Associations between reduction of hypotension and treatment intensity in hypotension prediction index studies

Abstract

BACKGROUND: The Hypotension Prediction Index (HPI) is developed for proactive prevention of hypotension and existing meta-analyses report reduced hypotension with HPI-guided management in comparison with a strategy based on MAP alerts of 65 mmHg. Still, validation studies suggest that HPI does not outperform simple mean arterial pressure (MAP)-based predictions and report that HPI alerts appear when MAP is ~70-75 mmHg. Existing HPI trials may therefore effectively compare two MAP-treatment thresholds that could be accompanied by different treatment intensity. However, meta-analyses do not consistently report treatment intensity differences. We hypothesized that reductions in hypotension are associated with greater treatment intensity. METHODS: Randomized trials evaluating HPI-guided hypotension prevention were identified from existing systematic reviews and citation tracking. Eligible trials were assessed for (1) statistically significant reduction in the primary hypotension outcome and (2) statistically significant increases in any hemodynamic treatment (fluids, vasopressors, or inotropes) in the HPI intervention group. We examined the association between hypotension reduction and treatment intensity using 2×2 tabulation and Fisher's exact test. Furthermore, for trials reporting time-weighted average (TWA) of MAP< 65 mmHg, we calculated the correlation between the TWA difference and treatment intensity categories (lower, neutral, or increased). RESULTS: Twenty randomized trials (2342 patients) were identified of which 18 reported both hypotension and treatment outcomes. Among 13 trials demonstrating hypotension reduction, 9 (69%) reported significantly greater treatment intensity in the HPI intervention group. In contrast, none of the five trials that failed to reduce hypotension reported increased treatment intensity. Reduction of hypotension was significantly associated with greater treatment intensity (p = 0.029). TWA differences were significantly correlated with treatment intensity (p < 0.001). CONCLUSIONS: Across randomized HPI trials, reduced hypotension was frequently accompanied by increased treatment intensity. These associations suggest that successful reduction of hypotension in HPI trials generally occurred together with more intensive hemodynamic treatment.