Purpose: To assess whether invasive and non-invasive blood pressure (BP) monitoring allows the identification of patients who have responded to a fluid challenge, i.e., who have increased their cardiac output (CO).
Methods: Patients with signs of circulatory failure were prospectively included. Before and after a fluid challenge, CO and the mean of four intra-arterial and oscillometric brachial cuff BP measurements were collected. Fluid responsiveness was defined by an increase in CO ≥10 or ≥15% in case of regular rhythm or arrhythmia, respectively.
Results: In 130 patients, the correlation between a fluid-induced increase in pulse pressure (Δ500mlPP) and fluid-induced increase in CO was weak and was similar for invasive and non-invasive measurements of BP: r² = 0.31 and r² = 0.29, respectively (both p < 0.001). For the identification of responders, invasive Δ500mlPP was associated with an area under the receiver-operating curve (AUC) of 0.82 (0.74-0.88), similar (p = 0.80) to that of non-invasive Δ500mlPP [AUC of 0.81 (0.73-0.87)]. Outside large gray zones of inconclusive values (5-23% for invasive Δ500mlPP and 4-35% for non-invasive Δ500mlPP, involving 35 and 48% of patients, respectively), the detection of responsiveness or unresponsiveness to fluid was reliable. Cardiac arrhythmia did not impair the performance of invasive or non-invasive Δ500mlPP. Other BP-derived indices did not outperform Δ500mlPP.
Conclusions: As evidenced by large gray zones, BP-derived indices poorly reflected fluid responsiveness. However, in our deeply sedated population, a high increase in invasive pulse pressure (>23%) or even in non-invasive pulse pressure (>35%) reliably detected a response to fluid. In the absence of a marked increase in pulse pressure (<4-5%), a response to fluid was unlikely.