Septic shock resuscitated with normal saline
This 65 year old gentleman presented to ED with fever and productive cough. For several days he has been too short of breath to eat or drink, and appeared severely dehydrated at triage. The chest Xray revealed an extensive right lower lobe pneumonia. BP on presentation was 78/40, and MAP remained below 60 in spite of large volumes of crystalloid. The ICU team were contacted after the seventh litre of fluid had not only failed to produce a sustained hemodynamic effect, but also catalysed an episode of acute pulmonary oedema. This ABG was taken while the patient was still conscious, being pre-oxygenated for intubation with a Laerdel bag-mask device.
Data to be interpreteted
Interpretation
Discussion and Interpretation
Assessment of tension-based and content-based oxygenation indices
| Alveolar oxygen | 650.25 |
| A-a gradient | 326.25 |
| a/A ratio | 0.50 |
| PaO2/FiO2 ratio | 324 |
Though the patient has a 99% oxygen saturation, there is obviously a serious problem here. The a/A ratio is around 0.5, which means that only half of the administered oxygen is getting into the arterial circulation; if the alveolar oxygen was 100mmHg (as in room air) the PaO2 would probably be 50mmHg. Note that the A-a gradient reflects an oxygenation defect, but as far as the PaO2/FiO2 ratio is concerned everything is just fine and there is no problem.
The change in pH
There is acidaemia; the pH is 7.202.
The change in pCO2
The pCO2 is high, likely contributing to the acidosis.
However, it is only raised by about 10mmHg. This means there will be little difference between the two different ways of estimating what the pH should be. If we use the flawed "0.008" rule to estimate the expected pH, we arrive at a pH value of 7.318. If we calculate the "expected" pH from the Henderson-Hasselbalch equation, using a predicted bicarbonate level (using the "1 for 10" bedside rule), the pH we get a pH of 7.317. The two methods vary only in the bicarbonate value they use: the "0.008" rule uses a bicarbonate level of 24mmol/L, whereas the Henderson-Hasselbalch equation uses an estimated bicarbonate level from the "1 for 10" rule. That rule estimates an expected bicarbonate of 25mmol/L; with such a small difference, both equations yield very similar results.
In any case, the measured pH is lower, making you think that some sort of metabolic acidosis is also present.
The change in Base Excess
The Actual Base Excess is strongly negative, suggesting a metabolic acidosis.
Assessment of compensation
Copenhagen interpretation of acid-base compensation:
With this slightly alkalotic ABE, one predicts that there should be no change to PaCO2- i.e. it should be normal. Since the measured PaCO2 is high, there must be a respiratory acidosis.
Boston interpretation of acid-base compensation:
Note that this ABG machine reports the actual bicarbonate rather than the standard bicarbonate, which saves the Boston supporter from having to calculate the actual bicarbonate themselves. The actual bicarbonate for this scenario is 19mmol/L.
Using the "1.5 plus 8" rule, the expected CO2 for this degree of metabolic acidosis is 36.5 mmHg; the actual CO2 is substantially higher, suggesting that there is a coexisting respiratory acidosis.
Assessment of the metabolic component of acidosis
The anion gap is 14.2
The albumin was 36. With this value, the "normal" anion gap should be 11.
The delta ratio is therefore 0.64
This suggests that some of the metabolic acidosis is due to the raised lactate, and some is due to the raised chloride. Close scrutiny of the fluid chart has revealed that most of the crystalloid used in this resuscitation was normal saline.
Assessment of oxygen-hemoglobin dissociation mechanics
There is an abnormally raised p50; the right shift can be accounted for by the acidosis and hypercapnea, given that the dyshaemoglobin levels are within the expected range. The fever was probably not contributing- at this stage, the patient had cooled down to 37.0°C.
Additional remarks
This is a mixed metabolic acidosis and acute respiratory acidosis with acute hypoxia. The effort of moving gas in and out of that pus-filled lung has exhausted the patient, and pulmonary compliance continued to decrease due to excess alveolar fluid contributed by the resuscitation process.
In addition to the lactic acidosis, extra chloride contributed by saline has produced a normal anion gap acidosis, which would not have been an issue had the resuscitation fluid choices been different. Perhaps the responsiveness to endogenous catecholamines would have been preserved had they used Hartmanns or Plasmalyte. With intact catecholamine receptor-ligand dynamics, less fluid might have been required to achieve the same haemodynamic goals, and intubation might have been averted, sparing this gentleman the Gigeresque horror of ICU admission.