‘Diagnostic Skills’ by Martin Docherty

Remember to check your blind spot

‘Diagnostic Skills’ by Martin Docherty

Thanks to Martin Docherty for an exceptional guest blog for the Clinical Maze

Martin is the Clinical Governance Manager Aviva UK Health ensurin implementation & review of clinical governance framework across rehabilitation services. He is also has a role with Partner Health & Care Professions Council as a Fitness to Practise Panel Member. Martin is well worth a follow on twitter here

‘The fact your patient gets well does not prove that your diagnosis was correct…’ (1)

Over the last year or so, we’ve seen an increased momentum behind the notion of physiotherapists taking a lead role in the management of musculoskeletal (MSK) disorders in primary care.  Talk of Vanguard projects, Sustainability and Transformation Plans, and the concept of therapists acting as First Contact Practitioners (FCPs) has featured regularly on social media forums.

For many in the profession, the adoption of a prominent primary care role will feel like a natural progression of the Extended Scope Practitioner (ESP) or Advanced Practice Physiotherapy (APP) posts which have been in existence for some time.  That said, the moves to develop and enhance our primary care remit in order to address the increasing burden of MSK disorders, has prompted some lively debate regarding the governance and medico-legal implications for physiotherapists.

The challenge to those responsible for implementing FCP roles is clear; they are tasked with delivering a specialist service which effectively triages patients with MSK disorders, that is capable of promptly identifying the possibility of serious pathology, and that reduces the volume of inappropriate referrals into secondary care.

A question which consistently appears during debates on this topic is whether, as a profession, we have a suitable framework to develop and maintain the skills required of an FCP clinician that is both credible and scalable at a national level?  More specifically, are we confident that our therapists are equipped with the diagnostic and clinical reasoning skills required to navigate what is an increasingly complex clinical environment?

My aim with this blog is to therefore provide some insight into diagnostic error rates in clinical practice, the factors which contribute to those errors, and the strategies we can use to improve our diagnostic skills and ultimately our clinical performance.  Although framed in the context of a physiotherapy FCP role, the blog may also be of interest to student and junior therapists, as well as other autonomous health professionals practising on the frontline.

What is the rate of misdiagnosis?

Managing MSK disorders in primary care is a colossal task.  The clinical environment in which we ask our clinicians to practise has all the ingredients required to make accurate diagnosis a challenge; multi-morbidity, a diverse range of conditions and age groups, short consultation times, and a relentless focus on cost efficiency (2).

Arthur Elstein, a cognitive psychologist who studied clinical decision making for much of his career, concluded that clinical diagnosis was wrong in 10-15% of cases (3). More recent studies suggest this percentage continues to hold true (4).  In a systematic review by Wallace et al (5) it was found that the most frequent cause of negligence claims in primary care related to a failure of or delay in diagnosis, accounting for 26-63% of all claims in their sample.  Earlier studies concluded that diagnostic error accounted for the greatest proportion of claims against GPs in the UK (6) and for a third of negligent adverse events in the US primary care (7).

The diagnoses most likely to be delayed or missed altogether are largely consistent across studies, namely cancer, circulatory disorders, fractures and infections (5).  Kostopoulou et al (8) found that typical reasons for misdiagnosing cancers was that they were rare (eg metastatic cord compression) and/or presented atypically (eg breast cancer presenting without a breast lump).

Before we consider the various methods used to identify diagnostic error rates, it is worth addressing two important medico-legal issues which invariably surface when discussing misdiagnosis.

Firstly, we should be clear that a delayed or missed diagnosis will not necessarily be proof of negligence per se; it may have been perfectly reasonable for a therapist to have formed an alternative, albeit erroneous, diagnosis based on a patient’s presenting symptoms and objective assessment.  In a study by Keillar (9), three patients were misdiagnosed due to ‘no-fault’ errors.  All three had presented with acute low back pain and had a history of cancer documented in their treatment records.  All three had normal blood tests, biochemistry tests and lumbar spine X-Rays prior to physiotherapy referral.  Moreover, all three had been discharged from physiotherapy reporting a significant improvement or full resolution of their symptoms.  The diagnosis recorded in each instance was mechanical low back pain.  Nothing odd about that, you might think.

