ABG Interpretation for FRCEM SBA: Compensation, Anion Gap and Mixed Disorders

An arterial blood gas can contain several abnormal results without representing several separate disorders. Equally, a pH within the reference range can conceal important acid–base abnormalities. The challenge in FRCEM SBA revision is to move beyond recognising a low pH and explain how the results fit together.

This guide develops a systematic approach to ABG interpretation, with worked calculations in UK units and five original practice questions. The emphasis is on identifying the primary disturbance, checking compensation and recognising when a second process is present.

A systematic approach to ABG interpretation

Use the same sequence each time: check the clinical context and sample, assess oxygenation, identify the acid–base disturbance, calculate expected compensation, assess the anion gap where relevant, and integrate the findings. A reliable method is more useful than memorising isolated patterns.

Check the sample, oxygen delivery and units first

Confirm whether the sample is arterial or venous. Record the oxygen device and delivered or estimated inspired oxygen concentration. A PaO₂ of 10 kPa on room air has a different significance from the same result on a high inspired oxygen concentration.

Venous pH and carbon dioxide may help answer selected clinical questions, but venous and arterial results are not interchangeable. In particular, venous oxygen tension cannot assess arterial oxygenation. Differences also become less predictable in some critically ill patients. See GGC Medicines’ blood-gas guidance.

Throughout this article, gas tensions are reported in kPa, and bicarbonate and electrolytes in mmol/L. Some compensation equations were developed using mmHg: 1 kPa is approximately 7.5 mmHg. Always check the units before calculating.

Describe the disturbance before naming the diagnosis

Acidaemia means a blood pH below 7.35; alkalaemia means a pH above 7.45. An acidosis or alkalosis is a physiological process tending to lower or raise pH. More than one process can occur at once, so an acidosis does not require the final pH to be below 7.35.

Begin with a description such as “metabolic acidosis with a raised anion gap and an additional respiratory acidosis”. Then use the history, examination and other investigations to identify the causes. A blood gas supports a diagnosis; it does not replace the clinical assessment.

Recognise the four primary acid–base disorders

The following table describes simple disorders and the expected direction of compensation. It is a starting point rather than a substitute for calculating whether the response is appropriate.

Respiratory acidosis and respiratory alkalosis

Carbon dioxide reflects the balance between production and alveolar ventilation. Inadequate ventilation raises PaCO₂, producing respiratory acidosis. Increased ventilation lowers PaCO₂, producing respiratory alkalosis. The same PaCO₂ can be associated with different pH values depending on the metabolic component and duration of the disturbance.

Renal adaptation develops over days, while buffering begins much earlier. Therefore, a person with established hypercapnia may have a higher bicarbonate than someone whose carbon dioxide has risen abruptly. Compare previous gases where available and avoid diagnosing chronic retention from bicarbonate alone.

Metabolic acidosis and metabolic alkalosis

A primary reduction in bicarbonate produces metabolic acidosis; a primary increase produces metabolic alkalosis. Respiratory compensation changes carbon dioxide in the same direction as bicarbonate, reducing the effect on pH.

For example, low bicarbonate and low PaCO₂ may represent metabolic acidosis with appropriate respiratory compensation. However, the carbon dioxide could also be lower than expected because an additional respiratory alkalosis is present. The direction is only the first clue.

Is compensation appropriate—or is there a mixed disorder?

Expected compensation provides a benchmark against which to compare the observed results. These equations are approximate physiological relationships, not exact targets. Acute evolution, treatment and additional disease processes can affect the result.

Winter’s formula without the kPa trap

For metabolic acidosis, Winter’s formula estimates the expected arterial carbon dioxide tension:

Expected PaCO₂ in mmHg = (1.5 × bicarbonate) + 8 ± 2

Bicarbonate is entered in mmol/L. To express the answer in kPa, divide the predicted value and its tolerance by 7.5:

Expected PaCO₂ in kPa = [(1.5 × bicarbonate) + 8] ÷ 7.5, with a tolerance of approximately ±0.27 kPa.

Consider bicarbonate of 12 mmol/L. The predicted PaCO₂ is (1.5 × 12) + 8 = 26 mmHg, with an expected range of 24–28 mmHg. Dividing by 7.5 gives 3.2–3.7 kPa. A measured PaCO₂ of 3.5 kPa is compatible with appropriate respiratory compensation.

