Continuous subcutaneous insulin infusion was initially developed as a research procedure in the 1970s but quickly became a routine treatment for selected people with Type 1 diabetes. Continuous subcutaneous insulin infusion and other diabetes technologies, such as continuous glucose monitoring, are now an established and evidence-based part of diabetes care, but there has been some confusion about effectiveness and best use, particularly because of conflicting results from meta-analyses. This is because literature summary meta-analyses (including all trials) are inappropriate for therapeutic and economic decision-making; such meta-analyses should only include trials representative of groups likely to benefit. For example, for continuous subcutaneous insulin infusion, this would be those with continued disabling hypoglycaemia or elevated HbA1c levels. Alternatively, individual patient data meta-analysis allows modelling of covariates that determine effect size, e.g. in the case of continuous glucose monitoring, baseline HbA1c and frequency of sensor usage. Diabetes technology is therefore an example of personalized medicine, where evaluation and use should be both appropriate and targeted. This will also apply to future technologies such as new ‘patch’ pumps for Type 2 diabetes, closed-loop insulin delivery systems and nanomedicine applications in diabetes that we are currently researching. These include fluorescence lifetime-based non-invasive glucose monitoring and nanoencapsulation of islets for improved post-transplant survival.