Shih et al. evaluated hypotension prediction index-guided haemodynamic management during elective caesarean section under spinal anaesthesia 1. They reported a statistically significant reduction in time-weighted average hypotension as compared with oscillometric monitoring and continuous non-invasive arterial pressure. However, several methodological considerations merit discussion before widespread adoption can be recommended. Contemporary consensus guidance recommends routine, preferably prophylactic, vasopressor administration following spinal anaesthesia for caesarean delivery, with a variable-rate infusion commenced immediately after intrathecal injection and minute-by-minute non-invasive blood pressure monitoring to maintain systolic arterial pressure ≥ 90% of baseline 2. Meta-analysis shows that prophylactic phenylephrine infusion reduces hypotension, nausea and vomiting as compared with reactive bolus therapy 3. More recent data suggest that noradrenaline as an infusion provides greater haemodynamic stability than bolus dosing 4. In contrast, clinicians in the present study were restricted to intermittent 5–10 μg noradrenaline boluses without prophylactic infusion. This approach is unlikely to reflect contemporary practice in many centres. It therefore remains unclear whether hypotension prediction index-guided haemodynamic management is superior to an optimised prophylactic infusion strategy, rather than simply to reactive bolus therapy. The oscillometric group was monitored at 2.5-min intervals, whereas many departments employ 1-min blood pressure measurements immediately post-spinal to mitigate the lag inherent in intermittent monitoring. When hypotension is defined as mean arterial pressure < 65 mmHg, treatment triggered at that threshold may already be temporally behind the physiological nadir. Targeting a slightly higher mean arterial pressure to create a buffer might have reduced the incidence of hypotension in patients allocated to the oscillometric group. The longer measurement interval may therefore exaggerate the apparent inferiority of intermittent non-invasive blood pressure measurement. The definition of hypotension also warrants reflection. Obstetric consensus guidance emphasises relative changes from baseline systolic arterial pressure rather than fixed absolute mean arterial pressure thresholds 2. Maintaining systolic arterial pressure ≥ 90% of baseline acknowledges that perfusion requirements vary between individuals. A uniform mean arterial pressure target does not account for women with higher baseline pressures, chronic hypertension or pre-eclampsia, in whom organ and uteroplacental perfusion may depend on higher driving pressures. Conversely, healthy women with lower baseline pressures may tolerate values <65 mmHg without adverse consequence. Blood pressure goals in obstetric patients should therefore be individualised rather than uniform. With respect to clinical outcomes, maternal nausea and vomiting were reduced when compared with patients allocated to the oscillometric group, yet no significant difference was observed between patients allocated to the continuous non-invasive and hypotension prediction groups. This raises an important question: does continuous monitoring alone confer most of the benefit, without the added complexity and cost of a predictive algorithm? Neonatal outcomes, assessed by Apgar scores, were similar across groups. In the absence of umbilical cord blood gas analysis, it is difficult to conclude that reduced time-weighted hypotension translated into improved fetal perfusion 5. If neonatal outcomes remain unchanged, the clinical significance of modest reductions in mild hypotension becomes less certain. Finally, while predictive monitoring is attractive conceptually, its future may lie in automation. Consensus guidance describes computer-controlled, closed-loop vasopressor systems combined with continuous monitoring show promising precision in blood pressure control and favourable outcomes 2. A hypotension prediction index-driven closed-loop system may represent a more transformative advance than decision-support alone.
O'Sullivan et al. (Tue,) studied this question.