A PaCO₂ above that range suggests an additional respiratory acidosis; a value below it suggests an additional respiratory alkalosis. Do not compare the answer “26” directly with a PaCO₂ reported in kPa. Winter’s formula predicts carbon dioxide from bicarbonate; it is not a general equation for every acid–base disorder.

Acute versus chronic respiratory compensation

Common teaching estimates use a baseline PaCO₂ of 40 mmHg, approximately 5.3 kPa, and bicarbonate of 24 mmol/L. The table below uses the usual approximate ranges summarised in the Merck Manual compensation table.

Ten mmHg is approximately 1.33 kPa. If PaCO₂ is 8.0 kPa, it is approximately 60 mmHg: 20 mmHg above the baseline. Bicarbonate around 26–28 mmol/L fits the acute respiratory-acidosis estimate; around 30–32 mmol/L fits the chronic estimate. A substantially different result should prompt consideration of another metabolic process.

For metabolic alkalosis, PaCO₂ generally rises by approximately 0.6–0.75 mmHg for each 1 mmol/L increase in bicarbonate above baseline. Hypoventilatory compensation is limited; a markedly elevated PaCO₂ should not automatically be attributed to metabolic alkalosis.

A normal pH can conceal significant disease

Consider a patient with pH 7.40, PaCO₂ 2.7 kPa and bicarbonate 12 mmol/L. The pH is within range, but both carbon dioxide and bicarbonate are markedly reduced. Winter’s expected PaCO₂ is 3.2–3.7 kPa; the observed value is lower.

This is compatible with metabolic acidosis plus respiratory alkalosis. Clinical possibilities include salicylate toxicity or sepsis, depending on the wider presentation. “Normal acid–base status” would miss the opposing processes. These values are rounded teaching examples, not diagnostic criteria for a particular disease.

Calculate and interpret the anion gap

The anion gap helps investigate metabolic acidosis by estimating the difference between routinely measured cations and anions. It does not identify a particular toxin or metabolite by itself.

State whether potassium is included

This article uses the potassium-excluded calculation:

Anion gap = sodium − (chloride + bicarbonate).

For sodium 140, chloride 104 and bicarbonate 12 mmol/L, the gap is 140 − (104 + 12) = 24 mmol/L. If potassium were 4 mmol/L and included, the result would instead be 28 mmol/L. Both conventions exist; they require their corresponding reference intervals.

Use the laboratory’s stated range or the range supplied in the question. For the worked delta calculations below, an assumed normal gap of 12 mmol/L is explicitly specified. This is a teaching assumption, not a universal laboratory cut-off. The RCPA anion-gap reference describes both conventions.

Account for albumin

Albumin contributes substantially to unmeasured anions. When albumin is low, an apparently ordinary measured gap can conceal an increase relative to that patient’s expected baseline.

A commonly used approximate correction is:

Albumin-corrected gap = measured gap + 0.25 × (40 − albumin in g/L).

If the measured gap is 12 mmol/L and albumin is 20 g/L, the corrected gap is 12 + 0.25 × 20 = 17 mmol/L. If albumin is expressed in g/dL instead, use 2.5 × (4 − albumin). Mixing these units creates a tenfold error in the correction term.

The correction is an aid to interpretation, not a diagnosis. A corrected elevation should lead you to consider the history and measure relevant substances, such as lactate or ketones, rather than assume their presence.

Raised-gap versus normal-gap metabolic acidosis

A gap can change during illness and treatment, and more than one cause may coexist. For disease-specific assessment and management, see our guides to DKA and HHS and AKI and electrolyte emergencies.

Recognise mixed acid–base disorders in clinical cases

Use the same method for each case. Establish the primary abnormality, calculate the expected response, then explain any additional process before considering its cause.

Metabolic acidosis with an additional respiratory disorder

A drowsy patient has pH 7.10, PaCO₂ 5.3 kPa and bicarbonate 12 mmol/L. The bicarbonate indicates metabolic acidosis. Winter’s expected PaCO₂ is 3.2–3.7 kPa, so 5.3 kPa is inappropriately high even though it falls within a typical population reference range.

The interpretation is metabolic acidosis with an additional respiratory acidosis. The important distinction is between “normal for the laboratory” and “appropriate for this patient’s metabolic disturbance”. Assess ventilation and the cause of the metabolic acidosis urgently in clinical practice.

Raised anion gap with a second metabolic process

For an elevated gap, compare its rise with the fall in bicarbonate. Using assumed normal values of anion gap 12 and bicarbonate 24 mmol/L:

Delta ratio = (anion gap − 12) ÷ (24 − bicarbonate).

Suppose sodium is 140, chloride 90 and bicarbonate 20 mmol/L. The anion gap is 30 mmol/L; the delta ratio is (30 − 12) ÷ (24 − 20) = 4.5. The gap has risen much more than bicarbonate has fallen, suggesting a concurrent metabolic alkalosis or a previously raised bicarbonate. Vomiting alongside ketoacidosis is one possible clinical setting.

Conversely, a disproportionately large bicarbonate fall can suggest an additional normal-gap acidosis. Delta relationships depend on baseline values, albumin, timing and treatment. They become unstable when the denominator is small and should not be treated as rigid diagnostic boundaries. The American Thoracic Society’s ABG teaching resource discusses this additional assessment.

Separate oxygenation from acid–base interpretation

A gas with an appropriate compensatory response may still show dangerous hypoxaemia. Always interpret PaO₂ alongside inspired oxygen, the patient’s clinical condition and oxygen-saturation monitoring.

Likewise, a normal oxygen saturation does not exclude hypercapnia. Oxygenation and ventilation answer different questions. For disease-specific respiratory assessment, see our asthma and COPD exacerbation guide.

Five original FRCEM-style blood-gas SBAs

These are original educational examples, not official or recalled RCEM questions. Gas values are rounded and bicarbonate refers to actual bicarbonate. Interpret each case before reading the options.

Questions: interpret before looking at the options

Question 1. An adult develops hypoventilation after receiving an opioid. An arterial sample shows pH 7.26, PaCO₂ 8.0 kPa and bicarbonate 26 mmol/L. Which is the best interpretation?

• A. Primary metabolic acidosis

• B. Acute respiratory acidosis with an approximately appropriate bicarbonate response

• C. Chronic respiratory acidosis with expected renal compensation

• D. Primary respiratory alkalosis

• E. Primary metabolic alkalosis

Question 2. A patient with metabolic acidosis has bicarbonate 10 mmol/L, PaCO₂ 4.8 kPa and pH 7.07. Using Winter’s formula, which additional disturbance is suggested?

• A. Appropriate respiratory compensation only

• B. Additional respiratory alkalosis

• C. Additional respiratory acidosis

• D. Additional metabolic alkalosis, proven by these values

• E. No acid–base disturbance

Question 3. Sodium is 138, potassium 4, chloride 100 and bicarbonate 14 mmol/L. What is the anion gap using the potassium-excluded formula?

• A. 24 mmol/L

• B. 28 mmol/L

• C. 20 mmol/L

• D. 38 mmol/L

• E. 10 mmol/L

Question 4. A patient has a measured anion gap of 11 mmol/L and albumin of 16 g/L. Using a reference albumin of 40 g/L and a correction of 0.25 mmol/L per 1 g/L albumin reduction, what is the corrected gap?

• A. 5 mmol/L

• B. 11 mmol/L

• C. 13.4 mmol/L

• D. 17 mmol/L

• E. 71 mmol/L

Question 5. An adult has pH 7.40, PaCO₂ 2.7 kPa and bicarbonate 12 mmol/L. Sodium is 140, chloride 104 mmol/L and albumin 40 g/L. The stated normal anion gap is 12 mmol/L, excluding potassium. Which interpretation best fits?

• A. Normal acid–base status

• B. Isolated metabolic acidosis with appropriate respiratory compensation

• C. Isolated acute respiratory alkalosis

• D. Metabolic alkalosis with respiratory compensation

• E. Raised-anion-gap metabolic acidosis with an additional respiratory alkalosis

Answers: show every calculation and explain the distractors

Question 1: B. PaCO₂ of 8.0 kPa is approximately 60 mmHg, 20 mmHg above baseline. An acute bicarbonate increase of approximately 2–4 mmol/L gives an expected value of 26–28 mmol/L. The measured 26 fits. A primary metabolic acidosis would not explain the raised carbon dioxide with this bicarbonate response; chronic respiratory compensation would usually produce a larger bicarbonate increase. Respiratory alkalosis requires reduced carbon dioxide, and primary metabolic alkalosis does not explain this acidaemic pattern.

