Prognostic Value of Admission Serum Chloride and of the Composition of Metabolic Acidosis in Postoperative Intensive Care: A Single-Centre Retrospective Cohort Study
- 1Department of Internal Medicine, Division of Intensive Care Medicine, Faculty of Medicine, Ondokuz Mayıs University, 55139 Samsun, Türkiye
- 2Department of Anaesthesiology and Reanimation, Division of Intensive Care Medicine, Faculty of Medicine, Ondokuz Mayıs University, 55139 Samsun, Türkiye
Abstract
Background/Objectives: Serum chloride is measured in almost every postoperative intensive care unit (ICU) admission. We examined whether it carries prognostic information once operative case-mix and the composition of the metabolic acidosis are accounted for. Methods: In 1136 adults admitted from the operating theatre to a single 20-bed ICU, admission chloride was modelled with restricted cubic splines in multivariable logistic regression, and three acid–base features (lactate > 2, chloride > 108, anion gap > 16 mmol/L) were counted in patients with base excess <−3 mmol/L. The primary outcome was ICU mortality. Results: A total of 95 patients (8.4%) died in the ICU. Chloride showed no association with ICU mortality on the overall test (χ2 = 4.98, 2 df; p = 0.083), and its removal left discrimination unchanged (ΔC +0.003, 95% CI −0.004 to +0.021), although the odds ratio for 115 versus 105 mmol/L was 1.91 (1.08–3.38). ICU mortality rose from 6.6% with one acid–base feature to 58.6% with three (Firth’s penalised odds ratio 10.50 (4.65–23.70) versus no metabolic acidosis), a contrast dominated by the depth of the acidosis; adjusted for that depth, each feature carried an odds ratio near 2.0. The anion gap, uncorrected for albumin, and bicarbonate, but not the strong ion difference, were associated with death. Conclusions: Admission chloride added no detectable prognostic information, though a modest association is not excluded. Coexisting acid–base abnormalities identified high-risk patients and place the signal in the bicarbonate deficit and unmeasured anions, but this is a descriptive marker needing external validation, not an established risk-stratification tool.
Related
- Prognostic performance of the 3-h COBRA score for ICU admission in comatose poisoned patients: a retrospective study.
- Preoperative metabolic–inflammatory profiles and early postoperative recovery after immediate breast reconstruction in patients with and without neoadjuvant chemotherapy: a retrospective cohort study
- Association of weekend admission on immediate postoperative outcomes following non-elective cardiac operations
- Early hematoma expansion provides incremental prognostic information beyond admission characteristics for 6-month functional outcome after adult traumatic brain injury
- Admission blood glucose and neutrophil-to-lymphocyte ratio provide independent and complementary prognostic value for in-hospital mortality in acute type A aortic dissection
BibTeX
@article{colak2026prognostic,
title = {{Prognostic Value of Admission Serum Chloride and of the Composition of Metabolic Acidosis in Postoperative Intensive Care: A Single-Centre Retrospective Cohort Study}},
author = {Özkul Yılmaz Çolak and Melda İşevi and Tuğçehan Sezer Akman and Neslihan Ünal Akdemir and Özgür Kılıç},
journal = {Journal of Clinical Medicine},
year = {2026},
month = oct,
doi = {10.3390/jcm15218235},
url = {https://www.mdpi.com/2077-0383/15/21/8235},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Prognostic Value of Admission Serum Chloride and of the Composition of Metabolic Acidosis in Postoperative Intensive Care: A Single-Centre Retrospective Cohort Study
AU - Özkul Yılmaz Çolak
AU - Melda İşevi
AU - Tuğçehan Sezer Akman
AU - Neslihan Ünal Akdemir
AU - Özgür Kılıç
PY - 2026
DA - 2026/10/22
JO - Journal of Clinical Medicine
DO - 10.3390/jcm15218235
UR - https://www.mdpi.com/2077-0383/15/21/8235
AB - Background/Objectives: Serum chloride is measured in almost every postoperative intensive care unit (ICU) admission. We examined whether it carries prognostic information once operative case-mix and the composition of the metabolic acidosis are accounted for. Methods: In 1136 adults admitted from the operating theatre to a single 20-bed ICU, admission chloride was modelled with restricted cubic splines in multivariable logistic regression, and three acid–base features (lactate > 2, chloride > 108, anion gap > 16 mmol/L) were counted in patients with base excess <−3 mmol/L. The primary outcome was ICU mortality. Results: A total of 95 patients (8.4%) died in the ICU. Chloride showed no association with ICU mortality on the overall test (χ2 = 4.98, 2 df; p = 0.083), and its removal left discrimination unchanged (ΔC +0.003, 95% CI −0.004 to +0.021), although the odds ratio for 115 versus 105 mmol/L was 1.91 (1.08–3.38). ICU mortality rose from 6.6% with one acid–base feature to 58.6% with three (Firth’s penalised odds ratio 10.50 (4.65–23.70) versus no metabolic acidosis), a contrast dominated by the depth of the acidosis; adjusted for that depth, each feature carried an odds ratio near 2.0. The anion gap, uncorrected for albumin, and bicarbonate, but not the strong ion difference, were associated with death. Conclusions: Admission chloride added no detectable prognostic information, though a modest association is not excluded. Coexisting acid–base abnormalities identified high-risk patients and place the signal in the bicarbonate deficit and unmeasured anions, but this is a descriptive marker needing external validation, not an established risk-stratification tool.
SN - 2077-0383
ER -