Hypotension prediction in elective caesarean section: a reply
Year: 2026
Type: article
Abstract: We thank O'Sullivan and Bartlett for their comments [1] on our study [2]. Although prophylactic vasopressor infusions are recommended, recent meta-analyses have reported an increased risk of maternal reactive hypertension and bradycardia associated with this strategy [3]. Moreover, designing an appropriate protocol to compare hypotension prediction index-guided management with continuous blood pressure monitoring in the context of vasopressor infusion remains challenging. At present, no studies have established an optimal prediction index protocol for titrating vasopressor infusions during caesarean delivery. Implementing this for comparison with a prophylactic vasopressor infusion was therefore not considered. In contrast, vasopressor bolus administration is effective and familiar to most. We are currently conducting further investigations comparing prediction index-guided treatment with a noradrenaline infusion protocol during caesarean delivery. Regarding the interval for oscillometric blood pressure measurements, our institutional protocol uses a longer interval than the 1-min interval recommended by many centres. This practice reflects frequent reports of maternal discomfort and difficulty holding a newborn due to repeated cuff inflation [4]. We nonetheless observed less hypotension in the prediction index group compared with the continuous non-invasive arterial pressure group. Although a shorter oscillometric interval might have improved performance, it would not explain the lower hypotensive burden observed in the prediction index group. Regarding the definition of hypotension and treatment targets, this represents an inherent limitation in the clinical application of the prediction index, as the algorithm predicts a fixed mean arterial pressure threshold rather than personalised blood pressure targets. We agree that relative systolic blood pressure targets are clinically important in obstetric anaesthesia. We selected a mean arterial pressure < 65 mmHg because this threshold is linked to the algorithm and therefore allowed internal methodological consistency. This should not be interpreted as evidence that a fixed mean arterial pressure target is superior to individualised systolic targets in obstetric practice. However, previous studies have shown that a mean arterial pressure < 65 mmHg is associated with increased risk of organ hypoperfusion and postoperative complications in patients having non-obstetric surgeries [5]. We agree that continuous monitoring alone confers substantial benefit and that predictive information may provide additional haemodynamic refinement. Our study was powered to detect differences in time-weighted average hypotension rather than adverse event incidence, which would require a substantially larger sample size. A common concern regarding predictive algorithm-based haemodynamic management is the potential risk of overtreatment, but this phenomenon was not observed in our study. One possible explanation is the high prevalence of hypotension during caesarean delivery, which may increase the positive predictive value of the prediction index [6]. Baseline systolic blood pressure was comparable between groups suggesting that baseline haemodynamic differences did not bias our results. We agree that the absence of umbilical cord blood gas analysis limits the ability to infer fetal perfusion and this limitation was acknowledged. We agree that automation represents an important future direction. Closed-loop vasopressor systems are highly promising and a prediction index-informed closed-loop strategy may represent a logical next step in proactive obstetric haemodynamic management. However, development of a reliable closed-loop protocol requires accumulation of evidence from studies across different clinical protocols in caesarean delivery. Our study represents an important initial step towards establishing improved haemodynamic management strategies for caesarean delivery.
Source: Anaesthesia
DOI: 10.1111/anae.70217
Authors: Po‐Yuan Shih, Chun‐Yu Wu
Institution: National Taiwan University Hospital
Language: English
Subfield: Surgery
Field: Medicine
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W7156808596