Home Print this page Email this page Small font sizeDefault font sizeIncrease font size
Users Online: 271
About us Editorial board Search Ahead of print Current issue Archives Submit article Instructions Subscribe Contacts Advertise Login 

 Table of Contents  
Year : 2017  |  Volume : 6  |  Issue : 4  |  Page : 245-249

Comprehensive, noninvasive saturation, oxygen, and pressure index: Does it reflect the severity of acute respiratory illness in neonates on continuous positive airway pressure? a prospective study

Department of Pediatrics, JSS Medical College Hospital, Mysore, Karnataka, India

Date of Web Publication17-Oct-2017

Correspondence Address:
Sushma Krishnegowda
Anagha #197, 80 Feet Road, KBL Enclave, Vijayanagar 4th Stage, 2nd Phase, Mysore - 570 032, Karnataka
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/jcn.JCN_68_17

Rights and Permissions

Background: The severity of respiratory illness in neonates on invasive ventilatory support is assessed by oxygenation index and alveolar–arterial oxygen gradient (A-aDO2). Both these parameters need arterial blood gas estimation which is an invasive procedure with attendant complications. Neonates with less severe respiratory disease are managed on continuous positive airway pressure (CPAP). The progress of the disease is generally assessed by noting the changes in FiO2and positive end-expiratory pressure (PEEP) provided. Blood gas analysis is done for objectively assess the babies who progress to more severe disease. A noninvasive tool such as saturation, oxygen, and pressure index (SOPI) helps in reducing the need for invasive blood gas estimation. A good correlation of SOPI with A-aDO2can provide near-continuous bedside assessment of the respiratory disease. Objective: To determine the correlation between SOPI and A-aDO2. Materials and Methods: All babies admitted to our neonatal unit requiring CPAP were considered eligible for this study. The adjustments in FiO2, CPAP pressures, and arterial blood gas were done as per unit protocol. A-aDO2was calculated. SOPI was calculated as (PEEP × FiO2)/SpO2. The two values were then correlated. Results: Seventy-five babies were recruited. SOPI correlated positively (r = 0.847) with A-aDO2(P < 0.0001). Coefficient of determination (R2) was 0.71. SOPI value of 1.6 had a sensitivity of 80% and specificity of 90% in predicting the A-aDO2of 70 which was considered as the value indicating severe illness. Conclusion: SOPI is a noninvasive monitoring tool for babies on CPAP support which has a good correlation of 84.7% with A-aDO2.This can be used as an objective currency to report the severity of respiratory illness in neonates on CPAP.

Keywords: Preterm, respiratory distress, saturation, oxygen, and pressure index

How to cite this article:
Krishnegowda S, Doreswamy SM, Thandaveshwar D. Comprehensive, noninvasive saturation, oxygen, and pressure index: Does it reflect the severity of acute respiratory illness in neonates on continuous positive airway pressure? a prospective study. J Clin Neonatol 2017;6:245-9

How to cite this URL:
Krishnegowda S, Doreswamy SM, Thandaveshwar D. Comprehensive, noninvasive saturation, oxygen, and pressure index: Does it reflect the severity of acute respiratory illness in neonates on continuous positive airway pressure? a prospective study. J Clin Neonatol [serial online] 2017 [cited 2021 Dec 2];6:245-9. Available from: https://www.jcnonweb.com/text.asp?2017/6/4/245/216913

  Introduction Top

Respiratory illness is the common cause for neonates getting admitted to neonatal units. Many of these babies are initially managed with noninvasive respiratory support such as continuous positive airway pressure (CPAP). Few of them progress to more severe illness needing mechanical ventilation. Progress in the severity of respiratory illness is monitored by SpO2 and blood gases. CPAP pressures and FiO2 are adjusted based on these parameters. Threshold and sequence of changing these respiratory parameters, initiating mechanical ventilation or administration of surfactant differs between the individuals and between different neonatal units. This variation in clinical approach is often justifiable but leads to nonuniformity among the individuals and the neonatal units in reporting the severity of illness and its management. Considering clinical parameters in isolation for the assessment of severity and progression of the parenchymal disease downplays the importance of complex interaction between adjustable treatment parameters and the dynamic pathophysiology of the neonatal lung. Uniformity in assessment and reporting of the severity of respiratory disease can be achieved if all the modifiable parameters and its outcome can be accounted in a single index. A comprehensive tool which incorporates and objectively represent all these factors is very desirable both from clinical and research point of view.

