explore alternative careers and to have a back-up plan in mind that they could easily pursue.

I do this for three distinct, but related reasons.

1) Medical school admissions are competitive and unpredictable

Except for a few super-human applicants who have perfect GPAs, amazing MCAT scores, stellar ECs, and an amazing interview approach, no one has a guarantee of acceptance to medical school. Many individuals still have a very good chance of admissions, so the idea of having a back-up isn't a way of saying that the person isn't cut out to be a physician.

Hope for the best, plan for the worst. If you've got an interesting back-up career plan in mind, the worst ends up not being too bad. That brings me to my next point...

2) Medicine isn't the only good career out there

I like medicine as a vocation. That's why I'm pursuing it. But medicine is hardly a uniquely good profession.

Some aspiring future physicians focus in on medicine without bothering to consider the alternatives. These alternatives might actually be more enjoyable, fulfilling careers for some people. They're worth checking out!

Every career has its pros and cons. Medicine is no different. Knowing what else is out there helps provide perspective on why medicine is actually worth pursuing - and maybe, why it might not be.

3) Developing a back-up makes students better medical school applicants - and better physicians

Applicants who are clearly well-positioned to pursue another career often make better applicants. Their ECs are typically unique and show skill development. They have meaningful experiences to discuss in their essays and interviews.

Basically, it's clear they bring something to the table that other applicants won't. When these individuals start in medical school, residency, or practice, that time spent developing a back-up career isn't lost, it's applied to the field of medicine. One way medicine's great is that it's a very diverse field with many opportunities to draw on virtually any area of expertise. Having some additional, unique expertise going into medicine only helps - it's much harder to get that after getting admitted to medical school than it is beforehand!


For the most part, I tell people to explore alternative careers not because they can't or shouldn't become physicians, but because I think it's beneficial no matter what the outcome. It's helpful if a student doesn't get into medical school, if they choose not to go to medical school, or if they get in. To me it's a win-win-win.

Edmonton, Goaltending and Devan Dubnyk

Obligatory, random hockey post now that the NHL playoffs have started, the NHL draft order is set, and the awards are slowly being announced.

Devan Dubnyk earned a nod for the Vezina trophy this year, after basically willing Minnesota into the playoffs. Doubt he'll win it (it's Price's trophy this year), but it's well-deserved recognition.

On the same note, the team that cut him loose for not being good enough, Edmonton, lucked into the first pick in next year's draft... again. Except this year, they don't just get a good player, they get an amazing one. And so, after yet another disappointing finish to the year, Edmonton's in a rebuild, trying to turn their wealth of talent into a successful team. What does this have to do with Dubnyk? Well, once again, the goaltenders in Edmonton are taking the heat and many are calling for better goaltenders to be brought in.

But that was the same story with Dubnyk and he did amazingly well... once he got outside of Edmonton. Scrivens, who is arguably Edmonton's best goaltender right now, did pretty well in LA. So maybe the problem isn't the goaltender, maybe it's the team around the goaltender, whether it's the other players or the coaching staff.

Edmonton lucked into yet another first overall pick, and they've already largely squandered three. To keep from doing that again, I really hope they take a good, hard look at what absolutely needs to get fixed, and what could simply use some improvement. When it comes to goaltending, I think it's the latter. There's no point bringing in another decent goalie, just to have him underperform because the defense around him is terrible, or because the system put in place by the coaches doesn't provide enough support. Edmonton needs better D-men - no point worrying about the goalie until that aspect is at least half-way fixed.

Dog Ears

So, my dog has an ear infection. Nothing too worrisome, just a bit of a fungal infection from jumping in dirty water, which he does at every available opportunity.

But the fun outcome for me is that now I get to tackle him twice a day to shove this goopy anti-fungal medication down each of his ears. That wouldn't be too tough if he was a normal-sized dog, or if he was a calm older dog, but he's a 90 pound puppy who likes to run away at the slightest hint of something bad happening. When you get a dog, you never think that, to be a good puppy parent, you'll have to put your dog in a choke-hold and shove a syringe down his ear, but for the next week or so, that's what I'm doing!

Exam Studying

I'm in the middle of studying for my neurology exam right now... not a fun time. What's really weird about this exam is that there are some elements I feel like I know implicitly, without ever having studied them, while some subjects I'm not only struggling to learn, I'm struggling with how to learn them.

