Jump to comment:

• Implementation of an automated oxygen Control system- Are we ready?
  Jogender Kumar, Sankalp Dudeja, Pradeep Debata, Khurana Supreet, Abhishek Aradhya and Venkataseshan Sundaram
  Published on: 18 August 2017
• Implementation of an automated oxygen Control system- Are we ready?
  Jogender Kumar, Sankalp Dudeja and Venkatseshan Sundaram
  Published on: 18 August 2017

• Published on: 18 August 2017

  Implementation of an automated oxygen Control system- Are we ready?

  • Jogender Kumar, Neonatologist Post Graduate Institute of Medical education and Research, Chandigarh India 160012
  • Other Contributors:
    • Sankalp Dudeja, Neonatologist
    • Pradeep Debata, Neonatologist
    • Khurana Supreet, Neonatologist
    • Abhishek Aradhya, Neonatologist
    • Venkataseshan Sundaram, Neonatologist

  We read with great interest the article by Van Zanten HA et al., published in this journal and found the results impressive.[1] However, we have certain observations about the conduct of the study
  Even though the authors state that this report was part of a quality improvement initiative in their NICU, the authors have neither reported the results in the format suitable for a quality improvement study nor have clearly stated the design; at the end of introduction they seem to mention that this was a retrospective data analysis; whereas, in the first line of the methods they state the design as a prospective observational study. Even though the automatic oxygen controller group would not have been affected much by any one of the design, the impact would have been in the manual group, keeping especially the training of the NICU staff in mind. It’s also worth emphasizing here that the authors mention about the local guidelines practiced for manual titration of supplemental oxygen based on the saturations, for the sake of external validity.[2]
  Minute wise data points used in this study may have significantly underestimated the hypoxemic episodes and thereby the proportion of times an infant remained in the ‘below target range’ saturations. In a logical sense, manual titration would have happened sooner than expected for a hypoxemic event and hence would not have been captured if more frequent data points are not considered. Using the same technology and a lesser in...

  Show More

  We read with great interest the article by Van Zanten HA et al., published in this journal and found the results impressive.[1] However, we have certain observations about the conduct of the study
  Even though the authors state that this report was part of a quality improvement initiative in their NICU, the authors have neither reported the results in the format suitable for a quality improvement study nor have clearly stated the design; at the end of introduction they seem to mention that this was a retrospective data analysis; whereas, in the first line of the methods they state the design as a prospective observational study. Even though the automatic oxygen controller group would not have been affected much by any one of the design, the impact would have been in the manual group, keeping especially the training of the NICU staff in mind. It’s also worth emphasizing here that the authors mention about the local guidelines practiced for manual titration of supplemental oxygen based on the saturations, for the sake of external validity.[2]
  Minute wise data points used in this study may have significantly underestimated the hypoxemic episodes and thereby the proportion of times an infant remained in the ‘below target range’ saturations. In a logical sense, manual titration would have happened sooner than expected for a hypoxemic event and hence would not have been captured if more frequent data points are not considered. Using the same technology and a lesser interval (5 seconds) Van Kaam et al had shown aptly that automated control reduced hypoxemia.[2]
  Comparative data on the clinical morbidities such as pulmonary artery hypertension and patent ductus arteriosus would be important as these morbidities if differentially distributed in the comparison groups, would have affected the hypoxemic episodes and proportion of time below the target range.
  While concluding, the authors seem to have conveniently avoided describing the significant time spent below the target range in the automated control group. Even though this could have been due to the inherent human behavioral pattern of responding faster to lower saturations, an inbuilt error in the closed loop algorithm with a differential sensitivity towards lower saturations could have also played a role in such phenomenon.[3] For a reader, a strong message would be that even though the control is automated, a close look into the algorithm of the manufacturer is immediately required to avoid spending more time below the target range, especially when one does not know the impact of such ‘below target range’ saturations on long-term neurodevelopmental and pulmonary outcomes.
  Competing interests: None
  Source of funding: None

  References:

  1 Van Zanten H, Kuypers K, Stenson B et al. The effect of implementing an automated oxygen control on oxygen saturation in preterm infants. Archives of Disease in Childhood - Fetal and Neonatal Edition 2017fetalneonatal-2016-312172. doi:10.1136/archdischild-2016-312172
  2 van Kaam A, Hummler H, Wilinska M et al. Automated versus Manual Oxygen Control with Different Saturation Targets and Modes of Respiratory Support in Preterm Infants. The Journal of Pediatrics 2015;167:545-550.e2. doi:10.1016/j.jpeds.2015.06.012
  3 Bancalari E, Claure N. Control of Oxygenation During Mechanical Ventilation in the Premature Infant. Clinics in Perinatology 2012;39:563-572. doi:10.1016/j.clp.2012.06.013

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top
• Published on: 18 August 2017

  Implementation of an automated oxygen Control system- Are we ready?

