Treating Hypotension in the Preterm Infant: When and with What: a Critical and Systematic Review
Total Page:16
File Type:pdf, Size:1020Kb
Journal of Perinatology (2007) 27, 469–478 r 2007 Nature Publishing Group All rights reserved. 0743-8346/07 $30 www.nature.com/jp STATE-OF-THE-ART Treating hypotension in the preterm infant: when and with what: a critical and systematic review EM Dempsey1,2 and KJ Barrington1,2 1Departments of Pediatrics and OB/GYN, McGill University, Montre´al, QC, Canada and 2NICU, Royal Victoria Hospital, Montre´al, QC, Canada Several studies have attempted to determine normal ranges of A very large proportion of extremely preterm infants receive treatments for blood pressure (BP) in the newborn infant.6–13 However, the hypotension. There are, however, marked variations in indications for ranges presented vary considerably for a number of reasons treatment, and in the interventions used, between neonatal intensive care including: retrospective data collection, small numbers of patients units and between neonatologists. included, combined invasive and noninvasive measurements, Methods: We performed systematic reviews of the literature in order to inclusion of small for dates and appropriate for gestational age determine which preterm infants may benefit from treatment with infants, collection of only a few data points and averaging data interventions to elevate blood pressure (BP), and which interventions points over wide time ranges. In some instances, a single absolute improve clinically important outcomes. mean BP value has been chosen and applied over a wide range of gestational and post-natal ages14 which ignores the increase in Results: Our review was not able to define a threshold BP that was usual BP with increasing gestational age, and the usual significantly predictive of a poor outcome, nor whether any interventions spontaneous increase in BP over the first few post-natal days. The for hypotensive infants improved outcomes, nor which interventions were Joint Working Group of the British Association of Perinatal more likely to be beneficial. Medicine has recommended that the mean arterial BP in mm Hg Conclusions: There is a distinct lack of prospective research of this should be maintained at or greater than the gestational age in issue, which prevents good clinical care. It is possible that a simple BP weeks.15 There appear to be no published data to support this threshold that indicates the need for therapy does not exist, and other recommendation; however, it is one of the commonest criteria used factors, such as the clinical status or systemic blood flow measurements, to define hypotension. may be much more informative. Such a paradigm shift will also require Definitions of what is normal, in an inherently abnormal careful prospective study. population, the very preterm infant, require a different conceptual Journal of Perinatology (2007) 27, 469–478; doi:10.1038/sj.jp.7211774 framework. Should definitions of normal ranges be statistically defined, and based on the entire population or only on relatively Keywords: hypotension; preterm infant; systematic review; inotropes; fluid boluses healthy infants with an eventual good outcome? It would be more useful to know what is a ‘safe’ BP, rather than a statistically defined ‘normal’ range. To determine a ‘safe’ BP, adequate prospective studies reporting the positive predictive value or likelihood ratio for adverse Introduction outcome depending on whether the BP was below the threshold would be necessary. Even more useful would be a ‘treatment’ BP, that is, a The diagnosis of hypotension and its subsequent management are threshold below which intervention had been proven to improve controversial areas in the care of the very low birth weight 1 clinically important outcomes; this requires prospective randomized (VLBW, <1500 g) newborn infant. The frequency of hypotension trials powered for clinical outcomes. varies between neonatal intensive care units (NICUs) even when 2 The principle concern regarding hypotension is an assumed the same diagnostic criteria are applied. The frequency of detrimental effect on end organ perfusion. However, perfusion is intervention also varies dramatically,3,4 from 29 to 98% among 5 dependent both on pressure gradient and on vascular resistance infants less than 28 weeks gestation. and is only directly correlated with BP if resistance and outflow 16 Correspondence: Dr KJ Barrington, Departments of Pediatrics and OB/GYN, Royal Victoria pressure are constant. The critical clinical questions, therefore, Hospital, McGill University, 687 Pine Ave., West, Room C7.68, Montre´al, QC, Canada are whether low BP (however defined) results in an adverse clinical H3A 1A1. E-mail: [email protected] outcome, including adverse short-term outcomes (increased Received 14 February 2007; revised 13 April 2007; accepted 8 May 2007 incidence of brain injury or intestinal injury in particular) and Treating hypotension in the preterm infant EM Dempsey and KJ Barrington 470 adverse long-term neurodevelopmental outcome and, if so, does requirements stated above. One particular limitation that should be intervention to treat hypotension result in improved outcomes? emphasized is the importance of using gestational age and The objectives of this review were to determine if there exists post-natal age appropriate normal values. If for example a single sufficient evidence to determine which preterm infants may benefit threshold for hypotension of 30 mm Hg is used to define from interventions to elevate BP, and which interventions improve hypotension, then the more immature infants (who are at greater clinically important outcomes. These objectives created three questions: risk of severe IVH) will be more likely to be deemed hypotensive, this will artifactually appear to show that hypotension is related to (1) Is there a defined BP threshold, or other clinical characteristics IVH. Either a statistical correction for gestational age or birth in preterm infants, which accurately identify those at risk for a weight, or better, the use of normal values appropriate for the poor outcome? individual baby is required to prevent this artifact. (2) Is there evidence that infants with hypotension (either as The four studies that appeared to satisfy the largest proportion of defined in #1 or arbitrarily) may have improved clinical our criteria are described below in more detail. They each had outcomes if they receive intervention? repeated measurements of BP and reliable descriptions of cranial (3) Is there evidence regarding which interventions are most ultrasound findings. Miall-Allen14 was the only study to use effective? masked evaluation of the head ultrasound findings, they also had continuous invasive BP monitoring. They found an excess of IVH/PVL in hypotensive preterm infants, however the threshold Methods value was a single value of a mean BP less than 30 mm Hg, and We performed systematic reviews of the literature in order to the sample size was only 33 infants of 26 to 0 weeks gestation, 9 of address the three objectives above. whom developed either a major IVH or PVL. Bada et al.31 found To address objective #1 we sought prospective inception cohort that infants in whom grade 2 to 4 IVH developed had lower BP studies of very preterm or VLBW infants, with regular assessment of values for their post-natal age than matched control infants BP using reliable methodology, and evaluation of ultrasound without IVH. However, the normal values were only derived from findings by investigators unaware of the BP findings. The infants 16 infants less than 1 kg and it was not possible to calculate birth should have received no therapy for their BP. The outcome weight or gestation-specific BP thresholds from the presented data. variables required were death, severe intraventricular hemorrhage Watkins8 reported a retrospective study which appears to have (IVH), either intraventricular with ventricular dilatation or included the entire cohort of VLBW admitted to the NICU. Having intraparenchymal hemorrhage, or other parenchymal cerebral taken post-natal age and birth weight into account, they identified injury such as cystic periventricular leukomalacia (PVL). We were an association between prolonged duration of a BP below the 10th also interested in studies that examined long-term outcomes, such percentile for birth weight and post-natal age, and the frequency of as Bayley scores of infant development, IQ testing or other IVH. However, there was no correction for other risk factors. indicators of long-term outcome. Cunningham27 in a retrospective cohort study, reported that IVH In November 2006, we performed a search using PubMed and was associated with a ‘low BP’ the day before IVH, but no details the terms ‘BP or hypotension’ ‘premature or preterm’ and ‘IVH or are given regarding the schedules of echoencephalography or what PVL or development’. The search was run both with and without was meant by low BP for this result. Furthermore, any infant with limits (clinical studies, humans and newborn infants, using the a mean arterial BP less than the gestational age in weeks received PubMed ‘limits’ facility) and using both MESH terms and intervention with a colloid bolus followed by inotrope infusion, and unrestricted searches. The articles found were screened to other infants with BP above this threshold were treated if perfusion determine potential applicability, and any study which appeared was poor. They excluded periods when infants received >10 ml/kg from this initial screen to be an evaluation of risk factors for IVH or of colloid, which occurred in 83% of the infants under 750 g. They PVL, or an analysis of normal BPs in preterm infants, was found no association between the development of IVH and the examined in its entirety. We also searched personal article files and duration of time when the mean BP was less than gestational age. the reference lists of recent review articles. This process produced All of these previous studies were confounded by the fact that 16 article,s which appeared likely to provide information about the pressor agents were used in at least some of the hypotensive relationship between BP and ultrasound brain abnormalities in the infants.