Learning in medicine is a lot like putting together a giant puzzle. Except, you don't know what the puzzle looks like. Or even it's shape. Or size.

When you learn a new fact in medicine, it's like getting an additional puzzle piece that you're expected to put into your puzzle. Except you don't really know the colour of your piece (or at best, you have a vague idea). It's shape also starts out a bit fuzzy - you don't really know what other pieces that piece connects to. This makes it very difficult to know where to place it, so you guess and hope it's in the right spot.

Over time, as the facts you learn once get repeated, the pieces become a bit clearer. The second time a fact is mentioned, maybe the colour comes into focus a bit better. The third time, an edge stands out clearly. The fourth time, another edge pops more plainly into view. With each repetition, it becomes more obvious where the piece should fit.

Neurology is hitting both ends of the spectrum for me. Cranial nerves I feel like I know, without ever having really studied them. They were introduced in my first year, several times. They've now been introduced again, in my second year, several times. I've never really sat down and studied them, but I also don't feel like I have to.

On the other end, I'm struggling with headaches (the topic, not me having headaches - though that second part may come in time). Headaches have really only been introduced once in my two years. I'm studying the material as best I can, looking for additional resources and trying to figure out all the details, but I don't feel like I have a good grasp of even the basics of headaches.

It really makes me wish more topics were taught like cranial nerves at my school - repeated instances, separated in time and in slightly different contexts. I know that I have to study and not everything can be adequately taught in school, but the divide in confidence I have in understanding these two subjects really strikes me. I'd love to see a lot more longitudinal learning with repetition of material than we currently receive.

Milestones

Medicine is a lot about milestones. The day-to-day activities, while important, seem like drops in an ocean compared to the big events that lead to becoming a physician. Writing the MCAT, Med School applications, CaRMS, qualifying exams... at times it seems like we're lurching from one milestone to another, just hoping we'll happily pass to the next stage.

I bring this up because of two events that have just past or are soon to happen.

First, what just happened. CaRMS just went through its second iteration, filling any spots left over from the first round. My school had an uncharacteristically large number of students in that match - I really hope it went well for as many of them as possible. From what I'm told, most if not all, are quality future physicians.

On the other hand, we have what's coming. Medical school admission decisions are only a few weeks away. Two years ago I was anxiously awaiting this time, barely able to think about anything else. Last year, I had a chance to help those applying along, to see the great number of stellar applicants. Many of them were admitted and are now my future colleagues (assuming I don't do something stupid like fail an upcoming exam... which I have totally studied for, sort of). I've had the great fortune of working with many of them over the past year. Perhaps unsurprisingly, I'm continually impressed with the dedication and competence I've seen in the first year students I've worked with - and I hope I've shown a similar level of ability!

It saddens me a bit, moving into clerkship, that I won't have much opportunity to work with the incoming students. As I've said many times, I really like the challenges of medical education - I'd love to teach in some capacity in the future - my experiences as a second year student have really been enjoyable. I may not have been teaching, exactly, but there were many times where I felt I could guide or advise others, just as more senior medical students guided and advised me. I hate to give that up.

Yet, clerkship is my next milestone, so it's something I have to accept. Moving from the classroom to the clinic is very exciting and is a progression in my training that I've been looking forward to. Then it'll be my electives, then CaRMS. More milestones to lurch to ahead.

I guess this is just my roundabout way of wishing everyone luck at enjoying the time in between the milestones - they're exciting and come more often than I initially thought - but it's the bits in the middle that I'll remember most fondly.

Talent

Every year, our school puts on a bit of a variety show, with each class contributing a decent-length skit to the show. I've not have had the pleasure of being involved due to lack of time and more importantly, complete lack of talent.

However, I got to see the finished product this year and it was nothing short of incredible. I don't have much talent for singing, dancing, acting, or theatrical production, but wow, my classmates certainly do. I get impressed on a reasonably regular basis at the medicine-related strengths of my fellow medical students - it's great to be reminded not only that they're capable future physicians, but capable people in general.

They put on a highly entertaining show with fairly strong production values, all the proceeds went to a deserving charity, and all those involved are going to be showing up for class on Monday to get back at it with the rest of us. Seeing their talent on display was both humbling and motivating - I like to think I work hard and do what I do well (I believe that's mostly true), but it's too easy to neglect the hard work and competence of others. It was on full display at last night's show.

