ABG INTERPRETATION

DR AYUSH AGARWAL
JU N I OR R ES I DENT
I N T ERNA L M E DI CI NE, P G I M ER
Outline of presentation
1. VBG vs ABG
2. Sources of error
3. Few concepts in Physiology
4. Terminology
5. Steps of acid-base analysis
6. Individual acid base disorders
7. Oxygenation and ventilation
8. Examples
9. Drawbacks of ABG
• Shows close correlation in pCO2 and pH
VBG ABG
• VBG pCO2 > 45 mmHg has 100 % sensitivity and pH ↑ (by ~ 0.04) pH ↓
NPV for predicting arterial hypercarbia
pCO2 ↑ (by 4-5 mmHg) pCO2 ↓
(i.e. if VBG PaCO2 < 45 → rules out arterial
hypercarbia)
• VBG pCO2 can be used to screen for arterial
hypercarbia
• VBG is less painful
• VBG is acceptable if only pH and pCO2 are
required
Sources of error
Time delay Hypothermia
1. Escape of cellular contents – K+ 1. pO2 and pCO2 decreases
increases
Air bubbles
2. Compartment shift – Na+ decreases
1. pO2 increases and pCO2 decreases
3. Metabolism by RBCs – PaO2 decreases
and PaCO2 increases. pH increases as 2. iCal decreases – because pCO2 decreases,
cells use HCO3- there is release of H+ from albumin and in
turn calcium is bound to albumin

Leucocyte larceny
1. Spurious hypoxemia due to very
high TLC
Sources of error
Excess heparin
1. Dilutional decrease in HCO3- and pCO2
2. Effect is increased with increased size of needle
(increased dead space) or decreased volume of blood
3. Recommended dose ~ 0.05 mL of low concentration
(1000 IU/mL) heparin per mL of blood
4. A 2mL blood sample should have 0.1 mL of heparin
which is roughly the dead space of the needle
A few concepts in Physiology
Q Why does pH matter? → Physiologic pH is essential to prevent enzyme inactivation and
denaturation
Net acid production = Net acid elimination

Produced in normal metabolism → Intravascular transport → Eliminated via lungs + kidneys

“buffers” to prevent large pH fluctuations (incompatible with life)

Most important buffer is HCO3- /CO2 buffer system
 A few concepts in Physiology
Henderson-Hasselbalch equation
A few concepts in Physiology
How is acid eliminated?

Lungs Kidneys

Expiration of CO2 PCT DCT

Reabsorption of HCO3- Excretion of H+ as titratable acid (H2PO4-)

Excretion of NH4+
A few concepts in Physiology
o K+/H+ exchange (transcellular shifts)

Acidemia → Hyperkalemia H+ K+
K+ H+
Alkalemia → Hypokalemia

o All acid-base disorders can coexist except respiratory acidosis and respiratory alkalosis
 Terminology
TERMINOLOGY
1. Acidosis vs Acidemia “-emia” → Actual pH of blood What is normal?
2. Alkalosis vs Alkalemia “-osis” → process
• pH 7.35 – 7.45
3. Respiratory vs Metabolic
4. Acute vs Chronic • pO2 80 – 100 mmHg

• pCO2 35 – 45 mmHg
A patient can be Acidemic or Alkalemic but not both
• HCO3- 22 – 26 mmol/L
A patient can have one or more acidosis, or one or more alkalosis
or both.
Terminology
1. Respiratory disorders – Pathologies which disrupt acid-base balance due to their effect on
lung

Respiratory acidosis → ↑ pCO2 Respiratory Alkalosis → ↓pCO2

2. Metabolic disorders – Pathologies which disrupt acid-base balance due to their effect on
anything other than the lung (like kidney and GIT)

Metabolic acidosis →↓ HCO3- Metabolic alkalosis → ↑ HCO3-

STEPS OF ACID-BASE ANALYSIS
 pH [H+] (nmol/L)
(from ABG (check from
report) formula)

Step 0 – Check internal validity 7.00

7.05
100
89
7.10 79

[H+] = 24 (pCO2) Compare this value 7.15 71

7.20 63
[HCO3-] against the pH table
7.25 56
7.30 50
Example: 7.35 45
• Calculated [H+] = [24][31.3]/[9.5] 7.40 40
7.45 35
= 79.07
7.50 32
• Corresponding pH from table is 7.55 28
7.10 which is comparable to pH 7.60 25
calculated from ABG. 7.65 22
• The ABG is valid
Step 1 – Check pH
pH < 7.35 → Acidemia pH > 7.45 → Alkalemia
(and at least 1 acidosis is present) (and at least 1 alkalosis is present)

