Key Points
- Not defining the system and thinking it through and instead just resorting to the animal needing a minimum database which may confuse the issue.
- Know what’s common. “Common things do occur commonly but that doesn’t mean uncommon things don’t occur.” This can be difficult to know as a new graduate but you will learn this with experience.
- Do not over or under-interpret the clinical pathology by not analysing it with clear questions. You need to be able to differentiate the causes of vomiting from the consequences of vomiting.
- Look at the clinical pathology systematically. The clinical pathologist’s comments can be helpful but remember they have not seen the animal and things are changing.
- Be careful about spending too long treating symptomatically and not recognising when an animal needs further investigation. “Delay in further investigation can sometimes be due to client reluctance but clients will be reluctant if you don’t explain clearly – they are more likely to follow your advice if they understand why you are making the recommendations and the obvious caveat of this is that you must understand what you are recommending and why.”
Professor Jill Maddison is a world-renowned small animal medicine clinician. She is currently the Professor of General Practice and the Director of Professional Development and Clinical Extramural Studies at the Royal Veterinary College, London where she lectures on clinical reasoning and clinical pharmacology. She is also the chairperson of the World Small Animal Veterinary Association’s continuing education committee, the senior editor of Small Animal Clinical Pharmacology and senior author and editor of Clinical Reasoning in Small Animal Practice. Her distance education course, Internal medicine: a problem solving approach, through the Centre for Veterinary Education is routinely sold out.
Today, Professor Maddison will talk to us about clinical reasoning and how you can apply it effectively to work up your cases.
“One of the things that veterinarians do naturally is the process called pattern recognition,” Jill says.
As soon as we have any sort of knowledge or experience, we start to try to match patterns of clinical signs with diseases we know about.
“It’s something that doesn’t have to be taught. It’s really based on knowledge and experience.” “When it’s correct it’s great, it’s quick, it’s cost-effective.”
However, problems occur when:
– We cannot match the pattern to a known disease or diseases, or
– We match the pattern to an incorrect diagnosis.
If we rely heavily on pattern recognition, when it fails we do not have an intellectual framework to fall back on to solve the problem. Instead we have to start again, and according to Jill, “we go fishing and we get more and more data and we do more and more tests but now we’ve really lost sight of what we’re looking for.”
Jill says that some veterinarians may think that, “all I need to do is get more knowledge and get more experience and then pattern recognition will solve everything. But it’s been shown time and time again that it doesn’t.”
The alternative approach to pattern recognition is a structured clinical reasoning process that can be applied to any clinical sign or scenario. “At its essence we’re always asking questions that drive what we do.” Jill says, “a structured logical reasoning process doesn’t always work – nothing always works – but it works pretty well most of the time.”
The Structure of Clinical Reasoning
- Define the Problem
“What are the problems that this animal has? I don’t mean, what’s the diagnosis? I mean what are the clinical signs? What is it showing? How important do I think each one is? And, am I sure I know what the clinical sign is?” This is the time when you ask yourself if you are sure, “the animal is vomiting as opposed to regurgitating, the red urine is blood as opposed to haemoglobinuria, the animal is fitting as opposed to having syncope, the animal is constipated as opposed to having large bowel diarrhoea.” It seems simple but if this question is not asked and answered the investigation can go very wrong from the start.
Construct a prioritised problem list which involves listing the clinical problems and prioritising them into what is important diagnostically and therapeutically.
- Define the System
Once the problem list has been created the next question to ask is: “What system is involved with each important or specific clinical sign?” For example, vomiting involves the gastrointestinal system. This can be further refined into primary or secondary causes. For example, “a problem directly involving the gastrointestinal tract is called primary structural gastrointestinal disease. Something outside the gastrointestinal tract, like liver disease, kidney disease or hypocalcemia, is called secondary gastrointestinal disease.
- Define the Location
Some problems (not all of them) can be further defined by the location within a system. For example, the small or large bowel within the gastrointestinal system, intra- or extracranial disease, upper or lower respiratory tract. For secondary gastrointestinal diseases Jill recommends just taking a mental walk through the abdomen: liver, kidney, adrenal glands etc. Diagnostic tests may help you focus on a location.
