VBG vs ABG Interpreter
VBG vs ABG Interpreter
You have a venous gas. Can you act on it, or do you need an arterial one? That depends entirely on what you are trying to establish, so this page asks. It deliberately does not convert a venous pCO₂ to an arterial one — the limits of agreement are about ±20 mmHg, which makes a point estimate dishonest. For the one value that is worth converting, see the venous to arterial blood gas converter.
Can I act on this venous gas?
Sample + purpose → is an arterial gas needed?A peripheral venous gas on a breathless patient: pH 7.32, pCO₂ 52 mmHg, bicarbonate 26 mmol/L. The question being asked is whether the patient is retaining CO₂.
The published agreement figures, which are the whole argument
Bicarbonate — peripheral: bias −1.20 mmol/L, limits of the order of ±5 (Kelly). Central: bias 0.52, limits −1.81 to +2.85 (Middleton). Use it to classify, not to track.
pCO₂ — peripheral: bias 6.2 mmHg, 95% limits of agreement of the order of ±20 mmHg (Kelly); bias prediction interval −10.7 to +2.4 mmHg (Byrne). Never as a number. Below 45 mmHg, as a screen: NPV 100%, 95% CI 89–100%.
pO₂ — arterial typically 36.9 mmHg higher, 95% CI 27.2–46.6 (Byrne). Not convertible at all.
Mixed venous — no sourced bias or limits against arterial for any of these. Not established.
- bias
- the average difference between the two samples. A small bias means the venous value sits close to the arterial one on average, which is necessary but nowhere near sufficient
- limits of agreement
- the interval containing 95% of individual differences, roughly the bias plus or minus twice the standard deviation of the differences. This is the number that decides whether a value is usable in one patient, and it is the number usually left out of a summary. A bias of 6.2 mmHg with limits of ±20 mmHg is not a small correction with a little noise; it is an unusable measurement with a well-characterised average
- why pH survives and pCO₂ does not
- pH is logarithmic, so the extra CO₂ that venous blood carries shifts it by very little, and the bicarbonate buffers most of the rest. The pCO₂ records that same extra CO₂ linearly, and how much of it there is depends on local blood flow, extraction and transit time — which vary between patients far more than the mean difference suggests
- 45 mmHg
- the screening threshold, equivalent to 6.0 kPa, derived in 196 paired samples and validated in 107 patients with chronic obstructive pulmonary disease against an arterial PaCO₂ above 50 mmHg. Sensitivity 100% (95% CI 91–100%), negative predictive value 100% (95% CI 89–100%), specificity 47% (95% CI 35–59%). Good at ruling out, poor at ruling in
- not in shock
- the condition attached to Kelly’s interchangeability conclusion, and the one most often dropped when the finding is quoted. Rudkin and colleagues, studying hypovolaemic shock, found the differences no wider but still concluded a peripheral venous gas is not a surrogate for tissue acid-base status in circulatory failure; Prasad and colleagues, in hypotensive emergency patients, concluded the opposite. Treat it as unsettled and take the arterial sample
Worked example
A peripheral venous gas on a breathless patient: pH 7.32, pCO₂ 52 mmHg, bicarbonate 26 mmol/L. The question being asked is whether the patient is retaining CO₂.
