# Predictive Hypoxemic Threshold for Tolerating the Apnea Test While Assessing Death by Neurological Criteria

## TL;DR

This study identifies a pre-apnea test oxygen level threshold that can predict whether a patient will tolerate the test safely during brain death evaluation.

## Contribution

The study introduces a predictive PaO2 threshold of 300 mmHg to help avoid dangerous oxygen drops during the apnea test.

## Key findings

- Patients with pre-AT PaO2 ≤ 60 mmHg had significantly lower initial PaO2 (243.7 mmHg) compared to others (374.8 mmHg).
- Pre-AT PaO2 showed good predictive ability with an ROC AUC of 0.76 for low post-AT PaO2.
- A pre-AT PaO2 cutoff of ~300 mmHg may help prevent oxygen saturation from dropping below 90%.

## Abstract

The apnea test (AT) plays a vital role in diagnosing brain death by evaluating the absence of spontaneous respiratory activity. It entails disconnecting the patient from mechanical ventilation to raise the CO2 partial pressure and lower the pH. Occasionally, the AT is aborted because of safety concerns, such as hypoxemia and hemodynamic instability, to prevent worsening conditions. However, the exact oxygen partial pressure level needed before commencing AT, indicating an inability to tolerate the test, is still uncertain. This study seeks to determine pre-AT oxygen levels linked with a heightened risk of test failure.

We conducted a retrospective cohort study involving patients suspected of having brain death at the Tel Aviv Medical Center from 2010 to 2022. The primary outcome was defined as an arterial partial O2 pressure (PaO2) level of 60 mmHg or lower at the conclusion of the AT. This threshold is significant because it marks the point at which the saturation curve deflects, potentially leading to rapid deterioration in the patient’s oxygen saturation.

Among the 70 patients who underwent AT, 7 patients met the primary diagnostic criteria. Patients with a PaO2 ≤ 60 mmHg at the conclusion of the AT exhibited a significantly lower initial median PaO2 of 243.7 mmHg compared with those with higher pre-AT PaO2 levels of 374.8 mmHg (interquartile range 104.65–307.00 and interquartile range 267.8–444.9 respectively, P value = 0.0041). Pre-AT PaO2 levels demonstrated good discriminatory ability for low PaO2 levels according to the receiver operating characteristic (ROC) curve, with an area under the curve of 0.76 (95% confidence interval 0.52–0.99).

PaO2 values at the conclusion of the AT are closely associated with PaO2 values at the beginning of the test. Establishing a cutoff value of approximately 300 mmHg PaO2 at the onset of AT may assist in avoiding saturation drops below 90%.