Alveolar–arterial oxygen gradient (A-aDO2) is the objective measure of severity of the lung disease. Value of <10 is considered normal in healthy adults.[1] There are no cutoff values suggested for neonates. However, due to higher physiological dead space, this can be as high as 25.[2] Arterial blood gas (ABG) values are necessary for calculating this parameter. Multiple arterial punctures for assessing the severity of lung disease becomes less desirable in neonates on CPAP support alone due to clinical, resource, and financial implications.

Saturation, oxygen, and pressure index (SOPI) is a single all-inclusive, noninvasive index proposed to assess the severity of parenchymal lung disease in neonates.[3] SOPI encompasses the interplay of the disease severity and the respiratory support received by the neonate. It can be used on near continuous basis for the assessment of the lung disease. Earlier study on ventilated babies demonstrated high correlation between SOP and oxygenation index suggesting this to be a potential tool for the assessment of the severity of respiratory illness in neonates on CPAP support.[3]

Demonstration of SOPIs correlation with an indicator of disease severity such as A-aDO2 would strongly support its use in day-to-day practice.

Primary objective

To determine the correlation between SOPI and A-aDO2.

Secondary objective

To determine the cutoff value of SOPI which represents abnormal A-aDO2 of 70.

  Materials and Methods Top

Study design

This was a prospective observational study conducted in a tertiary care neonatal unit between October 2015 and March 2016 in Mysore.

Inclusion criteria

All newborns admitted in JSS hospital neonatal unit with respiratory distress and were managed on CPAP support were included in the study.

Exclusion criteria

  1. Neonates who proceeded to mechanical ventilation within half an hour of commencing CPAP
  2. Babies with their saturation (SpO2) of 98% or more if their FiO2 administered was greater than 21% at the time of blood sampling.

Study procedure

Management of babies with respiratory distress followed the protocol of our unit. Babies who were recruited were placed on either bubble CPAP (Fanem India Ltd–Baby pap I-IMAJ-110050) or ventilator CPAP (Maquet critical care AB Ref: 6691287, SN: 170769). CPAP was delivered either through nasal prongs or masks as tolerated. Initial CPAP was set at 5 cm H2O and FiO2 of 0.3. Stepwise increments were made with FiO2 and CPAP pressure depending on the clinical assessment of severity of respiratory distress and SpO2. Maximum FiO2 and CPAP pressure tolerated before escalating the support to mechanical ventilation was 0.4 and 7 cm H2O, respectively. Chest X-ray, other relevant investigations and rest of the management regarding fluids, antibiotics, etc., were as per the unit protocol.

SpO2 monitoring was done using Philips Sure sign (VM4–SNCN42788815) monitor which employs Massimo technology for measuring the oxygen saturation. SpO2 probe was placed on the right upper limb. The limb was stabilized manually to avoid motion artifacts. Airway toilet was avoided for at least 5 min before recording the SpO2. Once a reliable consistent plethysmograph wave was noted on the monitor, SpO2 displayed was recorded. CPAP pressure and FiO2 needed were recorded at the same time.

Following this, the arterial puncture was performed. The right radial artery was preferred for sampling. Only in cases where it was technically not possible or attempt at sampling failed, the left radial artery was used for sampling. Half milliliter of arterial blood sample was obtained using preheparinized syringe (BD 364356) and was immediately manually transported to biochemistry laboratory. The sample was processed within 15 min of receipt in the laboratory.

SOPI was calculated as (positive end-expiratory pressure [PEEP] × FiO2)/SpO2. Both FiO2 and SpO2 are in decimals (1). A-aDO2 was calculated using the formula A-aDO2= (FiO2(%)/100) × (Patm-47 mmHg) − (PaCO2/0.8) − PaO2, (where, FiO2 room air = 21%, atmospheric pressure = 760 mmHg at sea level, water vapor pressure pH2O [mmHg] = 47 mmHg at 37°C, respiratory quotient [VCO2/VO2] = 0.8).[4]


Sample size

Sample size was calculated with an assumption of alpha error of 0.05 and power of 0.95. In an earlier study, Doreswamy et al. observed a correlation of 0.94 between SOPI and oxygenation index. A-aDO2 being a more exhaustive parameter needing all the blood gas value to compute, we assumed the correlation with SOPI to be at least 0.5. We used the online calculator statstodo.com (https://www.statstodo.com, STATSTODO TRADING PTY LTD. Brisbane, Queensland, Australia) for calculation. We needed to recruit a minimum of forty patients.


Baseline numerical variables are summarized as mean and standard deviation or median and interquartile range (IQR) depending on the distribution. Categorical variables are summarized as proportions. The association between SOPI and A-aDO2 is calculated as Pearson product moment correlation (r). The SOPI value cutoff for A-aDO2 values of 45 and 70 is calculated using received operative characteristics (ROC) curve. All the statistical computing is done using Microsoft excel 2016 and analyzes it version 4.65.


