Managing Diabetes in Residential and Nursing Homes

The prevalence of diabetes in the elderly is around 10% and it imposes an enormous burden on healthcare systems. In America in 1992 nursing home care for people with diabetes cost $1.83 billion.Elderly diabetics have

much microvascular and macrovascular disease and are two to three times more likely to need hospital admission than their non-diabetic counterparts.One might expect a higher prevalence of diabetes and its complications in residential or nursing homes. Several American studies have found diabetes in 20% of nursing home residents,and in one almost 90% of diabetic residents had coronary artery disease, strokes, or peripheral vascular disease—with 6.4 major diagnoses compared with only 2.4 in non-diabetic residents.In Alabama (and probably England) nursing home patients generate a disproportionately large number of out of hours calls.
In a recent issue of the BMJ Benbow and coworkers surveyed 44 residential and nursing homes in Liverpool comparing 109 diabetic residents with 107 age and sex matched controls. Almost 10% of residents had diabetes, and, although not as sick as their American counterparts, they were more likely to be admitted to hospital and needed more visits from their general practitioner. Three quarters had regular chiropody and eye examinations, but 64% had no evidence of formal diabetic care. A similar problem has been documented in most American studies. For example, in 17 nursing homes in Michigan diabetes care did not meet local or national standards. Protocols—for example, for treating hypoglycaemia—were often not available and, while all homes had blood glucose meters, none had protocols for adjusting insulin doses or calling the doctor. Had Benbow et al in their British study asked to see the care plans of individual patients or the diabetes protocols they would probably have found neither.
Thus one of the most difficult groups of diabetic patients (in terms of both morbidity and complexity of management) seems to be receiving little organised care. What can be done? In an ideal world when elderly patients are transferred from hospital to a home members of the diabetic team should agree a problem list and management plan that is then communicated to the home. Current practice is often unsatisfactory, in that hospital discharge summaries usually concentrate on the illness that led to admission and rarely include plans for the future. Also they often go (late) to the wrong person, to the patient's previous general practitioner rather than the home or the new general practitioner. Among other things, such plans should specify the aims of treatment, which will usually include a safe level of diabetic control that minimises the risk of hypoglycaemia.
Staff in homes are often isolated and feel bewildered by the complex problems of elderly diabetics, and establishing channels of communication with local diabetes services may help. Benbow et al suggest that districts might appoint a specialist nurse responsible for elderly diabetics whose duties would include education and continuing support of carers. Another possibility is to identify a link nurse in each home who could attend regular meetings with local diabetes nurse specialists. Nursing home staff should be able to contact the diabetes team for advice.
Protocols covering aspects of diabetic management may help. In the early 1980s Colorado devised an educational programme with 125 written policies and procedures. For two years workshops and follow up consultations were organised for home staff and, at the end of this arduous programme, an improved level of care was documented by a 56% increase in the number of criteria met, although acute admissions did not fall. One problem is a lack of general agreement about those aspects of diabetes care that are most important for the staff of homes to know. Symptoms of hypoglycaemia and hyperglycaemia, diet, and foot care might be a good start. Unfortunately, protocols are often self defeating because they are too complicated, lack flexibility, and get submerged in a pile of other protocols.
Benbow et al found that monitoring of diabetic control was inadequate and often inappropriate, raising the question of how and how often it should be done. For those on diet or tablets a daily urine test or a weekly fasting blood glucose test and weight measurement might be reasonable. For insulin dependent diabetes, however, monitoring should be tailored to the insulin regimen, and here it is important to distinguish between residential and nursing homes. Staff in residential homes do not undertake invasive procedures, and a district nurse may have to give insulin and monitor diabetic control. Metabolic control is best where one motivated knowledgeable person is in charge. Problems such as "brittle diabetes" are likely if the carers do not have the experience or authority to change insulin doses. The problem of caring for diabetics in homes is an important one, for which diabetes services need to take more responsibility.
Robert Tattersall, Professor of clinical diabetes, Simon Page, Consultant physician