|Year : 2019 | Volume
| Issue : 3 | Page : 282-286
Effect of chloride level on outcome in critically ill patients
Mohamed M Megahed, Dina H Zidan, Ziad M Abdelhalim
Department of Critical Care Medicine, Faculty of Medicine, Alexandria University, Alexandria, Egypt
|Date of Submission||16-Apr-2018|
|Date of Acceptance||08-Feb-2019|
|Date of Web Publication||29-Aug-2019|
Ziad M Abdelhalim
MS critical Care
Source of Support: None, Conflict of Interest: None
Background Chloride (Cl) abnormalities in the critical care units have received great attention, especially hyperchloremia as a cause of metabolic acidosis and hypochloremia as a cause of metabolic alkalosis. However, Cl abnormalities themselves have not been studied sufficiently.
Aim The aim of this study was to show the effect of Cl abnormalities in critically ill patients.
Settings The study was conducted in Critical Care Medicine Department in Alexandria University Hospital.
Patients and methods A Prospective clinical study was conducted on 375 critically ill adult patients admitted to Alexandria University hospitals. All patients included were subjected to demographics, diagnosis and cause of admission, relevant medical history, calculated APACHE-II score, and laboratory data (days 1 and 3 after admission), including serum electrolyte level, ABG analysis (anion gap–gap/gap ratio–corrected anion gap), and serum albumin level. According to the recorded Cl levels at days 1 and 3 from ICU admission, patients were defined as normochloremic (99–110 mEq/l), hypochloremic (<99), or hyperchloremic (>110). Outcome measures were all-cause mortality and length of stay (LOS) in ICU.
Results Cl abnormalities at day 1 were associated with increased mortality. The mortality rate was significantly higher in hyperchloremic (67.3%) compared with hypochloremic (72.1%) and normochloremic patients (36.3%) (P<0.001). Cl abnormalities (at days 1 and 3) were associated with increased ICU LOS. The mean of ICU LOS in days was significantly higher in hyperchloremic (21.29) and hypochloremic (16.38) than normochloremic (13.81) patients (P<0.001).
Conclusion In a mixed general intensive care population, after careful controlling for confounders, Cl abnormalities (outside 99–110 mEq/l) were independent predictors for poorer outcomes such as ICU mortality and LOS.
Keywords: abnormalities, chloride, critical, electrolytes, mortality
|How to cite this article:|
Megahed MM, Zidan DH, Abdelhalim ZM. Effect of chloride level on outcome in critically ill patients. Res Opin Anesth Intensive Care 2019;6:282-6
|How to cite this URL:|
Megahed MM, Zidan DH, Abdelhalim ZM. Effect of chloride level on outcome in critically ill patients. Res Opin Anesth Intensive Care [serial online] 2019 [cited 2020 Jan 18];6:282-6. Available from: http://www.roaic.eg.net/text.asp?2019/6/3/282/265716
| Introduction|| |
Chloride (Cl) is the electrolyte with the highest concentration in human plasma after serum sodium, and it is the most abundant anion in the extracellular fluid and constitutes approximately one-third of the extracellular fluid tonicity. It plays an important role in many body functions including acid–base balance, muscular activity, and osmolarity ,.
Cl also plays a vital role in neutrophil function. Neutrophil phagosomes need a continuous influx of Cl through different Cl channels and cotransporters to provide substrate for hypochloric acid generation by myeloperoxidase. Minimal or absent extracellular Cl concentration has been associated with diminished neutrophil function. It is not known whether high extracellular Cl generates an increase in neutrophil response, which could further contribute to the proinflammatory imbalance present in sepsis and hyperchloremic acidosis ,,.
Cl abnormalities in the critical care units have received considerable attention, especially hyperchloremia as a cause of metabolic acidosis and hypochloremia as a cause of metabolic alkalosis. However, Cl aberrancies by themselves have not been studied sufficiently ,. Despite the high prevalence of abnormal Cl levels in ICU, research on the possible detrimental effects is limited .
