Effectiveness of remote continuous glucose monitoring on adverse outcomes among patients with diabetes complicated with COVID-19.

This prospective study provided an effective way of glucose monitoring for patients with diabetes complicated with coronavirus disease 2019. The use of an intermittently scanned continuous glucose monitoring system was significantly associated with better outcomes of coronavirus disease 2019 in patients with pre-existing diabetes.

Glucose monitoring plays a crucial role in glucose control, which has been shown to be significantly related to the mortality risk of the new coronavirus disease 2019 (COVID‐19) . It is worth mentioning that point‐of‐care tests (POCTs) of blood glucose were used for glucose monitoring in most of the studies investigating the relationship between diabetes and COVID‐19 outcomes . However, during hospitalization under the COVID‐19 pandemic, POCTs might inevitably increase the risk of nosocomial infections and staff exposure, which could presumably lead to inadequate quality of care (e.g., lower frequencies of POCT) and suboptimal management of diabetes. In contrast, in a study comprising four patients with diabetes and COVID‐19, glucose control was significantly improved by using a remote glucose monitoring system . Nevertheless, whether the use of such systems contributes to better outcomes of COVID‐19 remains unknown.

We prospectively recruited 35 patients with diabetes complicated with COVID‐19, who were equipped with intermittently scanned continuous glucose monitoring (isCGM; Figure 1a) during their whole hospitalization. Another 35 patients with diabetes complicated with COVID‐19 who were undergoing regular POCTs of blood glucose were matched by propensity score matching. The caliper was set as 0.001. All the patients were confirmed without use of glucocorticoid and followed up for adverse outcomes of COVID‐19, which were defined as either the need for admission to an intensive care unit, need for mechanical ventilation or death. The index date was defined as the date of admission, as well as the date of starting isCGM. This study was approved by both Shanghai Jiaotong University Affiliated Sixth People’s Hospital and Zhongnan Hospital of Wuhan University. This study was also registered with the Chinese Clinical Trial Registry (ChiCTR2000030436). We obtained informed consent from all patients involved in this study.

Figure 1

(a) Study design and flow chart. (b) Cumulative survival rates of adverse outcomes of coronavirus disease 2019 (COVID‐19) by multivariable adjusted Cox proportional hazards regression analysis. CRP, C‐reactive protein. ESR, erythrocyte sedimentation rate; isCGM, intermittently scanned continuous glucose monitoring; POCT, point‐of‐care test; WBC, white blood cell count.

The mean age of both groups was 63.8 ± 10.5 and 63.7 ± 10.1 years, respectively. The baseline characteristics, including age, sex, fasting glucose levels, the overall severity of COVID‐19, white blood cell count, erythrocyte sedimentation rate, C‐reactive protein, body temperature, oxygen saturation and interleukin‐6 of patients with and without the use of isCGM, showed no differences (all P > 0.5). The baseline characteristics were well balanced and appropriate. The mean duration of diabetes was 10.1 years for all patients. A total of 23.2% of the patients received insulin therapy, whereas others received oral glucose‐lowering drugs. During a mean follow‐up period of 15.0 ± 7.24 days, 15 outcome events occurred in patients with the use of isCGM, and 21 outcome events occurred in patients without the use of isCGM. Zero death events occurred in the isCGM group, whereas three death events occurred in the POCT group. The total incidence rate of the adverse outcomes of COVID‐19 was 0.3 events per 10 person‐days. The multivariable adjusted hazard ratio of the adverse outcomes of COVID‐19 for patients without the use of isCGM was 2.76 (1.40–5.43) when compared with those using isCGM. The survival curve by multivariable adjusted cox regression analysis was illustrated as Figure 1b.

This prospective study provided an effective method of glucose monitoring for patients with diabetes complicated with COVID‐19. By using isCGM for glucose monitoring during hospitalization, the outcomes of patients with diabetes complicated with COVID‐19 eventually improved. One of the major characteristics of isCGM is that calibrations are unnecessary. Compared with traditional POCTs of blood glucose, isCGM can provide comprehensive information on glycemic fluctuations, and the exposures to both hyperglycemia and hypoglycemia . Additionally, isCGM can reduce the risks of biological exposures during hospitalization. Limitations included the relatively small sample size with older age and multiple comorbidities. The patients who were enrolled into both groups were matched using propensity score matching rather than randomization, which could inevitably cause selection bias. Only adverse outcomes were assessed, whereas changes of glucose levels before and after glucose monitoring were unavailable. In conclusion, the use of an isCGM system was significantly associated with better outcomes of COVID‐19 with pre‐existing diabetes. Further experience and evidence with a large sample size are necessary to validate the present findings.

Disclosure

The isCGM and cloud platform‐based system is partly supported by Abbott Diabetes Care Inc. Dr Jian Zhou receives funding from the Shanghai Municipal Education Commission‐Gaofeng Clinical Medicine Grant (20161430). The other authors declare no conflict of interest.