Team-Based Learning

Over the course of my too-many years of undergraduate education, I've had multiple opportunities to engage in team-based learning (TBL). TBL is exactly what it sounds like - learning in teams. The basic premise is that students know more collectively than they do individually and pooling that knowledge in a team setting allows it to be disseminated between students within that team more effectively than a traditional lecture format or with a self-study approach.

I've seen this used to good effect (my most recent interaction with TBL was quite positive), but in more instances, my experiences with TBL have been frustrating or unhelpful.

Like any approach to education, TBL is a tool, one which is appropriate in some situations, but not so appropriate in others. You wouldn't use a screwdriver to hammer in a nail; likewise, TBL shouldn't be used indiscriminately. That's no knock against TBL - just like screwdrivers are great for putting in screws, TBL is great when used where it's most effective.

TBL has greatest value in application problems. To answer these questions, details need to be shared between participants. If there are any differences in interpretations of core concepts, they also need to be contrasted or discussed to come to a solution. A single person knowing a detail leads to everyone knowing it within the team while interpretations of concepts can be refined and disseminated among group participants.

Where TBL falters is in the introduction of new information, or the regurgitation of facts without application. No matter how big a group is, if no one within that group knows a piece of information, it won't get shared. Data needs to come from somewhere, so TBL cannot be used as a substitute for teaching core concepts or fundamental data. A shared framework is necessary for TBL to work. Likewise, using TBL to test simple knowledge retention is fairly inefficient. There's no impetus for discussion - either group members know a fact or they don't, so deciding on the "correct" answer to a question often comes down to a matter of democracy (the most popular answer wins) or confidence (the most assured person's answer wins). There's no real value to collaboration in this case.

TBL was primarily introduced into physics instruction, where it works quite well. Physics is a conceptually challenging field, where the primary difficult comes from application of concepts rather than memorization of those concepts. Since it has worked rather well in a troublesome area of study, it has naturally received a fair bit of attention from educators.

However, as too often happens in education (and medicine), a good idea in one area has been applied in areas to which it is ill-suited. Medicine can be taught either in a wrote-memorization format, or through applications. The former is a poor fit for TBL. The latter works well for TBL. I've seen both. My hope is that, moving forward, medical educators try to be more discriminating about the use of TBL or at least adapt their TBL sessions to better take advantage of its virtues.

OSCE Prep

One of the major tests in medical school is the dreaded OSCE. It stands for something, but I've honestly forgotten what. The basics of an OSCE involves simulating patient interactions in a short amount of time. This can be interviewing, doing a physical exam, maybe a combination of both, and then answering some questions on that fake patient's case.

I would argue it's one of the higher-yield activities we do, at least in pre-clerkship. They're great learning tools - they reinforce a lot of the book-learning we do and force us to think about that knowledge practically. They might be the only training we get in doing the simple things for patients, like introducing ourselves, draping patients appropriately for exams, and washing our hands. These elements to medicine matter as much as the memorization of small textbooks worth of medical information that we seem to do on a regular basis.

We had a practice session for our first real OSCE held by the graduating class at my school. They were great, giving excellent advise and tips along with the opportunity for extra OSCE practice. I can't say how much I appreciate the support from people who are basically done and have no real further obligations to the school.

However, one thing that was emphasized for me during these practice sessions was that I need to talk more. Not to the patient - apparently I'm great on that front. No, I need to talk more to the examiner. I need to say everything I'm doing, why I'm doing it, and I need to use the big doctors words too. That annoys me. It saddens me. It disappoints me. Not because the graduating class told me that, not at all. In terms of helping me pass my OSCEs, it's probably the best piece of advice I got and I'm glad they were willing to be upfront with me about it.

What troubles me about this recommendation is that to pass the OSCEs, you need to behave in a way that really isn't optimal or even appropriate for real interactions with patients. Since the OSCE is the major way in which medical students prepare their interviewing and physical examination skills, it encourages some habits which are far from ideal and may not be broken later down the line. I think most of my colleagues recognize the big things that you do in an OSCE but you shouldn't do in real life (like not explaining everything you're doing in complex medical terms), but there are some subtleties as well which may not get picked up by a decent number of medical students (like briefly explaining what you're doing in simple, non-medical terms).