Except that within five months all three were diagnosed with spinal metastases.  The reason(s) for the absence of any clues suggestive of sinister pathology and the apparent improvement with physiotherapy treatment is not fully understood.  Greenhalgh & Selfe (10) state that the relationship between time and development of signs and symptoms with serious spinal pathology is not always linear and that temporary improvements can occur.  Nevertheless, the uncomfortable conclusion is this: serious pathology may be missed or impossible to identify when presenting to a physiotherapist if the condition is early on in its evolution (11).

The second point relates to the Bolam test which we considered in my previous blog.  The key issue here is that, when applying Bolam to determine liability for negligent diagnosis, a therapist will normally be judged in accordance with the standard expected from practitioners of the same rank and experience.  So the standard expected of a Band 5 therapist will be that of another Band 5 performing a similar role rather than that of an ESP.  The caveat though is that a therapist must ensure she practises within the confines of her own knowledge and experience.  As soon as she steps into a more demanding role, a higher standard of care may be applied.  Be aware that where a therapist elects to perform the duties of a specialised role but lacks the necessary expertise, inexperience will be no defence to an allegation of negligence (12).

Measuring Diagnostic Error Rates

At present there is no mandatory reporting of missed diagnoses in the NHS or private sector so the true scale of misdiagnosis is unknown (9).  Existing research methods used to analyse diagnostic error rates tend to be retrospective in that they consider missed or delayed diagnoses which have already occurred (13).

Although increasingly rare, autopsy is considered the gold standard in providing the most definitive data on diagnostic accuracy, identifying major diagnostic discrepancies in 10-20% of cases (14).  The main drawback of this approach is that, much like our modern-day imaging techniques, autopsies discover a large number of incidental findings that are clinically irrelevant.  Equally, they only provide us with the diagnostic error rate in patients who have actually died, happily a rare event for MSK therapists.

Standardised patient studies can also be used.  Think of these patients as medicine’s version of ‘secret shoppers’.  Here, real or simulated patients with classical presentations of common diseases are sent anonymously into real practice settings.  Looking back at my time in clinical practice, I’m not aware of having been assessed in this way but perhaps that says more about my levels of vigilance than the true value of this approach in measuring diagnostic accuracy.  In any event, errors are inevitably detected in this way (15) but perhaps more alarming is that this approach reveals clinicians can even disagree with their own diagnosis when presented with the same case for a second time.  That awkward finding aside, a key benefit of this approach is that it provides an invaluable insight into the interaction between clinician and patient in a real-life setting, thereby giving us a sense of the non-clinical, system-related factors that can impact a therapist’s thinking (14).

Reviewing clinical negligence claims databases is another way of gaining some insight into diagnostic error rates.  However, analysing closed claims can only provide relative data on error rates. Estimating absolute rates is not possible because so few true errors result in claims, and of course not all claims reflect true errors.

Peer reviews, diagnostic testing audits, case reviews and voluntary error reporting systems are additional methods but, again, it is questionable whether any of these in isolation provide us with sufficiently useful data.  While voluntary reporting does not provide us with absolute error rates, it does provide us with some insight into the relative incidence of diagnostic errors compared to, for example, medication or treatment errors.  One Australian study (16) found that 34% of voluntary reports submitted over a 20 month period related to diagnostic error.  And interestingly, when compared with medication and treatment errors, misdiagnosis was judged to have caused the most harm but was considered the least preventable.

The inability to truly quantify error rates is even more acute in MSK physiotherapy as much of what we deal with will not have fatal consequences when we get things wrong.  But the implications for our profession, our patients and those who fund physiotherapy services are still profound.  Equally, existing research methods do not provide us with an accurate picture of the percentage of errors which lead to adverse events.  The prospective tracking methods required for such research are not easily executed and consequently rarely performed (4).  One hopes that, with the increasing use of electronic medical records, analysis of diagnostic errors in prospective studies ought to become easier.

What are the causes of misdiagnosis?

Of course, identifying that a diagnostic error has occurred only gets us so far.  Understanding why the error occurred is of equal importance as it enables us to take remedial action to prevent a recurrence.

We know that diagnostic error is due to multiple causes with cognitive factors being the most prevalent (17) (18).  Cognitive errors are where a clinician believes he has the correct diagnosis but is in fact wrong.  Rarely is misdiagnosis due to not having knowledge of the patient’s condition per se.  Instead, the cause is often attributable to a breakdown in clinical reasoning.  This may arise where an assessment is not conducted in sufficient depth, where a therapist fails to recognise the significance of information provided by the patient, or where there is a fundamental failure to draw all the assessment findings and hypotheses together to form a comprehensive picture of what might be happening with the patient.

Indeed hypothesis generation is key; selecting the right questions to test them even more so.  An American study found that failure to gather sufficient information during the assessment stage of a consultation was responsible for most diagnostic errors (17).  Perhaps more disconcerting is that, in the majority of misdiagnosed cases, the correct hypothesis had not even been considered (19).

The assessment process, where we integrate our clinical knowledge with the patient’s history and clinical findings, is largely subconscious and automatic.  These rules of thumb, or heuristics, facilitate rapid clinical reasoning and are particularly suited to common and uncommon conditions presenting in a typical way.  Using pattern recognition, experienced clinicians such as FCP’s will often recognise what is wrong with a patient from just a few features very early on in the consultation or will entertain a small number of hypotheses.  This approach has obvious benefits such as more efficient consultations, a reduced need for investigations and will lead to the correct diagnosis and management most of the time.

The risk, however, is that it discourages careful consideration of alternative hypotheses which require testing nonetheless.  The process may also fall down where a patient presents atypically or where the correct diagnosis mimics a more common and less serious condition.  Think cervical artery dysfunction presenting as neck pain. Or shoulder pain in the presence of a Pancoast tumour.

Graber et al (4) took a detailed look at the attitudinal and cognitive aspects of overconfidence which impact diagnostic accuracy.  If you only have time to read one of the papers which helped inform this blog, let it be this one.  It is a sobering read.  With regard to attitudinal aspects of overconfidence, even where a clinician has easy access to resources which can assist in determining a diagnosis, it has been demonstrated that such information is rarely reviewed (20).

Kostopoulou et al (8) argue another route to misdiagnosis is the misattribution of symptoms to an apparently obvious explanation or readily available cause, especially in the absence of typical features.  This is known in the mental health literature as “diagnostic overshadowing”, where an existing illness ‘overshadows’ and leads to the under-diagnosis of other co-morbidities.  In the world of MSK physiotherapy, a corresponding example might be the misdiagnosis of paediatric bone cancers where the therapist mistakenly assumes non-specific back or leg pains are due to growth spurts or sports injuries.

Minimising The Risks

Graber et al (4) point out that clinicians grossly underestimate their own diagnostic error rates.  While acknowledging the possibility of error, the tendency is to believe that mistakes are only made by others.  Addressing this dichotomy has to be the starting point.

They propose a two-pronged approach to reducing error rates; individual and system-based strategies.  Individual strategies will focus on a therapist’s skills of clinical reasoning, reducing the influence of bias, and helping them recognise the limits of their knowledge and expertise.  A clear emphasis is placed on reflective practice, not just on what has gone well but specifically where improvements are required.  I particularly like the idea of using the technique of prospective hindsight – having reached your diagnosis, ask yourself what the actual diagnosis might be if, at some point in the future, you were to discover that your initial thinking was wrong.

System-based strategies will compliment those above and typically fall into three categories; prevention, making errors visible, and mitigating the effects of those errors.  Access to clinical decision making support tools or peer support, and providing therapists with the time and skills to use these resources, will assist in preventing mistakes.  Mandatory diagnostic error reporting and accurate and timely feedback systems should help to address the gap between a therapist’s perceived and actual performance.  The key issue here is that the feedback includes an explanation of why an alternative diagnosis or treatment would have been more favourable, not simply whether the initial diagnosis was correct or not.

