Issue 42 / 10 November 2014

WHEN it comes to tackling quality and safety, modern health care is often compared unfavourably to the aviation industry.

In the past 50 years aviation has seen a dramatic decrease in passenger deaths with the introduction of standardised protocols and processes. At the same time in our health care system, up to one in every 10 hospital admissions still results in an adverse event.

However, as clinicians at the coalface well know, there is a world of difference between flying a plane and treating a patient. Both are complicated procedures, but while aviation is amenable to applying formulae and analyses of cause and effect, improving health care is a far more difficult proposition.

Health care is a complex, adaptive system. It is multifaceted and constantly changing. Delivering health care is more like raising a child — it takes a community to nurture it and help it grow, and to work together so that it can achieve its best.

That’s a key reason why the billions of dollars spent internationally in the past 25 years have failed to put a dent in the high rate of errors in medical care. We have been looking at things from the wrong angle.

We have traditionally waited for health care providers to make a mistake, then analysed what went wrong and attempted to put in place provisions to prevent it from recurring.

That might be fine for policymakers and managers remote from the clinical front line, who can only imagine how clinical practice is performed and rely on producing policies and instructions, all based on procedural standardisation and rooting out error.

But as every clinician knows, hundreds of clinical guidelines and procedures can be worth very little in the busy environment of a hospital or doctor’s surgery. How many clinicians remember — or are even aware of — the policy documents that are designed to promote consistency and prevent mistakes?

In reality, everyone is self-managing their risk. Surprisingly to some people, clinicians get things right most of the time. Yet, although safe and effective health care encounters are in the majority despite the inherent complexity, we have never studied them systematically.

To focus on things going wrong is to understand only a small proportion of what gets done. Shifting the focus to what goes right as well means we can begin to understand how clinical practice works both when it performs and when it underperforms.

If we can understand how mistakes aren’t made as well as how they are, then this might provide a blueprint for us to start having much more of an impact on quality and safety in health care.

We need to support clinicians who get it right, to study their practices and develop theories about why these work, and then encourage others to follow their lead. Essentially the emphasis would change: we should be striving to make more things go well in the future.

The Australian Institute of Health Innovation, now based at Macquarie University, Sydney, endeavours to understand what clinicians understand about the health system. We want to methodically and rigorously evaluate how clinicians get things right, rather than trying to prevent mistakes from being repeated.

Working with the Australian School of Advanced Medicine, the new Faculty of Medicine and Health Sciences, and Macquarie University Hospital, we have Australia’s largest research group studying clinical governance, health informatics, health services, and systems and safety.

This research can make a significant difference. In some ways it’s more important than splicing genes or developing new treatments for cancer.

After all, understanding how health systems operate and building on their strengths are the best ways to deliver health care sustainably and effectively into the future.

Professor Jeffrey Braithwaite is Director of the Australian Institute of Health Innovation at Macquarie University.

7 thoughts on “Jeffrey Braithwaite: The right way

  1. Steve Flecknoe-Brown says:

    Couldn’t agree more!  The Clinical Governance movement has gone off the rails in recent years, focusing on retrospective incident analysis with all its Severity Assessment Codes, Root Cause Analyses and so on. Sure enough, every time a problem is defined, the result is to build more complexity in the system in an attempt to prevent the problem recurring.

    Any engineer you speak to will affirm that they constantly strive for elegance. To an engineer, an elegant device or system is one with the minimum number of moving parts. Provided it does the job it was designed for, a simple system or device is far less likely to fail than a needlessly complex one.

    We need to get back onto the track of building a better, less fault-prone health system. That involves studying what works then making it work with less complexity. The way to do that is to look forward, making constant clinical practice improvement part of our core business.

  2. Dr Jack Dascalu says:

    I wholeheartedly agree with Prof J Braithwaite about the idea of understanding and gaining experience on how mistakes ARE NOT made in order to improve the rate of post admission complications. The simpler and most user friendly protocols could be more utilised in the atmosphere of busy clinical practice. Also upskilling of clinicians in the evidence based reduction of “misadventure” as it is now accepted would help. As part of our problem based learning it is now widely accepted that the above is taught and examined in the case of GP registrars.

  3. University of New South Wales says:

    Contratulations on your new move to Macquarie. Your points are well founded. There are many frontiers in patient safety that can still be improved through system changes and require the eyes of many different professions before the obvious area for design change is identified. Hand hygiene is a classic example of blaming the healthcare worker instead of changing the environment  to assist behaviour change; imagine the only location of a sanitizer dispenser is on the wall outside the patient’s room but the healthcare worker is expected to sanitize their hands multiple times while inside the room; e.g. after touching the patient’s immediate zone and prior to touching the patient –  the clincian would need to leave the room and come back inside many times. Yet, this is what I have seen in some private healthcare settings where the sanistizer is on the outside wall not at every bedside. It is preferable to use designing for safer patient care than relying on the slow and often imperfect behaviour modification. 

