Objective: To evaluate the effect of insulin pump on perioperative patients with type 2 diabetes and the role of health economics by applying data envelopment analysis (DEA).

Method: This retrospective study involved 639 patients with type 2 diabetes undergoing surgery in Guangdong Provincial Hospital of Traditional Chinese Medicine from January 2009 to November 2017. All patients were treated with insulin during the perioperative period. They were divided into a insulin pump group for diabetes (perioperative blood glucose control with insulin pump, 369 cases) and a Non-insulin pump group (perioperative blood glucose control with subcutaneous insulin injection, 270 cases). Data Envelope analysis was applied to compare the input indicators of the two groups to total cost and total length of hospital stay. The output indicators of the two groups were blood glucose control time before surgery, mean fasting blood glucose, blood glucose in the first sky after surgery, days of antibiotic use, and hospital stay after surgery.

Result: In the aspect of comparison of scale efficiency, the efficiency values of the pump set in the C2R model and BC2 model are better than those of the unused pump set. In Another aspect of the relaxation variables, the higher the surgical level, the closer the pump group is to the ideal state compared with the non-pump group, in terms of improving average fasting blood glucose, antibiotic use days, preoperative blood glucose control time, first day postoperative fasting blood glucose, and postoperative hospitalization days.

Conclusion: Insulin pump can quickly and stably control blood glucose levels and shorten the length of hospital stay during the perioperative period of type 2 diabetes patients. It is suggested that the application of insulin pump in perioperative period may have better benefits and is worthy of clinical promotion.


W. Song: None. J. Wen: None. L. Zhao: None. G. Fan: None. M. Luo: None. J. Liang: None. J. Zhang: None. C. Yang: None. Q. Wang: None. Z. Gao: None. L. Ma: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license.