Methodological Definition and Validation of the Intracranial Pressure-End-Tidal CO2 Slope (IESlope): Characterization...

The metabolic component of cerebral autoregulation is crucial in traumatic brain injury (TBI), yet continuously monitored indices are limited. This study introduces the ICP-EtCO2 Slope (IESlope), quantifying the dynamic relationship between end-tidal CO2 (EtCO2) and intracranial pressure (ICP), and evaluates its association with 12-month mortality and unfavorable outcome (Glasgow Outcome Scale-Extended, GOSE≤4). Second, IESlope visualization and its ability to predict short-term ICP changes during EtCO2 variations are assessed.

Intensive care unit records of 218 adult and pediatric TBI patients were retrospectively analyzed (median age 44, IQR: 24-66; 76% male). IESlope was calculated as the angle of the linear regression between EtCO2 and ICP over 60-minute moving windows and visualized using streamline mapping (ICP-EtCO2 space). IESlope values were averaged to obtain a single patient-level metric used in univariate and multivariable analyses (area under the curve, AUC). Predictive capability was assessed by comparing predicted versus observed ICP changes at 5, 10, and 20 minutes during significant EtCO2 variations.

Lower IESlope values were associated with mortality and GOSE≤4 (median: 7.76 vs. 14.8 and 10.5 vs. 18.6, respectively, P<0.001), independent of age or decompressive craniectomy. Streamline maps showed expected CO2 reactivity patterns, with peak values at moderate ICP and EtCO2 levels and reductions at extremes and after decompressive craniectomy. IESlope maintained significant associations with outcomes (AUC ∼0.900) and accurately predicted short-term ICP responses to EtCO2 changes (mean absolute error, MAE=0.600-1.686 mm Hg).

IESlope provides a continuous, quantitative measure of CO2-related ICP reactivity and may support individualized ventilatory management in TBI, pending prospective validation.