Jump to comment:

• Single night oximetry may be inadequate
  Michael D Shields, Dara O'Donoghue
  Published on: 31 August 2017
• Pulse oximetry in children under the magnifying glass
  Martino Pavone, Elisabetta Verrillo and Renato Cutrera
  Published on: 3 August 2017

• Published on: 31 August 2017

  Single night oximetry may be inadequate

  • Michael D Shields, Respiratory paediatrician Queen's University Belfast
  • Other Contributors:
    • Dara O'Donoghue, Respiratory paediatrician

  We would like to thank the authors of the Pavone paper for their interest in our paper (1,2). We are sorry for not quoting their paper in our study report but do confirm that we were aware of it (2). In our introduction we selected several papers to quote in order to introduce the uncertainty with respect to the need to record 1, 2 or 3 nights of overnight oximetry and the Pavone paper was not one we selected. The Pavone paper claims excellent night to night consistency in oximetry and that only one night of oximetry measurement is necessary while our study did not find this to be the case (2).
  While we agree that the Pavone study used a pulsed oximeter with some superior properties (Radical Masimo) compared to that which we used (Nonin 9600) we do not believe that this is one of the most important reasons why our results differ from the Pavone study.
  We believe the main reasons for differences between the two papers include;
  1] Different primary aims - our study was aimed to determine whether doing 2 or 3 nights oximetry would increase the chances of getting adequate traces to make a report. We therefore included all studies (whether satisfactory or not). In the Pavone study only those with 2 nights each with > 6 hours satisfactory tracing were included and about one third of the children initially identified were therefore excluded. We do not know what then happened to these children – i.e. whether further studies had to be rescheduled. Clearly selec...

  Show More

  We would like to thank the authors of the Pavone paper for their interest in our paper (1,2). We are sorry for not quoting their paper in our study report but do confirm that we were aware of it (2). In our introduction we selected several papers to quote in order to introduce the uncertainty with respect to the need to record 1, 2 or 3 nights of overnight oximetry and the Pavone paper was not one we selected. The Pavone paper claims excellent night to night consistency in oximetry and that only one night of oximetry measurement is necessary while our study did not find this to be the case (2).
  While we agree that the Pavone study used a pulsed oximeter with some superior properties (Radical Masimo) compared to that which we used (Nonin 9600) we do not believe that this is one of the most important reasons why our results differ from the Pavone study.
  We believe the main reasons for differences between the two papers include;
  1] Different primary aims - our study was aimed to determine whether doing 2 or 3 nights oximetry would increase the chances of getting adequate traces to make a report. We therefore included all studies (whether satisfactory or not). In the Pavone study only those with 2 nights each with > 6 hours satisfactory tracing were included and about one third of the children initially identified were therefore excluded. We do not know what then happened to these children – i.e. whether further studies had to be rescheduled. Clearly selecting only those with ‘good’ adequate tracings could improve repeatability. However, as stated in our paper, we did not find this to be the case when we analysed the repeatability for those with 3 adequate nights recording.
  2] Pavone in their study excluded children with complex conditions (ie children with neuromuscular, complex or genetic conditions were not included). In our study this group represented close to 50% of those studied (Down Syndrome, Neuromuscular disorders and craniofacial structural problems, neurological disorders). We were especially worried about differential repeatability in children with Down Syndrome when compared with neuromuscular disorders. It is thus quite possible that better correlation will be noted in a specific group of otherwise normal children who also have performed well and allowed 2 nights of ‘good’ recordings to be made.
  3] A 3rd and very important difference between our paper and that off Pavone relates to the method of statistical analysis for the continuous variables. We were particularly interested in determining the agreement between measures addressing the question whether one night could be replaced by the value of the second night and what the repeatability was. We used Bland & Altman 95% limits of agreement with plots and the intraclass correlation. In the Pavone paper only Pearson and Spearman correlation coefficients are quoted for relating night 1 with night 2 continuous variables and thus we were unable to estimate how good the agreement between these measures actually was. Bland & Altman have previous shown that two measurements of the same quantity (as occurs in repeated measures) are highly likely to closely correlate and especially so if a wide range of values are included (3). It is now well known that while repeated measures variables may correlate well they often do not agree to the extent that one value could be used to replace another (3). Eye balling the scatter plots (with regression lines) in the Pavone paper would not suggest that agreement was necessarily good. However, this is impossible to determine without seeing the Bland & Altman plots (with 95% limits of agreement) or the scattergrams (night 1 versus night 2) with the line of identity included rather than the regression line.
  4] Pavone compared agreement between the two nights McGill Score (MOS) which is used to quantify the severity of Obstructive Sleep Apnoea (OSA). We did not use this as a significant proportion of our children especially those with neuromuscular disease were being studied for sleep disordered breathing with potential hypoventilation rather than detecting OSA. It would not have been appropriate to use the MOS score in this situation.

  Unfortunately, for these reasons, we do not find that the Pavone paper provided reassurance that a single nights oximetry is adequate when screening children for the variety of sleep disordered breathing problems that we studied.