Normal pH ≠ No acid-base disorder

Step 2 – Check pCO2 ROME
pH and pCO2 in Opposite directions pH and pCO2 in same (Equal) directions

Respiratory pathology Metabolic pathology

 Primary acid-base abnormality Compensation
Step 3 – Metabolic acidosis Hyperventilation to decrease
pCO2
Compensation Metabolic alkalosis Hypoventilation to increase
pCO2
1. Aim is to bring pH back to baseline (but
generally does not return pH back to
normal) Compensatory mechanism for Metab alkalosis is relatively
poor as compared to others
2. Respi pathology → Compensation is by (Hypoventilation → Dec PAO2 → stimulus to breathe more)
kidneys (takes time – 12 hours - 5 days)
3. Metab pathology → Compensation is by Respiratory acidosis Increased reabsorption of
lungs (relatively rapid – 1-24 hours) HCO3- and excretion of H+ by
kidneys
4. If compensation is not adequate → it is
Respiratory alkalosis Decreased reabsorption of
likely that another acid-base abnormality is
HCO3- and excretion of H+ by
present [Mixed disorder] kidneys

Overcompensation never occurs

Different approaches are there

Boston approach Copenhagen approach

•Bicarbonate based Standard base excess (SBE) based
•Commonly used
•Contains the 6 bicarbonate
based “bedside rules” for
compensation
 Formulae – Bicarbonate based Boston interpretation
1. All are approximations Acute Chronic
2. Multiple formulae may exist for the same Respiratory HCO3- increases by 1 HCO3- increases by 4
disorder
acidosis mEq/L for each 10 mEq/L for each 10
mmHg pCO2 above 40 mmHg pCO2 above 40
mmHg mmHg
Metab acidosis:
Winter’s formula pCO2 = 1.5(HCO3-) + 8 ± 2 Respiratory HCO3- decreases by 2 HCO3- decreases by 5
alkalosis mEq/L for each 10 mEq/L for each 10
mmHg pCO2 below 40 mmHg pCO2 below 40
Metab alkalosis: mmHg mmHg
pCO2 = 0.7(HCO3- – 24) + 40 ± 2
“1-4-2-5”
 Electroneutrality is

Step 4 – Anion gap Total cations = Total anions always maintained in

the body

Q Why do we have to
calculate anion gap? Cations (Measured + Unmeasured) = Anions (Measured + Unmeasured)
A To narrow the differential
[HAGMA vs NAGMA]
Unmeasured anions – Unmeasured cations = Measured cations – Measured anions

Why is potassium not a part of this?

Anion gap = Na+ – [HCO3- + Cl-]

Maybe because the absolute contribution to changes in

potassium is so small that it can be neglected
Inclusion of K+ is variable in the literature and it does not
affect the interpretation in most cases
Normal NAGMA HAGMA
High AG Low AG Negative AG

HAGMA Lab error (underestimate Na, Lab error

Overestimate Cl, HCO3-)

Hyperalbuminemia
Hypoalbuminemia Dec in Multiple myeloma →
unmeasured pseudohyponatremia
Hyperphosphatemia Monoclonal IgG/Polyclonal anions
gammopathy
Inc in
Anionic paraprotein (IgA) unmeasured
cations Bromide/Iodide
HyperCa/HyperMg intoxication
Interfere with chloride
Bromide (Myestin,
analysis and produce
Grilinctus)/Iodide intoxication “pseudohyperchloremia”
Lithium chloride
Metabolic alkalosis intoxication
Lithium chloride intoxication
Albumin correction of AG
1. Albumin is the main contributor of AG → if there is reduction in albumin then AG baseline
has to be adjusted appropriately
2. Low albumin → Falsely low AG
For comparison

3. AGadjusted = AGmeasured + 2.5 [4 – Albumin] Caadjusted = Cameasured + 0.8 [4 – Albumin]

Osmolal gap
•Difference between measured serum osmolality and calculated serum osmolality
•Calculated serum osmolality (mOsm/kg) = 2 [Na+] + BUN + Glucose
2.8 18

•Normal = – 10 to + 10 mOsm/kg
•Major causes:
1. Ethanol
2. Toxic alcohols and glycols – Methanol and Ethylene glycol
3. Advanced CKD
4. Pseudohyponatremia (due to hyperlipidemia or hypertriglyceridemia)
 Step 5 – Δ ratio = Δ AG = AGmeasured – AGnormal = AGmeasured – 12
24 – HCO3-measured
Delta ratio Δ HCO3- HCO3-normal – HCO3-measured

