In this open-label multicentre randomised controlled trial, we investigated three peri-operative treatment strategies to lower glucose and reduce the need for rescue insulin in patients aged 18-75 years with type-2 diabetes mellitus undergoing non-cardiac surgery. Patients were randomly allocated using a web-based randomisation program to premedication with liraglutide (liraglutide group), glucose-insulin-potassium infusion (insulin infusion group) or insulin bolus regimen (insulin bolus group), targeting a glucose < 8.0 mmol.l . The primary outcome was the between group difference in median glucose levels 1 h after surgery. We analysed 150 patients (liraglutide group n = 44, insulin infusion group n = 53, insulin bolus group n = 53) according to the intention-to-treat principle. Median (IQR [range]) plasma glucose 1 h postoperatively was lower in the liraglutide group compared with the insulin infusion and insulin bolus groups (6.6 (5.6-7.7 [4.2-13.5]) mmol.l vs. 7.5 (6.4-8.3 [3.9-16.6]) mmol.l (p = 0.026) and 7.6 (6.4-8.9 [4.7-13.2]) mmol.l ) p = 0.006, respectively). The incidence of hypoglycaemia and postoperative complications did not differ between the groups. Six patients had pre-operative nausea in the liraglutide group, of which two had severe nausea, compared with no patients in the insulin infusion and insulin bolus groups (p = 0.007). The pre-operative administration of liraglutide stabilised peri-operative plasma glucose levels and reduced peri-operative insulin requirements, at the expense of increased pre-operative nausea rates.

Continuous Glucose Monitoring (CGM) systems could improve glycemic control in critically ill patients. We aimed to identify the evidence on the clinical benefits and accuracy of CGM systems in these patients. For this, we performed a systematic search in Ovid MEDLINE, from inception to 26 July 2016. Outcomes were efficacy, accuracy, safety, workload and costs. Our search retrieved 356 articles, of which 37 were included. Randomized controlled trials on efficacy were scarce (n = 5) and show methodological limitations. CGM with automated insulin infusion improved time in target and mean glucose in one trial and two trials showed a decrease in hypoglycemic episodes and time in hypoglycemia. Thirty-two articles assessed accuracy, which was overall moderate to good, the latter mainly with intravascular devices. Accuracy in critically ill children seemed lower than in adults. Adverse events were rare. One study investigated the effect on workload and cost, and showed a significant reduction in both. In conclusion, studies on the efficacy and accuracy were heterogeneous and difficult to compare. There was no consistent clinical benefit in the small number of studies available. Overall accuracy was moderate to good with some intravascular devices. CGM systems seemed however safe, and might positively affect workload and costs.

No abstract

No abstract

Background Pancreatic panniculitis is characterized by subcutaneous fat necrosis and is a rare presentation of an underlying pancreatic disease, appearing in approximately 2–3% of all patients with a pancreatic disease. The nodules usually involve the lower extremities. Pancreatic panniculitis is commonly associated with acute or chronic pancreatitis, and occasionally with pancreatic cancer, especially acinar cell carcinoma. Case presentation A 77-year-old Caucasian woman with no significant medical history was referred to our center with multiple painful, itchy, and warm red/blue cutaneous nodules on the left lower leg. These skin lesions were consistent with the clinical diagnosis of panniculitis. The skin biopsy obtained showed a predominantly lobular panniculitis with fat necrosis of which the aspect was highly suspicious for pancreatic panniculitis. Further analysis revealed high lipase serum of > 3000 U/L (normal range < 60 U/L), and on computed tomography scan a mass located between the stomach and the left pancreas was seen. Endoscopic ultrasonography-guided fine-needle biopsy confirmed the diagnosis of acinar cell carcinoma. After discussing the patient in the pancreatobiliary multidisciplinary team meeting, laparoscopic distal pancreatectomy including splenectomy and en bloc wedge resection of the stomach due to tumor in-growth was performed. The cutaneous nodules on both legs disappeared 1–2 days after surgery. No long-term complications were reported during follow-up. One year after surgery, the patient presented with similar symptoms as preoperatively. Computed tomography scan showed local recurrence and distal metastases, which were subsequently confirmed by biopsy. She started with palliative folinic acid–fluorouracil–irinotecan–oxaliplatin chemotherapy but stopped after two cycles because of disease progression. The patient died 2 months later, 13 months after surgical resection. Conclusion This case illustrates the importance of clinically recognizing cutaneous nodules and pathological recognizing the specific microscopic changes as sign of a (malignant) pancreatic disease.