more_vert
I HAVE been thinking and writing about overdiagnosis for some time, and sometimes thoughts and information can come together to make things clearer.
It’s good to see the Choosing Wisely campaign has come to Australia — it’s a positive development. However, in my view, it’s not ambitious enough.
Another stir of interest came along last month with the Four Corners report called “Wasted”.
The program discussed the lack of evidence surrounding a number of diagnostic tests and surgical procedures that continue to be done, and the links to the Medicare Benefits Schedule. The role of GPs in referring patients into diagnostic or treatment “journeys” was presented.
What the program didn’t include was a cogent discussion of the motivating factors behind overdiagnosis.
Most of us who order tests and referrals know intuitively what motivates us. If we’re honest, it’s often fear.
We may not feel scared but, deep down, we know that we can be crucified for “missing something”. It’s bad enough to be called out by a colleague over an error, or complained about by a patient — or even sued — but that’s not entirely it.
We are fearful that we might actually harm someone.
For all the accusations against doctors of corruption or greed, we care about our patients and the idea that we might miss an important diagnosis, and cause that person harm, is terrifying.
Fear is a strong motivator — stronger than insight or knowledge. That’s why education just isn’t enough.
Even though we might know what the guidelines say, or what the evidence shows, just one anecdote about an unpredictable event that went horribly wrong can have us retreat to primal emotion. Unless we understand and acknowledge these dynamics, we won’t be able to change them.
Much of the discomfort with the Four Corners program focused on two areas. First, the evidence surrounding effectiveness in subgroups of patients, and second, the role of the GP (and, I would add, emergency department doctor) as referral agent.
The way I see it, good medicine relates to basic skills: knowing your pathophysiology and being able to communicate with your patient. If we don’t trust our own clinical skills, and keep referring for tests or consultations with other doctors, then we teach our patients not to trust us.
We teach them to want a test, or to see a specialist, “just to make sure”.
The problem is, sending patients off for a test or another consultation usually doesn’t help to make us sure. Sending a low-risk patient for a test increases the risk of false-positive results and the harms that that can cause. Sending a patient with a benign condition to another specialist sends the message that we are worried that there might be something serious going on.
These fears transmit to the patient and can influence the subsequent testing and treatment offered by the consulting doctor.
If we, as referring doctors, inappropriately send a patient on a journey of investigation or treatment that results in inappropriate action, then we are complicit in that action.
What to do, then?
We need to be honest and kind to ourselves and each other, and to our patients. We have to accept that, although clinical judgement isn’t perfect, tests and procedures aren’t perfect either.
We also have to own all of our complications — not just the harms of “missing something”, but also the harms of “doing too much”.
Most of all, we need to make clear to our patients that, in doing less, we are caring for them more.
Choosing to do less isn’t always withholding something good from patients. Sometimes, it’s saving them from something bad.
We live in a great country, and we have some of the best health outcomes in the world. Let’s not be afraid.
Dr Sue Ieraci is a specialist emergency physician with 30 years’ experience in the public hospital system. Her particular interests include policy development and health system design, and she has held roles in medical regulation and management.
Loading ...
Don’t you think it is time “evidence based medicine” takes the blame for the problems it has caused. While excellent in principle, in practice this means that one cannot diagnose and treat a simple common condition such as a urinary tract infection or a pneumonia without “evidence” – a urine mc&s or a CXR to prove the diagnosis – tests which would not have been necessary when a general practitioner could diagnose and treat without having to validate themselves. Sending the patient to ED ? – additional insurance.
Glenn – as I said, my suggestions apply to any clinician who has to diagnose the undifferentiated patient. If you choose to displace the risk to ED, the principles still apply. “Missed diagnoses” are not the only cause of harm, over-diagnosis and over-treatment cause harm too. A prudent GP, or ED doctor, or any clinician assessing the undifferentiated patient will explain their cognitive process to the patient, document it logically, and provide a pathway for review and follow-up.
Sue, the gist of your article is… “If we, as referring doctors, inappropriately send a patient on a journey of investigation or treatment that results in inappropriate action, then we are complicit in that action.” “…make clear to our patients that, in doing less, we are caring for them more.” It is only in your answers to my challenge did you state: “we need to return to clinical skills, outline our cognitive strategy, and not rely on inappropriate imaging.” “we need to value detailed clinical assessment and reassessment.” In the case in question, had Dr Ferguson formally excluded photophobia, and made an assessment of neck stiffness she might have had a better defence. But ’no neck stiffness’ would have been legally challenged – in the presence of cervical spondylosis – and any admission of ‘some’ neck stiffness would have drawn the question ‘is not that a positive finding’ – and the next question ‘why did you not consider meningitis’? I challenge your statement: “ Applying my suggestions might not have changed the clinical outcome… but it would almost certainly have changed the legal evaluation.” Which suggestions? The law now expects us to formally consider – and competently exclude – meningitis with every presentation of headache and neck soreness. (There will be stiffness.) Every case of meningitis may well attract a multimillion dollar settlement. The prudent GP will document their findings of headache and neck stiffness, and IMMEDIATELY refer the case to your colleagues in ER, so they can carry the can for a missed diagnosis. Defensive medicine is expensive, but necessary.
I certainly agree that over-medicalisation is a problem and that much screening does more than harm.
Nevertheless, i think the arguments could have been put better than in this video. For example, you say that half of back images are unnecessary (very true), but that invites the question, what about the other half?
If I may be permitted an n=1 anecdote by way of example, in 1996 I suddenly found myself barely able to walk. Imaging (actually a simple X ray) revealed a spondylolisthesis. L3 and L4 overlapped by only 50%. Much more and my spinal cord would have been severed. So I opted for vertebral fusion, despite the long term risks. It gave me 12 more years of pain-free life. I walked across the Alps 5 years after the fusion (and with one prosthetic hip). It’s true that I’m now paying the price (the predicted kyphoscoliosis has set in and it’s not fun). But the alternative would probably have been much worse.
I’ve written about the dubious statistics of breast cancer screening too. But my wife was recalled after mammography, and biopsy revealed a malignant tumour (insofar as the present knowledge of pathology allow that to be judged). She lost a breast but no chemo or radio. Five years on she’s fine.
The problem with saying, quite correctly, that most people don’t benefit, is that there is no way to predict which people will benefit. The basic problem is that so little is known about medicine.
Sue, we evaluate clinical practice ‘in hindsight’. We learn (a cognate process) from our mistakes. Law’s claim is that whatever we learn after the event, should have been known before. The financial security of GP is under attack by legal advocacy. In the name of justice, it seeks to profit from our alleged mistakes. Compensation for clients, a substantial living for itself. Ultimately, the solution must be ‘no-fault’ compensation – as in New Zealand. We must campaign for this. Until then, it is prudent to be pro-active with pathology, imaging – and referral – to obtain as early and definitive a diagnosis as is possible. The other problem is the resource of general practice. Money is time – lack of money means lack of time. Medicare’s own statistics cite 41 as the median patient number seen a day. In 7 hours – with no time wastage (an impossibility) a consult is 10 minutes. The 6 to 8 minute consultation gives no time for medical intuition to speak. Not even time for medical intuition to develop. Is your response that – as a consequence – we must take more time? That means we must accept the inevitable cut in income. Having accepted a 50% reduction in the purchasing value of the GP fee schedule since 1986, we must accept further reductions in income. Income is inevitably inverse to quality. But medical practice has become ‘big business’, vertically integrated corporates want pathology, imaging referrals, and prescriptions for their pharmacies. Reasons for their GPs to spend no more than 8 minutes in a consultation. That present reality sets the benchmark. Primary care must change. We need competent ‘general physicians’. That requires adequate resource. Government must understand this.
Glenn, it is a congitive error to modify our clinical practice according to an unavoidable clinical error, or a legal error. It is also a cognitive error to evaluate clinical practice using hindsight. These two things happen all the time, but they are dysfunctional. Applying my suggestions might not have changed the clinical outcome – because cryptococcal meningitis in the immunocompetent is very rare and difficult to diagnose – but it would almost certaily have changed the legal evaluation. Our clinical evaluation and notes need to reflect a logical clinical process, including risk assessment.
I am not saying that malpractice litigation has no effect – I;m saying that, at least in Australia – it is not the main influence. ALso, despite your attempt to paint my article as an attack on general practice, the opposite is true – I am suggesting these principles for all diagnostic environments.
If you think my campaign is misguided, feel free to suggest a better alternative.
Sue, while both GP and ER can be ‘first responders’, ER is the usual referral resource for GP, particularly in emergency. GPs expect both better resourcing and higher emergency skills – also immediate access to specialist care – from ER, when they refer cases. For Ms Mules, that higher level of competence in the first ER session was not there. Or was the clinical picture still so ambiguous that the ER staff did not consider lumbar puncture was indicated? Using their ‘good clinical skills’? In cryptococcal meningitis, only LP that can establish the diagnosis. That is an invasive procedure. It had to be done.
Your ‘campaign’ is misguided. We must acknowledge that malpractice litigation affects medical practice. That is not ‘primal emotion’, it is cold, hard commercial fact. You admit ‘the judge erred’. Is the law a simpleton? Or is it looking after its own? ‘No win – no fee’ puts a legal firm at great jeopardy if they finally fail to win the case. It will bankrupt their client – who highly deserving of care. If they win the case, the disabled client is cared for, and both legal teams will be paid their costs. All from doctor’s pockets. Our pockets. Money for jam. THAT is the precedent set in Mules vs Ferguson. Expect an ever increasing flood of cases. The judgments – in sequence – may subtly twist and turn to achieve a desired end. Benefit for the claimant, and for the advocates of both sides. “You doctors are rich, and can afford to be charitable.” Will you be able to afford your malpractice premium in 5 years time? Consider, in Florida: ‘High-Spending Doctors [are] Less Likely to Be Sued”. Cited Medscape: http://www.medscape.com/viewarticle/854034
Glenn Rosendahl – you have misunderstood my comments. If you re-read my article, you will see that I am not setting up a battle between general practice and emergency medicine. ON the contrary, I am putting us both in the same “basket” as first-responders to undifferentiated presentations.
My comments are aimed at all settings where there is diagnostic risk: we fool ourselves if we think that more testing is a substitute for good clinical skills. As I said in the article: “Even though we might know what the guidelines say, or what the evidence shows, just one anecdote about an unpredictable event that went horribly wrong can have us retreat to primal emotion. “. We all know people who have dramatically altered their practice, or increased their risk-aversion, due to bad outcomes from unpredictable, and possibly unavoidable, bad outcomes. These reactions, though understandable, do not inspire good practice.
I agree that the judge erred. My campaign is to resist changing medical practice into what the judge says, but keeping it to what is best for the patient. In my view, this means that our patient assessment must be thorough and well-documented, our notes must reflect the cognitive process, and we must teach the community the limitations of tests. These prinicples can apply to all settings where there is diagnostic uncertainty. As a community, we must understand that there is no area of life with perfect outcomes and no human error – whether this is in the law, medicine, or any other aspect of everyday life.
I consider I am entitled to make a response. And that it be printed. Ms Mules DID subsequently attend a public hospital ER some days before the diagnosis in a private hospital ER. Did your public hospital colleagues ‘breach’ their ‘duty of care’ in failing to make the diagnosis of meningitis? Did they also fail to test for nuchal rigidity, take the history of headache, consider ‘facial flushing’, and note these symptoms were prolonged, unusual, and worsening. They packed her up and sent her home. By your own acknowledgement it is those ER staff who should have been held accountable. They had the resources, there and then, to do the necessary preliminary brain CT, and then the lumbar puncture. Why did they not do these things? Where was their ‘cognitive strategy’? I have been in general practice for 40 years. A year in the Solomon Islands, 7 years in Canada, the rest in Australia, from Norfolk Island to Northam WA, from Canberra to Weipa, Qld. From that experience I say that we GPs are patronised, disparaged, undervalued, grossly underpaid – and expected to work miracles. We do not deal with acute emergencies all the time. It can take perspicacity to recognise them. Not uncommonly – serendipity. You expect us to separate out the 98% of minor and trivial problems, deal with them. NOT expend resources on them – all in 6 to 10 minutes! ‘Red flags’ are inconsistent, commonly unclear and ambiguous in interpretation. You know that. The judge’s decision was wrong – and much will flow from it. We GPs will be practicing more defensive medicine, not less. And make no apology for it. We cannot afford escalating malpractice premiums.
There is quite insufficient space provided here to discuss this matter. So, very briefly, nowhere in the literature is ‘facial flushing’ cited as evidence of meningitis. It leads nowhere. ‘Neck stiffness’ is commonly found in cervical spondylosis and is difficult to qualitatively assess, let alone determine as a ‘cut-off’ ‘red flag’ for meningitis. Are you assessing the ease with which the head can be lifted (having instructed the supine patient to relax the neck)? Is ‘positive’ the chin touching the chest, ‘negative’ if not? ‘Meningism’ is the trilogy of headache, neck stiffness and photophobia – this last never mentioned in the case. A patient told me that interior lights bothered her eyes, but she had no problem outside. Is this ‘photophobia’? I quote a relevant, specific, extensive study. “…Sensitivity for clinical signs such as headache, vomiting, or fever was low, generally less than 30%, neck stiffness could reach 45%, but the absence of two signs among fever, headache, neck stiffness, and altered mental status eliminated meningitis with a negative predictive value of 95%.” http://www.ncbi.nlm.nih.gov/pubmed/19632074 ‘Clinical skills’ would not have picked meningitis at this point. ‘Serendipity’ might have. Meningitis is a problem of the very young, and of treated HIV. In 2008, statistically one case of cryptococcal meningitis in Qld that year. The imaging performed was not ‘inappropriate’. It identified a very likely cause for the patient’s complaint. So what should have been Dr Ferguson’s ‘cognitive strategy’?
