Accuracy of cuffless blood pressure devices compared with invasive arterial line monitoring in high-acuity settings: a random-effects meta-analysis
A Malik, R De Monte Furtado, R Sukumaran, Y Charavanamuttu, U Khan, S Pradeep Kundur, B Young, K CivelloAbstract
Background
Cuffless blood pressure (BP) monitoring technologies are increasingly proposed for continuous and remote cardiovascular assessment. However, validation against the highest physiological reference standard, invasive arterial line (ART) monitoring, remains limited and has not been systematically synthesised.
Purpose
To perform a pooled analysis of studies evaluating cuffless BP devices against invasive ART measurements in acute care settings.
Methods
As part of an ongoing systematic review of cuffless BP technologies, we performed a predefined subgroup analysis of studies directly comparing cuffless devices with invasive ART monitoring. Eligible studies were conducted in intensive care unit (ICU) or immediate post-operative settings and reported within-patient mean differences (MD) with dispersion measures. Where paired standard deviations were not explicitly reported, they were derived from Bland–Altman limits of agreement, and corresponding standard errors were calculated as SD/√n. Random-effects meta-analyses were performed separately for systolic BP (SBP) and diastolic BP (DBP) using restricted maximum likelihood (REML) with Hartung–Knapp adjustment. Risk of bias was assessed using the QUADAS-2 tool. The systematic review was registered in PROSPERO (CRD420261326740).
Results
Four studies (n = 164) conducted in ICU or immediate post-operative settings with continuous monitoring (≈4–25 h) met inclusion criteria. For DBP, the pooled mean difference was 0.39 mmHg (95% CI 0.09–0.69) with no heterogeneity (I² = 0%). For SBP, the pooled mean difference was 1.42 mmHg (95% CI −3.44 to 6.28) with moderate heterogeneity (I² = 64.9%). Leave-one-out analyses confirmed robust DBP findings and identified a single outlier as the primary driver of SBP heterogeneity, with its exclusion eliminating heterogeneity and attenuating the pooled effect. Overall risk of bias was low to moderate.
Conclusion
In pooled analysis of invasive arterial line validation studies, cuffless BP devices demonstrated minimal mean bias for both SBP and DBP in acute care environments. These findings represent synthesis within the highest physiological reference context currently available and support the feasibility of cuffless monitoring in high-acuity settings. Given the risks and resource burden associated with invasive arterial lines, accurate noninvasive continuous monitoring may offer a safer and more scalable alternative for haemodynamic assessment in the ICU. Ongoing analyses will evaluate performance across broader ambulatory and outpatient populations.DBP cufflessSBP cuffless