  Finally, while the mean duration of overnight oximetry in our study was 8.4 hours (8.3 hours in the Pavone study) we included > 4 hours of technically satisfactory trace as adequate as our lower limit of acceptability. We agree that this may differ from the ATS recommendations but we think this is an important point for further discussion – and ask the question: Should we be discarding traces that are of good quality but only of 5.5 hours, 5 hours or even 4 hours duration in this group of children? We do not have the answer as to what is the minimum duration of good quality trace that should be regarded as useable. Instead of using a rather arbitrary value of > 6 hours our team considered that 3 hours was too short but that 5 hours was likely adequate and as a group agreed that if the trace was > 4 hours we would like and want it to be reported. It would be interesting and important to determine whether reporting studies of > 4 hours duration would accurately increase the diagnostic pick up in children who present with possible OSA.

  References
  1. Night-to-night consistency of at-home nocturnal pulse oximetry testing for obstructive sleep apnea in children. Pavone M, Cutrera R, Verrillo E, Salerno T, Soldini S, Brouillette RT. Pediatr Pulmonol. 2013 Aug;48(8):754-60
  2. Night-to-night variation of pulse oximetry in children with sleep-disordered breathing.
  Burke RM, Maxwell B, Hunter C, Graham D, O'Donoghue D, Shields MD. Arch Dis Child. 2016 Dec;101(12):1095-1099
  3. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–10

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top
• Published on: 3 August 2017

  Pulse oximetry in children under the magnifying glass

  • Martino Pavone, Pediatric Pulmonologist Pediatric Pulmonology & Respiratory Intermediate Care Unit Sleep and Long Term Ventilation Unit Academic Department of Pediatri
  • Other Contributors:
    • Elisabetta Verrillo, Pediatric Pulmonologist
    • Renato Cutrera, Pediatric Pulmonologist

  Dear Authors.

  In your paper [1], you did not mention a previous prospective study performed by our group [2], on 148 otherwise healthy children referred to a Sleep Center for suspected OSA. In our study, pulse oximetry metrics were similar on the two consecutive nights. The McGill Oximery Score (MOS) on the two nights showed excellent night-to-night consistency when analyzed as positive for OSA versus inconclusive. We highlighted that the findings may not apply to younger infants, to adolescents, or to children with complex comorbidities.

  Our conclusions were different from yours for at least two main reasons.

  Firstly, you accepted oximetry recording lasting ≥ 4 hrs. On our opinion, this cut off is too low and it cannot be sufficiently representative of an overnight study. In our study [2], we accepted recordings lasting ≥ 6 hrs according to the ATS guidelines for sleep study.

  Secondly, you used a Nonin 9600 Pulse Oximeter with Nellcor neonatal-adult SpO2 sensor. In our study [2], we used a motion-resistant Radical 5 Masimo Pulse Oximeter. Previous studies [3,4], demonstrated the superiority of the Radical Masimo technology.

  We believe that your study was performed using suboptimal technology and the criteria for minimum acceptable recording time didn’t respect the ATS guidelines. Therefore, your results should be considered with caution.

  Convincing data already exist on pulse oximetry and the analysis of MOS as a useful tool for...

  Show More

  Dear Authors.

  In your paper [1], you did not mention a previous prospective study performed by our group [2], on 148 otherwise healthy children referred to a Sleep Center for suspected OSA. In our study, pulse oximetry metrics were similar on the two consecutive nights. The McGill Oximery Score (MOS) on the two nights showed excellent night-to-night consistency when analyzed as positive for OSA versus inconclusive. We highlighted that the findings may not apply to younger infants, to adolescents, or to children with complex comorbidities.

  Our conclusions were different from yours for at least two main reasons.

  Firstly, you accepted oximetry recording lasting ≥ 4 hrs. On our opinion, this cut off is too low and it cannot be sufficiently representative of an overnight study. In our study [2], we accepted recordings lasting ≥ 6 hrs according to the ATS guidelines for sleep study.

  Secondly, you used a Nonin 9600 Pulse Oximeter with Nellcor neonatal-adult SpO2 sensor. In our study [2], we used a motion-resistant Radical 5 Masimo Pulse Oximeter. Previous studies [3,4], demonstrated the superiority of the Radical Masimo technology.

  We believe that your study was performed using suboptimal technology and the criteria for minimum acceptable recording time didn’t respect the ATS guidelines. Therefore, your results should be considered with caution.

  Convincing data already exist on pulse oximetry and the analysis of MOS as a useful tool for detecting and grading severity of paediatric OSA, to prioritize the most severely affected children for expeditious adenotonsillectomy, to predict the likelihood of having adverse peri-operative respiratory events.

  We reinforce the concept that pulse oximetry is useful when performed with appropriate motion-resistant technology, interpreted by expert physicians, and targeted to otherwise healthy children.

  Children with complex pathologies deserve a different approach as suggested by ERS Guidelines and poly(somno)graphy should be performed.

  References
  1) Burke RM, Maxwell B, Hunter C, et al. Night-to-night variation of pulse oximetry in children with sleep disordered breathing. Arch Dis Child 2016;101:1095–9.

  2) Pavone M, Cutrera R, Verrillo E, et al Night-to-night consistency of at-home nocturnal pulse oximetry testing for obstructive sleep apnea in children. Pediatr Pulmonol. 2013 Aug; 48(8):754-60.

  3) Trang H, Boureghda S, Leske V. Sleep desaturation: comparison of two oximeters. Pediatr Pulmonol. 2004 Jan;37(1):76-80”.

  4) Brouillette RT, Lavergne J, Leimanis A, et al. “Differences in pulse oximetry technology can affect detection of sleep-disorderd breathing in children. Anesth Analg. 2002 Jan;94(1 Suppl):S47-53”

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top