The estimated time fraction within range may mask prognostically distinct glycaemic profiles in critical illness: a three-database landmark study
Authors' conclusion
The estimated time fraction within range alone may not fully describe mortality associations.
Study context
- Population
- Adults alive in the ICU at 48 h with adequate glucose measurement coverage
- Participants
- 10,732 / 17,444 / 4,699
- Exposure
- Compositional contrast: 5 percentage point lower time fraction above 180 mg/dL and higher fraction below 70 mg/dL, at constant fraction within range
- MIMIC-IVn = 10,732
- eICUn = 17,444
- AmsterdamUMCdbn = 4,699
- Outcome
- Post-landmark ICU mortality
Methodological details48-h landmark study, three databases (MIMIC-IV, eICU, AmsterdamUMCdb), Time fractions estimated by linear interpolation of intermittent measurements
- 48-h landmark study
- Three databases (MIMIC-IV, eICU, AmsterdamUMCdb)
- Time fractions estimated by linear interpolation of intermittent measurements
Main finding
As reported in the abstract
Post-landmark ICU mortality for the compositional contrast: 5 percentage point lower time fraction >180 mg/dL and corresponding higher fraction <70 mg/dL, fraction within range held constant · after 48-h landmark
- The composition contrast was associated with higher mortality
- Cohort values given in the order MIMIC-IV, eICU, AmsterdamUMCdb (as listed in Methods)
- Per_group events are deaths in each cohort
Observational data: associations, not effects.
Further findings
Timing of association
- Associations strongest in days 0-2 and stronger for recent than remote exposure
Distributional analyses
- Association localized near and below 54 mg/dL
Parts of this Brief are withheld: the automatic checks could not confirm them against the abstract.
Original publication
The estimated time fraction within range may mask prognostically distinct glycaemic profiles in critical illness: a three-database landmark study. Diabetes Res Clin Pract. Published 23 September 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →