BACKGROUND: Dysfunctional cerebral blood flow autoregulation plays a crucial role in the secondary damage after traumatic brain injury. The pressure reactivity index (PRx) can be used to monitor dynamic cerebral blood flow autoregulation indirectly.
OBJECTIVE: To test different versions of the long pressure reactivity index (LPRx), which is based on minute-by-minute data and calculated over extended time windows, and to study their predictive ability and examine whether "long" and "short" pressure reactivity indices could improve predictive power.
METHODS: PRx and 3 versions of the LPRx calculated over 20-, 60-, and 240-min time windows were assessed in relation to outcome at 6 mo in 855 patients with traumatic brain injury. Predictive power and discriminative ability of indices were evaluated using area under the operator curves and determination of critical thresholds. PRx and LPR indices were combined to evaluate whether LPR indices could improve outcome prediction by adding information about static components of autoregulation.
RESULTS: Correlation of each LPRx with the PRx decreased with increased time windows. LPR indices performed successively worse in their predictive and discriminative ability from 20-min to 240-min time frames. PRx had a significantly higher predictive ability compared to each LPRx. Combining LPRx and PRx did not lead to an improvement of predictive power compared to the PRx alone.
CONCLUSION: The critical threshold and predictive value of the PRx for unfavorable outcome and mortality have been confirmed in one of the largest so far published patient cohorts. LPRx performed significantly worse, and its discriminative and predictive abilities decreased with an increasing calculation window.