Accuracy of verbal self-reported blood glucose in teenagers with type I diabetes at diabetes ski camp

Background While there have been considerable advances in diabetes management, self-monitoring of blood glucose remains vital. A number of studies, predominantly in adults, have confirmed that logbook entries are prone to a number of common errors. To date, no studies in either adults or children have looked at the accuracy of verbally reported self-monitored blood glucose levels (SMBG). Our aim was to determine the accuracy of verbally reported SMBG levels in adolescents at a diabetes camp. Methods Dual Data (verbally reported and meter-downloaded values) were obtained as part of camp safety monitoring from 20 adolescents (aged 13–18 years) attending a 3 day diabetes winter camp. Blood glucose values were classified as: accurate, absent/phantom, or modified - verbally reported value > / < meter downloaded value. No participant had prior awareness of the planned meter data download at camp conclusion. Results Discrepancies between verbally reported and meter downloaded values were observed in 14/20 (70%) participants and in 53/394 (13.5%) instances of testing. Absent/Phantom readings were the most common error at 30/394 (7.6%). Errors relating to hypoglycaemia were seen in 8/47 (17%) hypoglycaemia-related incidents of testing. No relationship with HbA1c was found between those with reporting errors and those without (p > 0.05). Conclusion While 70% of adolescents had errors, the overall error rate at 13.5% is lower than that previously reported for logbook studies. While this rate is lower than expected, misreporting remains a concern, particularly in the context of diabetes camp and exercise induced hypoglycaemia.