The institutional ethical committee approved this study. Parent consent was obtained on the hospital consent form. All the investigations were done as per the unit protocol and clinical need of the baby.

  Results Top

We recruited a total of 75 babies. Six of them were excluded because of wrong sampling (venous sample), another 6 were excluded due to preanalytical errors during sample collection. Six patients were excluded from the study as the SpO2 was above 98% while receiving oxygen at the time of collecting arterial blood sample. Fifty-seven babies remained for final analysis. Forty-seven babies had single ABG and rest 10 had multiple ABG analysis as per the clinical need. Fifty-two (91%) babies had arterial blood sampled from the right radial artery.

Thirty-two (56%) were preterm babies. Ten (17.6%) babies were less than 32 weeks with lowest being 28 weeks of gestation. Fifteen (26.3%) were female babies. The median birth weight of the preterm babies was 1620 g with IQR between 1410 and 2320 g. The median birth weight of the term babies was 2700 g. The IQR of birth weight in this group was between 2500 and 2920 g. Lowest and heaviest baby weight was 1020 and 3300 g, respectively. [Table 1] shows baseline characteristics such as age at recruitment, gestation in weeks, and respiratory support received.
Table 1: Baseline Characteristics

Click here to view

About 79% of the babies received CPAP for respiratory distress syndrome. [Figure 1] depicts various diagnosis encountered in our study. Forty-two (75%) of babies improved on CPAP. Fourteen (25%) of the babies failed CPAP requiring mechanical ventilation. Out of these fourteen, 10 had increasing respiratory distress with oxygen requirement and four had recurrent apnea. All the intubated babies received surfactant as per our unit protocol. The shortest duration of CPAP support was 7 h and the longest duration was 7 days.
Figure 1: Various diagnosis of respiratory distress of the individuals in the study

Click here to view

SOPI correlated positively with A-aDO2[Figure 2]. The Pearson correlation coefficient (r) was 0.847 with 95% confidence interval of 0.77–0.90. This correlation was significant with a P < 0.0001. The value of Y intercept (SOPI) can be ascertained by the equation Y = 0.011X + 0.751, where X is the A-aDO2 value. This equation can be rewritten to obtain A-aDO2 from SOPI value as A-aDO2= (SOPI – 0.75)/0.01. Coefficient of determination (R2) was 0.72. The distribution of the values was homoscedastic when the standardized residuals were plotted against the observed numbers.
Figure 2: Correlation of saturation, oxygen, and pressure index with alveolar–arterial oxygen gradient

Click here to view

There is no consensus or recommended value of A-aDO2 in normal and sick neonates. Available literature suggests a value of up to 25 in normal neonates. Hiremath et al.[4] studied the values of A-aDO2 in premature babies who could be successfully extubated and those who failed extubation after a period of mechanical ventilation. They reported a preextubation A-aDO2 value of 33 ± 14 in babies who could be successfully extubated compared to preextubation A-aDO2 of 55 ± 17 in those who failed extubation. Hence, we considered (mean ± standard deviation) rounded off value of 45 and 70 as two threshold A-aDO2 values representing the severity of illness with respect to need for mechanical ventilation. SOPI of 1.3 had a sensitivity of 82% and specificity of 70% in predicting the A-aDO2 of 45. To predict an A-aDO2 value of 70, SOPI of 1.6 had a sensitivity of 80% and specificity of 90%. [Figure 3] depicts the ROC curve for the same. The area under the curve is 0.896 with 95% confidence interval of 0.819–0.974.
Figure 3: Received operative characteristics curve of saturation, oxygen, and pressure index against alveolar–arterial oxygen gradient

Click here to view

In our study, median (IQR) SOPI and A-aDO2 in babies who failed CPAP and needed mechanical ventilation was 1.6 (1.5–2.2) and 94.8 (76.4–121.6), respectively. On the contrary, median (IQR) SOPI and A-aDO2 in babies who were successfully treated with CPAP alone was 1.3 (1.3–1.9) and 74.4 (54–107.5), respectively. Due to substantial difference in the sample size between these groups and nearly one-third of the babies in ventilation group needing mechanical ventilation for recurrent apnea, we did not attempt any statistical comparison between these groups.

  Discussion Top

Neonates with respiratory distress have a variable course. About 22% of the babies progress for mechanical ventilation.[3] Intense clinical monitoring and ABGs helps in identifying the babies who fail CPAP. ABG monitoring is an invasive procedure with a failure rate of 7.9%–8.9%.[5] Repeated arterial puncture can lead to ischemic complication ranging from 1.5% to 35%.[6],[7] Furthermore, it is both labor intensive and a financial burden. Although ABG is very valuable in guiding the management, repeated measurement increases the risk of complications. Having a reliable noninvasive tool which can guide the management will result in minimizing the need for arterial blood sampling.

