Recognition of esophageal intubation in Dutch prehospital emergency medical services using capnography analysis: a retrospective cohort study

Aim of this study Unrecognized esophageal intubation during prehospital cardiopulmonary resuscitation (CPR) leads to inadequate oxygenation and ventilation and often results in death. This study describes and analyzes, using mathematical calculations, how esophageal capnography differs from normal capnography in out-of-hospital cardiac arrest (OHCA) patients, in order to assess whether these differences could support future detection. Methods We conducted a retrospective analysis using the Rotterdam Helicopter Emergency Medical Service (HEMS) database from January 2015 to November 2024. Cases of unrecognized esophageal intubation after ground EMS airway management were identified and analyzed for capnography waveform patterns before and after HEMS reintubation. HEMS data were supplemented with ambulance and monitor data from the Corpuls Manager database. Capnography waveforms were manually annotated and analyzed for CO 2 levels, trend, area under the curve, and upslope. Results Out of 5830 cases with intubation prior or during HEMS deployment, 92 were identified as esophageal intubations, with 14 cases involving OHCA in our region of interest with usable capnography. The median maximal esophageal CO 2 was 2.5kPa, and the median maximal tracheal CO 2 was 10.8kPa. Esophageal intubations showed a downward trend approaching near-zero values of CO 2 within eight ventilations. Significant differences were found in the median corrected area under the curve and median upslope between esophageal and tracheal intubations. Conclusion This study uses analysis of waveform morphology to distinguish esophageal from tracheal intubation. While etCO 2 values can appear numerically similar, the waveform morphology differs significantly. These findings represent an important first step toward developing automated tools for detecting inadequate ventilation by esophageal intubation. Larger datasets and prospective validation are needed before clinical implementation.