Employers are an integral part in this quality improvement process.  They must provide an environment in which physiotherapists are enabled to comply with their professional standards as set out by the Health and Care Professions Council (HCPC).  In relation to the patient safety implications of misdiagnosis, the standards on reporting concerns about safety (Standard 7) and being open when things go wrong (Standard 8) feel the most relevant.

And of course no clinical governance blog can ever pass by without mention the importance of accurate record keeping.  A lack of continuity in therapists can increase the risk of misdiagnosis so we need to ensure that our clinical reasoning is clearly evident to any other therapist who assumes responsibility for our patients.   Setting out why you have chosen to order specific investigations or highlighting the differential diagnoses you have considered or ruled out will act as an additional safety net that can only enhance patient safety.

I’m mindful of the fact that there is the potential for some unintended consequences here.  Encouraging therapists to reflect more at the point of diagnosis could realistically result in prolonged consultation times, an increase in requests for additional investigations and the inevitable discovery of incidental findings.  Ironically, it is feasible that diagnostic error rates could actually increase if we subject what are largely automatic and subconscious decision-making processes to conscious scrutiny (21).  Multiply a therapist’s potential options for diagnosis and, arguably, you increase the chance of him selecting the wrong one.

My own view is that it is a risk worth taking.  The spotlight is on our profession to ease the MSK burden on GP’s and evidence our ability to safely and effectively triage patients in a complex primary care setting.  In that sense, we need to demonstrate appropriate leadership by addressing the risk of misdiagnosis head on and implement the necessary controls to keep patient safety foremost in our minds.


It is not difficult to see why diagnostic error rates have remained relatively static over the years.  Pattern recognition will, in the majority of cases, lead to the correct diagnosis.  But its success, and the lack of a mandatory diagnostic error reporting, is also our undoing when it comes to improving our current performance.  The important point for those developing physiotherapy FCP services is this; ensure you have effective feedback mechanisms in place so that, should a patient suffer a missed or delayed diagnosis, the therapist can quickly identify and learn from such events.    In the absence of information that the diagnosis is wrong, the assumption inevitably is that it must have been correct.

“Doctors think a lot of patients are cured who have simply quit in disgust.” (22)

Above all, we should remind ourselves that the real risk of misdiagnosis is not necessarily limited to the complex cases.  Such patients naturally raise our levels of vigilance and prompt us to dedicate more time and attention to getting things right.  Instead, the real danger lies with the cases which appear routine but which we unknowingly have misdiagnosed.  It is these patients who are at the greatest risk of harm and it is their experiences which will inevitably dictate public confidence in our profession.

What is beyond doubt is that therapists’ clinical reasoning skills, their ability to reflect on their thinking processes, and their willingness to critically examine their assumptions, beliefs and conclusions are fundamental to the accurate diagnosis of patients.  Leaders of clinical teams must consistently review how their therapists are making decisions and make every effort to support the team to make even better ones.

The challenge of course is achieving the balance between supporting the accurate diagnosis of cases without adversely affecting the natural reasoning process.  No-one will argue with the need to be vigilant for the serious conditions.  But we cannot afford to let the pendulum swing too far in the opposite direction by adopting too defensive a stance.

Before making your diagnosis, consider performing a simple check of your clinical reasoning in the same way you would clear your blind spot before overtaking another vehicle.  In the majority of cases you will find there is nothing there and will be good to go with your initial instinct.  But on those few occasions where the correct diagnosis is lurking in the shadows, you will be glad you took the time to look.

(Martin Docherty is a chartered physiotherapist, Fitness to Practise Panel Member for the Health & Care Professions Council and Clinical Governance Manager for Aviva UK Health.

The views and opinions expressed in this blog are my own or as referenced below and do not represent those of my employers.  Anything posted on this blog is for general information only and nothing contained within it is intended to be or constitutes legal advice on any general or specific legal matter.)