- Define the Lesion
This involves identifying the lesion or pathological process causing the problems – this essentially becomes your differential list. Some people like using acronyms like “DAMN IT V” (Degenerative, Anomalous, Metabolic, Neoplastic, Inflammatory, Traumatic, Vascular categories of disease).
Jill says the goal is to reach a stage where her clinical reasoning process is used unconsciously. “It’s not something that you can pick up only when you want to use it for a complex case,” she says. Jill recommends using the logical clinical reasoning process for every case, no matter how simple. “If you don’t practice it and get the approach embedded for the main clinical problems, when you get faced with the cases where it really helps you, then it’s going to take you longer.”
Jill uses an ice skating analogy to explain the learning process:
- Unconscious incompetence
When you are little and you have never seen ice skating before. “You can’t do it but you don’t know you can’t do it.”
- Conscious incompetence
When you try skating for the first time and you fall over. “You can’t do it but now you know you can’t do it.”
- Conscious competence
You start practicing. You can skate but you fall over when people touch your shoulder. “You can do it but you have to think about it.”
- Unconscious competence
You practice more and now you can skate easily, talk while skating and turn around easily and you do it without thinking about it. “You’re in a state of unconscious competence and on top of that you can learn to dance. You can do the hard stuff.”
Malcolm Gladwell in his book, Outliers: The Story of Success, talks about achieving mastery in any skill after 10, 000 hours of correct practice. “The problem in veterinary practice is your 10,000 hours of experience can be wasted if there’s no structure underneath how you’re thinking,” Jill says. “Your experience might not help you solve the complex stuff.”
The Balance of Probabilities
In general practice, vets often do not have a full data set and need to make decisions with incomplete information. “The legal analogy is that if you get charged with murder you can be found guilty but only if you’re guilty beyond reasonable doubt,” says Jill. “The burden of proof is very high.” “If you’re charged with a civil offence, the burden of proof is lower so you can be convicted on the balance of probabilities. Medicine is about the balance of probabilities and the evidence when we are in general practice can be quite scant on which we have to make decisions”.
Jill says she thinks young vets struggle with this because as students they wanted everything to be black and white. It is important to understand making decisions based on probabilities does mean errors can happen. Jill says, “even the most experienced vets can make errors and will make errors.”
To minimise the risk of this happening, after finding the most likely explanation for a patient’s presentation, we also need to consider other explanations especially causes which if we missed or did not treat now would lead to disastrous consequences. We also need to know when a definite answer is required. For example, if you suspect a dog has hypoadrenocorticism you want to prove it with an ACTH stimulation test because it requires life-long treatment.
Case Study 1: Applying the balance of probabilities
You are presented with a 16-year-old cat which, “occasionally goes a bit funny and weak. It has seizures every so often and it’s a little bit ataxic in its hindlimbs,” Jill says.
The key question is, “Do I have primary or secondary neurological disease?”
There are no significant findings on the blood results hence it is likely to be a primary rather than secondary neurological problem. The presence of seizures also indicates there may be an abnormality in the forebrain. The neurological examination may be normal or may indicate deficits consistent with forebrain disease.
The cat may have neoplasia, immune mediated inflammatory disease, an abscess or toxoplasmosis for example. Gold standard diagnostics would include a CSF tap and magnetic resonance imaging (MRI). As MRI cannot distinguish between neoplasia and inflammation a brain biopsy may be needed for ultimate diagnostic proof. However, if the owner cannot afford more sophisticated tests you will need to treat based on the balance of probabilities.
In this situation there is a small probability of toxoplasmosis. You may specifically test for this disease or treat any potential infections with a course of clindamycin. “Next, if I am sufficiently comfortable that infectious disease is very unlikely I’m going to treat symptomatically (anticonvulsants) and I may give my patient steroids in case there is a tumour or inflammatory disease present,” Jill says. “Client communication of course is paramount so that they are informed about the pros and cons of treating this case on the basis of probabilities rather than the greater chance that a certain diagnosis will be reached if the case was referred to a specialist.”
Case Study 2: Episodic weakness, PU/PD, Pallor patient
You are presented with an animal with reduced exercise tolerance and short periods of weakness for one week. It has been drinking a lot of water for 3 weeks and seems to be urinating a lot. It has not been eating well for 3 days. The clients are most concerned about its inappetence.
