Many older research trials do not include older people or only include relatively healthy older people and those using none or one medicine, thus it is difficult to generalise the findings to frail older people [3
]. Excluding older people from clinical trials is not a new phenomenon [21
]. There is strong evidence that many medicines commonly prescribed for older people should be used with caution or not prescribed [9
]. Table 2 outlines some medicines commonly used to manage diabetes in older people and their associated risks.
Key issues include:
is a significant risk with insulin and some sulphonylureas, especially long acting preparations [2
]. Older people with both T1DM and T2DM are particularly vulnerable to hypoglycaemia: it is the most frequent metabolic complication in older people and may account for one in five hospital admissions in older people with diabetes aged 80 years and older [25
]. The GLM prescribed, often the long acting sulphonylurea glibenclamide, may have been inappropriate in a large number of people who had dementia and/or renal failure in Greco et al’s study [25
Hypoglycaemia is associated with longer hospital stay and increased costs in T2DM [26
]. It is often difficult to detect in older people because neuroglycopaenic symptoms predominate and can be misinterpreted for confusion and not treated. Significantly, hypoglycaemia is associated with short term changes in delayed and working memory [27
] and dementia in the longer term [28
] and can precipitate myocardial infarction [30
], which might be ‘silent’ (present with atypical symptoms). The ability to mount a counter-regulatory response to hypoglycaemia declines over time and the glucagon response is virtually absent in many older people [26
] and when the individual is malnourished and has limited glucose stores. Macroalbuminuria predicts severe hypoglycaemia [31
]. Common causes of hypoglycaemia are shown in box 2.
Hypoglycaemia risk factors in older people
• Prescribed GLMs especially some sulphonylureas and/or insulin.
• Long duration of diabetes, which is associated with progressive changes in the counter-regulatory response to hypoglycaemia, in particular diminished secretion of glucagon and growth hormone, which contributes to hypoglycaemia unawareness.
• ‘Tight’ blood glucose control.
• Renal and liver disease.
• Nutritional deficits.
• Cognitive impairment and dementia, which make it difficult for the individual and health professionals to recognise hypoglycaemia.
• Multiple diabetes complications and other comorbidities that cause functional deficits and compromise diabetes-self care.
• Most current hypoglycaemia education programs and policies are not tailored for older people e.g. the focus on adrenergic hypoglycaemia symptoms such as sweating and trembling when neuroglypaenic symptoms such as confusion and behaviour change are more common in older people and contribute to hypoglycaemic unawareness.
• History of severe hypoglycaemia.
If health professionals, the person with diabetes or family carers do not consider medicine side effects
as the cause of symptoms, medicines can be prescribed to treat the symptom (prescribing cascade) and increase the likelihood of adverse events. Determining the cause of symptoms can be difficult because they are often non-specific and atypical.
Declining renal function
and chronic kidney disease is common in older people with diabetes and is a predictive risk factor for medicine AEs [24
] especially hypoglycaemia and falls [11
when they are prescribed to manage behavioural problems associated with dementia.
that carry risk of postural hypotension and can lead to falls.
Some lipid lowering agents
are contraindicated if liver disease is present. Liver function declines with age. Therefore, monitoring liver function is an important aspect of medicine management.
Sliding insulin scales
to manage hyperglycaemia [15
]. Sliding insulin scales might lower the blood glucose, but they do not address the preceding factors that led to hyperglycaemia such as infections (urinary tract, foot infections), diet, stress, pain, depression that need to be treated otherwise hyperglycaemia could lead to hyperosmolar states (HHS), confusion, delirium, falls, reduced quality of life and symptomatic discomfort [4
]. That is, it is essential to treat the cause rather than just reacting to the symptom, hyperglycaemia. For these reasons the trend is not to use sliding insulin scales in routine care, although they may still be indicate in acute illnesses such as DKA and HHS in an insulin infusion.
Blood glucose is often tested infrequently
in ACHs and/or the testing schedule is not related to key medicine safety factors such as meal times and glucose lowering medicine action profiles, especially peak action times for insulin.
Many medicines prescribed to manage diabetes and its complications are classed as high-risk medicines because of their side effects and the way they are used and metabolised in the body, for example insulin and anticoagulants [32
]. High risk medicines have a significant risk of causing catastrophic harm when used in error. However, as the information presented so far shows, high risk medicines can also cause significant harm when used appropriately if their effects are not monitored closely and when the dose and/or dose regimen is not safe for the individual.
Box3: Examples of some commonly prescribed medicines that can increase or lower blood glucose. Medicine availability changes as older medicines are discontinued or removed by regulatory authorities. The information in this box was current at the time of publication. Detailed information can be obtained from:
• Medicines that increase blood glucose
are known as diabetogenic medicines. Diabetogenic medicines also increase blood glucose in older people at risk of diabetes. Some of these medicines are PIMs/AIMs but they may be appropriate treatment for the individual’s illness at the time. The individual and their carers should be informed about the potential effect on blood glucose when medicines that increase blood glucose are commenced and what to do if the blood glucose is affected. Blood glucose monitoring is useful to detect changes early. Medicines should be used at the lowest effective dose for the shorted possible time and in the least diabetogenic dose form. It is important to manage the hyperglycaemia to reduce the associated risks such as dehydration, delirium, falls, incontinence and candida infections e.g. corticosteroids. Examples include:
• Corticosteroids especially long acting oral preparations.
• Antipsychotics especially atypical antipsychotics.
