Jump to comment:

• Cardiac Screening: Pulsatility Index, Population Studies and False-Positive Rates
  Jonathan Searle, Devangi Thakkar and Jayanta Banerjee
  Published on: 21 December 2018
• Perfusion index and undiagnosed critical left-sided obstructive heart defects.
  Itamar Nitzan, Robbert Koppel and Meir Nitzan
  Published on: 27 November 2018

• Published on: 21 December 2018

  Cardiac Screening: Pulsatility Index, Population Studies and False-Positive Rates

  • Jonathan Searle, Paediatric Registrar Imperial College Healthcare NHS Trust, London, UK
  • Other Contributors:
    • Devangi Thakkar, Neonatal Registrar
    • Jayanta Banerjee, Neonatal Consultant, Honorary Senior Clinical Lecturer

  We thank Nitzan et al for their comments in relation to our article on use of pulsatility index (PI) in screening for critical congenital heart disease (Searle 2018). In particular, we are grateful that they draw further attention to the potential for current screening to miss critical lesions, such as coarctation of the aorta. Given the progressive nature of these pathologies, it is an extremely difficult challenge to design an acceptable screening tool, which highlights all affected babies in appropriate time. Despite strong biological plausibility, current evidence is unclear whether pulsatility index can translate into such a tool.

  We fully agree that the local quality of both antenatal and postnatal screening significantly affects the measured benefit of pulsatility index. Several articles draw the distinction here between ‘tertiary’ and ‘non-tertiary’ units, though it may be more accurate to distinguish ‘better resourced’ from ‘less resourced’ settings, particularly in relation to antenatal scanning. As described in our original article, the apparent potential of PI screening in ‘less resourced’ settings seems strong, especially since many pulse oximetry sensors already measure it. Both Schena et al (2017) and Granelli & Ostman-Smith (2007) highlight a small but important population of babies not detected by standard screening, but with abnormal pre-morbid pulsatility indices. It seems incongruous, however, to extrapolate a single extra case detected by th...

  Show More

  We thank Nitzan et al for their comments in relation to our article on use of pulsatility index (PI) in screening for critical congenital heart disease (Searle 2018). In particular, we are grateful that they draw further attention to the potential for current screening to miss critical lesions, such as coarctation of the aorta. Given the progressive nature of these pathologies, it is an extremely difficult challenge to design an acceptable screening tool, which highlights all affected babies in appropriate time. Despite strong biological plausibility, current evidence is unclear whether pulsatility index can translate into such a tool.

  We fully agree that the local quality of both antenatal and postnatal screening significantly affects the measured benefit of pulsatility index. Several articles draw the distinction here between ‘tertiary’ and ‘non-tertiary’ units, though it may be more accurate to distinguish ‘better resourced’ from ‘less resourced’ settings, particularly in relation to antenatal scanning. As described in our original article, the apparent potential of PI screening in ‘less resourced’ settings seems strong, especially since many pulse oximetry sensors already measure it. Both Schena et al (2017) and Granelli & Ostman-Smith (2007) highlight a small but important population of babies not detected by standard screening, but with abnormal pre-morbid pulsatility indices. It seems incongruous, however, to extrapolate a single extra case detected by the Schena et al (2017) study into a real 50% improvement in detection rates, over pulse oximetry screening alone. With such small numbers, the difference between an important finding and statistical or technical anomaly is impossible to determine. We hope this clarifies our original statement that ‘current evidence does not support inclusion of PI into newborn screening’.

  Moving forward, the inclusion of pulsatility index within current screening practice would require further large population studies, particularly looking at babies within ‘less resourced’ settings. To make such a study acceptable, the methodology must either i) reduce the high false-positive rate (typically 5%) associated with previous approaches or ii) have robust and pragmatic mechanisms to safely minimise the impact of a false-positive screen on a newborn and their family. Use of different technologies may address the former requirement. Pre-to-post ductal pulse delay, for example, is an innovative approach, though it is untested within a screening context (Palmeri 2017). Within this study, all ‘case’ group babies already demonstrated significant clinical signs at time of testing. Similarly, Nitzan et al’s suggestion to repeat PI measurement after a few days has merit. Establishing the optimal timing of measurements however would be a challenge, to maximise detection without missing early cases. Granelli & Ostman-Smith (2007) could not answer this point, despite study of 10,000 babies up to day 5 of life.

  It seems unacceptable to remain in the status quo, where a significant portion of babies are only detected following life-threatening collapse. PI could be an important addition to the screening armoury in ‘low resourced’ settings, but true population-specific evidence would be needed. We very much welcome an open discussion of how to overcome the challenges this would pose.

  REFERENCES
  Granelli A, Ostman-Smith I. Noninvasive peripheral perfusion index as a possible tool for screening for critical left heart obstruction. Acta Paediatr. 2007;96:1455–9
  Palmeri L, Gradwohl G, Nitzan M, et al. Photoplethysmographic waveform characteristics of newborns with coarctation of the aorta. J Perinatol. 2017;37:77–80
  Schena F, Picciolli I, Agosti M, et al. Perfusion index and pulse oximetry screening for congenital heart defects. J Pediatr. 2017;183:74–9

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top
• Published on: 27 November 2018

  Perfusion index and undiagnosed critical left-sided obstructive heart defects.