Question 2: C. Expected PaCO₂ is (1.5 × 10) + 8 = 23 ± 2 mmHg, or 2.8–3.3 kPa. The observed 4.8 kPa is too high, indicating an additional respiratory acidosis. It is neither appropriate compensation nor a respiratory alkalosis. A separate metabolic alkalosis cannot be established from these data, and the markedly low pH excludes normal acid–base status.

Question 3: A. The gap is 138 − (100 + 14) = 24 mmol/L. Option B includes potassium, contrary to the question. Option C incorrectly subtracts potassium. Option D omits bicarbonate, while option E subtracts bicarbonate twice. Always write the specified equation before inserting values.

Question 4: D. Albumin is 24 g/L below the reference value. Add 0.25 × 24 = 6 mmol/L to the measured gap: 11 + 6 = 17 mmol/L. Option A subtracts the correction; B omits it; C uses an incorrect factor; E applies the g/dL coefficient to a g/L difference. The correction estimates the effect of low albumin but does not establish the underlying cause.

Question 5: E. The gap is 140 − (104 + 12) = 24 mmol/L, supporting a raised-gap metabolic acidosis. Winter’s expected PaCO₂ is 26 ± 2 mmHg, or 3.2–3.7 kPa. The measured 2.7 kPa is too low, indicating an additional respiratory alkalosis. The normal pH reflects opposing processes. Isolated acute respiratory alkalosis would not explain the elevated gap and this degree of bicarbonate reduction; metabolic alkalosis is inconsistent with the low bicarbonate as the primary explanation.

Common blood-gas interpretation mistakes

A wrong answer often comes from skipping one step rather than lacking a formula. Use the following checks when reviewing practice questions.

Mixing units and reference intervals

Calling every secondary change “compensation”

Calculate whether the magnitude is plausible, not just whether the direction is correct. In severe metabolic acidosis, a PaCO₂ in the usual reference range can represent inadequate ventilation. Conversely, a very low PaCO₂ may exceed the expected compensatory response.

Avoid “fully compensated” as a shortcut for any normal-pH gas. Describe the actual processes and support the interpretation with calculations, clinical timing and previous measurements.

Treating the numbers without checking context

Air contamination, sampling delays, specimen mislabelling and line-fluid contamination can affect results. If the findings conflict with the patient or with one another, consider sample validity and repeat testing where clinically appropriate rather than forcing an explanation.

Also distinguish actual bicarbonate from standard bicarbonate and chemistry total CO₂. For these teaching calculations, use actual bicarbonate; for anion-gap assessment in practice, interpret contemporaneous measurements and the laboratory’s method. Do not combine unrelated samples taken before and after major treatment changes.

Frequently asked questions

Can the pH be normal in a mixed acid–base disorder?

Yes. Opposing processes can bring pH into the reference interval while carbon dioxide, bicarbonate and the anion gap remain abnormal. A normal pH therefore does not complete the interpretation.

How do I use Winter’s formula with kPa?

Calculate (1.5 × bicarbonate) + 8 ± 2 in mmHg, then divide by 7.5. For bicarbonate 12 mmol/L, the expected range is 24–28 mmHg, equivalent to approximately 3.2–3.7 kPa.

Should potassium be included in the anion gap?

Either convention may be used. Follow the laboratory or examination question and use the corresponding reference interval. This article’s worked calculations exclude potassium unless explicitly comparing the two methods.

How should I practise blood-gas questions for FRCEM SBA?

Write your interpretation before looking at the options. Show the equation and units, identify any mixed disorder and then compare your conclusion with the choices. Record whether errors arose from arithmetic, units, compensation or clinical interpretation, and revisit that specific weakness.

Put your interpretation skills into practice

StudyMedical’s FRCEM SBA question bank includes over 3,000 questions, topic-based custom quizzes, timed practice and the option to revisit previously incorrect answers. Use focused practice to strengthen weak areas, then move to mixed sessions to test your reasoning in context.

Explore the StudyMedical FRCEM SBA question bank and make regular, carefully reviewed question practice part of your revision.