Kidney Week

Abstract: FR-PO628

Pseudohypobicarbonatemia Induced by Severe Hypertriglyceridemia

Session Information

• Fluid and Electrolytes: Clinical - Acid-Base, Magnesium, Calcium, Phosphorus
  November 08, 2019 | Location: Exhibit Hall, Walter E. Washington Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: Fluid and Electrolytes

• 902 Fluid and Electrolytes: Clinical

Authors

• Varghese, Vipin, University of Queensland - Ochsner Clinical School, New Orleans, Louisiana, United States

Vipin Varghese,

• Velez, Juan Carlos Q., University of Queensland - Ochsner Clinical School, New Orleans, Louisiana, United States

Juan Carlos Q. Velez,

Background

Reports of falsely low serum carbon dioxide (sCO₂) concentration, i.e., pseudohypobicarbonatemia (psHypoHCO₃^-) in patients with severe hypertriglyceridemia (hyperTG) have emerged. This phenomenon results from lipid interference in some spectrophotometric analyzers. Our aim was to assess the magnitude and implications of psHypoHCO₃^- in a tertiary care hospital.

Methods

We searched for cases of serum triglycerides (TG) > 1000 mg/dL with a concomitant (measured <24 hrs apart) sCO₂, between 2017 and 2018. We extracted those with sCO₂ ≤ 12 mEq/L to focus on the more clinically relevant cases. Each measured sCO₂ was compared with the calculated bicarbonate (HCO₃^-) from an arterial blood gas (ABG) obtained within 6 hrs of the venous blood draw. PsHypoHCO₃^- was defined as: erroneous HCO₃^- (eHCO₃^-) gap = (calculated HCO₃^- - measured sCO₂) > 5 mEq/L.

Results

We identified 1698 events (652 patients) of TG > 1000 mg/dL and a sCO₂ measured on the same day. TG inversely correlated with sCO₂ (R=-0.38, p=0.00001). We found 179 events (59 patients) with sCO₂ < 12 mEq/L. In 104 of those, an ABG was either not available or performed > 6 hrs apart from the venous blood draw. The remaining 75 events included 30 instances (11 patients) of true hypobicarbonatemia and 45 instances (24 patients) of psHypoHCO₃^-. Among those with psHypoHCO₃^-, the median values of sCO2, calculated HCO₃^-, anion gap and eHCO₃^- gap were 8 (<5 – 12), 20 (10 - 28), 21 (15 – 29) and 13 (5 – 19) mEq/L, respectively, whereas the median pH was 7.37 (7.14 - 7.56). True metabolic acidosis was either absent (42%) or spuriously magnified (58%). TG directly correlated with the eHCO₃^- gap (R=0.59, p=0.00004). Acute pancreatitis (56%) and diabetic ketoacidosis (38%) were the most common concomitant disorders but they did not fully account for the eHCO₃^- gap in the psHypoHCO₃^- cases. Additionally, unnecessary HCO₃^- therapy was initiated in 16% and serum lactate was measured in 80% of the psHypoHCO₃^- events (lactate was normal in 72%).

Conclusion

Severe hyperTG can lead to spuriously low sCO₂. The degree of hyperTG correlates with the magnitude of PsHypoHCO₃^-. Clinicians and laboratory personnel should be aware of this phenomenon to prevent incorrect interpretation of acid-base status and medical mismanagement.