An apneic, hypothermic newborn with glucose of 28 mg/dL and ammonia of 850 μmol/L needs resuscitation before a metabolic label. Yet the blood gas can immediately shift the differential. The board shortcut—ketones mean organic acidemia, no ketones mean urea cycle defect—is too brittle. Read pH, bicarbonate, respiratory compensation, anion gap, glucose, lactate, and ketones together.
Consider a term infant on day 3 of life, recently discharged, who returns with poor feeding, lethargy, hypotonia, and apnea. After initial airway support, the gas shows pH 7.15, PCO₂ 30 mmHg, and bicarbonate 10 mmol/L; lactate is 6 mmol/L. The ammonia is markedly elevated. This pattern raises concern for an inborn error of metabolism, but it does not establish a specific diagnosis. The sodium and chloride are not provided, so calculate the anion gap from the measured chemistry rather than assuming it is high.
Read the gas before reaching for a label
A pH of 7.15 is acidemia, and a bicarbonate of 10 mmol/L points to a metabolic acidosis. With that bicarbonate, Winter’s estimate gives an expected PCO₂ of about 23 ± 2 mmHg. A measured PCO₂ of 30 is higher than expected, suggesting inadequate respiratory compensation or an additional respiratory acidosis—not respiratory alkalosis.
That distinction matters. An early urea cycle disorder (UCD) crisis may produce hyperventilation and respiratory alkalosis. But apnea or worsening encephalopathy can change the respiratory pattern, and poor perfusion or infection can add acidosis. A later gas should not be treated as if it captures the infant’s earliest physiology.
Lactate also needs context. A level of 6 mmol/L may contribute to an increased anion gap during shock or hypoxia; it is not, by itself, proof of organic acidemia. Check perfusion and interpret the lactate alongside the gap, ketones, glucose, and the infant’s clinical course.
Compare the patterns, not one test
| Pattern | What it favors | What it cannot settle |
|---|---|---|
| Increased anion gap metabolic acidosis with ketones and hyperammonemia | An organic acidemia, such as methylmalonic or propionic acidemia | Ketones alone do not identify the disorder; lactate from shock may also raise the gap |
| Marked hyperammonemia with early respiratory alkalosis and no primary increased-gap acidosis | A UCD | The gas may change as illness progresses; sepsis or shock can add metabolic acidosis |
| Hypoglycemia with inappropriately few ketones | A fatty-acid oxidation or other fuel-use disorder | Hypoglycemia alone does not distinguish among metabolic conditions |
| Lactic acidosis, temperature instability, and poor perfusion | Sepsis or another cause of shock remains possible | Sepsis can mimic, trigger, or coexist with an inborn error of metabolism |
Ketonuria is a useful clue in a sick newborn: it commonly accompanies methylmalonic and propionic acidemias during decompensation, along with anion-gap acidosis and hyperammonemia. But ketones reflect the infant’s metabolic state; they are not a stand-alone diagnostic test. A positive result supports the pattern, while an absent result should not overrule the rest of the evidence.
Likewise, respiratory alkalosis with hyperammonemia makes a UCD more likely, but it is supportive rather than required. Metabolic acidosis with an increased gap often points elsewhere, yet it does not exclude a UCD when shock or infection is also present. The exam-ready distinction is a shift in probability, not a rule that makes either diagnosis impossible.
Let the pattern direct testing—without slowing care
An ammonia of 850 μmol/L is a metabolic emergency in its own right. Activate the local hyperammonemia protocol, involve metabolic, critical-care, and nephrology teams, begin specialist-directed ammonia-lowering treatment, and urgently assess and prepare for extracorporeal ammonia clearance. A properly collected repeat ammonia can be obtained promptly if needed, but treatment and transfer planning should not wait for repeat confirmation or diagnostic profiles.
For this infant, manage airway, glucose, and perfusion while treating neonatal sepsis as a live possibility. Cultures and empiric antimicrobial decisions should proceed in parallel with the metabolic evaluation when clinically indicated. An inborn error of metabolism does not make infection less urgent.
Contact a metabolic specialist or the receiving metabolic center early. If it does not delay stabilization or treatment, collect acute samples for plasma amino acids, plasma acylcarnitines, urine organic acids, urine orotic acid, and urine ketones. A urine organic-acid profile and acylcarnitine profile help identify organic acidemias; plasma amino acids and urine orotic acid help characterize UCDs. The results should be interpreted together, not as isolated yes-or-no tests.
Do not wait for a newborn-screen result before escalating a critically ill infant. Severe disease may become symptomatic before results are available, and newborn screening does not detect every UCD. If the sample handling or test sequence is uncertain, ask the metabolic team—but do not let that conversation postpone resuscitation, consultation, or transfer planning.
Common traps worth correcting
- Treating ketonuria as the answer: It supports organic acidemia in the right setting, but does not name the disorder or replace the full biochemical pattern.
- Assuming any metabolic acidosis excludes a UCD: Shock, infection, and late deterioration can alter the acid–base picture.
- Calling a gas respiratory alkalosis because the ammonia is high: Read the pH, bicarbonate, and PCO₂ together; apnea can produce a mixed or changing pattern.
- Attributing every abnormality to sepsis: Sepsis remains urgent, but it does not close the metabolic differential.
Practical takeaways
- In neonatal hyperammonemia, combine acid–base status, the calculated anion gap, ketones, glucose, and lactate.
- Increased-gap metabolic acidosis with ketones favors organic acidemia; early respiratory alkalosis with hyperammonemia favors a UCD.
- Neither ketonuria nor a single blood gas is definitive, especially in an infant with apnea, shock, or possible infection.
- Send targeted metabolic studies and involve specialists while resuscitation and sepsis care proceed—not after they are complete.
