Most of a doctor's training happens after medical school. A general practitioner (GP), emergency physician or surgeon can spend years longer in post-graduate training than they ever spent at university. That training is run by specialist colleges, and colleges have a different problem from medical schools.
A medical school has its students in its own building. It writes the timetable, employs the teachers and runs the exams. When it wants to know how a student consults, it can put them in front of an examiner and watch.
A college can't do most of that. It trains doctors it doesn't employ, in hospitals and practices it doesn't run, through supervisors who teach between their own patients. So the hard question for a college isn't what to teach. It's how to know its training is reaching the standard in places it can't see.
More doctors are heading into specialist training
New Zealand is training more doctors. Funded medical school places rise from 614 to 639 a year from 2026 (Ministry of Health, 2025), and to 759 a year from 2028, when the University of Waikato takes its first students (Ministry of Health, 2025). Those graduates reach the colleges a few years later.
Government is funding that next step too. From 2026, Health New Zealand pays the full training costs of an estimated 400 second- and third-year GP registrars each year (Royal New Zealand College of General Practitioners, 2025).
We've argued before that supervised practice, not student numbers, is the real limit on training more doctors. You can add seats. You can't add time. After graduation, that limit lands on the colleges.
How a college differs from a medical school
| Medical school | Specialist college | |
|---|---|---|
| Where learning happens | Its own teaching, plus placements it arranges | Hospitals and practices it accredits but doesn't run |
| Who employs the learner | No one; they are students | A hospital, a practice or health service |
| Who teaches | Faculty, plus clinicians on placement | Its own members, teaching around their patients |
| How it sees performance | Assessments and clinical examinations it runs itself | Examinations, supervisor reports and teaching visits |
| How long it is responsible | Until graduation | A whole career, through recertification |
None of this is a flaw. It is what specialist training is: learning the job by doing it, in real services, alongside people who already do it. The Medical Council of New Zealand accredits colleges on how well they manage it, including how they accredit training sites and supervisors (Medical Council of New Zealand, 2022). But it means a college vouches for training it mostly sees second-hand.
Most of a registrar's learning happens unobserved
A registrar spends most of the day seeing patients on their own. In an Australian study of 589 GP registrars across 70,412 consultations, registrars asked their supervisor for help in 8.8% of them (Magin et al., 2021, Health Education in Practice). The other consultations they handled without calling on anyone.
That is how independence is built. It also means help reaches the problems a registrar knows they have. The gaps they don't know about are the hardest to find.
Direct observation is how those gaps surface, and registrars value it. But their supervisors are seeing patients too, so observation happens less often than training organisations expect (Amarasekera et al., 2022, Australian Journal of General Practice). In the United Kingdom, only 52% of trainers could use their allocated training time for training (General Medical Council, 2025).
This is a time problem, not a quality problem. In Australia, 93% of GP registrars are satisfied with their supervisor's support (Australian Council for Educational Research, 2025). Where supervision happens, it works. There just isn't enough of it to go round.
General practice is the clearest example, but not the only one. An emergency registrar on a night shift or a physician trainee in a regional hospital is in the same position.
Good feedback doesn't need the supervisor in the room
There is a hopeful finding here. In Australian GP training, the odds of a registrar rating a teaching visit very useful were 12.8 times higher when the feedback was focused, specific and easy to act on. Visits done remotely were rated as useful as face-to-face ones (Fielding et al., 2025, BMC Medical Education).
So distance isn't the barrier. What limits learning is how much useful feedback a trainee gets between the moments a supervisor can be there.
A college is responsible for a doctor's whole career
A medical school's job ends at graduation. A college's carries on until the doctor retires. In New Zealand, recertification includes an annual structured conversation with a peer, colleague or employer about the doctor's practice, a professional development plan and continuing professional development (Medical Council of New Zealand, 2019).
Those conversations go further with something concrete to talk about. Doctors find it hard to judge their own performance: a systematic review concluded that "physicians have a limited ability to accurately self-assess" (Davis et al., 2006, JAMA). As artificial intelligence (AI) takes on more clinical work, knowing that a specialist can still reason through a case on their own matters more, not less.
What colleges need from a practice tool
This is the problem Gestalt is built for. We build consultation practice with formative feedback. A learner works through a realistic consultation and gets specific feedback on how they reasoned and communicated.
In a medical school, that adds practice to a timetable the school already runs. For a college, the job is different. Practice needs to:
- reach every trainee wherever they are placed, so learning doesn't depend on what each site happens to offer
- give feedback without taking time from a supervisor's list
- follow the college's own curriculum and standard, under the college's control
- give the college a cohort-level picture of where training is strong and where it is thin, while each trainee's practice stays their own unless they choose to share it
- carry on past fellowship, so recertification has something real to draw on
None of this replaces the supervisor or the examiner. Feedback from practice is formative. Every decision about whether a trainee passes, progresses or is ready to work unsupervised stays with the college. The aim is simpler: when a supervisor does sit in, the trainee has had more practice, and the supervisor's time goes on judgement rather than first corrections.
The extra medical students are already enrolled. In a few years they will be the colleges' trainees, and the colleges will need to see more of how their training is landing without asking more of the members who already give it.