  1. Samuel J Meltzer; US Physician & Researcher 1851-1921.
  2. Wallace E, Lowry J, Smith SM, et al The epidemiology of malpractice claims in primary care: a systematic review BMJ Open 2013;3:e002929. doi: 10.1136/bmjopen-2013 002929.
  3. Higgs J Elstein A: Clinical reasoning in medicine. In Higgs J, ed: Clinical reasoning in health professionals. Oxford, England: Butterworth-Heinemann Ltd, 1995, 49-59.
  4. Berner E, Graber M: Overconfidence as a cause of diagnostic error in medicine. Am J Med 2008; 121 (5 Suppl Diagnostic Error in Medicine): S2-23.
  5. Wallace E, Lowry J, Smith SM et al; The epidemiology of malpractice claims in primary care: a systematic review. BMJ Open 2013;3:e002929.
  6. Silk N. What went wrong in 1000 negligence claims. Health Care Risk Rep 2000; 13–16.
  7. Phillips RL Jr, Bartholomew LA, Dovey SM, Fryer GE Jr, Miyoshi TJ, Green LA. Learning from malpractice claims about negligent, adverse events in primary care in the United States. Qual Saf Health Care 2004; 13: 121–126.
  8. Kostopoulou O, Delaney BC, Munro CW; Diagnostic difficulty and error in primary care – a systematic review. Family Practice 2008; 25: 400-413.
  9. Keillar E; Are we missing any patients with serious spinal pathology? Int Journal of Therapy & Rehabilitation; Oct 2013, Vol 20, No 10, 487-494.
  10. Greenhalgh S, Selfe J (2010) Red Flags II: A Guide to identifying serious pathology of the spine. Churchill Livingstone, Edinburgh.
  11. Greenhalgh S, Selfe J (2003) Malignant myeloma of the spine. Case report. Physiotherapy 89(8): 486–8.
  12. Mardell J, Serfozo K; College of Law Publishing; Personal Injury & Clinical Negligence Litigation 2014.
  13. Car et al; BMC Family Practice, (2016) 17:131. Clinician identified problems and solutions for delayed diagnosis in primary care: a PRIORITISE study.
  14. Graber ML; BMJ Qual Saf 2013; 22: ii21-ii27. The incidence of diagnostic error in medicine.
  15. Gorter S, van der Heijde DM, van der Linden S, et al. Psoriatic arthritis: performance of rheumatologists in daily practice. Ann Rheum Dis. 2002;61:219 –224.
  16. Bhasale A, Miller G, Reid S, Britt HC. Analyzing potential harm in Australian general practice: an incident-monitoring study. Med J 1998;169:73–76.
  17. Gandhi TK, Kachalia A, Thomas EJ et al. Missed and delayed diagnoses in the ambulatory setting: a study of closed malpractice claims. Ann Intern Med 2006; 145: 488–496.
  18. Graber ML, Franklin N, Gordon R. Diagnostic errors in internal medicine. Arch Intern Med 2005; 165: 1493–1499.
  19. Kostopoulou O, Devereaux-Walsh C, Delaney BC. Missing celiac disease in family medicine: the importance of hypothesis generation. Med Decis Making (in press).
  20. Rosenbloom ST, Geissbuhler AJ, Dupont WD, et al. Effect of CPOE user interface design on user-initiated access to educational and patient information during clinical care. J Am Med Inform Assoc. 2005;12:458–473.
  21. Dijksterhuis A, Bos MW, Nordgren LF, van Baaren RB. On making the right choice: the deliberation-without-attention effect. 2006;311:1005–1007.
  22. LaFee S. Well news: all the news that’s fit. The San Diego Union-Tribune. March 7, 2006.
Clinical Maze

The Clinical Maze APP is coming soon

The Clinical Maze APP will be made available to all course participants. Check back soon for more updates and information

Coming soon!

We Value Your Feedback

Spread the word, improve our courses or even add another course to the Clinical Maze.

Feedback form

Host Your Own Clinical Maze Course

If you or your organization are interested in hosting a course please click here

Go to form