The quality of evidence is downgraded by study limitations.
Real-time continuous glucose monitoring is suggested for patients who do not achieve target level of HbA1c or have a tendency to hypoclycaemias.
The recommendation is weak because of questionable cost-effectiveness.
A Cochrane review [Abstract] 1 included 22 RCTs with a total of 2883 patients to assess the effects of continuous glucose monitoring (CGM) systems compared to conventional self-monitoring of blood glucose (SMBG) in patients with diabetes mellitus type 1. All age groups were included.
After six months there was a significant larger decline in HbA1c level for real-time CGM users who started insulin pump therapy compared to patients using multiple daily injections (MDI) and SMBG (MD in change in HbA1c level -0.7%, 95% CI -0.8% to -0.5%, 2 RCTs, n = 562, I2 =84%).
For patients starting with CGM only (on MDI), the average decline in HbA1c level six months after baseline was also statistically significantly larger for CGM users compared to SMBG users, but much smaller than for patients starting using an insulin pump and CGM at the same time (MD change in HbA1c level -0.2%, 95% CI -0.4% to -0.1%, 6 RCTs, n = 963, I2=55%). On average, there was no significant difference in risk of severe hypoglycaemia or ketoacidosis between CGM and SMBG users (table 1).
| Outcome (follow-up 6 months) | Participants (studies) | Assumed risk(control) | Corresponding risk (intervention) | Relative effect (95% CI) | Quality of evidence (GRADE) |
|---|---|---|---|---|---|
| Severe hypoglycaemia | 689 (5) | 75 per 1000 | 79 per 1000 | RR 1.05(0.63 to 1.77) | ⊕⊕⊝⊝low |
| Ketoacidosis | 689 (5) | 23 per 1000 | 20 per 1000 | RR 0.85(0.32 to 2.26) | ⊕⊕⊝⊝low |
A single-technology assessment 2 included 2 RCTs of unclear to high risk of bias comparing FreeStyle Libre (Libre) to SMBG. Main findings from these trials were that Libre may slightly improve treatment satisfaction, time spent with glucose in range 3.9 to 10 mmol/L, number of nocturnal events with glucose levels <3.1 mmol/L within 7h, and time spent with glucose levels >13.0 mmol/L in comparison to SMBG. Libre lead to little or no difference in quality of life and HbA1c level in comparison to SMBG. The evidence was uncertain about whether Libre leads to an improvement in time and events with glucose <3.9mmol/L within 24 h, time with glucose <3.1 mmol/L at night within 7 hours, and time with glucose >10 mmol/L. From a healthcare perspective, the submitter has calculated a budget impact for type 1 DM to have a total added cost the fifth year after adoption of the technology. The submitter did not calculate a budget impact for type 2 DM only. From a Norwegian healthcare perspective it was estimated, the annual costs five years after introduction would be NOK 186 million added cost and NOK 91.7 million saved cost for type 1 and 2 DM alone, respectively, and NOK 94 million added cost for type 1 and 2 DM combined.
A review 3 studied advanced technologies for the management of diabetes as measured by %CV and time spent with glucose < 70 mg/dL (a surrogate for hypoglycemic burden).
The GOLD study enrolled 161 participants with T1D on MDI therapy and baseline HbA1c of 8.6% in a randomized cross-over study to either CGM or SMBG first. Treatment periods were 26 weeks each with 17-week washout periods between assignments. CGM use was associated with statistically significant improved HbA1c by − 0.4% (P < 0.001). CGM use was also associated with reductions in time spent with glucose < 70 mg/dL during the day by 40%, evening by 48% (P < 0.001), and %CV 37% vs 40% (difference − 3% (− 5 to − 2%) P < 0.001).
The DIAMOND study enrolled 158 participants with T1D on MDI therapy and baseline HbA1c of 8.6%, and randomized them to either CGM or SMBG. The treatment period was 24 weeks. CGM use was associated with statistically significant improved HbA1c by − 0.6% (P < 0.001). CGM use was also associated with statistically significant reductions in time spent with glucose < 70 mg/dL by 46% (P = 0.002) and %CV 38% vs 42% (difference − 4% (− 6 to − 2%) (P < 0.001)
The IMPACT study enrolled 328 participants with T1D and baseline HbA1c 6.7% in a randomized trial of SMBG vs CGM. 70% of participants utilized MDI and 30% subcutaneous insulin infusion (CSII). The treatment period was 6 months. CGM use was not associated with statistically significant improved HbA1c. CGM use was associated with statistically significant reductions in time spent < 70 mg/dL by 38% (P < 0.0001) and %CV 37.6% vs 41.8% (difference − 4.4% (− 5.0 to − 3.8%) (P < 0.0001).
The IN CONTROL study enrolled 52 participants with T1D and baseline HbA1c of 7.5% in a randomized cross-over trial to either SMBG or CGM first. Insulin delivery methods included 44% CSII and 56% MDI. Inclusion criteria were notable for hypoglycemia unawareness. Treatment periods were 16 weeks with a 12 week of washout period. No significant difference in HbA1c was observed. CGM use was associated with statistically significant reductions in time spent with glucose < 70 mg/dL by 40% (P < 0.001) and %CV 39.5% vs 46.3% (difference − 6.7% (− 8.0 to − 5.5%) P < 0.0001). There was no difference for time spent in a normoglycemic state between MDI and CSII.
The HypoDE study enrolled 149 participants with T1D and baseline HbA1c 7.4% in a randomized trial of SMBG vs CGM. Inclusion criteria were notable for MDI therapy and hypoglycemia unawareness. The treatment period was 22 weeks. CGM was not associated with a statistically significant reduction in HbA1c. CGM was associated with statistically significant reductions in time spent with glucose < 70 mg/dL by 75% (P < 0.0001) and %CV 41.1% vs 34.1% (difference − 6.2% (− 7.5 to − 5.0%) P < 0.0001).
A systematic review and network meta-analysis 4 included 52 studies with a total of 3975 participants. Integrated insulin pump and CGM with low-glucose suspend or hybrid closed-loop algorithms resulted in HbA1c levels 0.96% (95% predictive interval 0.04 to 1.89) and 0.87% (95% predictive interval 0.12 to 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 HbA1c reduction utilizing the surface under the cumulative ranking curve (SUCRA-96.4). Overall risk of bias was moderate-high.
An RCT5 included 72 adults assessing the safety and efficacy of automated insulin delivery (AID) in adults with T1D at high risk for hypoglycemia. Compared with the sensor and pump (S&P), AID resulted in reduction of time below range by -3.7% (95% CI -4.8 to -2.6); an 8.6% increase in time in range (95% CI 5.2 to 12.1); and a -5.3% decrease in time above range (95% CI -87.7 to -1.8). Mean sensor glucose reading remained similar in the AID and S&P groups. Two severe hypoglycemic episodes occurred using AID.
Date of latest search: 2024-01-04
Primary/Secondary Keywords