Proactive haemodynamic management using the hypotension prediction index during caesarean section: a randomised controlled study

Citation:

Shih P-Y, Wei T-J, Lee C-T, Kang J, Shih M-C, Chen Y-H, Lee Y-S, Chen H-T, Wu C-Y. Proactive haemodynamic management using the hypotension prediction index during caesarean section: a randomised controlled study. Anaesthesia. 2026. doi:10.1111/anae.70161.

Study at a glance

- Design and setting: Single-centre, 3-arm randomised controlled trial (1:1:1) during elective caesarean section under spinal anaesthesia at National Taiwan University Hospital (Taipei, Taiwan). Screened 255; randomised 180; analysed 171.

- Population: Adults aged 20–50 years undergoing elective caesarean section under spinal anaesthesia. Key exclusions: pre-eclampsia; significant cardiovascular disease; inadequate spinal level (≤T6).

- Groups / monitoring strategies:

- Oscillometric group: Oscillometric BP every 2.5 min; CNAP/HPI recorded but blinded to anaesthetist.

- CNAP group: Continuous non-invasive arterial pressure (CNAP) displayed; HPI blinded.

- HPI group: CNAP + HPI displayed; treatment triggered by MAP <65 mmHg OR HPI ≥85 (even if MAP ≥65).

- Co-interventions (all groups): 1000 ml crystalloid co-load; hypotension treated with intermittent noradrenaline boluses 5–10 µg (as often as every minute); atropine 0.5 mg IV to keep HR >60 bpm.

Primary outcome

- Time-weighted average hypotension (MAP <65 mmHg), derived from CNAP MAP sampled every 20 s:

- Oscillometric: 0.89 mmHg (IQR 0.08–2.06), n=59

- CNAP: 0.30 mmHg (IQR 0.07–0.74), n=55

- HPI: 0.08 mmHg (IQR 0–0.43), n=57

→ Descriptively lowest hypotension burden with HPI-guided management.

Selected secondary / safety and process outcomes (Table 2)

- Area under the threshold for hypotension (mmHg·min):

- Oscillometric 68.7 (IQR 5.5–145.3) vs CNAP 22.7 (4.8–57.2) vs HPI 5.7 (0–32.7)

- Zero episodes of hypotension (no MAP <65 at all):

- 13/59 (22%) oscillometric vs 13/55 (24%) CNAP vs 26/57 (46%) HPI; p=0.009

- Noradrenaline dose (µg):

- 50 (30–90) oscillometric vs 70 (30–145) CNAP vs 90 (48–133) HPI; p=0.016 (higher in HPI arm)

- Maternal adverse effects:

- Nausea: 43/59 vs 29/55 vs 30/57; p=0.038

- Vomiting: 13/59 vs 6/55 vs 3/57; p=0.023

- Bradycardia: 10/59 vs 5/55 vs 4/57; p=0.199

- Neonatal outcomes: Apgar scores similar across groups (median 9 at 1 min and 5 min).

- Device agreement (exploratory): Bland–Altman CNAP vs oscillometric MAP showed small mean bias (+1.37 mmHg; 95% CI 0.71–2.03).

Important limitations / interpretation notes

- Post-randomisation exclusions: 9/180 (5%) were not analysed (protocol violation with ephedrine n=3; inadequate spinal level not reaching T6 n=6), so results reflect a modified analysed cohort rather than strict ITT for all randomised patients.

- Many outcomes are presented as medians/IQRs and p-values without effect sizes/CI for between-group differences; multiplicity adjustment is not described (many secondary comparisons in a 3-arm trial).