The following arterial blood gas result was obtained from a 70-year-old female with type 2 diabetes, presenting with acute exacerbation of asthma.


Measured Value

Adult Normal Range





7.35 - 7.45


60 mmHg (8.0 kPa)*

35 -45 (4.6 - 6.0)


55 mmHg (7 kPa)


23 mmol/L

22 - 27

Base Excess

-4 mmol/L*

-2 - +2


135 mmol/L

135 - 145


5.3 mmol/L*

3.5 - 5.0


100 mmol/L

100 - 110


9.2 mmol/L*

3.5 - 6.0


8.3 mmol/L*

3.5 -7.2


120 umol/L*

50 - 100


4.8 mmol/L*

< 2.0


11.0 mmol/mol*

50 -60

Describe the abnormalities in the above results, giving likely explanations             (30% marks)

[Click here to toggle visibility of the answers]

College answer

  • Normal A-a gradient (20) for 70 year old – hypoventilation
  • Respiratory acidosis – acute type 2 respiratory failure tiring from acute asthmatic attack
  • Normal anion gap metabolic acidosis  – underlying type IV RTA secondary to diabetic  nephropathy (or any reasonable cause)
  • Hyperlactataemia – beta-2 agonist induced
  • Hyperkalaemia and raised creatinine- renal impairment
  • Poor diabetic control (high BSL and HbA1C) 


Let us dissect this systematically;

  1. The patient is hypoxic (PaO2 of 55 mmHg) but the A-a gradient is minimally raised (~ 19) which suggests that alveolar hypoventilation is the main cause of the hypoxia.
  2. There is acidaemia
  3. The CO2 is contributing to the acidaemia
  4. The SBE is -4, suggesting that there is also a mild metabolic acidosis
  5.  The urea and creatinine are slightly raised, reflecting the possibility that in this asthma episode the patient has been too breathless to eat and drink normally.
  6. The lactate is raised, suggesting that salbutamol is being administered (or else, that hypovolaemia is quite severe)
  7. The glucose is slightly elevated and the HbA1c is raised, suggesting that this patient is a diabetic and their attention to their BSl control is patchy and haphazard
  8. Now, to the acid-base disturbance. The anion gap is (135) - (100 + 23) = 12, or 17.3 when calculated with potassium. Thus, this is a normal anion gap acidosis.

Let's explore that anion gap reference value. 

According to the RCPA, the anion gap range with potassium is 8-16, i.e. you'd take 12 as the middle reference value for delta ratio calculations. Without potassium the range is 4-13, i.e. the reference value would be 8.5. Unfortunately the RCPA use Sirker et al (2002) as their reference, an article which completely ignores this issue and gives no reference ranges. So, I am not sure where they got their numbers from; I only used them because of their relatively weighty authority (Royal College, etc).

However, they seem legit: the change in reference values from the higher values (16 with potassium, 12 without) is due to a change in measurement methodology and laboratory reference ranges which appears to have occurred in the late 1980s-early 1990s (Winter et al, 1990).  The reference ranges are always based on healthy volunteers who should have no acid-base disturbances, and those guys have not changed, but now we tend to use ion-selective electrodes instead of older photometric methods, a practice which has shifted the reference range for the anion gap into a lower range (mainly because of a drift in the chloride measurements).  There’s a few studies reporting this change in the last 20 years (and it’s always different, 5-10 mmol/L, or 3-11 mmol/L or, 4-12 mmol/L). A representative paper is Lolekha et al (2001) who got a range of 5-12 mmol/L.

However, this knowledge is of absolutely no use to the CICM trainee, because the college examiners continue using the pre-1990s reference ranges. 

If we use the (slightly different) modern reference ranges, we get significantly different results, because the numbers involved here are also quite small, near the borders of normality (obviously the change in reference ranges is going to play a minimal role whenever the acid-base disturbance is profound and obvious). Let's use the with-potassium formula for the anion gap. We get an anion gap of 17.3, and assuming the albumin is normal we would expect a normal value of 12, which means it has risen by 5.3.  The delta ratio is therefore 5.3 / 1.0 = 5.3, i.e it points to a co-existing metabolic alkalosis.   If you omit the use of potassium in the anion gap equation you get an anion gap of 12; with an expected normal anion gap value of  8.5 the delta ratio is still 3.5. 



Sirker, A. A., et al. "Acid− base physiology: the ‘traditional’and the ‘modern’approaches." Anaesthesia 57.4 (2002): 348-356.

Lolekha, Porntip H., Somlak Vanavanan, and Somsak Lolekha. "Update on value of the anion gap in clinical diagnosis and laboratory evaluation." Clinica chimica acta307.1-2 (2001): 33-36.

Winter, Sara D., et al. "The fall of the serum anion gap." Archives of internal medicine 150.2 (1990): 311-313.