Conventional medical training

Conventional training in medicine (and also the paramedical professions such as nursing, physiotherapy and occupational therapy) is based largely on scientific method and how this applies to health. From the outset of the training, medical ‘knowledge’ is taught as fact, with relatively little emphasis on the history of medicine or on the philosophy that underlies current medical thought.

The early years of medical school training are focused on the sciences of anatomy, biochemistry, physiology, pathology, pharmacology and psychology. As the training progresses, medical students are given increasing contact with patients in a hospital setting. It is in this clinical setting that the skills of case history taking, clinical examination, diagnosis and treatment are taught. Much of the learning in this clinical part of training is by example, as students accompany doctors in their practice in the clinic or on the ward.

One highly valued setting for learning in the UK is the ‘ward round’, in which the teaching doctor takes students to the bedside. The patient is the stimulus for the teaching, which is often in the form of encouraging rapid recall of relevant facts about the case in question. For example, on a ward round, the teaching doctor might ask: ‘This patient has renal failure. What are the causes of renal failure?’ The student is expected to reply with a systematic and complete list of causes of the condition in question. In this way, the diagnostic process becomes ingrained as a reflex in which, with the stimulus of characteristic symptoms and signs in a patient, appropriate diagnoses can be made and treatment selected.

The emphasis in medical training is that the skills of a fine clinician should include the ability to elicit the symptoms and signs of disease by careful questioning and sensitive clinical examination, and to have mental access to a complex series of lists of possible diagnoses and treatment options. To be successful, hospital clinicians require an analytical mind and an ability to structure facts in a rigorous way.

Interpersonal skills are also considered important to enable the rapport necessary to elicit symptoms from a patient, and the foundations of skills in communication are laid down in the early years of training of a medical student. More in-depth training in communication skills plays an important role in later aspects of medical training, particularly in the more person-centred specialities of general practice and psychiatry.

How disease is defined in conventional medicine

Most medical textbooks will describe a disease according to a system that lists the defining characteristics of the disease within a series of categories. For example, the characteristics of chickenpox could be presented as shown in Box 1.2a-I.

By having information structured in this way, each disease can be assessed in terms of its cause, symptoms and signs, investigations, etc. The differential diagnosis alerts the diagnosing doctor to other diseases that might have similar characteristics.

In forming a definition such as this one for chickenpox, the focus is on the characteristic pattern of the disease as it might affect any person, and not on the individual characteristics of the patient.

This model is often termed ‘reductionist’, because of the way in which the body is seen as a collection of smaller parts, and the characteristics of disease reduced down to tangible facts. The reductionist aspects of modern medicine are explored in more detail in Section 1.3.

How disease is diagnosed

Disease is diagnosed in three stages. These enable the doctor to match the individual patient with the descriptions of diseases that become so familiar through the course of medical training. The three stages of diagnostic information gathering are questioning for the presence of symptoms, examining the body for signs and checking the results of tests. Although a great deal of emphasis can be placed on the results of tests, medical students are taught that over 70% of diagnoses can be made with skilful history taking (questioning of the patient) alone. Adding in the results of physical examination will bring this diagnosis rate up to about 90%.

Thus, together, symptoms and signs will offer the necessary information for narrowing down the diagnosis. Medical students are trained to respond to characteristic symptoms and signs (e.g. central chest pain or a painless red rash) by generating a list of possible causes. To ensure the list of possible diagnoses is complete, students are taught to consider methodically the systems of the body and then important aetiological factors in turn. A full list will consider causes that might affect the circulatory, respiratory, digestive, urinary, nervous, endocrine, musculoskeletal and reproductive systems. Additional categories of causes that might be considered include infective, inflammatory, degenerative, malignant, social and psychological causes. This technique relies on memorisation of facts (see Q 1.2a-I and Q 1.2a-II).

With a list of possible diagnoses to hand, the doctor is then ready to ask more questions to narrow down the diagnosis and also to consider whether any investigations might provide further diagnostic information. However, it is noteworthy that test results are only essential in less than 10% of diagnoses, as the majority of diagnoses can be confidently made on the basic of skilful history taking and examination alone.

Tests should, ideally, be used to confirm a diagnosis and to help with the monitoring of treatment. An example would be the use of blood levels of thyroid-stimulating hormone (TSH) to confirm hypothyroidism in somebody with tiredness, weight gain and a slow pulse. An initial measurement that is well above the normal range would confirm the diagnosis. TSH levels would be expected to drop as the patient is given thyroid replacement therapy, thus showing a good treatment response (see Q 1.2a-3).

To summarise, the three steps to reaching a diagnosis are:

Performing tests

Medical investigations or tests are performed to narrow down a diagnosis and also to monitor the progress of treatment.

The diagnostic aspect of tests can be best illustrated by continuing with the case scenario given in Box 1.2a-II, as described in Box 1.2a-III. This case history illustrates the process of diagnosis and choice of treatment. Many doctors find this process a pleasing one to follow. The diagnosis is made by a logical process of elimination, and the treatment for a clear diagnosis is usually obvious from the current textbooks or guidelines.

However, in practice, diagnosis is rarely that straightforward. Some of the problems that beset doctors in the diagnostic process relate to the reliability of diagnostic signs as indicators of disease. Medicine is a discipline that is dogged with uncertainty. Not all symptoms, signs and positive test results indicate disease, as they can occur in healthy people (these are known as ‘false-positive’ findings). Similarly, not all normal measurements are indicative of health; in some people with disease, tests results are within the normal range (this would be described as a ‘false-negative’ result). Some of the uncertainty in diagnosis stems from limitations in the measurability of bodily functions.

