Students and teachers frequently operate on the assumption that learning merely comprises storing facts somewhere in the brain. This is hopelessly inadequate as a model for teaching and learning in clinical medicine. Schema theory specifically addresses the mechanisms whereby we actively make meaning of information. A schema is a mental structure for representing generic concepts stored in memory – a framework, plan, or script.
Imagine a post-intake round at King Edward VIII Hospital. Ask yourself whether a student would make any sense of, let alone learn anything useful from, them.
Registrar: The next patient is a 15 year old female. We admitted her in coma at 4 am this morning.
Consultant: Stiff neck?
Consultant: Any history of fits?
Consultant: Hm. Are you sure she hasn’t overdosed on something?
Registrar: Yes right. Her Mom attends psychiatric clinic for anxiety and depression. Toxicology came back positive for benzodiazepines and she woke up after a dose of flumazenil.
Registrar: Wentworth Hospital sent us this man last night in respiratory distress.
Consultant: HIV positive?
Registrar: Oral candidiasis.
Registrar: Diffuse bilateral infiltrates.
Consultant: Ground glass?
Registrar: Yes. I started Bactrim and prednisone.
Consultant: How is he?
Registrar: He’s desaturating on 60% facemask.
Consultant: ICU candidate?
Registrar: I thought so. The anaesthetic registrar is on his way.
Both the registrar and the consultant possess well-developed schemata which adequately encompass both these patients. In the first case, the consultant possesses a suitable schema for young woman in coma, which allows him to come to the most probable diagnosis with a just few well-directed questions. In the second case, both consultant and registrar clearly share a suitable schema for the patient with pneumocystis pneumonia. Hence they are perfectly in tune with each other’s thinking despite the absolutely minimalistic exchange of information. Indeed, the diagnosis has not even been mentioned but is just assumed to be shared. Note that their schema is comprehensive and encompasses context, presentation, diagnosis, treatment, recognition of severity and prognosis. There is no need to pull out each of these aspects either individually or sequentially.
Clearly none but the most precocious of students would be able to follow such a conversation. Lacking the same schema, all the nuances would be lost on them. Indeed, they would probably have to tug at a sleeve and ask: “Excuse me Prof, but what’s the diagnosis?”
As an aside, this is a reason why working ward rounds are not an ideal environment in which to teach students. If the student actually does benefit, then it is in fact more likely to have been a teaching ward round in which some work was done, rather than the reverse!
Properties of Schemata
In the context of our clinical learning, the features set forth by Stein and Trabasso (1982) might be illustrated as follows:
1. Schemata constitute a framework or template. They are composed of generic or abstract knowledge.
- In the example above, both doctors have a mental picture of the hypothetical patient with pneumocystis, the likely symptoms and signs, the possible causes, the differential to be considered, the exceptional presentations, the likely outcomes… And all of these not as a list, but weighted in shades of probability according to the context in which they occur.
2. Schemata are used to guide the storage, organization and retrieval of information.
- It is this which allows the expert to remember and retrieve a vast amount of information, to make a quick, accurate diagnosis with minimal information, and to proceed to a plan of action often without any conscious process of thought or deliberation at all.
3. Schemata reflect prototypical properties of experiences encountered by an individual, developed over many instances.
- Schemata are laid down by experiential learning, continually modified by reflection, reading, discussion and further experience. This allows the clinical to develop abstract, generic templates – the schemata – out of a number of individual and very specific experiences.
4. Schemata may be formed and used without conscious awareness.
- This is often the case. The examples above illustrate this. The doctors have a high-level conversation which comes completely naturally to them: they are probably unaware of the quite astonishing fact that they have successfully exchanged a textbook’s worth of information in just a few sentences.
5. Once formed, schemata are relatively stable over time.
- This is why older clinicians often remain outstanding diagnosticians, even if they fail to keep up with the literature. Fact-based knowledge will have changed, and they can’t tell the students all about the latest molecular tests or biological treatments, but the diagnostic schemata developed over 40 years of clinical practice will persist and they remain powerful and effective bedside teachers, at least until Alzheimer’s supervenes!