The aim of this study was to understand the effect of Cl abnormalities on critically ill patients. We analyzed the effect of abnormal serum Cl level on the outcome of critically ill patients.
| Patients and methods|| |
This prospective clinical study was conducted on 375 critically ill adult (age >18 years) patients admitted to Critical Care Medicine Department in Alexandria University hospitals from May to December, 2017, with complete records for clinically relevant data, and no exclusion criteria were defined.
The following data was collected. Demographics, diagnosis and cause of admission, relevant medical history, calculated APACHE-II score, and laboratory data (days 1 and 3 after admission), including serum electrolyte level, ABG analysis (anion gap–gap/gap ratio–corrected anion gap), and serum albumin level.
According to the recorded Cl levels at days 1 and 3 from ICU admission, patients were divided into three groups as normochloremic (99–110 mEq/l), hypochloremic (<99 mEq/l), or hyperchloremic (>110 mEq/l). Outcome measures were all-cause mortality and length of stay (LOS) in ICU.
After ethical approval for this clinical trial from the local committee of ethics in the Faculty of Medicine of Alexandria University and the Department of Critical Care, informed consents for participating and publishing were taken from the next of kin of patients after approval by Critical Care Department Committee.
Data were collected onto an electronic spreadsheet, and statistical package (version 24) was used for statistical analyses. Descriptive statistics were reported as raw percentages or means and SDs. A Student’s t test or Mann–Whitney test was used when appropriate to compare means for parametric or nonparametric data, respectively. A χ2 test or Fisher’s exact test was performed for comparison of categorical variables. Receiver operating characteristic (ROC) curve is generated by plotting sensitivity (TP) on y axis versus 1-specificity (FP) on x axis at different cutoff values. The area under the ROC curve denotes the diagnostic/prognostic performance of the test. Area more than 50% gives acceptable performance and area about 100% is the best performance for the test. The ROC curve allows also a comparison of performance between two tests. P value less than 0.05 was considered statistically significant.
| Results|| |
Regarding demographics, there were no any statistically significant differences between the three groups in their age (P=0.553) or sex (P=0.898) ([Table 1]).
|Table 1 Comparison between the different studied groups according to demographic data|
Click here to view
Among 375 patients, 143 (38.1%) patients were diagnosed with sepsis syndromes, 71 (18.9%) patients were diagnosed with respiratory failure, 27 (7.2%) patients were diagnosed with hypovolemic shock, 20 (5.3%) patients were diagnosed with cardiogenic shock, 35 (9.3%) trauma patients, and 79 (21.1%) patients had other diagnosis. There were no statistically significant differences between the three groups regarding diagnosis.
Regarding laboratory data at day 1, there were no statistically significant differences between the three studied groups in their sodium (P=0.400), potassium (P=0.121), albumin (P=0.786), corrected anion gap (P=0.260), gap ratio (P=0.985), or PCO2 (P=0.075). The mean of the measured pH values was significantly higher in hypochloremic (7.46) than in normochloremic (7.41) and hyperchloremic group (7.36) (P<0.001).
Regarding the measured primary outcomes, there were statistically significant differences between the three groups in their all-cause mortality (P<0.001) and LOS in ICU (P<0.001) ([Table 2]). Cl abnormalities at day 1 were associated with increased mortality. The mortality rate was significantly higher in hyperchloremic (67.3%) and hypochloremic (72.1%) than normochloremic patients (36.3%) (P<0.001). Cl abnormalities (at days 1 and 3) were associated with increased ICU LOS. At day 1, the mean of ICU LOS in days was significantly higher in hyperchloremic (21.29) and hypochloremic (16.38) than normochloremic patients (13.81) (P<0.001). Similar results at day 3, the mean of ICU LOS was significantly higher in hyperchloremic (21.33) and hypochloremic (15.55) than normochloremic patients (14.12) (P<0.001).
|Table 2 Comparison between the different groups according to the primary outcomes|
Click here to view
After plotting ROC curve to detect the ability of Cl level at different times (first and third day) in predicting mortality, the following data were obtained: it was found that area under the curve (AUC) = 0.595 and 0.556, respectively, and it was significant in the first day (P=0.001) at cut off less than or equal to 100.
