Accurate measurement of blood pressure (BP) is extremely important in the management of sick preterm newborns. The primary objective of this study was to compare non-invasive blood pressure measurement (NIBP) with invasive blood pressure measurement (IBP) using peripheral arterial cannulation (PAC) in preterm neonates < 37 weeks in the neonatal intensive care unit. Preterm neonates needing PAC were prospectively enrolled in the study. NIBP measurements were taken in the same limb as that of peripheral arterial line. Initially IBP was recorded followed by NIBP within 1 min using the same monitor. These were called as paired measurements since they are taken within 1 min of each other. Seventy-three preterm infants with 1703 paired measurements were included in the final analysis (median gestational age 32 weeks, IQR 30-34 weeks, median birth weight 1540 g, IQR 1160-2100 g). In preterm infants not receiving vasoactive agents (n = 51, 1428 paired measurements, Bland-Altman analysis for agreement between invasive mean blood pressure (MBP) and non-invasive mean BP revealed a bias of -2.9123 mmHg (SD 7.8074). The 95% limits of agreement were from -18.2157 to 12.3893 mmHg. In preterm infants with hypotension, we detected a bias of -3.9176 mmHg (SD 5.1135) between invasive MBP and non-invasive MBP. The 95% limits of agreement were from -13.9401 to 6.1048 mmHg. In normotensive preterm infants receiving vasoactive agents, we detected a bias of -0.7629 mmHg (SD 8.0539) between invasive MBP and non-invasive MBP. The 95% limits of agreement were from -16.5485 to 15.02274 mmHg. There is poor level of agreement between IBP and NIBP measurements in sick preterm neonates, leading to overestimation or underestimation of blood pressure. The bias was less for mean BP measurements as compared with systolic BP measurements and also for normotensive neonates as compared with hypotensive neonates. Hence, NIBP may be used as a screening method in haemodynamically stable preterm infants, but infants who are haemodynamically unstable and need to be commenced on vasoactive agents should have IBP monitoring.
Read full abstract