Arteriovenous blood gas agreement for intensive care unit patients with COVID‐19

Measurement of blood gas parameters is an important part of the assessment of the adequacy of ventilation. Previous research has shown that, in an unselected cohort of patients and in patients with chronic obstructive airways disease, venous values of pH are clinically interchangeable with arterial values while agreement of bicarbonate and lactate are close but less precise.1, 2 Agreement of pCO2 is not clinically acceptable; however, it has been shown to be a good screening test for significant hypercapnia.1, 2 There is a gap in the literature regarding arteriovenous (AV) blood gas agreement in patients with severe pneumonia. The COVID-19 pandemic provided an opportunity to address this issue. Additionally, the circumstances of such a pandemic and limited supply of blood gas syringes have made it impractical for all seriously ill patients to have an arterial line and/or undergo arterial blood gas (ABG) analysis, raising the question of whether venous blood gases (VBG) are a reasonable alternative in these circumstances.

We undertook a pilot project to compare venous and arterial values of pH, bicarbonate, lactate and pCO2 and oxygen saturation on ABG with transcutaneous oxygen saturation (SpO2) in patients with COVID-19 requiring intensive care unit (ICU) care.

This was a prospective cohort study of adult patients admitted to the ICU of a university-affiliated health service for COVID-19 related respiratory failure who required ABG analysis as part of their routine care during the so-called ‘second wave’ of COVID-19 in Melbourne in 2020. As close as practical to an ABG being taken (within 10 min), venous blood was taken for VBG. Up to five sample pairs per patient were taken. Analyses were performed on the same point of care analyser in the ICU. The research was stopped before planned because COVID cases overall (and hospital and ICU admissions) were eliminated for some months. Ethics approval was provided by Western Health low-risk ethics panel (QA 2020.66).

We analysed 28 matched sample pairs from seven patients (one pair excluded for protocol violation). Median age was 60 years, six underwent mechanical ventilation, three required inotropic agents and one died. Inspired concentration of oxygen varied from 30% to 100%. Median paO2 was 74 mmHg (interquartile range 62.3–86.8). Results are shown in Table 1. There were insufficient cases with paCO2 >50 mmHg to allow analysis of vpCO2 as a screening tool for hypercarbia.

TABLE 1. Arteriovenous agreement for blood gas parameters This study Previous research Parameter N Mean difference 95% CI of mean difference 95% limits of agreement Mean difference 95% limits of agreement pH 28 −0.036 −0.0416 to −0.0312 −0.063 to 0.010 −0.033 Approximately ±0.1 pCO2 28 5.9 mmHg 5.20 to 6.51 mmHg 2.5 to 9.2 mmHg 6.2 mmHg −17.4 to 23.9 mmHg Bicarbonate 28 1.5 mmol/L 1.13 to 1.87 mmol/L −0.4 to 3.4 mmol/L 1.3 mmol/L Approximately ±5 mmol/L Lactate 28 0.14 mmol/L 0.069 to 0.217 mmol/L −0.23 to 0.52 mmol/L 0.23 mmol/L −2 to 2.3 mmol/L N Mean difference§ 95% CI of mean difference 95% limits of agreement Mean difference 95% limits of agreement Oxygen saturation 28 0.16% −0.662 to 0.983% −4.00 to 4.32% 0.758% −8.2 to 6.7% ABG, arterial blood gas; CI, confidence interval; SpO2, oxygen saturation; VBG, venous blood gases.

Although this is a small sample, our results suggest that AV agreement for pH, pCO2, bicarbonate and lactate is similar to previously reported research. On that basis as long as clinicians are aware of the performance characteristics and weaknesses of VBG analyses and integrate VBG and SpO2 results with clinical features, it may be reasonable for VBG analyses to be used to guide management in patients with severe pneumonia when ABG is not available or feasible.

Acknowledgement

The authors acknowledge the assistance of Dr Romeo Restellini for data collection.

Competing interests

AMK is a member of the editorial board of Emergency Medicine Australasia.

留言 (0)

沒有登入
gif