Clinical background
The bedside blood gas analysis serves two decisions: identifying the primary acid–base disturbance, and judging whether compensation is adequate or whether an additional disturbance is hidden behind an apparently balanced pH. Without a systematic method, mixed disturbances risk being missed, particularly when the pH lies within the reference range but a metabolic acidosis and a metabolic alkalosis offset one another. A structured stepwise interpretation has been shown to double the detection of mixed disturbances compared with informal bedside interpretation [2].
Performing the blood gas analysis
The interpretation rests on three measured values and two calculated quantities. The arterial reference ranges are pH 7.35–7.45, PaCO₂ 35–45 mmHg and bicarbonate 22–26 mmol/L.
To judge whether respiratory compensation in metabolic acidosis is adequate, Winter's formula is used:
If the measured PaCO₂ lies within this range, compensation is appropriate. If it lies higher, a concurrent respiratory acidosis is present; if lower, a concurrent respiratory alkalosis.
The anion gap is calculated as:
The reference range is 8–12 mmol/L [3]. An anion gap above 12 suggests a high anion gap metabolic acidosis (HAGMA), and an anion gap of 20 mmol/L or more means that a metabolic acidosis is present whatever the pH or bicarbonate shows [3].
The systematic approach with predictive formulae for compensation was derived and formalised by Narins and Emmett in a 1980 review, in which they set out the relationships between primary disturbances and expected compensation for all the simple acid–base disorders [1]. The framework rests on physiological principles of how the kidneys and lungs compensate for one another and has been standard in textbooks and clinical guidelines ever since.
Interpretation in practice
Interpretation proceeds stepwise: first establish the primary disturbance from the direction of the pH and from which parameter (PaCO₂ or HCO₃⁻) moves in the same direction as the pH, then judge the compensation, and finally calculate the anion gap.
| Primary disturbance | pH | PaCO₂ | HCO₃⁻ | Assessment of compensation | Action |
|---|---|---|---|---|---|
| Metabolic acidosis | ↓ | ↓ (compensatory) | ↓ | Winter's formula; if the PaCO₂ is outside the range: mixed disturbance | Treat the underlying cause; consider buffering at pH < 7.1 |
| Metabolic alkalosis | ↑ | ↑ (compensatory) | ↑ | Expected PaCO₂ = 0.7 × HCO₃⁻ + 20 ± 5 | Correct volume depletion and hypokalaemia; distinguish chloride-responsive from chloride-resistant |
| Respiratory acidosis | ↓ | ↑ | ↑ (compensatory when chronic) | Acute: HCO₃⁻ rises ~1 mmol/L per 10 mmHg PaCO₂; chronic: ~4 mmol/L | Secure the airway and ventilation; treat the underlying cause |
| Respiratory alkalosis | ↑ | ↓ | ↓ (compensatory when chronic) | Acute: HCO₃⁻ falls ~2 mmol/L per 10 mmHg; chronic: ~5 mmol/L | Identify and treat hypoxia, pain, anxiety, sepsis |
When the pH is normal but both the PaCO₂ and the HCO₃⁻ are abnormal, a mixed disturbance is present in which acidosis and alkalosis offset one another. Here calculation of the anion gap is decisive: a raised gap reveals a hidden metabolic acidosis even when the pH looks well compensated [3].
Validation and performance
The systematic stepwise method has been evaluated against informal bedside interpretation in a cohort of 31 intensive care patients with chronic kidney disease, a population with a high frequency of complex acid–base disturbances [2]. The systematic method identified mixed disturbances in 50% of patients compared with 12.9% with informal interpretation. In the same study, the findings of systematic analysis could in most cases be correlated with the patient's clinical picture and provisional diagnosis, which was not the case for bedside interpretation.
The ability of the anion gap to detect occult tissue anions (lactate and other unmeasured anions) has been studied prospectively in 55 children in shock with 93 blood samples [4]. Hypoalbuminaemia was present in 76% of patients. The uncorrected anion gap detected only 48% of cases with clinically significant increases in tissue anions (> 5 mmol/L), whereas the albumin-corrected gap detected 87%. The uncorrected gap underestimated tissue anions with a bias of 10.2 mmol/L (limits 4.1–16.1) compared with 5.3 mmol/L (limits 0.4–10.2) for the corrected gap. The optimal cut-off for the corrected gap was > 15.5 mmol/L in this population.
The albumin correction is performed using Figge's equation:
that is, 0.25 mmol/L is added for every g/L by which albumin falls below 40 g/L [3,4].
Limitations
Winter's formula applies only to metabolic acidosis and must not be used in metabolic alkalosis or in respiratory disturbances. It also presupposes that the patient does not have a concurrent respiratory disturbance, which is precisely what the formula is meant to help exclude. In severe acidosis (HCO₃⁻ < 8 mmol/L) the formula becomes less reliable, since the linear approximation breaks down at the extremes.
The anion gap is strongly influenced by the albumin concentration. In hypoalbuminaemia, which is common in intensive care patients, a normal or only modestly raised gap can conceal a substantial metabolic acidosis [4]. If albumin is not available, a "normal" anion gap should be interpreted with caution in critically ill patients.
Pre-analytical errors are common and can lead to incorrect interpretation. Samples should be analysed within 15 minutes at room temperature, or within 1 hour if refrigerated [3,5]. Air bubbles in the sample lower the PaCO₂ and raise the PaO₂. With delayed analysis, platelets and leucocytes metabolise oxygen, which can give a falsely low PaO₂, particularly in leucocytosis. Salicylate can give a falsely raised chloride and hence a low anion gap, while ethylene glycol can give a falsely raised lactate [5].
The reference ranges apply to adults at sea level. In pregnancy the reference ranges for both PaCO₂ and HCO₃⁻ should be lowered, and in hypothermia temperature correction of the blood gas can exaggerate the PaCO₂ [5].
References
- Narins RG, Emmett M. Simple and mixed acid-base disorders: a practical approach. Medicine (Baltimore). 1980;59(3):161–87. PMID: 6774200
- Ghatak I, Dhat V, Tilak MA et al. Analysis of Arterial Blood Gas Report in Chronic Kidney Diseases, Comparison between Bedside and Multistep Systematic Method. J Clin Diagn Res. 2016;10(10):BC01–BC05. PMID: 27656429
- Habib T, Nair A, Murphy S et al. Mastering blood gas interpretation: A practical guide for primary care providers. S Afr Fam Pract. 2025;67(1):6058. PMID: 40336441
- Hatherill M, Waggie Z, Purves L et al. Correction of the anion gap for albumin in order to detect occult tissue anions in shock. Arch Dis Child. 2002;87(6):526–529. PMID: 12456555
- Carlton H, Shipman KE. Pitfalls in the diagnosis and management of acid-base disorders in humans: a laboratory medicine perspective. J Clin Pathol. 2024;77(11):772–778. PMID: 39025490