  4. Simon Knight says:

    It’s true that aviation safety has improved over the last 50 years, but probably due to technology (eg weather radar) rather than standardised protocols – although these helped. Most aircraft accidents result in multiple fatalities whereas medical adverse events tend to harm only one person, so stopping a few crashes can save thousands of lives

    There’s overlap between aviation and medicine, but I also agree there are differences. My military flying was as a solo pilot where ‘adverse events’ were managed by memorised immediate actions backed up by checklists. This was different as a commercial pilot where emergencies were 100%team decisions whlst implementing checklist items. Aviation emergencies are generally identifiable (if not always solvable). Also it didn’t matter which aircraft I was flying – if it went wrong it did so predictably. There was no normal variant of the hydraulic system! This eases diagnosis and management, making the checklist a powerful tool. More flexibility is required when diagnosing people and thus it must be expected errors will increase

    This doesn’t mean that there is nothing to be learned from aviation – there is. Besides checklists, which can be great, aviation is better with fault reporting. When a fast jet crashed there was always an investigation which gave useful feedback to avoid future similar happenings – and no doubt similar happens in medicine. Rather it is the ‘near misses’ that aviation handles better. Aviation is good at voluntary feedback – irrespective of how bad the transgression was. These are published anonymously in safety journals where others can learn from what almost happened. Medicine lacks this blame-free reporting culture

  5. Sue Ieraci says:

    Simon Kinight is right – while we can learn (and have learned) important principles from aviation, there are also serious limits to comparisons between aviation and health care. Aviation is entirely voluntary – flights are cancelled all the time. Health care is provided as needed and as requested, at all hours of the day, irrespective of resources available for service delivery.

    Many would say that health care is much more like traffic – highly complex, not totally predictable, essential to a funcition life, and highly operator-dependent. A society can minimise its road traffic fatalities, but never eliminate them, and the practicalities of use and cost call for many compromises.

    Read more about this approach here:

    Having said that, it;s great to see a positive approach to safety, as espoused in this article. The retrospective laborious analysis of every “incident report” , with an incentive to produce yet another policy for every specific situation, only adds to complexity and potential for error.

  6. David Henderson says:

    I was treating a man with moderate diabetic ketoacidosis.The BGL had fallen to 6.3 at the last reading 2 hours ago.  I asked for a new test to be told it was not due for another hour according to the protocol. This showed a lack of understanding of the management. The situation was rectified and the patient did well.

    There are many aspects to patient care. Many, but not all, can be made routine. Routine, protocols and guidelines are useful, but only with knowledge. They do not replace knowledge, and careful clinical processes.    

    Retrospective audit is useful, but only if analysed through the framework of knowledge. Analysis through a bureaucratic framework of adherence to guidelines may reveal some truth, but is really designed to maintain the status quo. Hence Dr Flecknoe-Brown’s observation; Clinical Governance can be used to avoid change.  

    Dr Patel was not identified as dangerous by bureaucratic process and he was rewarded for maintaining surgical activity. But the evidence suggests that he was unsafe becuse he lacked knowledge and judgement, was not collaborative, did not develop appropriate linkages and avoided scrutiny.   



  7. William Tarnow-Mordi says:

    One key strategy to improve patient safety is to use effective treatments that reduce severe illness and its complications. Effective treatments usually have only moderate benefits. Randomised controlled trials (RCTs) are the most reliable way to identify such treatments. Encouragingly, surveys show that at least half of treatments for patients in general medicine, psychiatry and other specialties seen in the Emergency Department are effective, based on evidence from  RCTs of new vs established treatments, or of two or more established treatments. 

    The result of any RCT is unpredictable but in a cohort study of Phase III RCTs with over 850,000 patients in various specialties, new treatments were, on average, more effective than established treatments just over half the time, and led to a small increase in survival.  This evidence was summarised by Djulbegovic et al in a commentary in Nature in 2013 entitled “Medical Research: Trial unpredictability yields predictable therapeutic gains”.

    In other words, to the extent that any specialty conducts RCTs of new or established treatments and applies their results, patient safety and other outcomes will improve. Acute Lymphoblastic Leukaemia is a dramatic example – 40 years ago 90% of these children died. Now, after continuous RCTs, 90% survive, with many fewer complications.

    As well as improving patient safety,  adequately powered RCTs provide perhaps the only viable strategy to ensure the sustainability of health systems, by reliably differentiating effective, cost-effective and ineffective care. Yet, we have barely begun to scratch the surface of the potential of RCTs. Many ordinary people don’t know what they are.

    How do we integrate RCTs into routine care?

Leave a Reply

Your email address will not be published.