For day 1, sensitivity was 60.57% and specificity was 51.50% whereas for day 3, sensitivity was 55.43 and specificity was 53.50 ([Table 3]).
|Table 3 Agreement (sensitivity and specificity) for chloride level to predict mortality|
Click here to view
| Discussion|| |
In this study, Cl abnormalities at day 1 were associated with increased mortality. The mortality rate was significantly higher in hyperchloremic group (67.3%) and hypochloremic group (72.1%) than normochloremic patients (36.3%) (P<0.001). However, in the study conducted by Tani and colleagues in 2012, the distribution of Cl levels was hyperchloremia 16.6%, normochloremia 74.6%, and hypochloremia 8.8%. Their study was conducted as a retrospective observational study in 488 ICU patients to investigate the incidence of Cl abnormalities.
After plotting ROC curve analysis, Cl level at day 1 at a cutoff value of less than 100 mEq/l was found to be a fair tool (AUC=0.595) to predict the mortality in critically ill patients [95% confidence interval (CI), 0.536–0.654; P=0.001]. These results were in agreement with the findings of Thongprayoon and colleagues who investigated the prevalence and outcome significance of Cl abnormalities on hospital admission and evolution within the first 2 days of hospital admission. It was a retrospective study of 76 719 admissions. Logistic regression models were built to determine the contribution of Cl abnormalities to the risk of mortality . They found that the odds ratio for hospital mortality were significantly elevated in hypochloremia (Cl <100 mEq/l) and hyperchloremia (>108 mEq/l). Cl abnormalities (outside 100–108 mEq/l) were independent predictor for hospital mortality .
Our findings were in concordance with Lee and colleagues who enrolled 266 adult patients with major trauma by retrospective chart review, who were admitted to ICU from January 2011 to December 2015. The overall 30-day mortality was 11.7% (n=31) . In the study by Lee and colleagues, hyperchloremia (Cl ≥110) after 2 days from admission occurred in 65 (24.4%) patients and the incidence was significantly different between survivors and nonsurvivors (19.6 vs. 61.3%, respectively, P<0.001). Multivariate logistic analysis identified hyperchloremia as independent predictive factor for 30-day mortality in patients with major trauma . Moreover, Boniatti and colleagues conducted a prospective cohort study on 175 ICU patients. They conclude that Cl at admission was a good tool to predict hospital mortality (AUC=0.67) (95% CI, 0.58–0.75; P˂0.05). They concluded that hyperchloremia was independently associated with increased mortality, although they cannot determine whether this association was clinically significant .
In contrast to these findings, Neyra and colleagues conducted a retrospective study on 1940 septic patients, where 615 (31.7%) had hyperchloremia (Cl ≥110 mEq/l) on ICU admission. Multivariate analysis showed that hyperchloremia at day 3 but not day 1 was independently associated with hospital mortality in the subgroup of patients with hyperchloremia on ICU admission (adjusted odds ratio for Cl at day 3 per 5 mEq/l increase=1.27; 95% CI, 1.02–1.59; P=0.03). For those who were hyperchloremic at day 1, every within-patient 5 mEq/l increment in Cl at day 3 was independently associated with hospital mortality (odds ratio for ΔCl 5 mEq/l=1.37; 95% CI, 1.11–1.69; P=0.003). These associations were independent of base deficit, cumulative fluid balance, acute kidney injury, and other critical illness parameters . Again the study by Tani and colleagues found opposing results as only hypochloremic group had significantly ICU and hospital mortality (P<0.0001). ICU mortality was 2.5% in hyperchloremic, 14.0% in hypochloremic, and 1.9% in normochloremic patients (P<0.0001) .
