Paul Dromgoole, RGN, RNMH, MSc Diabetes.
Lecturer Practitioner in Diabetes at the University of York and York Hospital
Illness in people with insulin-managed diabetes can necessitate adjusting their insulin regimen to meet altered needs. It is important that nurses understand the principles of insulin management during illness so that they can identify when a patient’s insulin dose needs to change, and to know when to refer them for specialist assessment. Overall principles of management discussed here are rooted in evidence, but because local practice and procedure may differ slightly, nurses should be aware of their own protocols and guidelines.
Effect of illness on blood glucose levels
Illness can have a range of effects on blood glucose levels, which in turn can affect recovery. For example, it can increase the levels of a range of ‘counter-regulatory’ hormones (catecholamines, glucagon, growth hormone and cortisol) that oppose the action of insulin or independently elevate blood glucose levels (Peters-Harmel and Mather, 2004) and cause a range of signs and symptoms (Box 1). Patients with insulin-managed diabetes may therefore require adjustment of their regimen. The main reasons adjustment may be required are to:
- Prevent escalating hyperglycaemia
- Reduce risk of diabetic ketoacidosis (DKA) and hyperosmolar hyperglycaemic state (HHS)
- Counter the delayed wound healing and increased infection risk associated with hyperglycaemia
- Reduce symptoms of hyperglycaemia and increase general well-being.
Aims of insulin management
The over-riding principle in managing insulin in illness is to prevent escalating blood glucose levels. Without an adequate supply of insulin, increased production of counter-regulatory hormones lead to an increase in blood glucose levels to a point where infection becomes harder to fend off and the risk of diabetic ketoacidosis (or in Type 2 diabetes, HHS) increases. Hyperglycaemia also puts patients at risk of a variety of sequelae, including increased infection risk, endothelial cell dysfunction and oxidative stress predisposing them to tissue damage (Clement et al, 2004).
Regular blood glucose monitoring is essential to provide the necessary information to determine patients’ insulin requirements. During illness blood glucose levels generally rise, and monitoring provides information on which dose adjustments can be based to maintain or return to acceptable levels. Diabetes management during illness follows guidelines generally known as ‘sick-day rules’ (Box 2).
Avoidance of extremes of blood glucose levels in pregnancy, even during illness, is vital for the health of both mother and baby. All pregnant women must be aware of the ‘sick-day rules’ and have 24-hour access to the specialist team (SIGN, 2001).
High meter readings
Most blood glucose meter scales measure up to 33mmol/L only. If the patient’s blood glucose levels are higher than this, he or she has no way of knowing the level, in which case the basic advice is washing the hands and repeating the test. If the repeat reading is similarly high, underlying causes should be considered, such as a missed injection, dietary indiscretion or illness. The test should be repeated after a couple of hours and if the reading is still high, medical advice should be sought from the diabetes centre, the GP, the GP out-of-hours service or NHS Direct.
Very occasionally, illness lowers blood glucose levels. While this may be contrary to accepted theory, it may relate to a patient’s lack of appetite, and shows the importance of individual assessment of the impact of illness on blood glucose levels and presentation of symptoms.
Guidance varies on how much insulin doses should be increased during illness, but the basic principle is that, if blood glucose levels are rising, insulin doses should be increased. The phrase to remember is: ‘sugars high: higher the insulin’. Some practitioners advocate increasing insulin in increments of 2-4 units. However, while this may be effective for people on very low doses of insulin, the higher the total daily dose the relatively less effective this increase will be. Others advocate increasing each dose of insulin by 10-20%, which is more likely to be an appropriate increase according to the individual’s own dose. It should be borne in mind, however, that some people struggle with calculating percentage adjustments.
If rapid-acting insulin is available, regular ‘stat’ doses should be administered until blood glucose levels fall towards single figures. For most people, one extra unit of insulin will cause blood glucose to reduce by around 1.5-2.0mmol/L (based on a usual total daily dosage of 40-60 units). If insulin has been increased in response to hyperglycaemia, doses should be promptly reduced as illness/hyperglycaemia resolves in order to prevent rebound hypoglycaemia.
Patients should be guided by specific individual advice from their specialist diabetes team, and should be aware of when, where and how to seek help if they have problems with blood glucose control generally and managing illness - particularly with regard to sick day rules.
Issues specific to Type 1 diabetes
Prevention of diabetic ketoacidosis
Diabetic ketoacidosis (DKA) occurs predominantly in Type 1 diabetes, where a profound lack of insulin means the patient is unable to negate the effect of counter-regulatory hormones that cause glucose levels to rise in response to illness. In basic terms, lack of insulin leads to the utilisation of alternative fuel sources (gluconeogenesis), and excessive fat metabolism (lypolysis) causes elevated concentrations of glycerol and free fatty acids (FFAs) and subsequently increased ketone body formation (ketogenesis). Since ketones are acidic, the more ketones the greater the acidity and consequently the lower the body’s pH. In general terms, profound acidosis (pH <7.0) is life-threatening. DKA mortality rates have remained unchanged over recent years and are still worryingly high in the UK at 1-5% (English and Williams, 2004).
