E.YEOH1, A. LIU2, J.L. SU2, L.Y. YEOH3, A. KOH3
Admiralty Medical Centre1, Khoo Teck Puat Hospital2, Sengkang General Hospital3
Assessment of glycemia in end-stage renal disease (ESRD) patients with diabetes mellitus (DM) on peritoneal dialysis (PD) is challenging with limitations of usual modalities like HbA1c and fructosamine. Continuous glucose monitoring (CGM) may be a better indicator of glucose excursions and overall glycemia. We aimed to correlate the estimated average glucose (eAG) derived from HbA1c and fructosamine against average glucose (AG) obtained from CGM and capillary blood glucose (CBG).
DM patients on stable PD regime for ≥ 3 months were subjected to 6-day retrospective CGM twice, 8-12 weeks apart, along with 4 calibration CBG daily.
Of the 15 patients [median(IQR)] age 57(46-64) years, DM duration 18(13-20) years, HbA1c 7.2(6.7-9.0)%; fructosamine 381(319-410) umol/l; Haemoglobin 9.6(8.9-11.2)g/dL recruited, 11 patients were on automated PD and 4 patients on continuous ambulatory PD; 8 patients were on insulin therapy and remainder on oral anti-diabetic agents.
Mean glucose level from CBG and CGM was 10.1 (8.3-14.8) and 10 (8.9-13.8) mmol/l respectively. There was good correlation between CGM AG and CBG (r=0.86) with Bland Altman plot showing bias of 0.286 (-1.341 to 1.912) mmol/l. There was better correlation between eAG derived from HbA1c with CGM AG (r=0.76) with bias of 1.254 (-5.226 to 2.721) mmol/l compared to fructosamine with CGM AG (r=0.45) with bias of -0.405 (-5.716 to 4.91) mmol/l.
Both HbA1c and fructosamine do not reflect AG accurately in ESRD PD patients with DM. Intermittent CGM provides an alternative method of glycaemic assessment which may enable safer titration of medications in these patients.