Abstract
Despite substantial advances in diabetes care over the previous decade, only a minority of people with type 1 diabetes achieve an HbA1c <53 mmol/mol.1 Automated insulin delivery (AID) systems have consistently demonstrated efficacy in reducing HbA1c.2 Omnipod 5 (Insulet Corp) is a tubeless AID system which was launched in the United Kingdom in 2023. Here, we report the first 1-year CGM, HbA1c and weight outcomes in people using Omnipod 5, under routine clinical care, in the United Kingdom.
This was a prospective, observational assessment based in a single Scottish centre which provides diabetes care for approximately 5000 adults with type 1 diabetes. As a service evaluation of routinely collected data, this project did not require ethical approval. We included all adults who transitioned from Omnipod DASH (standalone CSII) to Omnipod 5 (hybrid closed loop with Dexcom G6 CGM) between June and August 2023. Data were obtained from the electronic health record (SCI-Diabetes), LibreView, Dexcom Clarity and Glooko.
Paired CGM data were available in 45/50 (90 days prior to Omnipod 5 and 90 days after 1 year of use) and paired HbA1c data were available in 41/50 (measured at a median 200 days [202–336] after OP5 commencement). CGM metrics reported are consistent with those described in the international consensus document.3 Paired weight data were available in 28/50 (median 308 days after OP5 commencement [244–370]). Results are presented as median (IQR). Paired data were compared with Wilcoxon signed rank tests and correlations were assessed by Spearman correlation coefficient. p <0.05 were considered statistically significant. Statistical analyses were performed using R Studio.
Median age was 42 years (IQR: 30–53), duration of diabetes was 24 years (13–34) and 64% were female. Baseline HbA1c was 69 mmol/mol (61–75) and 22% had an HbA1c <58 mmol/mol. Thirty-eight per cent had BMI >30 kg/m2. Fifty-nine per cent were predominantly using the lowest glucose target (6.1 mM) at the end of follow-up and median time in auto mode was 94% (91–99).
In those with paired HbA1c data, median baseline HbA1c was 70 mmol/mol (63–76) and fell to 58 mmol/mol (52–63) during Omnipod 5 use (p < 0.001). Data summarising CGM changes are presented in the figure (Figure 1), including a change in TIR from 42% (33–58) to 60% (53–68, p < 0.001). GMI fell from 66 mmol/mol (57–70) to 58 mmol/mol (54–62, p < 0.001) and coefficient of variation for glucose fell from 37% (34–42) to 34% (32–38, p = 0.009). Change in TIR at 1 year was strongly negatively correlated with baseline TIR (R −0.581, p < 0.001). Percentage of insulin delivered as bolus was negatively correlated with increase in TIR (R −0.518, p < 0.001). Age, sex and socio-economic deprivation were not associated with TIR response to Omnipod 5.
This was a prospective, observational assessment based in a single Scottish centre which provides diabetes care for approximately 5000 adults with type 1 diabetes. As a service evaluation of routinely collected data, this project did not require ethical approval. We included all adults who transitioned from Omnipod DASH (standalone CSII) to Omnipod 5 (hybrid closed loop with Dexcom G6 CGM) between June and August 2023. Data were obtained from the electronic health record (SCI-Diabetes), LibreView, Dexcom Clarity and Glooko.
Paired CGM data were available in 45/50 (90 days prior to Omnipod 5 and 90 days after 1 year of use) and paired HbA1c data were available in 41/50 (measured at a median 200 days [202–336] after OP5 commencement). CGM metrics reported are consistent with those described in the international consensus document.3 Paired weight data were available in 28/50 (median 308 days after OP5 commencement [244–370]). Results are presented as median (IQR). Paired data were compared with Wilcoxon signed rank tests and correlations were assessed by Spearman correlation coefficient. p <0.05 were considered statistically significant. Statistical analyses were performed using R Studio.
Median age was 42 years (IQR: 30–53), duration of diabetes was 24 years (13–34) and 64% were female. Baseline HbA1c was 69 mmol/mol (61–75) and 22% had an HbA1c <58 mmol/mol. Thirty-eight per cent had BMI >30 kg/m2. Fifty-nine per cent were predominantly using the lowest glucose target (6.1 mM) at the end of follow-up and median time in auto mode was 94% (91–99).
In those with paired HbA1c data, median baseline HbA1c was 70 mmol/mol (63–76) and fell to 58 mmol/mol (52–63) during Omnipod 5 use (p < 0.001). Data summarising CGM changes are presented in the figure (Figure 1), including a change in TIR from 42% (33–58) to 60% (53–68, p < 0.001). GMI fell from 66 mmol/mol (57–70) to 58 mmol/mol (54–62, p < 0.001) and coefficient of variation for glucose fell from 37% (34–42) to 34% (32–38, p = 0.009). Change in TIR at 1 year was strongly negatively correlated with baseline TIR (R −0.581, p < 0.001). Percentage of insulin delivered as bolus was negatively correlated with increase in TIR (R −0.518, p < 0.001). Age, sex and socio-economic deprivation were not associated with TIR response to Omnipod 5.
Original language | English |
---|---|
Journal | Diabetic Medicine |
Volume | 41 |
Issue number | 12 |
DOIs | |
Publication status | Published - 11 Oct 2024 |