Issue 28 / 23 July 2018

RAPID advances in computing and artificial intelligence are forecast to change medical education. As part of the spectrum of medical education, continuing professional development (CPD) is likely to be similarly transformed over the next decade.

CPD is part of the medical education continuum between post-graduate medical training and transition to retirement. Arguably, CPD is the most important component of medical education as it covers all Australia’s medical practitioners working as clinicians (83 731 in 2015) for the many years of their professional practice, many more person-years of education than medical students and post-graduate trainees combined. Because of its long duration, CPD is more subject to change over time. CPD is increasingly recognised as an academic discipline with its own curriculum, evidence base, educators and governance structures. In Australia, the accrediting body for CPD is the Australian Medical Council (through accreditation of specialist medical colleges), the training organisation is the specialist colleges, and the training provider is health services, colleges and self-directed learning.

Computing power has been growing exponentially since Intel released the first microprocessor chip in 1971. The over 400 000-fold increase in computing power has enabled transforming advances in medical education and practice. For example, Microsoft HoloLens is a head-mounted self-contained holographic computer that enables mixed reality, where the user can interact with digital content and holograms. The smart glasses are available at a cost of $3000 each. Case Western Reserve University, in the United States, plans to use HoloLens to teach anatomy in the near future.

Computing power also enabled deep learning, the latest advance in artificial intelligence, a type of machine learning where networks are capable of unsupervised learning from data that are unstructured or unlabeled. Deep learning is now readily available from sources such as Google and is already being used in diagnosis, including photographic screening for diabetic retinopathy, where the results surpass human ability and speed. When coupled with big data analytics (extremely large datasets that may be analysed computationally to reveal patterns), with additional data sources, including the published medical literature, deep learning will offer significant improvement in health care decision making as well as a reduced cost of health care. My Health Record, Medicare and Pharmaceutical Benefits Scheme data, health care apps and smart wearables and patient collaboration via social media are all possible additional data sources.

In this rapidly changing environment, the role of medical professionals will change and CPD will be part of the change. Screening will be performed by machines rather than medical professionals or paramedical staff, and the role of doctors will be to interpret the machine-generated results. Arguably, humanistic aspects of medical care will become even more important in direct patient care. The curriculum of CPD will need to expand to include understanding of data analytics and deep learning. Medical professionals will need to control datasets and teach machines deep learning algorithms, keeping control of the process against “the rise of the machines”. Humanistic health care skills (“human ware”) will be highly valued and must be nurtured through appropriate CPD programs. Humanistic skills will also be necessary to avoid burnout in the new medical environment.

New surgical techniques and technologies continue to develop; many have a steep learning curve. New CPD programs will allow proceduralists to learn new technologies via mixed reality and patient avatars. Learning outcomes from CPD will be able to be assessed far more accurately by mixed reality assessment of competence (Moore’s level 4) instead of surveys of participation or satisfaction which are current practice.

Audit and feedback are thought to be critical in CPD-driven improvements in patient outcomes. Advances in data analytics will allow medical professionals to interrogate their patient records to understand patient outcomes and improve quality of their care. Likely, deep learning will detect patterns that doctors have not identified among their own patients.

Outcomes of CPD also ideally include measures of community health. Big data analytics will facilitate measurements of these outcomes but will require large and novel datasets to do so.

Medical regulatory bodies will also have access to advances in technology. The Medical Board of Australia, as part of its Professional Performance Framework, intends to “proactively identify […] doctors who are either performing poorly or are at risk of performing poorly, [and] assessing their performance […] [Age], professional isolation and multiple complaints are all risk factors for poor performance”. Such an approach will be facilitated by data analytics. Practising doctors will need excellent medical records and understanding of their patient population and comorbidities to ensure that external analysis of their data is accurate and interpreted correctly.

Thus, advances in technology will transform medical practice in ways that cannot as yet be clearly predicted. The role of CPD will expand to allow doctors to be experts in machine data and dataset interrogation, including much wider sources of data than today. Most importantly, humanistic aspects of medical care will become highly valued and CPD will play a critical role in allowing doctors to further develop these skills and prevent burnout.

Heather Mack is President-elect of the Royal Australian and New Zealand College of Ophthalmologists (RANZCO). She is an Honorary Senior Clinical Lecturer in the Department of Surgery (Ophthalmology), University of Melbourne. She practises in Melbourne and has a special interest in medical retina and retinal degenerations. She was Chair of the RANZCO CPD committee (2004–2011) and led the development of the current RANZCO CPD program.

I acknowledge Helena Prior Filipe, Chair of CPD for the International Council of Ophthalmology, my long term CPD collaborator.


The statements or opinions expressed in this article reflect the views of the authors and do not represent the official policy of the AMA, the MJA or MJA InSight unless that is so stated.

To find a doctor, or a job, to use GP Desktop and Doctors Health, book and track your CPD, and buy textbooks and guidelines, visit doctorportal.

