Perioperative diabetes regulation with the help of a standard protocol

According to a standard protocol 115 patients (96 type 2 DM, 19 type 1 DM) were treated with a 3 litre glucose 5%/day infusion, KCl and insulin, with the aim of obtaining near-normoglycaemia perioperatively. Blood glucoses of 6.7-10.0 mmol/l were scored as ideal, 3.5-15.0 mmol/l as acceptable, and the remaining values as unacceptable. In 32 patients (28%) the protocol was followed in all aspects, whereas in 45 patients (39%) the protocol was followed in therapeutic aspects (total n = 77.67%, Group 1). In the remaining 38 patients (33%) protocol violations occurred (Group 2). In 48 patients (62%) of Group 1 ideal or acceptable control was obtained perioperatively versus 11 patients (29%) of Group 2 (p < 0.05). In patients with good preoperative long-term regulation (HbA1 < or = 8.0%; n = 81) ideal or acceptable perioperative control was achieved in 58% (n = 47) versus 13% (n = 4) of those (n = 31) with preoperative HbA1 > 8.0% (p < 0.05). Stepwise multiple logistic regression analysis showed that preoperative long-term regulation (HbA1; p < 0.001) and adherence to the protocol (p = 0.022) were the only independent variables with prognostic significance for perioperative blood glucose control. In 17 patients (15%) minor electrolyte disturbances occurred. No patient had frank ketosis. Duration of disease, presence of complications, type of anaesthesia and operation did not affect perioperative diabetes control.(ABSTRACT TRUNCATED AT 250 WORDS)