Hyperglycaemia is common in critically ill patients and is associated with increased morbidity and mortality. Management in intensive care units frequently relies on continuous intravenous insulin infusion, a complex intervention that requires frequent glucose monitoring and substantial nursing involvement. Because implementation demands consistent monitoring and protocolised action, understanding current nursing-reported practice is important for patient safety, protocol development and resource planning.
The survey aimed to describe reported practices for continuous intravenous insulin infusion management across Italian intensive care units and to explore variability in three key domains: glycaemic targets, insulin initiation thresholds, and glucose measurement and monitoring practices. An additional objective was to assess whether reported practices differed according to healthcare setting.
A national cross-sectional survey using a structured questionnaire was distributed to intensive care unit nurses and nurse coordinators in Italy between September and November 2025. The questionnaire covered institutional characteristics, the presence and features of insulin infusion protocols, glycaemic targets, thresholds for initiating insulin infusion, methods used to measure blood glucose, and reported monitoring frequency before and after glycaemic stabilisation. Exploratory comparisons by healthcare setting were performed using cross-tabulations and exact tests as appropriate.
A total of 128 valid responses were included in the analysis. The vast majority of respondents (119; 93.0%) reported that a standardised insulin infusion protocol was available at their institution. Despite this high reported availability of protocols, the responses revealed considerable heterogeneity in how those protocols were applied or interpreted in routine nursing practice.
Reported thresholds for initiating continuous intravenous insulin infusion varied across respondents. The single most commonly reported initiation threshold was 180 mg/dL, cited by 50.0% of participants. The survey identified variability in glycaemic targets as well, though specific target ranges beyond the initiation threshold distribution were not detailed in the abstract. This variability indicates lack of uniformity in the point at which intravenous insulin therapy is commenced and in the explicit glycaemic goals nurses use in practice.
Respondents reported different methods for blood glucose measurement, and these differences varied by healthcare setting in exploratory analyses. Monitoring frequency before glycaemic stabilisation showed notable variability: the median reported interval for glucose checks during the unstable period was 120 minutes, with an interquartile range of 60 to 240 minutes. This median interval exceeds commonly cited recommendations for hourly monitoring during periods of glycaemic instability, suggesting potential gaps between recommended practice and reported local practice.
The authors report exploratory comparisons that suggested differences across healthcare settings in initiation thresholds, glucose measurement methods and monitoring frequency before stabilisation. The abstract notes these comparisons were exploratory; the data support observed differences but do not establish causation. Importantly, potential explanatory factors such as nurse staffing levels, workload, and actual protocol adherence were not directly measured in this study and therefore must be considered possible contributors rather than demonstrated causes.
Although standardised insulin infusion protocols were widely reported as available, substantial variability persisted in reported continuous intravenous insulin infusion management across Italian ICUs. Variability was most pronounced in initiation thresholds, glucose measurement methods and monitoring frequency during periods before glycaemic stabilisation. The authors recommend that local insulin infusion protocols be reviewed, with explicit instructions for monitoring intervals during glycaemic instability, and that systems for staff education, audit and feedback and escalation when glucose checks are delayed be implemented or reinforced. These steps are intended to improve consistency of care and adherence to recommended monitoring during unstable glycaemic periods.
The survey is based on self-reported nursing practice and institutional reporting; direct observation or measurement of actual practice, workload or protocol adherence was not performed. Because staffing, workload and real-world adherence were not measured, they remain plausible but unproven explanations for observed variability. The abstract does not report patient-level outcomes or detailed numeric breakdowns of all target ranges, so conclusions are limited to reported practices rather than measured clinical effects.