Failure of NICE to resolve the question about the management of hyperglycaemia in acute coronary syndrome
- 1National Institute for Clinical Outcomes Research, London, UK
- 2College of Medicine, Swansea University, Swansea, UK
- 3Department of Medicine, John Radcliffe Hospital, Oxford, UK
- Correspondence to Dr John Birkhead, National Institute for Clinical Outcomes Research, 170 Tottenham Court Road, London W1T 7HA, UK;
Contributors CW, MH and JB all contributed to the development and writing of this editorial.
- acute coronary syndrome
- clinical practice guideline
- coronary artery disease
- metabolic medicine
- emergency medicine
Acute coronary syndromes (ACS) that present with hyperglycaemia (≥11 mol) are associated with higher mortality across the whole spectrum of ACS, whether in the context of known diabetes or not.1 There is a near linear relation between admission glucose and the adjusted relative risk (RR) of death both in those with, and those without a prior diagnosis of diabetes mellitus, with the slope of the relationship being steeper in non-diabetic individuals.2
ACS presenting with hyperglycaemia, as defined above, are common. In England and Wales, 15% of patients presenting with troponin positive ACS are hyperglycaemic: 45% of those already known to have diabetes and 7.5% of those not known to have diabetes (data on file; Myocardial Ischaemia National Audit Project (MINAP) database 2010).3 Despite the recognised association between admission glucose and outcome, and the very high frequency of dysglycaemia and adverse events in survivors of ACS,4 evidence to support management strategies that cover all ACS is lacking.
Studies addressing the effect of control of hyperglycaemia with insulin on mortality are few, relatively small and underpowered; all experienced difficulties with recruitment.5–7 Only in DIGAMI-1 was a statistically significant reduction in mortality reported, and that was at 1 year rather than at the intended primary endpoint of 3 months.5 DIGAMI-2 failed to show any mortality benefit.6 The HI-5 study also failed to show benefit in the treatment arm, and in any event was not directed solely at hyperglycaemic subjects, as the mean admission glucose in the study was 11 mmol.7 While there have been many specific reasons advanced for the failure of these studies to show clinical benefit, a feature common to each study was a delay from onset of symptoms to treatment of …