  • Jogender Kumar, Neonatologist Post Graduate Institute of Medical education and Research, Chandigarh India 160012
  • Other Contributors:
    • Sankalp Dudeja, Neonatologist
    • Venkatseshan Sundaram, Associate Professor Pediatrics (Neonatology)

  We read with great interest the article by Van Zanten HA et al., published in this journal and found it very useful.1 The author rightly stated that the results reflect the real situation as data were collected for the duration infants were admitted, while nurses taking care of them and where workload varied. It will be very relevant for developing countries where nurse patient ratio is poor. But; at the same time would like to offer following comments, clarification to which would benefit the readers of this journal and will help in replication of these results in different settings also.
  It is not very clear whether it was a prospective study or retrospective. In Introduction section, in the end, the author mentioned that we performed a retrospective study in preterm infants to evaluate automated fraction of inspired oxygen (FiO2) control when it was used as standard care and thus for a longer period. While in “Methods” section it is mentioned that it was a prospective observational study. These contradictory statements create confusion to the reader.
  The author mentioned that during the manual period, the nurses manually titrated the supplemental oxygen following local guidelines. However; these guidelines are not given in the current paper. It would be better if clear guidelines would have been described like other studies to improve the external validity and generalizability.2
  In the present study, FiO2 and pulse oximeter saturation (SpO2) were sa...

  Show More

  We read with great interest the article by Van Zanten HA et al., published in this journal and found it very useful.1 The author rightly stated that the results reflect the real situation as data were collected for the duration infants were admitted, while nurses taking care of them and where workload varied. It will be very relevant for developing countries where nurse patient ratio is poor. But; at the same time would like to offer following comments, clarification to which would benefit the readers of this journal and will help in replication of these results in different settings also.
  It is not very clear whether it was a prospective study or retrospective. In Introduction section, in the end, the author mentioned that we performed a retrospective study in preterm infants to evaluate automated fraction of inspired oxygen (FiO2) control when it was used as standard care and thus for a longer period. While in “Methods” section it is mentioned that it was a prospective observational study. These contradictory statements create confusion to the reader.
  The author mentioned that during the manual period, the nurses manually titrated the supplemental oxygen following local guidelines. However; these guidelines are not given in the current paper. It would be better if clear guidelines would have been described like other studies to improve the external validity and generalizability.2
  In the present study, FiO2 and pulse oximeter saturation (SpO2) were sampled every minute, and it is quite common to have fluctuations within a period of one minute which will get missed if we have a data point of one-minute intervals and hence might give fallacious results. In this interval one might have done rapid intervention during hypoxemia in the manual group, hence that hypoxemia episode might get masked with intervention. A study by Van Kaam et al using the same technology and lesser interval (5 seconds) found that automated control reduced hypoxemia.2
  The author didn’t mention about sickness status of babies enrolled in the study. Babies with persistent pulmonary arterial hypertension and patent ductus arteriosus may have a difference in pre and postductal saturation. So, in such cases probe site will have implication. What was probe site chosen in babies in this study? Also; the author should mention whether sickness levels in both groups were similar or not.
  The study concludes that with the implementation of automated oxygen control there was a significant reduction in hypoxemia, but not hypoxemia. It seems to be oversimplification of results. Carefully looking at the results there will be concern that more babies were having SpO2<90%, (median (IQR) 8.6 (7.2–11.7) 15.1 (14.0–21.1) <0.0001) in the automated group. And this finding is consistent with most of the studies done on automated oxygen control.3 Possible reason for this finding may be the human behavior to accept high saturation or an inherent problem with the design of automated algorithm which tends to keep saturation on the lower side of target range and hence more prone to hypoxemia.4 The cumulative effect of preventing hyperoxemia (>95%) and allowing lower saturations (85-89%) in automated group, on clinical outcomes needs to be evaluated.
  As of now from various studies done using this algorithm, it seems that automated oxygenation control systems cannot prevent the occurrence of episodes of hypoxemia. Although studies have shown that it reduces the duration or severity and severity of hypoxemia episodes by increasing Fio2 but still there is need of development in an algorithm to prevent these episodes.
  Competing interests: None
  Source of funding: None

  References:
  1 Van Zanten H, Kuypers K, Stenson B et al. The effect of implementing an automated oxygen control on oxygen saturation in preterm infants. Archives of Disease in Childhood - Fetal and Neonatal Edition 2017;:fetalneonatal-2016-312172. doi:10.1136/archdischild-2016-312172
  2 van Kaam A, Hummler H, Wilinska M et al. Automated versus Manual Oxygen Control with Different Saturation Targets and Modes of Respiratory Support in Preterm Infants. The Journal of Pediatrics 2015;167:545-550.e2. doi:10.1016/j.jpeds.2015.06.012
  3 Claure NBancalari E. Oxygen Saturation Targeting by Automatic Control of Inspired Oxygen in Premature Infants. NeoReviews 2015;16:e406-e412. doi:10.1542/neo.16-7-e406
  4 Bancalari EClaure N. Control of Oxygenation During Mechanical Ventilation in the Premature Infant. Clinics in Perinatology 2012;39:563-572. doi:10.1016/j.clp.2012.06.013

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top