In this study, Cl abnormalities (at days 1 and 3) were associated with increased ICU LOS. At day 1, the mean of ICU LOS in days was significantly higher in hyperchloremic (21.29±4.42) and hypochloremic (16.38±4.93) than normochloremic patients (13.81±6.34) (P<0.001). Similar results were found for day 3, as the mean of ICU LOS was significantly higher in hyperchloremic (21.33) and hypochloremic (15.55) than normochloremic patients (14.12) (P<0.001). In the study by Thongprayoon et al. , Cl less than 100 and more than 108 mEq/l were independently associated with elevated LOS.
In Tani and colleagues study, the hypochloremic group had significantly longer stays in the ICU and hospital (P<0.0001). The mean of ICU stay in days was 4.4±2.5 in hyperchloremic, 14.3±13.3 in hypochloremic, and 7.3±9.6 in normochloremic patients (P<0.0001) .
There are limitations in this study including the small sample size. The study lacks granularity to examine clinical manifestations of Cl abnormalities; however, they were not a study objective. This study was held on heterogenous group of patients. The fluid resuscitation strategy before admission was not controlled. This study was liable to many confounders. Lactate levels were not reported. Renal function tests and the incidence of acute kidney injury at day 1 were not reported.
| Conclusion|| |
From the results of this study, in a mixed general intensive care population, after careful controlling for confounders, Cl abnormalities (outside 99–110 mEq/l) were found to be independent predictors for poorer outcomes such as ICU mortality and LOS. To a certain extent, Cl level at day 1 was a better tool more than at day 3 to predict mortality. Further larger studies should be conducted to minimize the effect of confounders. Fluid resuscitation should also be controlled.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
McCluskey SA, Karkouti K, Wijeysundera D, Minkovich L, Tait G, Beattie WS. Hyperchloremia after noncardiac surgery is independently associated with increased morbidity and mortality: a propensity-matched cohort study. Anesth Analg 2013; 117:412–421.
Neyra JA, Canepa-Escaro F, Li X, Manllo J, Adams-Huet B, Yee J et al.
Association of hyperchloremia with hospital mortality in critically ill septic patients. Criti Care Med 2015; 43:1938.
Aiken ML, Painter RG, Zhou Y, Wang G. Chloride transport in functionally active phagosomes isolated from human neutrophils. Free Radical Biol Med 2012; 53:2308–2317.
Sun YT, Shieh CC, Delpire E, Shen MR. K+-Cl− cotransport mediates the bactericidal activity of neutrophils by regulating NADPH oxidase activation. J Physiol 2012; 590:3231–3243.
Nauseef WM. Myeloperoxidase in human neutrophil host defence. Cell Microbiol 2014; 16:1146–1155.
Constable PD. Hyperchloremic acidosis: the classic example of strong ion acidosis. Anesth Analg 2003; 96:919–922.
Fencl V, Rossing T. Acid-base disorders in critical care medicine. Annu Rev Med 1989; 40:17–27.
Thongprayoon C, Cheungpasitporn W, Cheng Z, Qian Q. Chloride alterations in hospitalized patients: prevalence and outcome significance. PloS One 2017; 12:e0174430.
Lee JY, Hong TH, Lee KW, Jung MJ, Lee JG, Lee SH. Hyperchloremia is associated with 30-day mortality in major trauma patients: a retrospective observational study. Scand J Trauma Res Emerg Med 2016; 24:117.
Boniatti MM, Cardoso PR, Castilho RK, Vieira SR. Is hyperchloremia associated with mortality in critically ill patients?A prospective cohort study. J Crit Care 2011; 26: 175–179.
Tani M, Morimatsu H, Takatsu F, Morita K. The incidence and prognostic value of hypochloremia in critically ill patients. Sci World J 2012; 2012:474185.
[Table 1], [Table 2], [Table 3]