Aims Continuous subcutaneous infusion of insulin (CSII) via an insulin pump is believed to provide more physiological and flexible control of type 1 diabetes compared with injection regimens such as intensive multiple daily injection (MDI) regimens. However, a number of small underpowered randomised trials have failed to provide conclusive evidence for this. The authors undertook a systematic review and meta-analysis to better estimate the effectiveness and safety of CSII compared with MDI.

Methods A systematic review was undertaken to identify randomised controlled trials (RCT) and unrandomised prospective controlled studies comparing CSII with MDI in subjects ≤18 years with established type 1 diabetes (≥6 months) and of duration ≥3 months. Meta-analysis for haemoglobin A1c (HbA1c), hypoglycaemia and quality of life outcomes was undertaken using random effects model in Review Manager 5.0.22.

Results 10 RCT s (n=307) and five prospective studies (n=318) were included. Overall quality of trials was poor.

Glycaemic control: At 6 months CSII was shown to provide a minor benefit in reduced HbA1c over MDI in RCTs (mean pooled difference = −0.24% (95% CI −0.48 to −0.00); minimal heterogeneity (I2=0%)) and prospective studies (−0.32 (95% CI −0.62 to −0.03); I2=0%). These benefits remained at 12 months in prospective studies although they failed to reach significance. There were insufficient data for analysis in RCTs at 12 months.

Hypoglycaemia: CSII was associated with reduced risk of severe hypoglycaemia at 6 months in RCTs (RR 0.66 (95% CI 0.32 to 1.36); not significant) and at 12 months in prospective studies (RR 0.50 (95% CI 0.28 to 0.88); I2=0%).

Quality of life: Data were sparse and inappropriate for meta-analysis.

Conclusion CSII in children and adolescents is associated with small mean reduction in HbA1c and a decreased risk of severe hypoglycaemia when compared with MDI in relatively unselected populations during short-term follow-up. There is little evidence that CSII improves quality of life compared with MDI. Larger better quality trials with longer-term follow-up are needed to establish the benefit of CSII in paediatric populations.