6. Although they reflect an individual’s experience, schemata are frequently shared between individuals.
- This allows the consultant and registrar to work effectively despite the exceedingly brief, laconic and minimalistic interchanges illustrated above. These are completely opaque to a student since he or she does not yet share the schema.
Driscoll suggests that a schema is analogous to:
1. A play, in that it has a basic script, but each time it’s performed, the details will differ.
- Once the student has a generic schema for heart failure, he can quickly and confidently recognise, investigate, manage and treat many patients with heart failure, despite considerable variation in the actual clinical presentations and contexts of those patients.
2. A theory, in that it enables us to make predictions from incomplete information, by filling in the missing details with “default values.”
- As illustrated in the hypothetical exchanges above. The clinician whose patient has oral candidiasis, a serious respiratory complaint and comes from a population with a high HIV-prevalence is quite prepared to assume that his or her patient has AIDS in the absence of all the other evidence that a junior student might think essential, from lymphadenopathy to an actual laboratory HIV result.
3. A computer programme, in that it enables us to actively evaluate and parse incoming information.
- As illustrated in the examples above. The diagnostic programme is running as the information comes in. Data are not gathered first, then thought through, then a diagnosis made.
How are Schemata created and modified?
Three processes are proposed to account for the modification of schemata:
1. Accretion: New information is remembered in the context of an existing schema, without altering that schema. (The student picks up additional detail, which is added to the schema.)
- The student learns that viral myocarditis is another cause of heart failure (along with the ischaemia, hypertension etc that he already knows about).
2. Tuning: New information or experience cannot be fully accommodated under an existing schema, so the schema evolves to become more consistent with experience.
- A student understands from her reading that cardinal signs of heart failure include a raised JVP, an enlarged liver and pedal oedema. She now meets a patient who has no pedal oedema, no hepatomegaly and a normal JVP. The patient does however have a displaced apex, a sinus tachycardia, a gallop rhythm and extensive basal crackles. Initially she struggles with a diagnosis as she has difficulty equating this with her understanding of heart failure. After some thought, discussion and reading, she realises that this patient has predominantly left ventricular failure, and she comes to appreciate that her previous schema was incomplete.She now modifies her schema for heart failure to incorporate patients with predominantly left ventricular failure, as may be seen after myocardial infarction.When she meets another, similar patient with a tachycardia, basal crackles and a third heart sound, she will have no difficulty at all in recognising that the problem is one of heart failure.
3. Restructuring: When new information cannot be accommodated merely by tuning an existing schema, it results in the creation of new schema.
- The student examines a patient with the lateral medullary syndrome. He cannot account for the combination of ipsilateral cranial nerve and cerebellar signs in the context of contralateral long tract signs. This does not tie in with his concept of an internal capsule-type stroke with a facial nerve palsy and hemiparesis. He is forced to read, learn and construct a new schema, which require a whole new conceptualisation of the anatomy of the brainstem, the position of the cranial nerve nuclei and medulla and the connections between them, etc.
Schemata are powerful forces in learning. In an article on the role of schemata in story comprehension, Stein and Trabasso (1982) noted that:
1. Schematic knowledge has a significant effect on organization of ambiguous or disorganized stories.
2. Narrative schemata specify expected components of a story, such as the time-sequence of events, and causal relations that should connect the events; during encoding or retrieval of a story, missing events may be inferred to fill in omitted information, and events may be reordered to correspond to a real-time sequence.
Clinical tutors used to listening to student present their cases, particularly the history, will identify immediately with these two postulates. Weak students are unable to organise their disjointed information. They cannot rework the information to flow in a cause-effect or start-finish format, or recognise the gaps in the history which arose when they failed to ask the patients about certain key issues. This is because they see the elements of information in isolation, and are unable to assemble them into a coherent schema-based framework.