Diabetes care in the very old. Beyond the limits of evidence?
Dr. Shota Hamada is a Visiting Research Fellow in the Department of Primary Care and Public Health Sciences, King’s College London. Very old people have rarely been included in clinical trials in sufficient numbers and treatment recommendations for them are largely based on professional opinion informed by evidence generated from younger patients. Selecting drug treatment for very old people may be influenced by distinct concerns including co-morbidities, declining physical and cognitive functioning, and perceptions of limited life expectancy, that may be less relevant in younger people. Our recent research, published in Age and Ageing, investigates changing prescriptions for antidiabetic and cardiovascular medications for very old patients who were newly diagnosed with type 2 diabetes over a 20 year period. The study included a representative sample of nearly 13,000 patients from primary care in the UK. From 1990 to 2013, use of sulphonylureas declined rapidly from 94% of patients to 29%, while metformin became the mainstay among antidiabetic drugs, increasing from 22% to 86%. Prescribing of antihypertensive (46% to 77%), lipid-lowering (1% to 64%), and antiplatelet drugs (34% to 47%) also increased substantially in this period. These changes in prescribing seem to be motivated by evidence generated from clinical trials in younger people. From the 1990s onward, the results of several large clinical trials have been available, such as UKPDS, ACCORD and 4S, which promoted implementation of intensive multifactorial interventions for patients with diabetes to reduce mortality, cardiovascular diseases and other complications. Do these major increases in the intensity of pharmacological management really improve survival or well-being in very old people? We do not have a clear answer to this question. Our present study is surely an initial step to evaluate drug therapy in very old people with diabetes. One of the challenges in making treatment decisions may be so-called “lag time to benefit” or “payoff time”. For example, initiation or continuation of antidiabetic drugs should be determined carefully given that it requires several years to obtain benefits from antidiabetic drugs. Further research is needed to justify or adjust prescribing practices in terms of both effectiveness and safety outcomes. We should now prepare to develop evidence-informed strategies for diabetes management in very old people.