Sample is peripheral venous and the purpose is ventilation adequacy, so the screening rule applies
Venous pCO₂ 52 mmHg is at or above the 45 mmHg threshold → the screen is positive
That is not a diagnosis of hypercapnia. Specificity at this threshold is 47% (95% CI 35–59%), so roughly half of positive screens are not hypercapnic
Do not subtract the 6.2 mmHg mean difference to get "about 46 arterial". With limits of agreement around ±20 mmHg the true value could be anywhere from the low thirties to the low seventies
What the same sample does settle: pH 7.32 with bicarbonate 26 is an acidaemia with a raised bicarbonate, and both values transfer — so this is a respiratory acidosis with some metabolic compensation, which implies it is not of the last few minutes
Had the venous pCO₂ been 38 instead, the answer would have flipped to screen negative, and the arterial sample could reasonably have been avoided for this question — but not for oxygenation, which no venous sample addresses
What each sample answers, and how well
| Peripheral venous | Central venous | Mixed venous | |
|---|---|---|---|
| Acid-base status (pH, bicarbonate) | Yes — if not shocked | Yes — tightest limits of the three | Not established |
| Is the patient hypercapnic? (screen) | Yes — below 45 mmHg, NPV 100% (CI 89–100%) | Threshold not validated | Not established |
| A pCO₂ number to titrate on | No — limits about ±20 mmHg | No | Not established |
| Oxygenation (pO₂, saturation) | No — and this is the dangerous one | No | No — SvO₂ measures extraction |
The offsets in both units
| Value | Difference (mmHg or units) | Difference (kPa) | 95% limits of agreement |
|---|---|---|---|
| pH, peripheral | 0.034 (Kelly) / 0.03 (Byrne) | — | Narrow; Byrne’s CI on the bias 0.029–0.038 |
| pH, central | 0.03 (Middleton) | — | −0.07 to +0.01 |
| pCO₂, peripheral | 6.2 mmHg (Kelly) | 0.83 kPa | Of the order of ±20 mmHg (±2.7 kPa) |
| Bicarbonate, peripheral | −1.20 mmol/L (Kelly) | — | Of the order of ±5 mmol/L |
| Bicarbonate, central | 0.52 mmol/L (Middleton) | — | −1.81 to +2.85 mmol/L |
| pO₂, peripheral | 36.9 mmHg (Byrne) | 4.92 kPa | CI on the difference 27.2–46.6 mmHg |
Why this page refuses to convert a pCO₂, and what it does instead
The obvious form of this page is a venous-to-arterial calculator: enter the venous values, read the arterial ones. It should not be built that way, and the reason is in the numbers above. For pH it would work, and the arithmetic exists on the venous to arterial blood gas converter. For pCO₂ it would not. Kelly’s 95% limits of agreement are of the order of ±20 mmHg, so a page that took a venous pCO₂ of 55, subtracted the 6.2 mmHg mean difference and printed 48.8 would be presenting a figure whose true value could plausibly be 29 or 69. Nothing in the output would say so. A decimal point is a claim about precision, and that claim would be false.
So this page asks the question a reader actually has instead. Almost nobody with a venous gas in front of them wants an arterial number for its own sake; they want to know whether they can act on what they have or whether they need to do something more invasive. That answer depends on the purpose, and the same sample earns three different verdicts. For acid-base classification it is a good test. For the question of whether a patient is retaining CO₂ it is an adequate screening test with a sourced negative predictive value. For a pCO₂ to titrate ventilation on, and for anything to do with oxygenation, it is not a test at all.
The screening role deserves its own paragraph, because it is the most useful and least known finding here. Kelly, Kerr and Middleton took a threshold derived from 196 paired samples and validated it prospectively in 107 patients with chronic obstructive pulmonary disease. A venous pCO₂ below 45 mmHg ruled out an arterial PaCO₂ above 50 mmHg with a negative predictive value of 100%, 30 of 30, 95% confidence interval 89 to 100%, and caught all 43 hypercapnic patients. The specificity was 47%. That combination — excellent at ruling out, poor at ruling in — is exactly the profile of a good screening test, and it means a normal venous pCO₂ can reasonably spare a patient an arterial puncture when the only question is CO₂ retention. It also means a raised one settles nothing.
Two limits run underneath everything. The first is shock. Kelly’s interchangeability conclusion is restricted to patients who are not in shock, and the restriction is usually dropped when the finding is repeated. The evidence in circulatory failure is genuinely unsettled rather than simply negative: Rudkin and colleagues found the arterio-venous differences no wider in hypovolaemic shock but still concluded that a peripheral venous sample reflects the perfusion of the limb rather than the patient, while Prasad and colleagues, studying hypotensive emergency department patients, concluded the values were adequate. In practice, take the arterial sample when the patient is shocked. The second is the draw itself. A tourniquet left in place, a prolonged search for the vein or a repeatedly clenched fist all raise the local pCO₂ and lower the local pH, and no published offset corrects any of it. If the result does not fit the patient in front of you, repeat the sample properly before believing it.
For reading the gas once you have the right one, the arterial blood gas interpreter works through the acid-base pattern, the anion gap calculator and the base excess calculator quantify the metabolic component, and the A-a gradient calculator and the PaO₂/FiO₂ ratio calculator put an arterial pO₂ in context.
