CSII vs MDI: the evidence

Insulin regimens are traditionally known as “conventional”, with two injections of mixed insulin, or “intensive”, in which the basal [background] and bolus [meal] insulin doses are given separately. Intensive insulin therapy has been shown to reduce risk of complications. It can be delivered as multiple daily dose insulin injections (MDI) including long acting basal insulin with boluses of rapid acting insulin given pre-meal, or using continuous subcutaneous insulin infusion [CSII] via an insulin pump. CSII delivers the “basal” component of the insulin regime via a slow infusion, and the patient can deliver “boluses” as and when required in addition to this. While both CSII and MDI are effective methods of intensive insulin therapy, they each have advantages and disadvantages. This review evaluates the evidence for both these regimens, highlighting the limitations of the evidence as it currently stands.


In MDI, we separate out the basal and bolus components of insulin delivery. Basal insulin, delivered using long acting insulin [NPH, Determir, Glargine, Degludec] provides background insulin and ideally the dose is titrated to keep blood glucose stable if no carbohydrate is consumed. Fast acting insulin [Soluble insulin, Lispro, Aspart, Glulisine] is given pre-meal to cover the carbohydrate content of the meal and correct any readings that may be out of range. The DCCT study from the late 1980’s showed that MDI provides tighter glucose control than conventional insulin given with twice daily mixed insulin, with a significant reduction in risk of complications of diabetes[1]. In this study, there was however a penalty to pay in the form of three-fold higher rates of severe hypoglycaemia requiring third party assistance. However, when MDI is used after structured education in flexible insulin therapy [FIT] and patients are taught how to adjust basal insulin in response to events such as exercise and illness, and adjust the rapid acting insulin according to expected carbohydrate intake and pre-meal glucose, the evidence suggests lasting improvements in glycemic control with reductions in hypoglycaemia rates as well. Randomised trials and registry data of FIT such as the Dose Adjustment for Normal Eating [DAFNE] study in the UK[2] and the ITTP [Intensive treatment training programme][3]in Germany have demonstrated improved glycemic control with reduced hypoglycaemia and improvement in quality of life. The original DAFNE trial excluded patients with problematic control, and only included those with poor glycemic control, but subsequent registry data has demonstrated a 0.3% reduction in HBA1c maintained up to 7 years[4], with greater reductions in those with higher baseline HbA1c. Data from the DAFNE database also show a 3-fold reduction in severe hypoglycaemia, with hypoglycaemia awareness improved in 43% of those who had impaired awareness at baseline.


Fig 1: This is the usual insulin profile here in blue with a background amount of insulin always present and then rapid high spikes of insulin after meals - the pump allows variable doses of background insulin through the day and allows ferquent boluses with food if required. It has been shown in trials and metanalyses that there is about 05% reduction in HbAc on average with a 4-fold reduction inhypoglycameia. This line shows how we can vary the basal insulin as required to more closely match a physiological profile.
Fig 1: This is the usual insulin profile here in blue with a background amount of insulin always present and then rapid high spikes of insulin after meals - the pump allows variable doses of background insulin through the day and allows ferquent boluses with food if required. It has been shown in trials and metanalyses that there is about 05% reduction in HbAc on average with a 4-fold reduction inhypoglycameia. This line shows how we can vary the basal insulin as required to more closely match a physiological profile.
Continuous subcutaneous insulin infusion [CSII] via Insulin pumps was first used to deliver insulin almost 35 years ago. The pump delivers the basal component of the regimen by constant infusion of rapid acting insulin, with meal boluses delivered when required by button presses on the machine, or with some more recent pumps by the use of a hand held control. Delivering basal insulin in this way allows more precise adjustment of basal rates at different times of the day, theoretically providing a better fit to the diurnal profile of basal requirements. It also allows for changes in basal rates to cover exercise, illness or stress. Most modern pumps also have calculators built in that allow precise calculation of mealtime and correction boluses, and also allow the meal boluses to be delivered in a prolonged manner to cover high fat or low glycaemic index meals.

Comparison of MDI vs CSII

Surprisingly, there are no large-scale RCT data comparing CSII and MDI. Most of the available literature is restricted to small scale before and after studies. Meta-analyses of these studies show a reduction in HbA1c with CSII between 0.3-0.6% [5][6]. Those with the highest baseline HbA1c levels on MDI benefit the most from switching to CSII. The meta-analyses also show a significant reduction in severe hypoglycaemia, again with the greatest reductions in those with the greatest rates of hypoglycaemia at baseline. The meta-analysis by Pickup et al found that the frequency of severe hypoglycaemia is reduced by approximately 75% during treatment with CSII compared with MDI [rate ratio 4.19, 95% CI 2.86-6.13], although another meta-analysis by Yeh et al, that used different criteria to choose studies, reported no change in SH[7].

A major problem with studies comparing CSII with MDI is that most were done with NPH as the comparator basal insulin, which causes known problems with nocturnal hypoglycaemia due to the peak of the insulin action occurring around 3 am when requirements are naturally low. There have only been 4 studies comparing CSII to MDI based on modern “peak-less” basal insulin analogues, with a total of 111 patients, and these showed less impressive benefits for CSII. None of the studies have controlled for education and attention, which as seen with FIT, confers a benefit in both glycemic control and reduction in hypoglycaemia, although the current REPOSE trial is evaluating this question.

Studies of flexible insulin therapy and CSII both report improved quality of life, in particular improved freedom and satisfaction with treatment over conventional MDI, although these have not been compared head-to-head.