Measurability in medicine

From a conventional medicine perspective, all functions of the parts of the body should be measurable, because they are believed to stem from something that is physical in nature. Although conventional doctors appreciate that the physical basis of some aspects of function, such as the emotions, are still unclear, there is still an underlying fundamental belief that, in the course of time, and with the results of high-quality research, all aspects of the body are potentially quantifiable.

The use of symptoms and signs is a form of measurement. For some symptoms and signs, the measurement is crude and rather subjective, as for example in the visual assessment of the redness of a sore throat. For other signs, however, such as pulse rate or temperature, the measurement can be more precisely recorded and then expressed in comparison to a ‘normal range’. Normal ranges are derived by taking measurements of a variable, height for example, from a large population. For many variables, the measurements are found to cluster around the most common, or mean, value.

An example of this is shown in Figure 1.2a-I, which illustrates how the birthweight of babies varies according to the week of pregnancy in which they were born. The graph shows three almost parallel curves. The central curve (labelled 50th centile) indicates the average or mean birthweight at any length of gestation. For example, an average baby born at the normal time (the 40th week of gestation) will have a birthweight of about 3.5 kg. The two outer curves (labelled 90th and 10th centiles) are an indication of the range of birthweights of most babies (80% to be precise) born at a particular time. For example, at 40 weeks of gestation, 90% of babies will have a birthweight less than 4.0 kg, and only 10% will have a birthweight less than 2.75 kg.

Figure 1.2a-I • Graph illustrating mean birthweight according to stage of pregnancy at delivery
• The mean birthweight is indicated by the 50th centile line.

(From Lissauer and Clayden 1997, p. 78, with permission).

According to the graph, the normal range for birthweight at 40 weeks of pregnancy is 2.75–4.0 kg. In other words, 80% of babies will have a birthweight between 2.75 and 4.0 kg. Conversely, the 20% of babies who have birthweights either greater or less than this range are ‘outside the normal range’. This implies that 20% of babies born at the normal time are, according to their birthweight, abnormal in some way. Logically speaking, this statement has to be true if a normal birthweight has been defined as 2.75–4.0 kg. Of course, it does not follow that one-fifth of all babies born at 40 weeks of gestation are unhealthy in any way. Some might be, but others might be small or large because of a familial tendency to be of a small or large build. The important point to draw from this example is that ‘abnormal’ as defined by a measurement that falls outside the normal range does not necessarily mean unhealthy.

The value of such an approach is that a baby whose weight falls outside the normal range will be examined more carefully to check that there are no underlying problems that might account for the abnormality. Conversely, the disadvantage of such an approach is that some healthy babies will be labelled as ‘abnormal’, and this can be a cause of unnecessary anxiety.

Most commonly, normal ranges express the range of values into which 95% of a population might fall for any one variable. This means that, for every possible variable, only 5% of the population will be considered to be outside this normal range. However, whatever the percentage of the population that has been selected to express the normal range, it would be very unlikely for anyone to be normal in every respect.

To summarise, this approach to defining normality has consequences that are not always appreciated by either the patient or the doctor. These are:

Another problem that arises from health checks is the question of what should be done if a previously healthy person has an abnormal test result. This is complex territory, not only for patients but also for most practitioners. An abnormal test is always a worrying finding, despite the fact that in a large proportion of cases there may be no actual abnormality in terms of disease. Conversely, a negative test may give false reassurance as, despite a value in the normal range, the disease may still be present.

An additional problem with many tests is that they cannot be wholly reliable because of inevitable errors in the processing of the test. This is another reason for an abnormal test result in someone with no actual disease. Conversely, an unreliable test can fail to reveal actual disease. Both these types of error are well recognised to affect any screening procedure, and doctors have to account for them when they interpret test results.

The prostate-specific antigen (PSA) test for prostate cancer is a good example of these points. If the PSA test is positive, this suggests a cancer is present, but it can be positive in some cases of benign prostatic hypertrophy (BPH) (i.e. a false-positive result). The PSA test can also be negative in some cases of cancer (i.e. a false-negative result), so giving a false sense of reassurance. The same applies to the sign of a smooth prostate. A smooth prostate suggests a benign cause, but is not a reliable indicator that a cancer is not present (i.e. it could be a false-negative).

To return to the scenario of Ian, the 63-year-old man with nocturia (see Box 1.2a-II), although he may feel that he has been given the ‘all clear’ from cancer, Ian’s doctor cannot totally be sure that the smooth prostate is not hiding a small cancer, and that the PSA result is not a false-negative. Given Ian’s age, together with the signs and test result, cancer would seem unlikely, but remains possible. The doctor has a dilemma whether or not to discuss this small possibility with Ian. It would actually be a very difficult discussion, because most textbooks do not offer information on the certainty of symptoms, signs and test results as predictors of disease. Without undertaking some in-depth research of the literature, the doctor would not be able to put any figure on the probability of an undiagnosed cancer in Ian’s case. Therefore, most doctors choose not to worry the patient about a serious, but unlikely, outcome.

This aspect of the diagnostic process is not completely satisfactory. In Ian’s case it is very likely he has a benign problem, but in other cases the diagnosis may be far less certain (see Q 1.2a-5 and Q 1.2a-6).