Glenn Rosendahl – your case example illustrates exactly why we need to return to clinical skills, outline our cognitive strategy, and not rely on inappropriate imaging. Cryptococcal meningitis is extremely uncommon outside the setting of immunosuppression, such as HIV. Mules vs Ferguson acknowledged this. At the appeal, “The judge found that on 18 and 19 September Dr Ferguson breached her duty of care by not physically examining Ms Mules’ neck and enquiring further about her previously reported symptoms of headache and facial flushing.” Her symptoms were prolonged, unusual, and worsened.
SO, no, this case does not mean we need to exclude meningococcal meningitis in everyone with a headache – it means we need to value detailed clinical assessment and reassessment. Patients with headace and equivocal neck stiffness are referred for second opinions to EDs every day, but many others are successfully managed in general practice with a good explanation, reassurance and review, with referral if red flags are raised.
There is just one problem, medico-legal liability. Consider Mules vs Ferguson, decided by two High Court Judges in September. Ms Mules presented to Dr Ferguson in 2008 with headache, neck pain and a degree of neck pain and stiffness. No fever, no photophobia. A CT of the cervical spine demonstrated significant spondylosis. The trial judge considered varying facial flushing significant, and implicitly critiqued Dr Ferguson for not formally assessing ‘neck stiffness’. Ms Mules worsened, the public hospital ER sent her home without further assessment or treatment. Finally assessment at the private hospital identified cryptococcal meningitis. She has lost substantial sight and hearing, gained a settlement of $6.7 million. Allegedly, Dr Ferguson was remiss in not arranging a prompt consultation with a neurologist or general physician. The GP is expected to demonstrate the wisdom of hindsight before the event, competently take the history, examine, develop and prioritise the DD, act on it and record the entire event and process. There is no margin for error.Apparently there are 50 cases of this quantum now in process.A case today: headaches, major neck and shoulder pain with muscle stiffness, nausea, equivocal photophobia – apparently for months. I could not get her chin to touch her chest.This is technical meningism. I sent her to ER, my rationale for the referral explicitly citing Mules vs Ferguson. Warned them not to perform a lumbar puncture without first doing brain imaging.If we have to treat all such in this way, what will it cost?We now must treat all such in this way. That is the legal dictum.
I think there are a couple of aspects to this;
The first is ‘know your patient’ and this is where GPs have a big advantage ;eg John Smith comes to see me infrequently and when he does come there is usually something significantly wrong.
The second is medical intuition–to be differentiated from instinct which is innate unlearned behaviour–Intuition develops after many years of experience , is usually correct and in the presence of an equivocal clinical situation guides us on how to proceed.
The problem is that many of us will ignore our medical intuition and defensively order a series of investigations. A scan may then pick up an ‘incidentaloma’ which then leads to further anxiety and investigation.
Time is a good tool in medicine ; excluding emergencies there is little if any harm done by a short period of ‘watch and wait.’ From a purely medicolegal viewpoint it can be easier to defend a carefully considered non interventional approach than ordering a series of test but omitting the one test that would have given the diagnosis
Thanks for your comments, Paul. I am puzzled about these conditions you claim are “undetectable with standard medical testing protocols”. Could it be that these conditions you allude to are not underpinned by pathophysiology at all?
In my view, creating false certainty is no solution to a diagnostic dilemma. Understandiing the limitations of diagnosis, and accepting doubt and uncertainty, may be, however.
Perhaps the insecurity also stems from a lack of understanding/training in two of the major pathophysiological pathways to disease, 1. nutritional deficiencies, whether primary or secondary or conditional; and 2. functional somatic processes. These are often the earliest to present and yet are mostly unrecognised by GPs and specialists, and undetectable with standard medical testing protocols. An ability to understand the pathophysiological underpinnings of nutritional/functional impairments would go a long way to solving Dr Ieraci’s diagnostic dilemma.