= Esophageal pressure measurement
\b[Esophageal pressure measurement] records pressure in the <oesophagus>, commonly with an <esophageal balloon>, to obtain information about respiratory mechanics and <pleural pressure>. The measured signal is a local surrogate. It is not a direct map of pressure everywhere in the <pleural cavity>.
A primary validation experiment by <Ahmet Baydur> and collaborators compared changes in esophageal and mouth pressure during end-expiratory airway occlusion and spontaneous efforts. Ten subjects were studied in different postures; some discrepancies changed when the balloon was repositioned.<A Simple Method for Assessing the Validity of the Esophageal Balloon Technique by Ahmet Baydur and others>[Abstract: subjects, posture, occlusion and repositioning]{ref} The <esophageal balloon occlusion test> therefore evaluates pressure-change transmission under specified conditions, rather than proving all absolute-pressure values correct.
In an ideal local model, write $P_{\rm es}=aP_{\rm pl}+b$. If $a$ and $b$ remain fixed during an effort, then
$$
\frac{\Delta P_{\rm es}}{\Delta P_{\rm mouth}}\approx a
$$
when mouth and pleural pressure changes coincide under the test's assumptions. A ratio near one supports transmission gain near one, but leaves $b$ unidentified. Even perfect change agreement cannot determine that constant baseline offset.
The consequence matters when calculating <transpulmonary pressure> $P_L=P_{\rm alv}-P_{\rm pl}$. For an illustrative alveolar pressure of 15 and true local pleural pressure of $7\,\mathrm{cm\,H_2O}$, the difference is eight. Substituting an esophageal reading of nine with an uncorrected offset of two gives six instead. The arithmetic identifies an error mechanism; it does not supply a clinical correction or assume that all pleural regions have pressure seven.
<Instrument calibration>, balloon coupling, body position and the physiological assumptions of the comparison address separate sources of error. Reporting the waveform without those conditions can conceal which inference was actually tested.
Back to article page
