Abstract
There is a lack of evidence regarding the relationship between capillary blood glucose (CBG) >15 mmol/L and blood ketones in maintenance haemodialysis dialysis patients meaning the utility of ketone measurement in this group is uncertain. Despite this, guidance published by the Joint British Diabetes Society for Inpatient Care (JBDS-IP) in 2022 states that blood ketones should be measured in patients on maintenance haemodialysis with diabetes mellitus (DM) who have pre- or post-dialysis CBG >15mmol/L and symptoms consistent with diabetic ketoacidosis (DKA), acknowledging that these can be non-specific [1]. We sought to identify if there is a relationship between elevated CBG and blood ketone levels in haemodialysis patients, and if haemodialysis affects blood ketone levels.
Maintenance haemodialysis patients with DM admitted to a single ward over 2 separate 2 month periods were included. Nursing staff recorded CBG and blood ketone levels pre- and post-dialysis using a FreeStyle PrecisionPro ketone meter. For each patient, further demographic data were collected from the electronic patient record.
79 measures were collected from 30 patients, with between 1 and 7 measures for each patient. 57% patients (n = 17) were male, 37% (n = 11) had T1DM, and 80% (n = 24) were on insulin therapy. The median pre-HD CBG for all measures was 10.5mmol/L (IQR 8.15-15.05), median post-HD CBG was 9.1mmol/L (IQR 6.8-12.95), median pre-HD ketones were 0.1mmol/L (IQR 0.1-0.2) and median post-HD ketones were 0.1mmol/L (IQR 0.1-0.3). There was no significant difference in blood ketones pre-and post-HD for all measures (p = 0.09) (Figure 1). For measures where pre-HD CBG was >15mmol/L (n = 21) median pre-HD and post-HD ketones were 0.1 (IQR 0.1-0.2) and 0.2mmol/L (IQR 0.1-0.2) respectively (p = 0.45). For measures where pre-HD CBG was >20mmol/L (n = 6) median pre-HD and post-HD ketones were 0.1 mmol/L (IQR 0.1-0.13) and 0.2 mmol/L (IQR 0.1-0.2) respectively (p = 0.13). For measures where pre-HD CBG was <15mmol/L, median pre-HD ketones were 0.1 (IQR 0.1-0.2). There was no significant difference in blood ketones pre-HD between those with a pre-HD CBG >15mmol/L and those with a pre-HD CBG <15mmol/L (p = 0.69).
Graphs showing pre- and post- HD CBG and blood ketone measures for each patient. Where a patient has more than one measure recorded, the median is used.
These results demonstrate no significant relationship between elevated CBG measures (>15mmol/L) and blood ketone levels in haemodialysis patients, nor do they demonstrate a significant change in blood ketone levels post haemodialysis. Unwell haemodialysis patients with elevated CBG clearly warrant medical assessment, but there is a lack of evidence that blood ketones are useful to aid assessment and guide management of these patients.
Comments