• Sympathomimetics such as adrenaline and salbutamol.
• Thyroid and growth hormones.
• Thiazide diuretics.
• Some antihypertensive medicines such as atenolol, carvedilol and metoprolol.
• Some herbal medicines such as Chrysanthemum extract, Honey bee pollen, Tamarind. Note: people generally use these medicines to treat intercurrent illnesses not to manage blood glucose. Significantly, most of the information is based-on single case reports, which do not provide the botanical names of the herbs in the medicine.
• Medicines that lower blood glucose
in addition to GLMs Determine the individual’s hypoglycaemia risk before commencing medicines that can increase the hypoglycaemia risk or mask hypoglycaemia symptoms. The individual and their carers should be informed about the potential effect on blood glucose when medicines that increase blood glucose are commenced and what to do if the blood glucose is affected. Blood glucose monitoring is useful to detect changes early. Examples include:
• MAO inhibitors
• Herbal medicines that can lower the blood glucose include Aloe, Ginseng, Momordica charantia
(Bitter melon), Trigonella foenum graecum
(Fenugreek) and Opuntia
species (prickly pear) especially if they are combined with conventional GLMs. Note: there is some reasonable quality research to support the glucose lowering effects of these medicines.
• Note: some medicines can cause either hyop- or hyperglycaemia e.g. magnesium salicylate, Lithium, Lanreotide.
Strategies that can to Improve Medicine Safety and Reduce Medicine-Related Adverse Events in Older People with Diabetes
Who assess well, treats well
Regular comprehensive medicine assessments and medicine reconciliation are essential, especially when several doctors prescribe medicines for the same person [24
]. A medicine assessment must include information about the individual’s medical and medication history, health status, physical functioning as such as Activities of Daily Living (ADL) Instrumental Activities of Daily Living (IADL) Geriatric Discussion Scale, cognitive function such as Mini Mental State Examination (MMSE) and sensory status (vision and hearing), social circumstances and available support, especially if the individual lives in the community. Home medicines reviews can elicit important information about the social factors that affect medicine self-management.
Medicine assessments should aim to identify key related risks such as PIMs/AIMs, duplicate prescriptions, and CAM and over-the-counter (self-prescribed) medicine use, hypoglycaemia, hyperglycaemia, other adverse events, falls, pain, and self-care capacity. It is essential to maintain up-to-date medicine histories and medicine lists and communicate any changes to everybody involved in the individual’s care, the individual and his or her family carers in a timely manner and using appropriate language and design/format for written material.
As indicated, it is important to monitor liver and renal function, which decline with increasing age: most people over age 60 have liver and renal function changes [33
]. These changes affect medicine choices. It is important to stop or reduce the doses of medicines such as ACE, NSAIDs and Cox-2 inhibitors that contribute to declining renal function. Antituberculosis medicines and alcohol affect liver function. It is important to regularly monitor renal and liver function, especially when such medicines are needed.
The medicine regimen should be reviewed every time a medicine is started or stopped, an adverse event occurs or health status changes. An admission to hospital is an ideal time to undertake a medicines review and assess their understanding of their medicines and their medicine self-management capacity. Admission and discharge between wards/units and between care settings are high risk times: medicine reconciliation should occur at every transition [12
The general practitioner and other carers can contribute important information to medicine reviews and should be informed about the outcome of any medicine review and changes to the medicine regimen. The medicines review must include complementary and alternative therapies (CAM) because people with diabetes are high CAM users [34
]: some CAM are beneficial, safe non-medicine options.
Quality use of medicines (QUM) [35
] is a useful framework for using the information from comprehensive assessments to make decisions about the medicine regimen, doses and dose frequency. Importantly, QUM encompasses the entire medicine pathway (from bench to bedside), regulatory processes, labeling, and advocates using non-medicine options when they are safe and evidence-based.
Several decision support tools are available to help health professionals manage medicines for/with older people. These include:
BEERS criteria [15
Screening Tool to Alert doctors to the Right Treatment (START) [22
Screening Tool of Potentially inappropriate Prescriptions (STOPP) [23
Medication Appropriateness Index (MAI) [36
Australian Inappropriate Medication Use and Prescribing Indicators Tool [36
Guidelines such as National Institute of Clinical Excellence (NICE) [38
], National Prescribing Service (NPS) [24
] The McKellar Guidelines [4
] and the International Federation Global Guideline for Managing Older People with Type 2 Diabetes [3
Various medicine AE risk assessment tools such as Medicines Risk Screen (NPS) [39
], the GLM-related Adverse Event Risk Assessment Tool [4
High Risk Medicine Alerts [32
However, it is essential to prescribe medicines appropriate for the individual’s clinical context when using these decision aids and alerts. Making decisions in collaboration with the multidisciplinary team facilitates appropriate in dividualised prescribing [3
]. Where possible the individual is involved in such decisions to enhance medicines self-care.
These lists focus on high risk medicines but many AEs and side effects are due to commonly prescribed and self-prescribe medicines, not on lists such as BEERs. Miller et al. [40
] reported 11.6% of people had at least one AE in the preceding six months, most of which were mild to moderate; 11.8 were severe and 5.4% required a hospital admission as a result of using 13 commonly prescribed medicines not on lists such as BEERS Criteria [15
]. These tools were developed to aid medicine prescribing decisions, not replace reasoned clinical decision-making. However, STOPP does help identity possibly avoidable AEs [41
]. Some decision support tools are not appropriate for every country because they list medicines not commonly used or available in some countries [37