The patient population and methods of the USIC 2000 registry have been described in detail elsewhere.8,9
Briefly, the objective of the study was to gather complete and representative data on the management and outcome of patients admitted to intensive care units for definite AMI over a one month period in France, irrespective of the type of institution to which the patients were admitted (that is, university hospitals, public hospitals or private clinics). Of the 443 centres that treated patients with AMI at that time, 369 participated in the study (83%). One physician responsible for the study was recruited in each centre and filled in a case record form for each patient meeting the inclusion criteria and admitted to the intensive care unit during the study recruitment period. The physicians in charge of the patients took care of them according to their usual practice and independently of the study. The methods used for this prospective registry were similar to those of a previous survey carried out in France five years earlier,10
although more data were collected in the most recent registry.
All consecutive patients admitted to the participating centres from 1 November through 30 November 2000 were included in the registry if they had (1) raised serum markers of myocardial necrosis higher than twice the upper limit of normal for creatine kinase, creatine kinase MB fraction or troponins, and (2) any or all of symptoms compatible with AMI for at least 30 minutes, ECG changes on at least two contiguous leads with pathological Q waves (at least 0.04 seconds) and persistent ST elevation or depression > 0.1 mV. The time from the beginning of symptoms to admission to the intensive care unit had to be < 48 hours.
For the present analysis, all patients with glycaemia measured on admission and who had no history of or treatment for diabetes mellitus at entry were included. Diabetics diagnosed during the hospital stay were excluded. Of a total population of 2320 patients, 1833 had no known or recognised diabetes mellitus. From this group, the concentration of admission blood glucose was recorded in 1604 patients (88%), who formed the study population.
The patients' cardiovascular history, their medications at the time of admission, their risk factors, their in‐hospital clinical course, including maximum Killip class, and the initial diagnostic and therapeutic management were recorded for each patient. Furthermore, left ventricular ejection fraction, when assessed at any time during the first five days, was recorded.
We compared initial and outcome data according to sextiles of blood glucose concentrations at admission. However, as the baseline characteristics and outcomes in the patients in the first three sextiles were similar, the results presented here regroup all three first sextiles into a single category (that is, those with admission glycaemia below the median value). All continuous variables are described as their mean (SD). All categorical variables are described in terms of absolute and relative frequency distributions. Groups were compared by one way analysis of variance for continuous variables and χ2 tests for discrete variables. Multiple logistic regression analysis was used to determine independent correlates of in‐hospital mortality and Cox multivariate regression analysis was used to assess predictors of one year outcome. Variables with p < 0.10 on univariate analyses were included in the models. Survival curves were generated by the Kaplan–Meier method and compared by log rank tests. For all tests p < 0.05 was considered significant.