Maybe it comes from working in healthcare before medical school, but I've learned that how you do something with patients matters almost as much as what you're doing with them. Granted, there are marks for patient care and rapport on most OSCEs, since schools recognize their importance. The problem is, as long as you aren't being an overt @#$hole to the standardized patient during an OSCE, you're not going to fail because of missed patient care marks. You are going to fail if you don't talk to the examiner enough, which is why my graduating colleagues were so consistent about emphasizing that to me.

I realize this is somewhat unavoidable - to accurately evaluate someone, you need to know what they're doing and why they're doing it, so they ask students to verbalize their actions. I don't really have a good suggestion around this. All I'll say is that this situation gets back to the fundamental divergence in goals between teaching and testing. Testing should evaluate teaching and adapt the evaluation to what is being taught, but in many cases, including the OSCE, teaching tends to adapt to the testing.

It may be worth spending some more time working on our patient interactions without a marking scheme to worry about.

What Should Medical School Admissions Look Like? (Part I)

It's a bit hypocritical of me to fault current admissions processes without providing an alternative, so I thought I'd take a second to present one. This was intended to be a single post, but it's a bit too big of a subject for that, so I've split it up into two parts.

This first part will focus on the more overarching principles I believe would be helpful to guide the admissions process, while the second part will present an algorithmic framework for admissions, that incorporates those overarching principles.

My Principles of Medical School Admissions

1) Admissions to medical school is arguably the single most important part of training physicians

2) There are numerous attributes we want in physicians - intellect, work ethic, communication skills, empathy - and it's more important that applicants be good in all attributes than great in some and deficient in others

3) Applicants should be willing to challenge themselves, even if it means making some mistakes or experiencing a degree of failure - as long as those mistakes or failures are not too great and were learned from

Principle #1 - Admissions Are Important
We need to have a rigorous approach to medical school admissions. In this post and in Part II, I will make some suggestions, suggestions I hope are based on a logical analysis of the current state of medical school admissions and, to the extent possible, have some grounding in available research. However, each one of these suggestions should be scrutinized and evaluated.

Additionally, we need to devote a significant amount of resources to the selection process, even if it somewhat reduces spending on actual medical education. Current admissions processes are often designed to minimize the work done by schools, typically by using somewhat arbitrary hard cutoffs to eliminate large swaths of applicants. To an extent, this is necessary, but it's become overused - weak applicants get cut out, but so too do a lot of strong applicants, applicants who may actually be better suited to being physicians than those admitted.

Spending large amounts on admissions seems like a waste - after all, wouldn't it be more productive to better educate those admitted? Yet admissions may be the most productive aspect of the whole process. A study on Ivy League schools in the US showed that these highly competitive schools did produce very good graduates - but so did second-tier schools when the students were able to get into an Ivy League school and simply didn't attend it. The Ivy League's greatest feat is selecting students who were good coming in, not necessarily in making those good students better. I see no reason the same situation isn't occurring in Canadian medical schools - where graduating medical students do well not because they received an amazing education, but because they had the capability to do well from the start and their medical schools merely succeeded in not screwing them up. Conversely, students who struggle as physicians may have been likely to do so from the beginning.

An even better admissions process could therefore lead to stronger graduating physicians. Since admissions is the major chokepoint between becoming a physician and not becoming a physician, it may be the only real point where our medical education system can meaningfully change the type of physicians in our workforce. It's worth spending some money on.

Principle #2 - Get Well-Rounded Candidates

The current admissions process is great at getting intelligent students. I've yet to meet a single person who is clearly not intelligent enough to be a physician - and more than a few who I wonder whether they're not too intelligent to be a physician.

Most students I've met have a good work ethic, though there are certainly exceptions. There are more than a few stories floating out there of clerks, residents, or physicians who clearly don't put the effort into their work or learning. These are the exception, not the rule, and it's entirely possible for work ethics to change as students or physicians become overworked, jaded, or depressed, but some individuals with a poor work ethic could be identified from the start.

