Routine glucose assessment in the emergency department for detecting unrecognised diabetes: a cluster randomised trial

In reply: We thank Fourlanos and colleagues for their comments regarding our cluster randomised trial,1 which demonstrated that emergency department screening for diabetes and hyperglycaemia alone did not improve identification of diabetes, documentation of follow‐up plans, or hospital outcomes. The RAPIDS trial has shown that after identification through screening, intervention by a specialised inpatient diabetes team (IDT) is required to achieve improved outcomes.2 But this IDT comes with additional cost, which is a major barrier to the implementation of such teams around Australia.

However, there are data showing that the cost of IDTs is more than offset by the resultant savings, mainly through a reduction in length of stay (LOS) and hospital infections. A randomised controlled trial in the United Kingdom found that an inpatient diabetes specialist nursing service reduced the median LOS from 11 to 8 days (P < 0.01), with a cost saving of ₤436 per admission.3 Observational studies have also shown reductions in LOS with the introduction of an IDT. An American study reported a reduction in LOS from 6.01 (standard deviation [SD], 0.32) to 5.75 (SD, 0.38) days (P = 0.01), resulting in an estimated US$2.2 million saving to a tertiary care hospital in one year.4 This was a 467% return on investment.

Our study was conducted with the optimistic desire that the minimal cost of glucose and glycated haemoglobin screening would result in better management and outcomes for a large number of people with hyperglycaemia in hospital. The reality is that more than screening is required. For the management of many hospital‐acquired infections and for antimicrobial stewardship, there is already a well established model of care involving automatic and early participation of specialist infectious diseases teams.5 For optimal outcomes, management of many hospital‐acquired infections cannot be devolved to other clinicians. The same must be said for the challenge of inpatient diabetes. The time has come that the need for specialist IDTs is similarly recognised and adequately resourced.