There is no consensus on the efficacy of ketone testing, even in Type 1 diabetes management. Arguments against are that patients may not have the strips to hand when they need them and that testing does not necessarily change management during illness. The key principles are to monitor blood glucose more frequently and increase insulin to correct hyperglycaemia.
However, there is not necessarily a strong association between the degree of hyperglycaemia and the degree of ketonuria, as moderately raised blood glucose levels can be accompanied by high levels of ketones (Child, 2001). Ketone testing will aid the assessment of the clinical significance of illness and resultant hyperglycaemia and provide valuable information to specialist services advising (often initially by telephone) on the patient’s management plan.
Some patients appear particularly predisposed to recurrent DKA (Skinner, 2002). In some cases the underlying reason is obvious non-concordance with insulin injections, which either leads directly to ketogenesis or puts the patient at particular risk during illness owing to lack of sufficient background insulin. However, a small cohort appear susceptible to DKA during illness and these patients must be strongly advised of the importance of self-ketone testing during illness and encouraged to access 24-hour telephone support whenever necessary. This may be through the diabetes specialist services, a hospital medical ward or the emergency GP service.
Children and vulnerable adults
Parents and carers of people with diabetes must be aware of the general principles of sick-day rules and local arrangements for seeking support. Continuity of care from the same diabetes specialist professional is imperative.
Issues specific to Type 2 diabetes
Hyperglycaemic hyperosmolar state (HHS) occurs exclusively in patients with Type 2 diabetes; indeed it can lead to an initial diagnosis of diabetes. For these patients, prolonged hyperglycaemia as a result of illness may lead to beta cell toxicity and escalating blood glucose levels, resulting in a need for increased oral hypoglycaemic agents (OHAs) or insulin doses in those whose blood glucose levels are spiralling upward. Patients should be aware of this possibility and of whom to approach if requiring guidance.
Hospitalisation for non-diabetes reasons
Many people with diabetes are admitted to hospital for reasons not specifically related to their condition but which nevertheless impact on their diabetes control, normally causing blood glucose levels to rise. Assessing pre-admission glycaemic control can provide a reliable baseline against which to compare their in-hospital control. This may be especially useful if referring the patient to specialist diabetes services, and should also help the team managing the patient to make appropriate changes to diabetes treatment if required. Glycated haemoglobin (HbA1c) estimation is usually by laboratory analysis of a venous non-fasting blood sample for the degree of binding of glucose to a specific part of haemoglobin. The level, measured as a percentage, indicates mean glucose levels in the preceding six to eight weeks.
One of the most frequent complaints from patients with diabetes who are in hospital concerns their insulin equipment being withdrawn. If patients are able to continue self-injecting and monitoring (where appropriate), they should be allowed to continue. This reduces frustration and enables them to initiate dose changes and time injections around meals.
A randomised control trial, Diabetes Mellitus Insulin-Glucose Infusion in Acute Myocardial Infarction (DIGAMI), provided strong evidence that post-myocardial infarction (MI) insulin therapy significantly improved patient outcomes in terms of further coronary events (Cummings et al, 1999). Many patients are therefore started on insulin immediately post-MI and other acute coronary events and remain on it for perhaps three months, when a joint decision is made between clinician and patient regarding future treatment options. A subsequent trial (DIGAMI 2) attempted to identify the period over which insulin benefited the patient, and results were recently presented in Munich in 2004 (as yet unpublished); however, this study was less convincing with problems in subjects’ baseline characteristics compared to DIGAMI 1, and is less useful in informing practice.
Intravenous sliding-scale insulin
Sliding-scale insulin is perhaps the most abused therapeutic intervention in diabetes management, and causes significant problems for patients, ward staff and diabetes specialists (Lorber, 2001). To assist problem-free management, the following should be considered:
Is it necessary?
Intravenous sliding scale insulin (IVSSI) is often used in patients who are admitted with newly diagnosed diabetes or those who are incidentally diagnosed while hospitalised. If they are ketoacidotic, it is likely to be appropriate, but many are not. Although IVSSI will lower blood glucose levels, it also delays initiation of regular diabetes treatment (oral hypoglycaemic agents or subcutaneous insulin). It is also retrospective, acting on what has happened to blood glucose levels, whereas regular diabetes treatment will help to prevent glucose levels rising.
Is the dose effective?
Sliding-scale insulin must be regarded as a dynamic scale - one size does not fit all. If three units per hour repeatedly and alarmingly drops the patient’s blood glucose levels, the dose is too high for that patient and should be reduced to perhaps 2.5 units per hour, with a gentle adjustment of the other infusion rates to balance with this. Similarly, if blood glucose levels rise dramatically in a patient on one unit per hour, the dose is and should be increased, perhaps by one unit an hour, again balancing the other doses. Most sliding-scale problems are related to erratic blood glucose levels, which are due to lack of adjustment of the scale according to the individual patient’s response.