Frequently asked questions
Can a venous blood gas replace an arterial one?
For some questions, yes. For pH and bicarbonate in a patient who is not shocked, Kelly’s systematic review of 2,087 paired samples concluded the venous values have sufficient agreement to be clinically interchangeable, and Byrne’s independent meta-analysis agreed on the pH to the third decimal place. For the pCO₂ as a number, no — the 95% limits of agreement are of the order of ±20 mmHg. For oxygenation, no, and not by any margin: a venous pO₂ measures local tissue extraction rather than pulmonary gas exchange.
How much lower is venous pH than arterial?
About 0.03 units, and that figure is unusually well established. Kelly’s weighted mean arterio-venous difference across 2,087 paired peripheral samples is 0.034 units with narrow limits of agreement. Byrne’s meta-analysis of 18 studies and 1,768 subjects puts the arterial pH typically 0.03 higher, 95% confidence interval 0.029 to 0.038. Middleton’s central venous figure is also 0.03, with 95% limits of agreement of −0.07 to +0.01 — so the bias is essentially the same for central and peripheral samples, which is worth knowing because it is often assumed to differ.
Can a normal venous pCO₂ rule out hypercapnia?
Yes, and this is the single most useful thing on this page. Kelly, Kerr and Middleton validated a venous pCO₂ threshold of 45 mmHg in 107 patients with chronic obstructive pulmonary disease: against an arterial PaCO₂ above 50 mmHg, a value below 45 gave a negative predictive value of 100% (30 of 30, 95% CI 89 to 100%) and a sensitivity of 100% (43 of 43, 95% CI 91 to 100%). The specificity was only 47%, so it works as a rule-out and not as a rule-in. Note that the threshold was derived and validated on peripheral samples, so this page does not apply it to central venous blood.
Why will this page not convert the venous pCO₂ to an arterial one?
Because the scatter is too large for a point estimate to be honest. The mean difference is well characterised — Kelly gives 6.2 mmHg — but Kelly’s 95% limits of agreement are of the order of ±20 mmHg and Byrne’s 95% prediction interval for the bias runs from −10.7 to +2.4 mmHg, which Byrne calls unacceptably wide. Subtracting the mean from a single patient’s value produces a number that looks like a measurement and is not one. A corrected pCO₂ could read normal in a patient whose true value is 60.
Can I get an oxygen saturation or pO₂ from a venous gas?
No. This is the branch where getting it wrong is dangerous. Byrne’s meta-analysis found the arterial pO₂ typically 36.9 mmHg higher than the venous, with a 95% confidence interval on that difference of 27.2 to 46.6 mmHg, but the width is not the real objection. A venous pO₂ measures how much oxygen the tissue drained by that vein extracted, which depends on local flow and local metabolism, and no offset turns that into a measure of how well the lungs are working. Use pulse oximetry, or take an arterial sample. A mixed venous saturation from a pulmonary artery catheter is a genuine measurement, but of oxygen extraction rather than oxygenation, and a normal SvO₂ is compatible with severe arterial hypoxaemia.
Do the published offsets apply to a central or mixed venous sample?
Central venous, partly: Middleton and colleagues studied 168 matched pairs from 110 patients and found tight agreement for pH, bicarbonate, base excess and lactate, with limits narrower than the peripheral figures. Mixed venous, no — no bias or limits of agreement against arterial values could be sourced to primary literature for mixed venous samples, so this page states that the figure is not established rather than borrowing the peripheral numbers. In practice a patient with a pulmonary artery catheter almost always has arterial access as well.
Do these findings still hold if the patient is in shock?
This is genuinely unsettled and should be treated as such. Kelly’s interchangeability conclusion is explicitly limited to patients who are not in shock. Rudkin and colleagues, studying hypovolaemic shock specifically, found the arterio-venous differences no wider than in normotensive patients but concluded that a peripheral venous gas is still not a surrogate for tissue acid-base status in circulatory failure, because it reflects the perfusion of that limb. Prasad and colleagues, in hypotensive emergency department patients, reached the opposite conclusion. The practical answer is to take the arterial sample when the patient is shocked.
