Interpreting a blood gas with the Stewart approach
Read pH, SBE and compensation, then use chloride, lactate, albumin and residual charge to explain the metabolic findings.
Start with the gas, then explain the metabolic component
Stewart analysis describes pH through carbon dioxide, strong ion difference and weak acids, chiefly albumin and phosphate. It is most useful when several metabolic effects coexist: chloride-rich fluid, raised lactate and low albumin, for example.
1. Check what was measured
Enter pH and PCO₂ first. Select the sample type and confirm whether PCO₂ is in mmHg or kPa. Add chemistry from the same sampling time; a gas taken before treatment and electrolytes taken afterwards may not describe one acid–base state.
Full SIG calculation also needs sodium, potassium, chloride, lactate, ionised calcium, magnesium, albumin and phosphate. Use the stated units and magnesium assay. Leave unavailable measurements blank. Arterial compensation rules are withheld for venous, capillary and unknown samples.
2. Describe the pH
For an adult arterial sample, pH below 7.35 is acidaemia and above 7.45 is alkalaemia. These describe the net result. Acidosis and alkalosis describe processes, which can coexist even when pH is within range.
A pH of 7.40 does not end the analysis. Check the metabolic effects and respiratory response.
3. Separate the metabolic and respiratory findings
Standard base excess (SBE) summarises the net metabolic component. The default screening interval is −2 to +2 mEq/L; the app allows a local cutoff. An SBE below the selected range indicates a net acidifying metabolic component. An SBE above it indicates a net alkalinising component.
An abnormal SBE can represent renal compensation for a respiratory disorder. It does not automatically establish a second primary disorder. Select acute or chronic respiratory disease from the history; leave the time course unknown when it is uncertain.
4. Check the expected compensation
Compare the observed response with the expected range. In metabolic acidosis, Winter’s formula estimates PaCO₂ as 1.5 × HCO₃⁻ + 8 ±2 mmHg. A lower PaCO₂ suggests an additional respiratory alkalosis; a higher value suggests an additional respiratory acidosis.
Bicarbonate is a dependent variable in Stewart analysis. The app uses calculated bicarbonate for empirical compensation rules, not as the underlying cause of the metabolic disturbance. Appropriate compensation may leave pH abnormal. A normal pH alone does not establish complete compensation.
Compensation rules, borderline results and time-course assumptions
5. Identify the metabolic contributors
- Chloride relative to sodium: a lower sodium–chloride difference has an acidifying direction. Review saline exposure, intestinal losses, renal function and medicines. A higher difference has an alkalinising direction; review gastric losses, diuretics and potassium.
- Lactate: raised lactate lowers SID. Assess perfusion, oxygen delivery, adrenergic stimulation, seizures, clearance and relevant drugs.
- Albumin: low albumin has an alkalinising effect and may partly offset acidosis. It can also lower the uncorrected anion gap.
- Phosphate: a higher concentration adds weak-acid charge; a lower concentration reduces it.
- SIG: a raised residual gap may suggest additional unmeasured anions. Check the selected reference and method before assigning causes.
A high SID has an alkalinising direction. It is different from a high SIG. Measured SIDa, weak-acid effects and residual charge must be read together; neither SIDa nor a sodium–chloride estimate alone establishes the diagnosis.
6. Use the causes list to choose the next check
The app places metabolic groups with larger estimated charge effects first. That order describes the estimated contribution, not the probability of a particular disease. Within a group, use the history, examination and targeted measurements to distinguish causes.
For chloride-associated acidosis, check fluids, gastrointestinal losses and renal function. For residual anions, check blood ketones, kidney function and exposure history. A raised lactate does not exclude another acidifying process. Review opposing effects as well as the largest contributor.
Further reading
Adult blood gases at 37°C. Use and limitations.