Acid–base balance, serum electrolytes and need for non-invasive ventilation in patients with hypercapnic acute exacerbation of chronic obstructive pulmonary disease admitted to an internal medicine ward

The hypercapnic AECOPD is an alarming event that requires a careful management. Several factors determine the final outcome in terms of both survival and need for assisted mechanical ventilation in ICU. These factors are patient’s age, general health status, disease severity, hemodynamic stability, concomitant comorbidities, pulmonary function, respiratory acidosis degree, and underlying AB and HE balance disorders [10].

This is a retrospective study, in which patients with hypercapnic AECOPD, managed in an internal medicine ward of a General Hospital have been studied. We have shown that the group of patients with respiratory acidosis?+?metabolic acidosis (group C) had the most serious prognosis, both in terms of non-invasive ventilation, and ICU support. These patients often underestimate their clinical status so that an acute event easily leads to metabolic derangement, most of the times caused by renal failure or worsening of diabetes mellitus. However, in some patients, adequate medical treatment of the metabolic imbalance, and the correction of hypoxemia and hypercapnia by oxygen and NIV could allow a safe management with quick recovery. In patients with hypercapnic AECOPD, a good metabolic compensation and an adequate renal function significantly reduce the mortality [11].

In respiratory acidosis, if an excessive increase of bicarbonate concentration (higher than expected by renal compensation laws) occurs, a mixed disorder with autonomous metabolic alkalosis should be suspected. In patients with multiple comorbidities and multi-drug treatment, several circumstances could lead to metabolic alkalosis. In our patients, the use of diuretics and steroids and a relative reduction in circulating blood volume were the main causes of metabolic alkalosis. We observed that patient belonging to group B (respiratory acidosis?+?metabolic alkalosis) had a better prognosis than patients in group A (compensated respiratory acidosis) with a significantly lower need of NIV.

The admission to ICU was not significantly different in comparison to patients belonging to group A.

These data are conflicting with Terzano et al. who observed that metabolic alkalosis worsened the prognosis of respiratory acidosis in hypercapnic AECOPD patients. In their study, patients with mixed disorder (respiratory acidosis plus metabolic alkalosis) had a longer and higher incidence of NIV support than patients with compensated respiratory acidosis [12].

To explain such different finding, we can speculate that patients with hypercapnic AECOPD managed in an Internal Medicine ward of a General Hospital are quite different from patients managed in a Pulmonary Department of a Teaching Hospital. The latter have scheduled follow up visits and are hospitalized only when the clinical status progressively deteriorates. In contrast, hypercapnic AECOPD patients admitted to an Internal Medicine ward of a General Hospital usually come from the Emergency Department, where they are hospitalized because their clinical status worsened for sudden events (such as diarrhoea, vomiting, fever, sweating, decreased fluid intake) with additional autonomous metabolic alkalosis. Therefore, a diuretic and/or glucocorticoid induced metabolic alkalosis with hypovolemia is often detectable in these patients. These events can be easily managed by blood volume and serum chloride restoring (with saline solution) allowing the clinical improvement and avoiding the NIV support.

In an earlier study, it was shown that in COPD patients with peripheral edema treated with loop diuretics, furosemide withdrawing led to increased ventilation and an improvement in the respiratory acidosis [13].

Lactate concentration was higher in NIV+ group. Our result confirms previous studies in which blood lactate was considered a diagnostic hallmark of tissue hypoxia and respiratory muscle fatigue and, indirectly, of COPD severity in patients not requiring NIV [12]. However, hyperlactataemia (lactate values??2 mmol/L), plays a different role in the three groups of AB disorders. In fact, in group B it was present only in patients not requiring NIV (33 %); in group A was present both in patients requiring (25 %) and not requiring NIV (30 %); and in the group C it was present in 50 % of patients requiring NIV. In group B the lactate overproduction may reflect a positive response to the stress enabling an alternative energetic way through a carbon units transport between cells (“good lactate”); whereas in group C the lactate hyperproduction is probably due to hypoperfusion (“bad lactate”) and should be immediately recognized and treated [14, 15]. In the patients of the NIV-treated respiratory acidosis group we didn’t observe metabolic “post-hypercapnic alcalosis” since the concomitant metabolic component was simultaneously treated and corrected.

We acknowledge that many of the cross-associations between pCO₂ and serum electrolytes found in our study are expected, as the AB balance is closely related to the HE balance. Nevertheless, some remarks seem worthy of focus.

In Brackett’s classic experiment, on healthy volunteers, acute hypercapnia was experimentally induced by having them hyperventilating by breathing CO₂–enriched (at 7 % and at 10 %) air. In these subjects, a variation of both pH and bicarbonates occurred, with a very slight, but significant, increase in Na⁺ and K⁺, and no changes in Cl^?, indeterminate anions (anion gap: AG) and lactates concentrations [16]. In a following study with patients with alveolar hypoventilation and absence of clinical conditions potentially responsible for metabolic alkalosis, the response to the chronic hypercapnia was investigated. In this group a decrease in pH with a progressive increase in pCO₂ was seen; moreover, when pCO₂ values increased, Cl^? decreased whereas no significant association was found with HE balance (Na⁺, K⁺ and AG) [17].

In our study population composed by patients with hypercapnic AECOPD, we found a significant association between increase in pCO₂ and bicarbonate (which reflects the renal compensation) and a statistically significant relationship between the pCO₂ increase and serum chloride decrease. Such associations were present in all the subjects studied and was maintained also when dividing the subjects in the 3 subgroups classified according to the ABG disorder. Serum chloride decrease may be related to bicarbonate increase as an attempt to assure the electro-neutrality. We can interpret the data of serum chloride reduction also in the light of the Strong Ion Difference (SID) of Stewart’s theory (Stewart). In this case, in hypercapnic AECOPD, after the pulmonary response to increasing pCO_2, the renal response starts, by regulating Na⁺ and Cl^?, which are the major SID determinants. A relative (compared to Na⁺ concentration) hypochloremia would lead to a SID increase and would indicate a metabolic alkalosis with a bicarbonate increase. According to this theory, therefore, bicarbonate increase reflects a shift in bicarbonate equilibrium of the ratio H^+/ HCO3^?connected with the SID variation [18]. In fact, in the overall group and in the subgroups with compensated respiratory acidosis and respiratory acidosis?+?metabolic alkalosis, an inverse association between pCO₂ and indeterminate anions was found.