Related calculators
References
- Kelly AM. Agreement between arterial and venous blood gases in emergency medical care: a systematic review. Hong Kong J Emerg Med. 2013;20(3):166–171. doi:10.1177/102490791302000307. Weighted mean arterio-venous differences: pH 0.034 (n=2,087) with narrow limits of agreement; pCO₂ 6.2 mmHg (n=1,043) with "95% limits of agreement up to the order of ±20 mmHg"; bicarbonate −1.20 mEq/L (n=1,403) with "95% limits of agreement of the order of ±5 mmol/L"; base excess 0.4 (n=295) with conflicting data on the width of the limits. Conclusion: "for patients who are not in shock, venous pH and bicarbonate have sufficient agreement to be clinically interchangeable for arterial values."
- Byrne AL, Bennett M, Chatterji R, Symons R, Pace NL, Thomas PS. Peripheral venous and arterial blood gas analysis in adults: are they comparable? A systematic review and meta-analysis. Respirology. 2014;19(2):168–175. doi:10.1111/resp.12225. Eighteen studies, 1,768 subjects. Arterial pH "typically 0.03 higher than the venous pH (95% confidence interval 0.029–0.038)"; venous-arterial pCO₂ bias prediction interval "unacceptably wide, extending from −10.7 mm Hg to +2.4 mm Hg"; arterial pO₂ "typically 36.9 mm Hg greater than the venous with significant variability (95% confidence interval from 27.2 to 46.6 mm Hg)".
- Kelly AM, Kyle E, McAlpine R. Venous pCO₂ and pH can be used to screen for significant hypercarbia in emergency patients with acute respiratory disease. J Emerg Med. 2002;22(1):15–19. doi:10.1016/S0736-4679(01)00431-0. 196 sample pairs, 56 with significant hypercarbia defined as arterial pCO₂ above 50 mmHg. At a venous pCO₂ cut-off of 45 mmHg, sensitivity 100% and specificity 57%.
- Kelly AM, Kerr D, Middleton P. Validation of venous pCO₂ to screen for arterial hypercarbia in patients with chronic obstructive airways disease. J Emerg Med. 2005;28(4):377–379. doi:10.1016/j.jemermed.2004.10.017. 107 patients with complete data. Venous pCO₂ above 45 mmHg against arterial pCO₂ above 50 mmHg: sensitivity 100% (43/43, 95% CI 91–100%), specificity 47% (95% CI 35–59%), negative predictive value of a venous pCO₂ below 45 mmHg 100% (30/30, 95% CI 89–100%).
- Middleton P, Kelly AM, Brown J, Robertson M. Agreement between arterial and central venous values for pH, bicarbonate, base excess, and lactate. Emerg Med J. 2006;23(8):622–624. doi:10.1136/emj.2006.035915. 168 matched pairs from 110 patients. Mean differences and 95% limits of agreement: pH −0.03 (−0.07 to 0.01); bicarbonate 0.52 mmol/L (−1.81 to 2.85); base excess 0.19 mmol/L (−1.86 to 2.24); lactate 0.08 mmol/L (−0.27 to 0.42).
- Rudkin SE, Anderson CL, Grogan TR, Elashoff DA, Treger RM. Assessing acid-base status in circulatory failure: relationship between arterial and peripheral venous blood gas measurements in hypovolemic shock. J Intensive Care Med. 2020;35(5):511–518. doi:10.1177/0885066618762335
- Prasad H, Vempalli N, Agrawal N, et al. Correlation and agreement between arterial and venous blood gas analysis in patients with hypotension — an emergency department-based cross-sectional study. Int J Emerg Med. 2023;16:18. doi:10.1186/s12245-023-00486-z
- Bloom BM, Grundlingh J, Bestwick JP, Harris T. The role of venous blood gas in the emergency department: a systematic review and meta-analysis. Eur J Emerg Med. 2014;21(2):81–88. doi:10.1097/MEJ.0b013e31836437cf
Medical Disclaimer: The tools and content provided here are for educational and reference purposes only. They are not intended to substitute for professional medical advice, diagnosis, or treatment. Clinical decisions should always be based on the comprehensive assessment of a qualified healthcare professional.