Comparative effectiveness and safety of three different basal insulins in patients with type 2 diabetes in a real world setting (ORBIT): a multicentre, prospective, registry study
Linong Ji,Puhong Zhang,Dongshan Zhu,Juming Lu,Xiaohui Guo,Yangfeng Wu,Xian Li,Jiachao Ji,Weiping Jia,Wenying Yang,Dajin Zou,Zhiguang Zhou,Yan Gao,Satish K. Garg,Changyu Pan,Jianping Weng,Sanjoy K. Paul
DOI: https://doi.org/10.1016/S2213-8587(16)30389-8
IF: 44.867
2016-01-01
The Lancet Diabetes & Endocrinology
Abstract:Abstract Background The three most commonly used basal insulins in patients with type 2 diabetes are neutral protamine Hagedorn (NPH) insulin and the long-acting insulin analogues glargine and detemir. Although randomised controlled trials have shown equivalent efficacy of these three insulins, inconsistencies in safety exist. We aimed to assess the safety and efficacy of these three basal insulins in real-world clinical setting in the ORBIT (Observational Registry for Basal Insulin Treatment) study. Methods In this multicentre, 6 month, prospective, registry study, patients were recruited between Nov 25, 2011, and Sept 1, 2014, who were aged 18–80 years, with type 2 diabetes that was inadequately controlled (HbA 1C ≥7%, 53 mmol/mol) with oral antidiabetic drugs and who were willing to start treatment with basal insulin were enrolled from 209 hospitals in eight regions of China. Patients who were prescribed any type of insulin in the past 2 years, with clinically significant acute major organ or systemic disease or other condition judged by the investigator that would create difficulty for the 6 month follow-up, or current or planned pregnancy or lactating women were excluded. Type and dose of basal insulin were at the physicianu0027s discretion and at the patientsu0027 willingness. Assessments were done at baseline (0 month—visit 1), mid-term (3 months—visit 2), and at the end of the study (6 months—visit 3) by investigators during routine practice. During analysis, patients were separated into NPH, glargine, and detemir insulin groups according to their initial prescription of basal insulin. Outcomes included the change in concentration of HbA 1c , hypoglycaemia rate, and bodyweight change in the three groups from baseline to 6 months. Propensity scores with regression adjustment were used to balance the baseline differences among the three groups. Intention-to-treat analysis was used to assess the outcomes. Findings Our findings show that treatment with long-acting insulin analogues is associated with better glycaemic control and less hypoglycaemia and weight gain than treatment with NPH insulin in patients with type 2 diabetes. Interpretation 16 341 patients were analysed in the intention-to-treat group and 9002 were analysed in the per-protocol group. In the intention-to-treat population, there were 11 290 (69%) patients treated with glargine, 2135 (13%) patients treated with detemir, and 18% (2916) patients treated with NPH insulin. After controlling for the baseline and process variables in the propensity score regression, the intention-to-treat analysis showed that the reductions in HbA 1c in the glargine (2%, SD 2·1, p=0·0014) and detemir (2%, 2·1, p=0·0065) groups were higher than that in the NPH group (2%, 2·2). The detemir group had lower weight gain (−0·1 kg, SD 2·9) than the glargine group (0·1 kg, 3·0; p=0·032) and the NPH group (0·3 kg, 3·1; p Funder Sanofi (Shanghai, China).