BSPED2021 Poster Presentations Miscellaneous (6 abstracts)
1Faculty of Biology, Medicine and Health, University of Manchester, Manchester, United Kingdom; 2Department of Paediatric Endocrinology, Royal Manchester Childrens Hospital, Manchester, United Kingdom; 3Department of Computer Science, University of Manchester, Manchester, United Kingdom
Background: Congenital Hyperinsulinism (CHI) is characterised by dysregulated and excess secretion of insulin leading to severe hypoglycaemia. Monitoring of glucose levels is essential in this condition as prolonged hypoglycaemia can cause life-threatening complications such as permanent neurological impairment. Interstitial glucose monitoring by continuous glucose monitoring (CGM) devices can identify nocturnal hypoglycaemia retrospectively through data analysis. Analysis can be conducted quickly using bespoke computer code, however this technique requires significant work initially to develop the code. We have aimed to investigate if manual methods are equivalent in the interpretation of nocturnal hypoglycaemia by CGM.
Methodology: Nocturnal CGM (00-0700 h) was investigated by two different glucose thresholds (<3.0 and <3.5 mmol/l) and two different criteria for an episode of hypoglycaemia using both coding (Python) and manual (visualisation of spreadsheet) methods in 23 patients (9395 data points) with CHI. Hypoglycaemia interpretations were compared to markers of disease severity.
Results: Hypoglycaemia interpretations were consistent between coding and manual methods. Mean percentage time in hypoglycaemia was similar by coding and manual methods for a threshold of 3.5 mmol/l [Criteria One (5% vs 4.9%, chi-square=0.707, p=0.4), Criteria Two (5.5% vs 4.7%, chi-square=0.759, p=0.3)] and a threshold of 3.0 mmol/l [Criteria One (1.05% vs 0.9%, chi-square=0.817, p=0.4), Criteria Two (0.95% vs 0.8%, chi-square=0.721, p=0.4)]. In univariate analysis of variance age or severity markers (pancreatic surgery/genetic mutations) did not predict percentage time spent hypoglycaemic (P all > 0.2).
Conclusion: Manual methods are equivalent to coding methods in the interpretation of nocturnal hypoglycaemia demonstrating this methods utility in routine clinical practice. Severity markers for CHI did not predict individual time spent hypoglycaemic, highlighting the importance of personalised CGM analysis in the management of hypoglycaemia due to CHI.