Delta ratio in HAGMA

In a pure HAGMA, the increase in AG should be equal to the decrease in HCO3-
i.e. Delta ratio = 1 Delta ratio Interpretation
<0.4 Pure NAGMA
One would expect the ratio to be 1 for pure 0.4 – 0.8 HAGMA + NAGMA
HAGMA but in reality it is a range of values
because anions are not only buffered by 0.8 – 2 Pure HAGMA
HCO3- but by other buffers also
>2 HAGMA + Metabolic alkalosis
Normal pH ≠ No acid base disorder
Follow the sequence of steps
1. Check pH → Normal
2. Check pCO2
3. Check compensation → Since pH is normal, evaluating compensation is not appropriate,
rather the aim is to identify the balancing disorder. This disorder will have to be the complete
opposite of the first disorder to bring the pH to normal.
4. Calculate AG
5. Calculate Delta ratio if AG high
Normal
pH acid-
base
disorders
Individual acid-base disorders
 Metabolic acidosis Causes of HAGMA
Lactic acidosis Toxins
Causes of NAGMA Ketoacidosis Ethylene glycol
GI Bicarb loss Renal loss Diabetic Methanol
Diarrhea RTA – 1, 2, 4 Alcoholic Salicylates
Ureterosigmoidostomy Early CKD Starvation Propylene glycol
Volume expansion Pyroglutamic acid (5-oxoproline)
with NS
Renal failure

CKD can cause NAGMA in early phases HAGMA in late severe phase
Tubular dysfunction resulting in decreased HCO3- Decline in GFR contributes to raising AG by
reabsorption and impaired H+ excretion → retention of phosphate, sulfate, urate and
retained H+ buffered by HCO3- in ECF hippurate
Approach to HAGMA
Raised AGadjusted If without acidosis, consider causes of raised AG w/o acidosis

Ketones + Lactate Check for DKA/AKA/Starvation; Shock, Regional ischemia

(Bowel/Limb)

Renal function Renal failure

Osmolal gap Toxic alcohols

NAGMA/Hyperchloremic
Concept of Urine Anion Gap
Urine Anion gap = [ Na+ + K+ ] – [ Cl- ]

The major unmeasured cation in urine is NH4+

The higher the
(which usually can’t be measured directly) urine NH4+ , lower is
the UAG
Q Why do we need UAG?
A To indirectly estimate the Urine NH4+ excretion

Q Why do we need an estimate of Urine NH4+ excretion?

A To narrow down differential of NAGMA
 Renal acid secretion

Absorption of HCO3- Secretion of H+

Phosphate → H2PO4-
The 2 main urinary buffers
Ammonium → NH4+
(stimulated by intracellular acidosis)

Main adaptive response to chronic metabolic acidosis is to

increase H+ secretion in the form of NH4+
Excretion of acid load in urine requires urinary buffers to bind H+ (if
H+ were free and buffers were not present in urine, the urine pH
would fall to <2.5 and result in a high pH gradient between the cell
interior and tubular lumen, which cannot be achieved)
NAGMA/Hyperchloremic
Q How does Urine NH4+ narrow down differential of NAGMA?
Diarrhea causing NAGMA + hypokalemia CKD; RTA 1, 4

Increased renal acid secretion Impaired distal NH4+ excretion

Increased urine NH4+ Decreased urine NH4+

Low/Negative UAG High UAG

Indicates appropriate Indicates renal
acidification of urine acidification defect
 Metabolic alkalosis
Generation – Any factor causing loss of acid or gain of HCO3- will generate a Metab alkalosis
This can normally be corrected quickly by renal HCO3- excretion → To sustain
metabolic alkalosis, other factors are required

Maintenance – When there is failure to eliminate excess HCO3- from kidneys from ECF.

Occurs in these situations:

1. Chloride deficiency
2. Hypokalemia In combination with reduced eGFR
3. Volume deficiency
4. Hypokalemia due to autonomous hyperaldosteronism
Causes of metabolic alkalosis
Primary issue GI Renal
Loss of H+ Vomiting Loop/Thiazide
diuretics
NG suction Mineralocorticoid
excess
Congenital chloride Contraction alkalosis
diarrhea
Post hypercapneic

Bartter/Gitelman
syndromes
Gain of HCO3- Milk-alkali syndrome Contraction alkalosis
Ingestion of NaHCO3-
Approach to metabolic alkalosis
Differential diagnosis requires assessment of volume
status, BP, Serum K+ and assessment of RAAS.