  • Itamar Nitzan, neonatal fellow Monash medical center clayton, Victoria, Australia
  • Other Contributors:
    • Robbert Koppel, Neonatologist
    • Meir Nitzan, Professor of physics

  The review by Searle et al “Does pulsatility index add value to newborn pulse oximetry screening for critical congenital heart disease?" (Searle 2018), provides a comprehensive overview of the current evidence regarding the addition of perfusion index to the CCHD screening algorithm.
  The authors’ main concern is that adding pulsatility index (perfusion index, PI) will not significantly improve the current detection rate which is already quite high. As a proof, they cite the work of the large trial by Schena et al, (Schena 2017) that found one additional case of CCHD in 42,169 babies examined. The authors conclude that incorporating PI into current screening algorithms provides little additional benefit in detecting CCHD and confers a high false positive rate.
  We would like to voice several comments regarding this article:
  First, in the study of Schena et al, CCHD was suspected before screening in 36/38 cases in tertiary centers. This is the main reason that PI (and pulse oximetry screening) did not have any additional value in tertiary centers. In this study, only 23.6% of the neonates were born in non-tertiary center. We suggest that an alternative way to describe the results is that in non-tertiary centers, pulse oximetry detected 2 cases and PI detected an additional 1 case per approximately 10,000 screened neonates. Therefore, adding PI to the screening algorithm improved the detection rate by 50%. Moreover, the 2 cases detected by pulse oximetry...

  Show More

  The review by Searle et al “Does pulsatility index add value to newborn pulse oximetry screening for critical congenital heart disease?" (Searle 2018), provides a comprehensive overview of the current evidence regarding the addition of perfusion index to the CCHD screening algorithm.
  The authors’ main concern is that adding pulsatility index (perfusion index, PI) will not significantly improve the current detection rate which is already quite high. As a proof, they cite the work of the large trial by Schena et al, (Schena 2017) that found one additional case of CCHD in 42,169 babies examined. The authors conclude that incorporating PI into current screening algorithms provides little additional benefit in detecting CCHD and confers a high false positive rate.
  We would like to voice several comments regarding this article:
  First, in the study of Schena et al, CCHD was suspected before screening in 36/38 cases in tertiary centers. This is the main reason that PI (and pulse oximetry screening) did not have any additional value in tertiary centers. In this study, only 23.6% of the neonates were born in non-tertiary center. We suggest that an alternative way to describe the results is that in non-tertiary centers, pulse oximetry detected 2 cases and PI detected an additional 1 case per approximately 10,000 screened neonates. Therefore, adding PI to the screening algorithm improved the detection rate by 50%. Moreover, the 2 cases detected by pulse oximetry did not have ductal dependent systemic circulation. Therefore, the most important detection in this study was by PI. We agree with the authors that pulse oximetry CCHD screening, with or without PI, has a low detection rate in tertiary units. However, we believe that this study demonstrates the potential utility of PI as a supplementary screening method in non-tertiary units. This information may be of value in regions with limited access to fetal diagnosis which is expensive and requires advanced technical skills while PI is inexpensive and may be performed by nurses and midwives.
  Second, we would like to draw attention to the three missed cases of critical left-sided obstruction described in this study. Two of the neonates presented with signs of shock and cardiac failure. We believe that the current screening algorithms may be improved by measuring the pulse wave delay between extremities (Palmeri 2017) as was mentioned in the review. Detection rates may also increase by improving the method of derivation of PI and by comparing pre and post-ductal PI values (Palmeri 2017, Nitzan 2018). Another option that may be relevant in some regions, is a repeated screening test performed in the community clinic or during a home visit after hospital discharge in the hope of detecting critical left-sided obstruction during the process of duct closure before the onset of cardiac failure.
  We thank the authors for raising awareness of this important question, and we hope that future studies will address this issue.

  Searle J, Thakkar DD, Banerjee J.Does pulsatility index add value to newborn pulse oximetry screening for critical congenital heart disease? Arch Dis Child. 2018 Nov 9. pii: archdischild-2018-315891. doi: 10.1136/archdischild-2018-315891. [Epub ahead of print]

  Schena F, Picciolli I, Agosti M, et al. Perfusion index and pulse oximetry screening for
  congenital heart defects. J Pediatr 2017;183:74–9.

  Palmeri L, Gradwohl G, Nitzan M, et al. Photoplethysmographic waveform
  characteristics of newborns with coarctation of the aorta. J Perinatol 2017;37:77–80.

  Nitzan I, Hammerman C, Fink D, Nitzan M, Koppel R, Bromiker R. The effect of patent ductus arteriosus on pre-ductal and post-ductal perfusion index in preterm neonates. Physiol Meas. 2018 Jul 20;39(7):075006. doi: 10.1088/1361-6579/aacf25.

  Show Less

  Conflict of Interest:
  We are currently working on impruving PI measurements like we stated in the response.

  • Back to top