In older in-patients with type 2 diabetes and cognitive impairment, CGM showed superior glucose metrics compared to point-of-care tests.

Background: Older in-patients have a higher prevalence of diabetes and cognitive impairment compared to those living without diabetes or cognitive decline. Cognitive impairment can complicate glucose management more challenging, as patients may forget to measure their blood glucose or report symptoms. Investigating the accuracy of continuous glucose monitoring (CGM) compared to usual care will inform clinical decision-making in this vulnerable population. Aim: To compare CGM derived glucose metrics and point-of-care tests (POCT) in older in-patients with type 2 diabetes and cognitive impairment, and to investigate the analytical and clinical accuracy of CGM in the hospital setting. Methods: Thirty-two older people with comorbid type 2 diabetes and cognitive impairment were recruited within a UK tertiary care hospital. All participants were CGM-naive and wore blinded Dexcom G7 sensors for up to 10 days, while receiving. Usual care during their hospital stay including POCT. Key accuracy metrics comprised the mean absolute relative difference (MARD), median absolute relative difference (median ARD), Clarke Error Grid (CEG) and, correlation (R2). The proportion of CGM readings within +/-20% of reference glucose values when the reference was >5.6 mmol/L, or within +/-1.1 mmol/L when >5.6 mmol/L (+/-20%/1.1 mmol/L) was calculated to assess analytical and clinical accuracy. Results: Thirty participants completed the study. CGM-derived mean glucose for time in range (TIR, 4-10 mmol/mol) was 36.2% (range 0-90%), time above range (TAR>=10 mmol/mol) was 62.8% and time below range (TBR<=3.9 mmol/mol) was 1%. Mean POCT-based TIR was 41%, TAR was 57.2% and TBR 2%, showing broadly comparable glucose metrics. CGM accuracy analysis showed a MARD of 17.4%, median ARD of 12.2% and the outcome of +/-20%/1.1 mmol/L was agreement of 72%. CEG analysis revealed that 99.3% of paired readings fell within the clinically acceptable Zones A and B, with a strong correlation between CGM and POCT (R2=0.82). CGM identified more hypoglycaemic readings, particularly at night. Conclusion: CGM and POCT produced largely similar glucose metrics in older in-patients with diabetes and cognitive impairment. CGM demonstrated clinically acceptable accuracy, and captured additional hypoglycaemia, undetected by routine POCT. These findings suggest that CGM may be a viable and potentially advantageous alternative to POCT in this high-risk population.