About Me

My photo
I'm a radiologist and writing helps me make sense of the world.

"My method is to take the utmost trouble to find the right thing to say, and then to say it with the utmost levity" -George Bernard Shaw

Monday, 3 September 2012

"We are not in the business of exclusion"

An "exclusiometer"?
Although Professor Sir Howard Middlemiss died nearly 30 years ago, his memory lives on. A radiological giant of his time, I once heard a quote attributed to him that further elevated him in my mind. It was said that he wanted a sign put up in the reception area of the radiology department of the Bristol Royal Infirmary that said, “We are not in the business of exclusion”.

This adage could equally be a rallying cry for twenty-first century radiologists. Particularly if you have high clinical standards and loathe sloppiness of medical thought, word and deed. Specifically, if you believe that sloppiness of deed stems from sloppiness of word, which in turn derives from sloppiness of thought.

When I was a newly appointed consultant, the terms “Rule out x” or “Exclude y” on a radiology request used to be minor irritations. I was more forgiving. Nowadays I just can’t abide them. It seems that every other request now bears these poisonous words. So I asked clinicians why they write this. They always state, “Well obviously I didn’t mean that”. So why did you put it then? I just don’t understand why it is so commonplace in modern medicine. I don’t know what has changed.

Some of my colleagues frown at me, puzzled at clumps of hair that I’ve just torn out (again). Indeed many of those reading this may be wondering why I am getting all het up. Just why is ‘exclusion’ or ‘ruling out’ so bad? It isn’t pedantry, honestly; let me explain.

First up, it is an anathema to high standards of medical practice. It heralds a move away from medical diagnostic acumen, based on knowledge and expertise. The backbone of Western Medicine, probabilistic hypothetico-deductive reasoning, is rejected. Diagnostic indecision is embraced “just in case”. It is clinically and morally lazy.

Second, it reveals ignorance on behalf of the requester. Now I am no guru of clinical epidemiology. Quite the reverse. I’ve always found medical stats textbooks a reliable cure for insomnia. But ask a junior doctor about Bayesian probabilities and 9 out of 10 of them fish-mouth quite predictably. I remind them that virtually no radiological test is sufficiently sensitive and specific to ‘rule out’ anything. Ever. Occasionally there is a glimmer of recognition when I mention pre- and post-test probabilities. Not often, though.

Third, such language influences thoughts. I hear otherwise sensible doctors say things like, “Well the CT excluded a hepatorhubarboma”. To which I cannot help but say, “No, the CT is normal but it doesn’t exclude a hepatorhubarboma”. Clinicians develop inappropriate dependence on and faith in test results, forgetting the patient whose signs and symptoms are screaming the diagnosis.

Fourth, it is wasteful of taxpayer’s money. Inadequate triage leads to high negative rates and cost inefficiency. For example, an audited of our last 1000 CT urograms and the pick up rate for upper tract TCCs was just 3%. Similarly, ~25% of CTPAs were positive a decade ago; it felt like a worthy and interesting thing to do. Now the pick up rate is in single figures. Normal pulmonary arteries were once a rare and fascinating structure but the sheer volume of the blighters we all look at has rendered them distinctly mundane. Finding a significant thrombus these days brings the registrars running to gawp.

Lastly, and most importantly, “ruling out” is bad for patients. It is not that patients like having tests (they don’t). Nor is it that tests are risky (they most certainly can be). It isn’t even the diagnostic delay or personal inconvenience that multiple investigations bring. It is that the doctor: patient relationship is betrayed: inappropriately extensive investigation apes thorough and hence caring medicine. The truth is that the doctor is transferring risk to the patient under the guise of being thorough. “Ruling out” is defensive medicine, plain and simple.

The only way forward is a zero tolerance approach. All papers requests are stamped with Prof Middlemiss’s wise words and politely returned. It wouldn’t be too much to programme any ordercomms system to screen for the offending phrases (including the ugly “r/o”) and automatically reject them with the message, “Error 404: it looks like you are trying to phrase a clinical question using a negative diagnostic paradigm”. If they do it three times, the system crashes irreversibly. Clinicians will soon get the message.

Tuesday, 8 May 2012

Therapeutic Investigation

The great thing about a career in medicine is that everyone can find a niche. An old medical friend said radiology was definitely the best choice for me; locked in a dark room. Such was the cliché about radiology; a boffinish speciality for introverts who were bright enough but best kept away from patients. Imagine my surprise on day one of radiology school when most of the consultants were affable and ebullient.

When I was a young(er) radiologist, this ‘talking to patients’ notion was seen as a luxury extra. The implication was that we weren’t there for that kind of thing. Radiology was a technical speciality and we were there to learn the science behind it. Communication with patients was purely a means to an end. A handshake and a few kind words to break the ice before the enema tube went in; warnings of a “small prick in the back of the hand” as the cannula was inserted. The finale of the investigation was a smile and “the result will go to your doctor”; that was about it, really.

The topic of ‘communication in radiology’ exclusively focuses on the radiologist: clinician dynamic. The notion that a radiologist could have anything of use to say to a patient has been ignored. To me, this is daft. I think we are missing out on a trick here. I think radiologists can help patients feel better through communicating with them.

Like many radiologists (but not all) I have regular patient contact. Albeit mainly via ultrasound jelly rather than, say, a 20 gauge needle. Every patient walks into the room anxious and with an agenda. When I say agenda, I mean emotional and participatory needs as well as questions. Some seem disengaged; some only seek bland reassurance whereas some harbour specific questions. There is no greater source of professional satisfaction for me than for patients to smile and genuinely thank me on the way out. Its not that I need my ego massaging (quite the reverse, my wife is quick to point out). It simply means I have allayed their anxiety, sussed their agenda and provided answers to spoken and unspoken questions. I take a few moments to explain any pathology that I’ve found, be it innocuous or bad news. I also like to give an idea of further management steps. Even if I haven’t provided a definitive answer, I’ve explained the issues clearly enough for them to be happy.