3 thoughts on “CPD of the future: artificial intelligence and big data

  1. Anonymous says:

    The Problem with CPD – is that it becomes anti-social when the requirements become so structured and so demanding that society suffers by the professional not having the the time for family and pursuits to develop as a normal member of society. Governments will then take over the CPD by regulation – so the Colleges must restrain themselves from being counterproductive to Society by developing perspectives of medicine as part of the present ‘quality of life’ – so onerous CPD will become the next problem in society . Much as the College of GPs who have burrowed techniques from Communism e.g. self reflection – outmoded concept of brain washing from the Cold War period.

  2. Anonymous says:

    I admire your enthusiasm, Professor Mack, but feel, as with many similar articles in MJA Insight, it is from a “rarified air” of academia & specialist practice that you speak! It shows the dream of technological outcomes rather than the reality. As a practicing GP of over 30 years experience, I can advise you that the reality of interacting with the “computer world” is a very different experience at the GP level.

    PBS add-ons to office computer software slow consultations significantly because it relies on excellent broadband access and speeds…this is rarely achieved during usual practicing hours in most of Australia. Interacting with NBN is like trying to have intelligent dialogue with a stone, and internet provider promises, like political promises, rarely turn out to be fulfilled.

    My Health Record, like it’s precursors, is an absolute nightmare, and with the usual political intransigence & wilful dismissal of concerns concerning medico-legal & privacy issues, the Government has now adopted the “opt-out” model for this train wreck, no doubt hoping that most of the population won’t notice before it is too late to opt out.

    I feel this is possibly unconstitutional, and perhaps an attempt by the faceless bureaucracy to achieve The Australia Card by stealth.

    It will not achieve the projected health benefits in the vast majority of cases, and is yet another unpaid administrative burden being forced onto General Practice by bright “know-it-alls’ with little life/real-world experience or who’s part-time practice income is supplemented by outrageously inflated fees tor IT services. They are wittingly or unwittingly assisting the politicians in ruthless cost-cutting & avoiding paying real dollars for real services by real professionals! The PIP payments to general practice are another way of avoiding actually paying proper rebates for GP consultations that allow sufficient time for the patient to receive adequate care.

    As for Medicare…another impersonal nightmare! The appalling amalgamation of 2 Government agencies with totally different roles & philosophies is illustrative of the breathtaking indifference of politicians to human reality & needs. Many of my patients are not computer “savvy” or even don’t have access to a computer. They wait for hours in the grossly understaffed Medicare offices to get help (I use this term very loosely!) and generally end up with conflicting advice and sometimes feel like they are in some way a criminal!

    In my experience, patients really dislike being referred to get services on line, finding it confusing, complicated, time consuming & frustrating. As an example of a woefully designed website, just look at MyGov!! Indeed much of the population hate being unable to speak to a real person to clarify concerns…banks, telephone providers, government agencies, etc, etc. are illustrative of this, too. I spend a portion of most consulting days assisting patients with these problems, and in some cases having to provide stress counselling!

    The other thing that is overlooked in all this rush to technology is that patients are not very happy with the whole idea, especially the older patient. They appreciate the warmth & depth of human interaction. Most still resent the time GP’s spend typing notes into the computer, yet we are expected to keep a FULL, contemporaneous record, and are not paid for any work done if we are not face to face with the patient! Any supposed income supplement from PIP payments is a joke in monetary terms, and are often targeted to currently politically sensitive sections of health care.

    Computer resources as a tool, like references, diagrams and perhaps diagnostic algorithms are useful….though I still get better patient understanding of their problem if I use my old anatomy atlas to explain the problem. the current practice software programmes are beneficial to the consultation especially medication lists, allergies, results, etc., as long as they are properly maintained…the original computer adage applies..GIGO (Garbage in, garbage out)!!

    As for CME using computer/VR technology…I can see its use in procedural techniques, especially in your field, Professor. BUT, as a general tool…when the VR programme interacts as a real human experience (smells, vomits, bleeds, assaults or abuses you or dies in reality with real raw emotion on all sides) it will be useful, but not until then! There is no substitute for learning medicine & human interaction in the consulting room or at the bedside, and preferably including teaching from experienced individuals. Art as well as science!

    Politicians, Regulators, and Governing Bodies love the computerisation of things – it means they don’t have to do any real work, spend less money on wages, and appear to do things that often don’t occur in reality!

    This world, and especially Australia, is racing to computerisation, often with much more consideration to shareholder profit and tax budget ideals than human benefit, neglecting the ultimate result – no humans needed at all! Afterall, they cost too much, are inefficient and wear out. Much simpler to stay with machines!

  3. Geoff Chapman says:

    One of my earliest mentors (and he was very wise) told me that I would not find trouble by not knowing, but rather I would find a lot of trouble by not looking.
    Perhaps the computers can do Inspection, palpation, percussion and auscultation ??

Leave a Reply

Your email address will not be published.