Objectives: Tight glycemic control using intermittent blood glucose measurements is associated with a risk of hypoglycemia. Glucose concentrations can now be measured near continuously (every 5–15 min). We assessed the quality and safety of glycemic control guided by a near-continuous glucose monitoring system in ICU patients. Design: Prospective, cluster-randomized, crossover study. Setting: Thirty-five–bed medico-surgical department of intensive care with four separate ICUs. Patients: Adult patients admitted to the department and expected to stay for at least 3 days were considered for inclusion if they had persistent hyperglycemia (blood glucose > 150 mg/dL) up to 6 hours after admission and/or were receiving insulin therapy. Interventions: A peripheral venous catheter was inserted in all patients and connected to a continuous glucose monitoring sensor (GlucoClear; Edwards Lifesciences, Irvine, CA). The four ICUs were randomized in pairs in a crossover design to glycemic control using unblinded or blinded continuous glucose monitoring monitors. The insulin infusion rate was adjusted to keep blood glucose between 90 and 150 mg/dL using the blood glucose values displayed on the continuous glucose monitor (continuous glucose monitoring group—unblinded units) or according to intermittent blood glucose readings (intermittent glucose monitoring group—blinded units). Measurements and Main Results: The quality and safety of glycemic control were assessed using the proportion of time in range, the frequency of blood glucose less than 70 mg/dL, and the time spent with blood glucose less than 70 mg/dL (TB70), using blood glucose values measured by the continuous glucose monitoring device. Seventy-seven patients were enrolled: 39 in the continuous glucose monitoring group and 38 in the intermittent glucose monitoring group. A total of 43,107 blood glucose values were recorded. The time in range was similar in the two groups. The incidence of hypoglycemia (8/39 [20.5%] vs 15/38 [39.5%]) and the TB70 (0.4% ± 0.9% vs 1.6% ± 3.4%; p < 0.05) was lower in the continuous glucose monitoring than in the intermittent glucose monitoring group. Conclusions: Use of a continuous glucose monitoring–based strategy decreased the incidence and severity of hypoglycemia, thus improving the safety of glycemic control.

Vincent, Jean-Louis; Carlier, Eric; Brimioulle, Serge; Pinsky, Michael; Lejeune, Philippe; Naeije, Robert; Goldstein, Jacques; Kahn, Robert J.; Leclerc, Jean-Louis; Primo, Georges Author Information

Degaute, J. P.; Goldstein, J. P.; Mèlot, C.; Leeman, M.; Le Clerc, J. L.; Primo, G. Author Information

Serotonin can induce pulmonary hypertension, hypoxia and bronchoconstriction, and ketanserin has been shown to reverse these effects on various experimental models of acute respiratory failure. In the present study, the hemodynamic and gasometric effects of ketanserin were studied during acute respiratory failure induced by an air infusion at a rate of 10 ml/min in dogs. During a 60-min air infusion, 10 dogs received 4 mg of ketanserin i.v. and 10 dogs served as control. Ketanserin-treated dogs had similar pulmonary hypertension even though more significant decreases in arterial pressure and systemic vascular resistance characterized the systemic effects of ketanserin. Similarly, a marked increase in hematocrit observed in control dogs (from 36.9 to 43.8%, p less than 0.01) was totally prevented by ketanserin (from 40.3 to 40.4%, NS). Hypoxia was similar, although the increase in pulmonary shunt was attenuated (259 instead of 468%). Therefore, the influence of serotonin is very limited in acute respiratory failure secondary to air embolization. Serotonin might have a more important influence on the systemic than on the pulmonary vasculature in these conditions.

Air embolization has been shown to produce a reversible permeability-type of pulmonary oedema. The present study investigated the haemodynamic, gasometric and haematological changes associated with air infusion in the spontaneously breathing dog (weight 31 +/- 5 kg). Air was infused at a rate of 10 ml X min-1 for 60 min (10 dogs) or 180 min (5 dogs). During the air infusion, a dramatic increase in pulmonary artery pressure was associated with only a moderate increase in right atrial pressure and limited decreases in arterial pressure and in cardiac output. A marked decrease in end-tidal PCO2 reflected the increase in dead space. These changes were stable during air infusion, but rapidly reversed after the end of infusion. However, hypoxaemia, defined by a decreased PaO2/PAO2 ratio, deteriorated with time and was only partially reversible. At histological examination, interstitial pulmonary oedema was present around the pulmonary arterioles. Air infusion was associated with rapid decreases in circulating leukocytes and platelets and complement activation. Since leukotriene release might be associated with leukocyte activation in this model, seven additional dogs were pretreated by inhalation of 10 mg of the leukotriene inhibitor U-60,257. The increase in pulmonary vascular resistance and in pulmonary shunt were moderately reduced and the drop in circulating leukocytes and platelets was strikingly abolished in the treated animals. Air infusion in the spontaneously breathing dog represents a model of very stable and reversible pulmonary hypertension. It can reproduce important pathophysiological features implicated in the development of pulmonary oedema.(ABSTRACT TRUNCATED AT 250 WORDS)

Prostacyclin (PGI2) has been shown to present myocardial protective effects which could be beneficial during cardiac arrest. We tested this hypothesis in a closed-chest dog model in which electromechanical dissociation (EMD) can be predictably observed after 90 to 120 seconds of ventricular fibrillation whithout chest compression. Six dogs were pretreated with a PGI2 infusion at a rate of 1 mcg/kg/min and six other dogs served as control animals.

Calcium-entry blockers were administered in an attempt to protect myocardium during cardiac arrest due to ventricular fibrillation in mechanically ventilated dogs. An intravenous injection of verapamil, nifedipine, or lidoflazine was administered prior to successively prolonged episodes of ventricular fibrillation, during which no thoracic compression was performed. It is of interest that ventricular defibrillation was more easily obtained after treatment with the three drugs. As previously observed in this model, nine control dogs developed electromechanical dissociation (EMD) after 120 seconds of ventricular fibrillation. In contrast, six of the 11 dogs treated with 0.3 mg/kg of verapamil recovered mechanical systole after 120 seconds of ventricular fibrillation (p less than 0.05). Nifedipine administration also postponed the onset of EMD in three of four dogs. However, lidoflazine postponed the onset of EMD in only one of the eight dogs. The later onset of EMD after administration of verapamil or nifedipine in this model was attributed to myocardial protection by calcium-entry blockers during ventricular fibrillation. Decreased energy utilization during cardiac arrest was considered to be the principal protective mechanism. These observations indicate calcium-entry blockers, and especially verapamil and nifedipine, can be valuable drugs during cardiac resuscitation for ventricular fibrillation.