To start prioritising the list, Jill says, “on the balance of probabilities if I solve [the problems of] why this animal’s weak and why it’s drinking too much water, I’m going to know why it’s inappetent.” Hence the inappetence is not so important (even though it is important to the clients).
Prioritised problem list:
- Episodic weakness
- PU/PD
- Inappetence
On the exam, the patient is dehydrated with pale mucous membranes. Jill says although the dehydration needs to be treated it does not help her diagnostically so it is less of a priority. “The pale mucous membranes is an important clinical sign because it could be due to anaemia or poor peripheral perfusion.”
The Revised Problem list:
- Episodic weakness
- Polyuria/polydipsia
- Pale mucous membranes
- Inappetence
- Dehydration
To define and refine the problem and identify what systems are involved you should be asking your key questions about the first three problems on the list.
Prioritising the list takes training. Jill says she often sees veterinary students when they first start prioritising the list, “they’ll put dehydration and inappetence up the top because the inappetence is the first thing that is mentioned by the client and dehydration because they can fix it.” Reordering is critical otherwise you will focus on problems you do not have to worry about.
The questions you need to ask are: “Episodic weakness: is it a primary or secondary neuromuscular problem? The polyuria-polydipsia, is it primary polydipsia or primary polyuria? And if it’s primary polyuria, is it structural or functional renal disease? And the pale mucous membranes, is it anaemia or poor peripheral perfusion? These questions are what’s going to drive everything that you do.”
In contrast, the question you would ask yourself if you are using pattern recognition is:
“What diseases do I know of that would cause weakness, polyuria, polydipsia and pale mucous membranes?”
You may know some differentials that fit but you will miss others. By separating the problems out and having your key questions in relation to defining the problem and system, the pathway and the options become very clear.
The key question for episodic weakness is: “is this primary or secondary neuromuscular disease?” Jill says. “If it was primary structural neuromuscular disease, my next question is, why would it have PU/PD and pale mucous membranes?”
From this you can deduce it is likely to be a secondary neuromuscular disease and so you can start thinking about the causes for example, electrolyte, glucose and oxygenation abnormalities.
“For PU/PD, it’s an unwell animal so it’s not going to be primary polydipsia,” Jill says.
The next question is then, “does this animal have primary structural kidney disease or does it have functional kidney disease?”
Functional kidney diseases are diseases outside of the kidney which are impairing its function including hypercalcaemia, hyponatremia and liver disease. We might not know what is impairing the kidney but as you have asked the question your mind is now awake to the possibilities.
“The next question, for the pale mucous membranes, is it anaemia or poor peripheral perfusion?”
After asking these questions you know you need to check for evidence of:
– Secondary neuromuscular dysfunction
– Renal function
– Function (secondary) renal diseases
– the presence of anaemia
Some may say it’s a sick animal and so you would do a full biochemistry profile anyway. However, asking the questions first will allow your brain to assess the blood results in relation to the presenting problems, rather than in isolation.
“[In this example, let’s assume] the animal is profoundly azotaemic,” Jill says. “It’s got anaemia and it’s also hyperkalaemic because it’s got acute on chronic kidney failure and everything kind of fits.”
“If it’s a unique pattern of clinical signs then pattern recognition is brilliant.”
If there’s only two or three explanations for the pattern of clinical signs and they are easily tested pattern recognition is fine as well. However, when the problem is more complex clinical reasoning is important. Let’s say it is azotaemic, hyperkalaemic, hyponatraemic and the liver enzymes are also mildly elevated.
The liver enzymes are unlikely to be significant but from there you may have two possible explanations. One potential mistake would be to make a diagnosis of renal disease based on the polyuria/polydipsia and azotaemia.
Another approach would be to realise the azotaemia may be caused by a functional renal disorder. In this case the patient has hyperkalaemia and hyponatraemia and the sodium-potassium ratio is less than 24. Changes in potassium and sodium can cause azotaemia. Let’s also say it has hypoglycaemia, hypercalcaemia and it doesn’t have a stress leukogram. Addison’ disease is now a significant differential. Your next step would now be to perform an ACTH stimulation test.
Case Study 3: The Vomiting Patient
You are presented with a 4 year old Dalmatian called Poncho who has been vomiting for four days and is now not eating.
- Questions to ask:
“Is there anything else going on?” Other clinical signs like polyuria will make secondary gastrointestinal disease more likely.