Can be divided into

1. Saline responsive (Urine Cl- < 20 mEq/L)

2. Saline unresponsive (Urine Cl- > 20 mEq/L)

T/t
1. Underlying disorder
2. Treat the factors responsible for maintenance
(ECFV deficiency, Hypokalemia, Chloride
deficiency) → i.e. Normal saline
Respiratory acidosis
MECHANISM ETIOLOGY
Decreased minute Central Sedatives
ventilation hypoventilation CVA/Brainstem disease/Hypothyroidism
Neuromuscular Spinal cord + Thoracic cage + Metabolic disorders

Increased dead space Short shallow breathing

COPD

Increased CO2 production Fever

Thyrotoxicosis
Sepsis
 Pain/Anxiety?
Respiratory alkalosis
Assess oxygenation
Assess for PTE
(can trigger hyperventilation
even in the absence of
hypoxemia)

Consider causes
of hypoxemia

Rare etiologies
Causes of hypoxemia would cause
• Progesterone excess
respiratory alkalosis by stimulating
the respiratory drive • CNS lesions
Assessment of oxygenation and ventilation
Hypoxia and O2 delivery
Hypoxia Hypoxemia Hypoxic Anemic
hypoxia hypoxia
Decreased tissue oxygenation Decreased arterial oxygen concentration
Stagnant Histotoxic
Hypoxemia is just one of the causes of hypoxia hypoxia hypoxia

Measured Formula
O2 delivery as
Not a good measure of blood oxygenation
Dissolved in PaO2 PaO2 x 0.03 as it represents only ~1.5% of O2
blood (1.5%) Better for assessing “ventilation”

Bound to SpO2/SaO2 1.34 x Hb x SaO2 Better measure of arterial oxygenation

hemoglobin as it represents of 98.5% of O2 in blood
Pulse-ox ABG
(98.5%)
 A-a gradient
1. Check A – a gradient
2. Compare with predicted
Alveolar gas equation A-a gradient for age =
(Age/4) + 4
Room air and sea level 3. Check saturation gap
PI = 760
PH2O = 47
FIO2 = 0.21
Normal diet, RQ = 0.8
Increased A – a gradient Normal A – a gradient

V/Q mismatch Hypoventilation

(eg Pulmonary vascular disease, embolic disease) (a/w ↑pCO2)

Right to Left shunt Low PI

(severe form of V/Q mismatch in which V/Q is 0) (High altitude)

Anatomic shunt Physiologic shunt

AVM Atelectasis The only real utility of A-a

Hepatopulmonary syndrome Pneumonia gradient in most real-world
scenarios is to catch
hypoventilation as an etiology
Diffusion defect
(eg ILD)
Drawbacks of A-a gradient
1. Variability with age – Increases with age due to increasing V/Q mismatch. This needs to be
accounted for.

2. Variability with FiO2 – With increasing FiO2 (i.e. the patient is on supplemental oxygen) both
PAO2 and PaO2 will increase but PAO2 increases disproportionately and
there is a widening of the A-a gradient.

• Limited usefulness in critically ill patients in ICUs because of this, as almost all patients on
supplemental oxygen will have a widened A-a gradient.

• It also requires an exact value of FiO2 which we can only roughly estimate in most ICU situations
Saturation gap = SaO2 – SpO2
•Gap between SaO2 (on ABG) and SpO2 (on pulse oximetry)
•Should be <5%
•If >5% → significant difference

•Suggests that there is an abnormal hemoglobin which is detected by ABG but not by pulse
oximetry
1. Carboxyhemoglobin
2. Methemoglobin
3. Sulfhemoglobin (H2S)
PaO2/FiO2 ratio
• Measure of oxygenation
• Normal – 300 to 500 mmHg
• Can be used as an estimate of A-a gradient if the pCO2 is normal and no shunt is suspected
• Used in Berlin definition for ARDS severity
•Disadvantage → It tells the relationship between FiO2 and arterial O2 but does nothing to
discriminate among the potential causes
•Markedly dependent on FiO2
P/F ratio Mortality
200 – 300 Mild ARDS 27%
•Rule of thumb: PaO2 = 500 x FiO2
100 – 200 Moderate ARDS 32%
Example: FiO2 = 0.21
<100 Severe ARDS 45%
PaO2 = 500 x 0.21 = 105
Examples
Example 1
pH 7.34 Step 1 – Validity: 24 x 55/29 = 45.5 → corresponds to pH
of 7.35 → Valid
pCO2 55
Step 2 – pH → Acidemia

HCO3- 29 Step 3 – pCO2 → Respiratory acidosis

Step 4 – Compensation: as it is chronic, HCO3- should be

COPD since 5 years 24 + 4[(55-40)/10] = 30

Final Dx – Chronic compensated respiratory

acidosis due to COPD
Example 2
pH 7.25 Step 1 – Validity: 24 x 28/12 = 56 → corresponds to pH
of 7.25 → Valid
pCO2 28
Step 2 – pH – 7.25 → Acidemia

