Plasma chloride in the critically ill

According to the authors of a recently published study little is known of the clinical effects of abnormal plasma chloride concentration among the critically ill. Their retrospective study illuminates this apparent grey area of clinical knowledge.  The study involved retrieval and analysis of laboratory data, including plasma chloride measurements, relating to 488 patients admitted to mixed (surgical and medical) intensive care units over a year long period.  

All those included in the study were adults (more than 16 years) who stayed in the unit for more than 24 hours and had at least one full biochemical data set that comprised:   arterial blood gases (pH, pCO₂, bicarb), electrolytes (including chloride), lactate, glucose, ionized calcium, albumin  and phosphate.  With the exception of albumin and phosphate, all analytes were measured together on arterial blood using a blood gas analyzer. 

Derived calculated data included anion gap, standard base excess and three parameters used in Stewarts quantitative physiochemical approach to acid-base disorders:  apparent strong ion difference (SIDa), effective strong ion difference (SIDe)  and strong ion gap (SIG).  

Clinical and demographic data retrieved for each patient included age, sex, admission diagnosis and Acute Physiology and Chronic Heath Evaluation (APACHE) II score on admission.

The prevalence of abnormal plasma chloride was 25.4 %; reduced plasma chloride (hypochloremia) was found in 43 (8.8 %) of the 488 patients and increased plasma chloride (hyperchloraemia) in 81 (16.6 %).  The remaining 364 patients (74.6 %) had normal plasma chloride.  

Those with hypochloraemia had significantly higher pH (7.44 ± 0.065) than those who were normochloremic (pH 7.42 ± 0.035) and hypochloremic (pH 7.41 ± 0.04) in line with the received wisdom that hypochloremia can cause metabolic alkalosis. By analyzing SIDa values the authors were able to demonstrate that alkalemia in those patients with hypochloremia is due mainly to hypochloremia.    

A negative correlation between chloride and APACHE score indicated that those with most severe disease are likely to have lowest plasma chloride, although there was an unexplained difference in this regard between surgical and medical admission; the correlation was not present in medical patients. By several measures hypochloremia was shown to be associated with poorer outcome. 

For example, ICU and hospital mortality rate was higher (14.0 % and 23 % respectively) in those with hypochloremia, compared with hyperchloremic patients  (2.5 % and 3.7 %) and those with normal plasma chloride (1.9 % and 3.8 %).  The authors conclude that their study results indicate that ‘hypochloremia has clinical importance as an indicator of prognosis in critically  ill patients’.