Does it need to remain in situ?
In almost all cases, if the patient is successfully tolerating his or her diet, sliding-scale insulin should be discontinued, if necessary with advice and specialist assessment from the diabetes specialist team on the diabetes regimen to which the patient should revert.
How should it be discontinued?
Intravenous insulin has a half-life of around 15 minutes (Mincu and Ionescu-Tirgoviste, 1980). In patients who usually require insulin to manage their diabetes, discontinuing IVSSI without first putting in place their regular insulin injection (or possibly a revised dosage) will cause a rapid rise in blood glucose levels. There should be an overlap of 30 minutes between administration of the first insulin injection and discontinuation of IVSSI. Likewise, treatment for patients on oral hypoglycaemic agents must be started again when IVSSI is discontinued unless there is a clinical reason not to do so.
On occasion, insulin is required to manage hyperglycaemia during terminal stages of illness. For those with Type 1 diabetes it should be continued, although change of dosage and insulin type is often indicated. Those with Type 2 diabetes may require insulin therapy if maximum oral hypoglycaemic agents fail to maintain blood glucose levels below a level where the patient is free of osmotic symptoms. Many patients say that the worst symptoms of hyperglycaemia are dry mouth and thirst, which can be easily palliated by even a once-daily insulin dose. The insulin of choice is probably a gentle long-acting analogue insulin, started at a low dose (perhaps 0.2 units per kg body weight) and gently increased, with the aim of lowering blood glucose levels to a point where osmotic symptoms resolve. However, advice should be sought from the diabetes specialist team.
The importance of osmotic symptom relief in people who may well have other symptoms of their disease cannot be over-stated. Again, advice should be sought from the diabetes specialist team.
Effective insulin management during illness is vital to prevent a range of problems arising.
Nurses should understand the general principles of illness management for people using insulin but also be aware of when and how to access diabetes specialist support. Wherever possible, patients (or carers) in the community should be advised on illness management principles and, most importantly, where to obtain advice, especially out of hours. Those in hospital should be enabled to retain as much control as appropriate over their diabetes management.
Testing Times (Audit Commission, 2000)
Provides salutary insights into inpatient care of people with diabetes. States that over half of in-patients with diabetes were identified by ward staff as having problems with their diabetes control, but less than half of these were referred to specialist diabetes services. Of particular concern was the fact that less than 30% of wards allowed patients to keep control of their own insulin or other medication
National Service Framework for Diabetes (Department of Health, 2003)
Standards 3, 4, 5, 7 and 8 make clear the importance of specialist inpatient services and diabetes self-management. This is a landmark document for those living with or supporting people with diabetes in the UK.
- Illness can result in raised blood glucose levels in many patients, which may necessitate adjusting the doses of their insulin or oral hypoglycaemics
- Adjustments should be made on the basis of prior assessment and the monitoring of the patient’s response
- Intravenous sliding-scale insulin should be used only after careful consideration and discontinued as soon as possible.
Author’s contact details
Paul Dromgoole, Lecturer Practitioner - Diabetes, York Hospital, Wigginton Road, York YO31 8HE.
Audit Commission. (2000) Testing Times: A Review of Diabetes Services in England and Wales. London: AC.
Clement, S., Braithwaite, S., Magee, M. et al. (2004)Management of diabetes and hyperglycaemia in hospitals. Diabetes Care 27: 2, 553-591.
Cummings, J., Mineo, K., Levy, R., Josephson, R. (1999)A review of the DIGAMI study: intensive insulin therapy during and after myocardial infarctions in diabetic patients. Diabetes Spectrum 12: 2, 84-88.
Child, D.F. (2001)Euglycaemic diabetic ketoacidosis. Practical Diabetes International 18: 7, 239-240.
Department of Health. (2003)National Service Framework for Diabetes: Modern standards and service models. London: DH.
English, P., Williams, G. (2004)Hyperglycaemic crises and lactic acidosis in diabetes mellitus. Postgraduate Medical Journal 80: 253-261.
Lorber, D. (2001)Sliding scale insulin. Diabetes Care 24: 2011-2012.
Peters-Harmel, A., Mathur, R. (2004)Davidson’s Diabetes Mellitus. Philadelphia, Pa: Saunders.
Scottish Intercollegiate Guidelines Network. (2001)Management of Diabetes in Pregnancy. London: SIGN.
Skinner, T.C. (2002)Recurrent diabetic ketoacidosis: causes, prevention and management. Hormone Research 57: (suppl), 78-80.
Mincu, I., Ionescu-Tirgoviste, C. (1980)Half-life and hypoglycemic effect of intravenous insulin in patients with diabetic ketoacidosis. Medicine Interne 18: 3, 287-292.
Warren, R.E., Deary, I.J., Frier, B.M. (2003)The symptoms of hyperglycaemia in people with insulin-treated diabetes: classification using principal components analysis. Diabetes/ Metabolism Research and Reviews 19: 5, 408-414.