Communication skills are a point of major concern, something emphasized by current research and patient surveys. Physicians as a whole are not great at communicated with patients, with other healthcare professionals, or even with other physicians. That's troubling in a field that is so dependent on good communication - medical errors can and do occur when miscommunications occur. We could certainly use some stronger communicators in medicine. (As an aside, it's also a reason I want to keep going with this blog, to maintain and improve my communication skills!)

Empathy is the tricky element. It's something every patient wants in their physician. It's almost impossible to evaluate effectively or efficiently. It's certainly important during patient interactions in developing a rapport and establishing patient preferences. Yet, there's some research that indicates fake empathy produces much the same as genuine empathy. In that sense, an effort to be empathetic may be the important metric here - and to hope that a person's intelligence, work ethic, and communication skills can compensate where empathy might fail (as it does for everyone at some point).

Moving the focus off of simply getting smart medical students requires relaxing what most admissions committees consider "smart enough". This means relying less on hard cutoffs for academic attributes and taking a harder look at more candidates. As long as the cutoffs aren't lowered too much, we'll still get intelligent medical students - the reliably-intelligent students I've met so far have included people with rather low GPAs or MCAT scores, thanks to the different ways individual medical school evaluate applicants. It also means not letting academic strengths make up for weaknesses in other areas, at least not to the degree that they currently do at some medical schools.

Of course, we still want to get medical students who are as strong as possible with all these attributes, including intelligence. All other things being equal, I'd rather have someone with a 3.99 GPA and a 99th percentile MCAT than someone with a 3.70 GPA and a 30 MCAT. But if the second applicant always goes the extra mile and can communicate with patients exceptionally well while the first lounges around and can't connect with a patient to save their life, I'd rather have the second person. Finding that balance is tough, but doable.

Principle #3 - Mistakes are Acceptable, Challenges are Desirable

The main point I'd like to make here is that using GPA to evaluate candidates rather than a more fluid metric like a percentage average may not be the best idea.

GPA does have some advantages. Because a GPA system is very hard on low marks, it encourages consistency, which is a desirable trait in medical students. It also prevents significant difficulties in some courses from being masked by extreme strength in other courses. Under a percentage system, a 60% plus an 100% is equal to get two 80% marks, while in a GPA system those marks are equivalent to getting two low 70's. Since 60% is almost a fail, that probably indicates some degree of poor understanding or effort in that course and it's worth discouraging low performance.

However, the current standards take these principles too far. It's one thing to discourage a 60, but the current standards at many medical schools discourage even the odd 70. You basically need over 80% in everything to be competitive at many medical schools - some schools basically want almost every mark to be above 85%. Furthermore, hard shifts between what's considered good and what's not mean minor differences in candidates get exaggerated - a 79 is worth so much less than an 80 in a GPA format it's terrifying, even though these marks are essentially identical.

GPA systems encourages two undesirable phenomenon. First, grade inflation, which could be it's own full post. Second, it encourages prospective medical students to go for the easiest programs and courses. If there's a significant possibility of getting a 75 in a course, it's not worth taking, even though in some courses, a 75 demonstrates a very high level of intellect and work ethic.

Schools have recognized this and have often put in place weighting schemes which allow for particularly bad courses or years to be dropped from consideration. These weighting schemes reduce some of downsides of using a GPA, but they're a kludge - they accomplish that goal in a relatively ineffective, inefficient manner. You still need exceptionally high marks within that weighting scheme just to get considered at these medical schools and these weighting schemes suffer from much the same problem a percentage system - namely that low marks can be compensated with high marks. Overall, I'd rather see these weighting systems used than the alternative, but their necessity speaks to the inherent weakness of using GPAs.

A percentage system is more fluid and less punitive for difficult programs. It can also be weighted in ways to try to regain some of the advantages of using GPAs. Very low marks can be punished by re-scaling marks below 70%. For example, keep a 70 worth 70%, but make a 50 worth 0% and scale everything in between linearly (so a 60 would be worth only 35%). Very high marks can be discounted to prevent them from having an undue influence - make everything over a 90 worth only 90%, effectively the same as what is done in a GPA system.

Overall, the point I'd like to make is that medical schools should be as critical of their application process as they are on their applicants. It's one thing to set high standards, but if those standards eliminate good applicants in favour of other who may not be as well suited to being a physician, it's a disservice to the applicants, to the school, and to the patients who will be relying on the selected medical students in the future.