Respiratory failure & ventilation
Assessing the accuracy of ChatGPT in interpreting blood gas analysis results ChatGPT-4 in blood gas analysis
Authors' conclusion
ChatGPT-4 demonstrated potential as a supplementary tool for ABG interpretation in high-demand clinical settings, supporting rapid, reliable decision-making.
Study context
- Population
- Arterial blood gas (ABG) samples from ICU patients
- Participants
- 400 ABG samples
- Exposure
- ChatGPT-4 interpretation of ABG results assessed by experienced anesthesiologists
- ChatGPT-4n = 400
- Outcome
- Accuracy of ChatGPT-4 in interpreting ABG parameters (potential diagnoses and treatment)
Methodological detailsTwo anesthesiologists independently evaluated the model's predictions
- Two anesthesiologists independently evaluated the model's predictions
Main finding
As reported in the abstract
Accuracy by ABG parameter
| Group | Accuracy pH, oxygenation, sodium, chloride | Accuracy hemoglobin | Accuracy bilirubin |
|---|---|---|---|
| ChatGPT-4 | 100 % | 92.5 % | 72.5 % |
ChatGPT-4
Accuracy pH, oxygenation, sodium, chloride100 %
Accuracy hemoglobin92.5 %
Accuracy bilirubin72.5 %
- Overall performance statistically significant across most parameters
- P < 0.05
Observational data: associations, not effects.
Safety
Treatment recommendations
- Unnecessary bicarbonate treatment recommended in several cases
Original publication
Assessing the accuracy of ChatGPT in interpreting blood gas analysis results ChatGPT-4 in blood gas analysis. J Clin Anesth. Published 21 February 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →