What are the harms of a pre-diabetes label?

The treatment for pre-diabetes is, in essence, adopting a healthier diet and taking more exercise. If adopted and maintained, these lifestyle changes are likely to benefit most patients in multiple aspects of health, not just their risk of developing diabetes. However, although they may slightly delay the point at which a patient develops diabetes, studies of lifestyle-based diabetes prevention programmes show that most patients do not or cannot maintain long-term lifestyle changes.5 12 Weight loss is generally short term or minimal and patients usually slip back into old habits and routines. While there is undoubtedly an argument for informing younger patients who may receive a benefit from knowing they have pre-diabetes, the harms of informing increase with age.

Many elderly patients with comorbidities may struggle to increase physical activity. Dietary change and attempts to lose weight after a certain age can have detrimental health effects13 Labelling somebody as having a medical condition carries a psychological burden in itself, and being unable to engage in the behaviour change recommended may also have negative consequences, that is, engendering a feeling of being ‘a failure’.14–16 If the label leads to further follow-up this may also place a burden on patients. There are also considerable implications for the use of health resources if the labelling of individuals as pre-diabetic requires further follow-up and intervention. Annual blood tests are standard (£6.42), subsequent general practitioner (GP) or nurse (£30) appointments to discuss results frequently take place as do referrals on to the national Diabetes Prevention Programme (£270).17 There are roughly 3 million people in the UK aged 80 years or over.18 If one-third of them have pre-diabetes and, of those, half have an annual blood test, a quarter have a GP appointment and one in eight get referred to the National Health Service (NHS) Diabetes Prevention Programme that is an annual cost of around £37 million.

What is ideal practice and what is the reality?

While some patients may have been tested following screening for being at risk of diabetes, in the UK most patients in whom pre-diabetes is diagnosed have blood sugar level tests carried out as part of a battery of other blood tests that are performed as part of annual chronic disease monitoring for conditions such as hypertension.19 The contents of the battery are determined by individual practices and usually based on guidance and payment targets issued by the NHS.20 In theory, a patient should give informed consent before any test, including blood sugar and HbA1c testing. In reality many patients who are given a diagnosis of pre-diabetes are unaware that they had blood tests for diabetes/pre-diabetes.19 When checking blood glucose or HbA1c in an elderly patient, especially one without symptoms of diabetes, the clinician should talk through with them the potential outcomes of the test and the implications this may have to them. The patient can then make an informed decision as to whether they want to go ahead with testing or not. In routine clinical practice in the UK this happens rarely, if at all. This is likely due to the volume of blood testing, the automated nature of the process, the limited time a clinician has to devote to each individual patient and the priority that individual clinicians assign to such conversations.

As we discussed in a recent paper a more individualised approach to ‘routine’ blood tests needs to be taken.19 The utility of each test should be gauged for each patient as an individual, not as the average patient that has a particular disease. The reality, however, is that this change will, at best, be adopted slowly or, at worst, not at all. What then, should clinicians who are presented with a pre-diabetic blood result in an elderly patient do?

The see-saw model of paternalism

When faced with a series of test results for a patient, clinicians exercise judgement about what they consider ‘normal’ or ‘satisfactory’. They also exercise judgement in what they communicate to the patient about the results. In certain circumstances a patient may, for instance, have a mildly raised bilirubin or mildly decreased albumin and the clinician may file the result as ‘satisfactory’ and not inform the patient. Is this an act of paternalism or is it the act of a clinician filtering out the ‘noise’ that is generated from carrying out tests and using an individual patient’s circumstances to contextualise what is ‘normal’?

Should clinicians, therefore, assume that all new pre-diabetic blood results above a certain age should not be disclosed to patients? This is obviously an indefensible position as a general policy since patients have a right to information that concerns their health. However, while the blood result may be a factual piece of data, the labelling of a result as ‘satisfactory’, ‘acceptable’ or ‘abnormal’ is a clinical judgement. There is, in most circumstances, a moral obligation on the clinician to disclose to a patient that they are suffering with a disease. Pre-diabetes is not a disease and unless a patient fulfils the three criteria set out in the introduction to this paper the information is not likely to benefit the patient.

In younger patients, where the criteria related to a significant likelihood of progressing to diabetes with negative health effects are likely to be fulfilled, there is an onus on the clinician to inform patients they have pre-diabetes. In many younger patients it will be difficult to judge whether they fulfil the third criterion and can successfully change their lifestyle. In these cases the likely benefits of ‘diagnosis’ outweigh any potential drawback. However, as a patient ages and develops certain other comorbidities, a tipping point is reached where the criteria are very unlikely to be fulfilled and the harms of a ‘diagnosis’ will outweigh any potential benefits. At that point informing the patient becomes harmful and should arguably only be done if the patient explicitly requests the information.

Rather than having a full discussion of the pros and cons of a pre-diabetes label with each patient we would advocate a ‘see-saw’ model of paternalist considerations. Younger fitter patients are automatically informed of their pre-diabetes whether or not they have requested the information explicitly while those who are very elderly and have comorbidities and a limited life expectancy are not informed. In the middle is the group of patients for whom paternalism either way is not appropriate because the benefits and harms of a ‘diagnosis’ are uncertain. These patients in the middle of the see-saw are those for whom an in-depth discussion about the relevance and meaning of ‘pre-diabetes’ to them as an individual needs to take place, and also those patients where the blood test most strongly ought to have been discussed before it was performed.

It could be argued that a drawback to this approach is the effect that it may have on patient–physician trust. In modern medicine patients are frequently seen by multiple clinicians. Clinician one may choose, quite ethically, not to reveal to a patient that they are pre-diabetic. The patient may then see clinician two who tells them. This could then create a situation where the patient loses trust in clinician one and, indeed, the whole medical profession. However, pre-diabetes is not a disease state. The non-disclosure of pre-diabetes is markedly different to the non-disclosure of a disease. If the patient understands that clinician one did not disclose to them because pre-diabetes is a risk factor that is not relevant to them, and not a disease, then, hopefully, there would be no loss of trust. In primary care in the UK, there is frequently non-disclosure of other ‘pre’ conditions, such as chronic kidney disease.21 This non-disclosure takes place where the condition is of relevance to the patient and full disclosure would, generally, be in the best interest of the patient. This is ethically and professionally problematic. However, the response of patients who find out about non-disclosure in these cases is of interest. When interviewed, the response of patients to finding out about these non-disclosures is nuanced and varied.21 It does need lead to automatic loss of trust in the medical profession.

Wider use of this approach?

The purpose of the paper is to outline principles that could be applied, in an ethical manner to an unexpected blood test result of pre-diabetes. In theory, the principles outlined could be more widely applicable in other pre-conditions and other risk factors. To be applicable, a condition must have a fairly predictable trajectory, have a point where ‘pre-disease’ becomes ‘actual disease’ and be potentially reversible (or delayable). The principles could possibly be applied to early chronic kidney disease or early hypertension but may not be appropriate for other conditions or risk factors. The difficulty in other conditions is predicting whether a patient is going to convert from a pre-condition to a disease state, predicting when they are going to convert and predicting whether this is going to cause harm. In these cases, where there is doubt, this should always be discussed fully with the patient.