Photo Credit: Brastock Images
Researchers have identified new lung function and blood test cutoffs to improve timely chronic hypercapnia management in patients with COPD.
Researchers have identified new lung function and blood test cutoffs for identifying patients with chronic obstructive pulmonary disease (COPD) who have hypercapnia.
“Hypercapnia is a poor prognostic factor for survival in COPD,” explained Patricia J. Checinski, MD, and colleagues in their article in Respiratory Medicine. “Using non-invasive ventilation (NIV) in this patient population improves survival and decreases re-admission risk when PaCO2 level reduction is targeted. However, outpatient prescription of NIV for patients with chronic hypercapnic respiratory failure remains low.”
The researchers sought to address these gaps by creating a pragmatic hypercapnia screening tool based on serum bicarbonate and pulmonary function parameters.
Parameters Identified
The retrospective study relied on EHR data from 276 patients hospitalized for COPD exacerbations from February 2016 to May 2022. On average, patients were 68.1 years of age, and approximately half (50.4%) were male.
“We defined hypercapnia as a PaCO2 greater than or equal to 52 mmHg, the cutoff used by the Centers for Medicare & Medicaid Services to qualify for NIV,” the researchers said.
The participants had mean FEV1 of 44.1% predicted, HCO3− of 30.9 ±6.0 mmol/L, and PaCO2 of 51.0 ±13.5 mmHg. Overall, 39.9% of patients met the criteria for hypercapnia.
The study authors found an independent association between lower FEV1 % predicted and hypercapnia (OR, 0.93; 95% CI, 0.91–0.95; P<0.001 per 1-unit increase in FEV1 % predicted). There was also an independent association between higher HCO3− and hypercapnia (OR, 1.87; 95% CI, 1.60–2.17; P<0.001 per 1-unit increase in HCO3−).
“The best discriminative cutoff of FEV1 % predicted for hypercapnia was 40% with an AUC of 0.783, a sensitivity of 75.5%, and a specificity of 70.5%,” the researchers reported. “The best discriminative cutoff of HCO3− for hypercapnia was 33 mmol/L with an AUC of 0.951, a sensitivity of 87.3%, and a specificity of 93.4%.”
In addition, the AUC was 0.954 in a model incorporating both FEV1 % predicted of 40% or less and HCO3− of at least 33 mmol/L.
Implications for Clinical Practice
Given their findings, the researchers recommended screening patients with COPD for FEV1 less than or equal to 40% predicted and/or HCO3− greater than or equal to 33 mmol/L. The researchers advised clinicians to evaluate chronic hypercapnic respiratory failure by conducting an arterial blood gas test (ABG) in patients who meet these criteria.
“Pulmonary function tests and basic metabolic panels (BMP) are commonly obtained in routine clinical practice and are often readily available in the [EHR], making FEV1 and HCO3− both accessible tools to screen those at risk,” the researchers said.
The authors added that BMP results “are generally more available than ABG results [and] should be leveraged to identify [patients with COPD] likely to have chronic hypercapnia.”
The study was strengthened by including patients across all GOLD spirometric stages without exclusions for BMI or comorbidities, ensuring a representative sample reflective of real-world patients with COPD.
However, the researchers also reported several limitations. Renal dysfunction, a potential influence on HCO3− levels, was not assessed, and the focus on patients hospitalized for COPD exacerbations may limit applicability to stable cases. Additionally, the authors did not evaluate comorbid conditions and medication use, which could affect blood gas results. Additionally, the use of pre-bronchodilator FEV1/FVC values—due to missing post-bronchodilator data—might have led to some misclassification. The researchers recommended future studies on stable COPD populations to refine screening.
Nevertheless, Dr. Checinski and colleagues concluded that “using these results in the outpatient setting can increase recognition of chronic hypercapnia and improve NIV prescription for those who would benefit from it.”
