The Efficacy of Technology in Type 1 Diabetes: A Systematic Review, Network Meta-analysis, and Narrative Synthesis

Background: Existing technologies for type 1 diabetes have not been compared against the full range of alternative devices. Multiple metrics of glycemia and patient-reported outcomes for evaluating technologies also require consideration. We thus conducted a systematic review, network meta-analysis, and narrative synthesis to compare the relative efficacy of available technologies for the management of type 1 diabetes. Methods: We searched MEDLINE, MEDLINE In-Process and other nonindexed citations, EMBASE, PubMed, All Evidence-Based Medicine Reviews, Web of Science, PsycINFO, CINAHL, and PROSPERO (inception-April 24, 2019). We included RCT >= 6 weeks duration comparing technologies for type 1 diabetes management among nonpregnant adults (>18 years of age). Data were extracted using a predefined tool. Primary outcomes were A1c (%), hypoglycemia rates, and quality of life (QoL). We estimated mean difference for A1c and nonsevere hypoglycemia, rate ratio for severe hypoglycemia, and standardized mean difference for QoL in network meta-analysis with random effects. Results: We identified 16,772 publications, of which 52 eligible studies compared 12 diabetes management technologies comprising 3,975 participants in network meta-analysis. Integrated insulin pump and continuous glucose monitoring (CGM) systems with low-glucose suspend or hybrid closed-loop algorithms resulted in A1c levels 0.96% (predictive interval [95% PrI] 0.04-1.89) and 0.87% (95% PrI 0.12-1.63) lower than multiple daily injections with either flash glucose monitoring or capillary glucose testing, respectively. In addition, integrated systems had the best ranking for A1c reduction utilizing the surface under the cumulative ranking curve (SUCRA-96.4). While treatment effects were nonsignificant for many technology comparisons regarding severe hypoglycemia and QoL, simultaneous evaluation of outcomes in cluster analyses as well as narrative synthesis appeared to favor integrated insulin pump and continuous glucose monitors. Overall risk of bias was moderate-high. Certainty of evidence was very low. Conclusions: Integrated insulin pump and CGM systems with low-glucose suspend or hybrid closed-loop capability appeared best for A1c reduction, composite ranking for A1c and severe hypoglycemia, and possibly QoL. Registration: PROSPERO, number CRD42017077221.