The storm that was delayed: the deterioration of an in-hospital diabetes process-of-care metric during the COVID-19 pandemic

Background/Aims Networked glucose blood monitoring has been demonstrated as a useful process of care for improving glycaemia and clinical outcomes in hospital inpatients. However, these benefits are partly reliant on the accurate entry of patients' medical record numbers by healthcare staff. This study assessed the accuracy of such data entry, comparing the periods before and after the onset of the COVID-19 pandemic. Methods This retrospective observational study analysed glucose meter medical record number entries at a large hospital in Victoria, Australia. The study period spanned from September 2019, when the networked blood glucose monitoring system was introduced, to July 2020. The proportion of inaccurate entries were presented as a percentage of the total number of entries and comparisons were made between the pre-COVID-19 and post-COVID-19 onset periods. Data were analysed using an interrupted time series methodology and presented using a Quasipoisson distribution. Results A gradual decrease in the percentage of accurate medical record number entries was observed following the introduction of the networked blood glucose monitoring system. This decline in accuracy decreased further following the onset of COVID-19, despite the hospital serving a relatively low number of patients with the virus. Conclusions The ongoing decrease in accuracy of data entry into the networked blood glucose monitoring system is thought to be a result of insufficient training and time constraints, which were exacerbated by the COVID-19 pandemic because of protocol changes and furloughed staff. It is recommended that accurate use of the networked blood glucose monitoring system is allocated more regular training in hospital wards.