My partially retired GP father talks of “therapeutic investigation”. This is where investigation is a part of the consultative process, part of healing. Obviously those coming for radiological investigation should be treated with warmth and compassionate, rather than like pieces of meat in a sausage factory. That isn’t all. To extract the maximum clinical utility from any investigation, we need to understand the agendas of the clinician and patient alike. This way, not only does the clinician get a clinically useful report but the patient has a therapeutic experience. By the time patients leave the radiology department, they could already be back on the road to recovery.

I’m not advocating psychoanalysis on the ultrasound couch nor breaking bad news on the CT gantry. It isn’t the right place. The point is that any patient interaction has an element of consultation and every consultation has a potential therapeutic effect. You may well think, “That’s all well and groovy but I don’t have the time”. Neither do I. My lists always run late. Some of that could be teaching registrars to scan. When it takes them 5 minutes to find the portal vein (confidently labelling it “CBD”), this eats into your time still further. But I don’t want to cut corners. I want to demonstrate the value-added service that a consultant radiologist can offer; a service that patients and clinicians both appreciate.

This comes at a time when besides being overworked, there is immense pressure for ‘efficiency savings’. The problem is that quality measures are unquantified in radiology. To an external auditor or manager, they may as well not exist. They are only interested in how cheaply and quickly our work is done. Quality never enters into it.

I argue its virtually impossible in radiology to be simultaneously cheap, quick and good. You can have a fast turnaround that is cheap but it won’t be any good. You can have a fast turnaround and a good service but it won’t be cheap. Time to dig our heels in: put quality first every time.

Thursday, 12 January 2012

Threshold Concepts and the Ha-Ha Effect

We’ve all been to disappointing lectures by noted experts. The reasons vary but a mild anticlimax is commonplace. Perhaps your medical hero has an unexpectedly reedy voice and bad teeth. Or they come across a bit weird: nasty-and-sinister weird, not pleasant-yet-eccentric weird. It has been said that a professor is like an undescended testicle: difficult to find but when you do then they are often malignant.

Nevertheless, keynote speakers consistently pull in the crowds, flocking to gather the pearls of wisdom that drip from their cherry lips. However, what often follows is anaesthetic induction by PowerPoint. This is the main reason to dread talks by Big Names. The speaker, rapt with academic fervour, launches into their talk and promptly loses everyone within the first 5 minutes.

At worst, they are so self-absorbed that they fail to notice the audience move first from bemusement to disengagement, then from boredom to frank annoyance. Occasionally the speaker briefly looks away from their over-crowded PowerPoint slides and notices the fidgeting, frowning and yawning. But rather than pausing to reflect, they assume the audience are obviously ignorant and rude. Flushed with annoyance, they speak faster and faster, skipping slides just to get it over with as fast as possible.

The glorious inability of specialists to explain their work is well recognized. They have known so much about their field for so long that it is difficult to for them to even comprehend that normal mortals find such issues difficult to understand. Because they find it easy, they can’t see why others don’t ‘just get it’.

Audiences are normally polite and socially subservient. Consequently they wouldn’t dream of saying that they didn’t understand a single word. Feedback forms are curiously muted and faintly positive. The expert gets no inkling that they were unintelligible, and so the cycle continues.
The curiously named ha-ha (pictured above) is a useful metaphor for the differing perspectives of expert and those listening to them [1]. It is a deep ditch used in 18th-century country estates to separate gardens from surrounding pasture. It is deliberately invisible from the house so as to give the impression that the house is in the midst of untamed nature. However, from the pasture, the sheer brick wall is obvious and impassable.

The expert is standing in the garden; the novice (or audience) is in the pasture. The expert can see no reason why they shouldn’t come and join them. The novice clearly cannot access the garden without help.

This unscalable wall represents a threshold concept [2]. This notion describes “opening up a new and previously inaccessible way of thinking about something” and representing “a transformed way of understanding, or interpreting, or viewing something without which the learner cannot progress”. Coming to grips with a threshold concept is inherently difficult: extensive knowledge is a prerequisite and the integration of such knowledge can be counterintuitive. Such concepts take time to grasp; rarely are there epiphanic “Eureka!” moments.

For example, take the ‘silhouette sign’ on chest radiographs as a radiological threshold concept. Knowledge of physics, anatomy and pathology are united into a radiological diagnostic sign. You may exclaim, ‘Pfft: easy-peasy!”. But you may change your tune after trying to explain it to 3rd year medical students. The memory of the transformative process quickly fades. And it is an irreversible process. Once you have climbed over the ha-ha, there is no going back.

So, how can the expert help the novice over the ha-ha? The natural response is to create a simplified version of a complex problem. Whilst superficially attractive, this should be avoided as the novice is falsely reassured that they are fully trained. Unless the principles are grasped, novices can be unprepared for the complex realities of clinical practice.

The best we can do is to recognize that certain issues are difficult to grasp. Don’t expect ‘first-time’ learning: break the concept down into small parts, explain carefully the thoughts and actions, give simplistic examples, use visual aids, check for understanding and build up gradually.

Richard Feynman, the late Nobel Laureate in physics, prided himself on being able to explain even the most profound ideas to students. His lectures and writings are the epitome of lucidity. But even he knew his limitations. When a TV reporter asked him if he could briefly explain what he won his Nobel Prize for, he replied, “If I could explain it to the average person, I wouldn’t have been worth the Nobel Prize”.


[1] Academic Medicine 2009; 84: 954
[2] Higher Educ 2005; 49:373