HCO3- 12 Step 3 – pCO2 → Metabolic acidosis

Step 4 – Compensation: pCO2 = 1.5(12) + 8 ± 2 = 24 to 28

Post cardiac arrest (appropriate)

Final Dx - Compensated metabolic acidosis

?lactic acidosis
Example 3
pH 7.47 K 3.2 75/F

pCO2 25 Cl 110 Fever and SOB

HCO3- 17 Albumin 2.7 RR 35/min

HR 122/min
Na 148 BP 95/40 mmHg
SpO2 86% on RA

CXR – Pneumonia
Example 3
Step 1 – Validity: 24 x 25/17 = 35.3 → Step 5 – AG = 148 – (110+17) = 21 → Corrected AG =
Corresponds to pH 7.45 → Valid 24 → HAGMA

Step 2 – pH: Alkalemia Step 6 – Delta ratio = 12/7 = 1.7 → Pure HAGMA

Step 3 – pCO2: Respiratory alkalosis Final Dx – Respiratory alkalosis + HAGMA

Step 4 – Expected HCO3- is 21 → Metabolic Differential – Alkalosis due to hypoxemia and

acidosis is present acidosis due to lactic acidosis in sepsis
Example 4
pH 7.41 Na 133
61/M
pCO2 55 K 3.4
Morbid obesity
HCO3- 34 Cl 92 COPD and CHF

Routine ABG during pulmonary testing

Example 4
Step 1 – Validity: 24 x 55/34 = 38.8 → Step 5 – AG = 133 – (92+34) = 7 → Normal
Corresponds to pH 7.40 → Valid

Final Dx – Respiratory acidosis +

Step 2 – pH: Normal Metabolic alkalosis

Step 3 – pCO2 Elevated → Respiratory acidosis Differential – Acidosis due to COPD and
alkalosis due to diuretics

Step 4 – Compensation: Opposite disorder

must be present → Metabolic
alkalosis
Example 5
pH 7.41 Na 146
32/F
pCO2 42 K 3.2
Vomiting since 4 days
HCO3- 26 Cl 92
HR 145
BP 78/42
 Elevated anion gap
with negative delta HAGMA + Metabolic alkalosis

Example 5 ratio may be seen in

Step 1 – Validity: 24 x 42/26 = 38.7 → Step 5 – AG = 146 – (92+26) = 28 → HAGMA

Corresponds to pH 7.40 → Valid

Step 6 – Delta ratio = 28-12 = 16 = – 8 → Metabolic

Step 2 – pH: Normal alkalosis
24-26 -2

Step 3 – pCO2 Normal

Final Dx – HAGMA + Metabolic alkalosis

Step 4 – Compensation: none

Differential – Lactic acidosis due to shock
and alkalosis due to vomiting
 59/M, CAD s/p LAD PCI→ SOB with crepts and RR 30
Example 6 Given diuretics as CXR was s/o pulm edema

1. pH – Alkalemic
2. pCO2 – Low → Respiratory alkalosis
3. Compensation – Expected HCO3- is 22.8 Actual
bicarb is 28.3 → Metabolic alkalosis
4. AG – 8
5. Delta ratio – x
6. Final Dx – Respiratory alkalosis +
Metabolic alkalosis
7. Differential – Metabolic component due to
diuretics and respiratory component due to
hypoxia
Drawbacks of ABG
1. Pulse oximetry is a better measure of tissue oxygenation than ABG (as SpO2 is better than
PaO2 for oxygenation)
2. Checking A-a gradient may be misleading in ICU. A normal A-a gradient does not rule out PTE
3. Large random variability in PaO2. This means that a 2nd sample showing a slightly lower PaO2
may just be random variation and does not necessarily indicate that oxygenation is
worsening.
4. Captures only a “snapshot” of the patient status.
5. Painful and invasive
Summary
ACID BASE OXYGENATION
o Step 0 – Validity o Step 1 – A-a gradient
o Step 1 – pH o Step 2 – Compare with age corrected
o Step 2 – pCO2 to see primary disorder o Step 3 – Saturation gap (to not miss
out occult poisoning)
o Step 3 – Compensation
o Step 4 – Anion gap
o Step 5 – Delta ratio
o Step 6 – Formulate a differential
diagnosis
Sources used
1. Harrison’s Textbook of Internal Medicine
2. UpToDate – Acid Base disorders
3. Strong medicine – YouTube
4. The ICU book – Paul Marino
5. FOAMed– LITFL, EMCrit, PulmCrit, DerangedPhysiology
Thank you