Predictors of mortality during hospitalisation for hyperosmolar hyperglycaemic state
Kew, Tania Manta, Aspasia Sawlani, Jhanvi Pravesh Hallum, Abigail Sharma, Angelica Mann, Amar Rengarajan, Lakshmi Boden, Charlotte Dalzell, Joseph Qamar, Sulmaaz
Kew, Tania
Manta, Aspasia
Sawlani, Jhanvi Pravesh
Hallum, Abigail
Sharma, Angelica
Mann, Amar
Rengarajan, Lakshmi
Boden, Charlotte
Dalzell, Joseph
Qamar, Sulmaaz
Affiliation
Queen Elizabeth Hospital Birmingham; Sandwell and West Birmingham NHS Trust; The Dudley Group NHS Foundation Trust
Other Contributors
Publication date
2025-07-24
Subject
Collections
Research Projects
Organizational Units
Journal Issue
Abstract
Hyperosmolar hyperglycaemic state (HHS) is a life-threatening metabolic emergency of diabetes mellitus (DM). However, there is a paucity of data on predictors of mortality. A national surveillance system was established to monitor trends in HHS admissions and patient outcomes using the Digital Evaluation of Ketosis and Other Diabetic Emergencies (DEKODE) model. Determine key predictors of mortality and risk stratification for HHS; Assess the impact of biochemical markers on HHS outcomes; Compare institutional variability in HHS outcomes. We conducted a secondary analysis of data obtained from the DEKODE-HHS database across 12 hospitals in the UK between January 2021 and November 2024. Univariate and multivariate logistic regression analyses were employed to identify key predictors of mortality in HHS, focusing on demographic and biochemical factors. Biochemical markers, including glucose, pH, bicarbonate, potassium and insulin requirements, were assessed for their impact on patient outcomes. Institutional variability in mortality rates was examined by comparing two hospitals with demographically equivalent populations and evaluating performance standards per international HHS guidelines.3 Findings were used to inform risk stratification models to improve early clinical decision-making for high-risk patients. We included 218 HHS episodes in our analysis. The mortality rate during the index admission was 16.1% (35/218 patients). Patient demographics, such as gender, body mass index (BMI) and Charlson Comorbidity Index (CCI), were not significantly associated with increased mortality. Significant predictors for mortality on univariate logistic regression analysis included age (odds ratio (OR)=1.062, 95% confidence interval (CI) 1.025–1.100, p<0.001), sodium at diagnosis (OR=1.040, 95% CI 1.010–1.071, p=0.008), urea at diagnosis (OR=1.052, 95% CI 1.015–1.090, p=0.006) and serum osmolality at diagnosis (OR=1.016, 95% CI 1.004–1.028, p=0.010). Biochemical markers at admission, including glucose, pH, bicarbonate, potassium and insulin requirements for HHS resolution, did not predict mortality with univariate analysis. Multivariate analysis identified age and sodium at diagnosis as significant predictors of mortality. Each additional year of age increased the odds of mortality by 4.5% (OR=1.045, 95% CI 1.008–1.085, p=0.018), while each unit increase in sodium increased the odds by 6.8% (OR=1.068, 95% CI 1.003–1.136, p=0.038). Serum osmolality and urea on presentation did not demonstrate a significant association with the outcome. In comparing mortality between two similar institutions, while Hospital B demonstrated elevated odds of mortality compared with Hospital A, this difference did not reach statistical significance (OR=5.176, 95% CI 0.580–46.178, p=0.141). Age and serum sodium at diagnosis were the strongest predictors of mortality in HHS. No significant interhospital differences were observed, suggesting institutional factors alone do not influence outcomes. However, further evaluation of management practices could identify areas for improvement. Integrating age and sodium into risk stratification models may improve early identification and targeted interventions, optimising patient care and outcomes.
Citation
Tania Kew, Aspasia Manta, Jhanvi Pravesh Sawlani, Abigail Hallum, Angelica Sharma, Amar Mann, Lakshmi Rengarajan, Charlotte Boden, Joseph Dalzell, Sulmaaz Qamar, Alexandra Lubina Solomon, Elena Armeni, Gerry Rayman, Ketan Dhatariya, Punith Kempegowda, Predictors of mortality during hospitalisation for hyperosmolar hyperglycaemic state, Clinical Medicine, Volume 25, Issue 4, Supplement, 2025, 100446, ISSN 1470-2118, https://doi.org/10.1016/j.clinme.2025.100446
Type
Article