Despite the complexities and demands of CSII, the discontinuation rate for CSII (either by choice or determined by the healthcare professional), is low and estimated to be below ≤5% [8]. Both of these methods of delivering intensive insulin regimens provide increased contact with health care teams and a lot of education of patients, which may be a contributory factor in improved metabolic control as well as improved quality of life.

Type 2 diabetes

Many patients with type 2 diabetes are now insulin treated, and may be treated with a variety of insulin regimens. As patients with type 2 diabetes still have some residual beta-cell function, dosing may not need to be as accurate, or may be supplemented with oral agents. There have been a few studies comparing CSII with MDI in type 2 diabetes. Meta-analysis of these studies showed no benefit of CSII over MDI in improving glycaemic control in patients with T2DM [9].

For insulin resistant patients with suboptimal glycaemic control despite high insulin requirements, the use of U-500 (highly concentrated insulin preparation, 500 IU/ml) in CSII has shown promising results in sustained improvement in glycaemic control without additional weight gain although more studies are needed [10]. In summary, CSII may be beneficial in selected groups of patients with T2DM, particularly those with huge insulin requirements (in excess of 200 units daily). However, further RCTs investigating changes in glycaemic control, hypoglycaemia rates, weight changes and quality of life indicators are needed before definitive recommendations can be made.


In pregnant women with pre-existing type 1 or type 2 diabetes, suboptimal glycemic control is associated with poorer pregnancy outcomes, resulting in maternal, fetal and neonatal complications. Due to the tight glycaemic targets recommended during pregnancy, the majority of women with pre-gestational diabetes have complex insulin requirements that change rapidly through pregnancy. Although CSII is safe and effective in pregnancy, there is a lack of evidence due to relatively few RCTs or observational studies in this area. A Cochrane systematic review showed no significant differences in primary outcomes investigated (macrosomia, caesarean deliveries) and secondary outcomes - neonatal complications (perinatal mortality, fetal anomalies, gestation at delivery, neonatal hypoglycaemia and small for gestational age), and maternal complications (maternal hypo/hyperglycaemia, mean HbA1c and 24 hour mean blood glucose levels in each trimester) [11].

Despite this, CSII does have potential advantages that would be useful in pregnant women with both T1DM and T2DM [8][12]. These include improvements in HbA1c levels, flexible dosing of insulin with temporary basal rates, reduced frequency of hypoglycaemia and glycaemic variability, improvement in quality of life and additional benefits of weight maintenance in T2DM.


Both MDI and CSII are forms of intensive insulin therapy that are able to achieve optimal glycaemic control, with CSII showing advantage over MDI in selected patient populations and when used appropriately. For either to be successful, intensive insulin therapy should always be coupled with structured patient education and blood glucose monitoring. Current CSII can be coupled with continuous glucose monitoring systems (CGMS) and ongoing research has made the development of the closed-loop system soon-to-be a reality.


  1. ^ DCCT, The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. The Diabetes Control and Complications Trial Research Group. N.Engl.J.Med., 1993. 329(14): p. 977-986.

  2. ^ Hopkins, D., et al., Improved biomedical and psychological outcomes 1 year after structured education in flexible insulin therapy for people with type 1 diabetes: the U.K. DAFNE experience. Diabetes Care, 2012. 35(8): p. 1638-42.

  3. ^ Bott, S., et al., Intensified insulin therapy and the risk of severe hypoglycaemia. Diabetologia, 1997. 40(8): p. 926-932.

  4. ^ Gunn, D. and P. Mansell, Glycaemic control and weight 7 years after Dose Adjustment For Normal Eating (DAFNE) structured education in Type 1 diabetes. Diabet Med, 2012. 29(6): p. 807-12.

  5. ^ Pickup, J.C. and A.J. Sutton, Severe hypoglycaemia and glycaemic control in Type 1 diabetes: meta-analysis of multiple daily insulin injections compared with continuous subcutaneous insulin infusion. Diabet Med, 2008. 25(7): p. 765-74.

  6. ^ Misso, M.L., et al., Continuous subcutaneous insulin infusion (CSII) versus multiple insulin injections for type 1 diabetes mellitus. Cochrane Database Syst Rev, 2010(1): p. CD005103.

  7. ^ Yeh, H.C., et al., Comparative effectiveness and safety of methods of insulin delivery and glucose monitoring for diabetes mellitus: a systematic review and meta-analysis. Ann Intern Med, 2012. 157(5): p. 336-47.

  8. ^ Pickup, J.C., Management of diabetes mellitus: is the pump mightier than the pen? Nat Rev Endocrinol, 2012. 8(7): p. 425-33.

  9. ^ Monami, M., et al., Continuous subcutaneous insulin infusion versus multiple daily insulin injections in type 2 diabetes: a meta-analysis. Exp Clin Endocrinol Diabetes, 2009. 117(5): p. 220-2.

  10. ^ Lane, W.S., Weinrib, S.L., Rappaport, J.M., Hale, C.B. et al, The effect of long-term use of U-500 insulin via continuous subcutaneous insulin infusion on durability of glycemic control and weight gain in obese, insulin-resistant patients with type 2 diabetes. Endocrine Practice, 2012. 27: p. 1-18.

  11. ^ Farrar, D., D.J. Tuffnell, and J. West, Continuous subcutaneous insulin infusion versus multiple daily injections of insulin for pregnant women with diabetes. Cochrane Database Syst Rev, 2007(3): p. CD005542.

  12. ^ Reznik, Y. and O. Cohen, Insulin pump for type 2 diabetes: use and misuse of continuous subcutaneous insulin infusion in type 2 diabetes. Diabetes care, 2013. 36 Suppl 2: p. S219-25.


Nobody has commented on this article

Commenting is only available for registered Diapedia users. Please log in or register first.