“Is there any sign of diarrhoea?” It’s important to specifically ask about the faeces. If he has profuse diarrhoea, a primary gastrointestinal disease is more likely.
“Is he defecating normally?”
“Has there been any change in his water intake?
“How’s he been in himself? Is he depressed? Is he happy?”
From these questions let’s say, “he’s been bright for most of that time and that in the last day he’s become a bit sadder and he’s not eating very well.”
- “[Make] absolutely sure he is vomiting and not regurgitating or gagging after coughing.”
Firstly, ask these questions:
“When does the vomiting occur in relation to eating?” “If he is consistently vomiting in relation to eating then it’s more likely to be primary gastrointestinal disease or pancreatitis.” Regurgitation is still a possibility.
What is he vomiting? “Is there signs of mucus in it?” Mucus suggests it is more likely from the oesophagus and hence regurgitation.
“Is there any yellow material in it?” Yellow material is bile which indicates vomiting.
If Jill is still unsure about whether the patient is vomiting or regurgitating, she may observe him herself by performing a feeding trial. She could see what he does when he eats food or ask the owner to film him vomiting with their phone.
Identifying if the patient is vomiting, regurgitating or you-can’t-quite-tell is a crucial step.
“If the dog has regurgitation, regurgitation is a nasty clinical sign that means something bad. Animals who are truly regurgitating have often got [something] horrible like a foreign body in their oesophagus.”
- If the dog is definitely vomiting then the next question is: Is this due to primary or secondary gastrointestinal disease?
If there is primary gastrointestinal disease he could be systemically well or unwell, and may or may not be vomiting related to eating. However, if an animal is vomiting related to eating, has no systemic signs of disease and is otherwise well he is more likely to have primary gastrointestinal disease. If this is the case, “blood tests aren’t going to tell [you] why he’s vomiting. Blood tests don’t tell you anything about the gut in relation to why they’re vomiting. They might tell you the consequences of vomiting but not the cause.”
Jill may treat a patient symptomatically at this point if she suspects a transient primary upper gastrointestinal disease, for example dietary indiscretion.
If the patient been consistently vomiting after eating for four days, which is a long time, Jill would be more concerned of an upper structural gastrointestinal disease. This is when she starts defining the lesion: could it be a foreign body, tumour or inflammatory gastric disease that’s more than just a toxin? Imaging (plain radiographs, ultrasound and/or contrast radiographs) is indicated. Jill would still run blood tests to assess secondary electrolyte derangements, hydration and to rule out comorbidities but blood tests are not as diagnostically important in this case as imaging the gut.
Consider a scenario where the dog started vomiting at the same time he became inappetent, so there is no relation between vomiting and eating. He is vomiting yellow material and a bit of water and it happens without consistent association with eating. He is also getting progressively more depressed. “There’s no really strong clue that [he] has primary gut disease.” It could be primary or secondary and this is when blood tests (complete blood count, biochemistry including electrolytes and cPLi) are recommended. The secondary causes need to be ruled out and if they are then it’s more likely to be a primary gut disease problem. “There’s one complicating factor and that’s the pancreas.” “The pancreas [has] got a bit of fear of missing out (FOMO) syndrome. It wants to be the gut and it wants to be the liver. When it’s inflamed it behaves like primary gut disease – at least in dogs. In cats, not so much.”
“If it’s primary gastrointestinal disease and I need to investigate it then I know that I’m going to have to image the abdomen. Bloods aren’t going to help me. If it’s secondary gastrointestinal disease then I know that bloods are going to really help me but I’ve got to make sure it’s comprehensive. If I don’t know, it’s easier to rule out secondary GI causes of vomiting with blood tests than primary GI causes.”
“The only time I symptomatically treat a vomiting patient is when I have made a decision that this dog is (1) definitely vomiting and not regurgitating, (2) has primary gastrointestinal disease and (3) that the cause is of transient nature and will resolve spontaneously e.g. dietary indiscretion. I’m going to treat it symptomatically by withholding food or giving an anti-emetic while the gut is healing from whatever disgusting thing the dog just ate.”
Jill says if she treats symptomatically she expects to see an improvement within 24-48 hours. After this time, she recommends investigating further as the metabolic consequences of vomiting will start developing e.g. electrolyte disturbances and it is now more likely that the cause is not transient and will not resolve spontaneously.