With the intention of reducing invasive procedures, several noninvasive tools such as oxygen saturation index [8] and respiratory severity score [9],[10] have been studied in neonates and have been demonstrated to be useful with some limitations. However, these tools were designed to be used on babies receiving mechanical ventilation, and none are available for babies on CPAP support.

Assessment of severity of illness in babies on CPAP support alone has relied on the clinically adjustable parameters such as FiO2 and CPAP pressure. The pattern, sequence, and the threshold for changing these parameters have been a matter of clinical judgment. This results in nonuniformity in reporting the severity which is generally settled by blood gas analysis.

To circumvent these issues, a comprehensive noninvasive index (SOPI) was developed and studied by Doreswamy et al.[3] SOPI demonstrated a good correlation with oxygenation index in which ventilatory parameter of PEEP alone was considered for calculation. Given the fact that the study was conducted on neonates receiving ventilatory support, we had to demonstrate the usefulness of SOPI in babies on CPAP support to mitigate other confounding factors associated with mechanical ventilation.

A-aDO2 is a measure of ventilation–perfusion (VQ) mismatch in the lungs. Increase in the V/Q mismatch results in increased intrapulmonary shunt and hence hypoxemia.[1] In neonatal lung disease, there is impaired gas exchange resulting in the variable amount of V/Q mismatch and elevated A-aDO2 is an objective measure of the severity of lung parenchymal disease. Normal A-aDO2 in neonates is not well-established. However, this gradient is <10 in healthy adults. In neonates, due to higher physiological dead space, A-aDO2 can be as high as 25.[2] Increase in A-aDO2 indicate increasing severity of lung parenchymal disease resulting in decreased oxygenation of the blood.[11],[12] A-aDO2 is superior to oxygenation index with respect to the assessment of overall pulmonary disease as it measures the efficiency of the respiratory membrane in exchanging the gases.[1]

The study done by Hiremath et al. has reported the A-aDO2 values for babies with varying the severity of pulmonary illness noted as the ability to be successfully weaned or failed extubation. We believe it is reasonable to consider the preextubation A-aDO2 values of unsuccessful extubation babies as representing severe disease and less than that as moderate disease. The reported values in their study have wide variance, and hence, we considered to use mean plus one standard deviation as cutoff for our study. These A-aDO2 values turn out to be 45 and 70 for moderate and severe disease, respectively.

Considering the clinical utility of SOPI to be able to make interventional decisions such as intubation and surfactant administration, we preferred to choose SOPI with higher specificity as cutoff values. Hence, we recommend the SOPI values of 1.3 and 1.6 to indicate moderate and severe disease, respectively. A SOPI value of 1.6 should prompt the clinician to consider intubation and surfactant in babies deteriorating on CPAP support. Since SOPI is an assessment tool rather than a predictive tool, the individual clinician or unit can choose their own cutoffs of SOPI depending their comfort levels and resources available. SOPI has an added advantage that it can be used irrespective of the unit or individual practice of changing either PEEP or FiO2 in different sequence. This index can also serve as an objective way of communicating the progress of pulmonary illness in neonates and hence of immense value in research.

  Conclusion Top

SOPI is a noninvasive monitoring tool for babies on CPAP support which has a good correlation of 84.7% with A-aDO2. It has a high coefficient of determination (R2 = 0.71) with A-aDO2 and hence can be useful in tracking the progress of respiratory illness in neonates. A value of 1.6 indicates A-aDO2 of 70 with high specificity and hence can be considered for the escalation of respiratory support. Once validated with a bigger sample size, SOPI can be used as an objective currency to report the severity of respiratory illness in neonates.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

  References Top

Dargaville PA, Aiyappan A, De Paoli AG, Dalton RG, Kuschel CA, Kamlin CO, et al. Continuous positive airway pressure failure in preterm infants: Incidence, predictors and consequences. Neonatology 2013;104:8-14.  Back to cited text no. 1
Lovering AT, Romer LM, Haverkamp HC, Pegelow DF, Hokanson JS, Eldridge MW. Intrapulmonary shunting and pulmonary gas exchange during normoxic and hypoxic exercise in healthy humans. J Appl Physiol 2008;104:1418-25.  Back to cited text no. 2
Doreswamy SM, Chakkarapani AA, Murthy P. Saturation oxygen pressure index for assessment of pulmonary disease in neonates on non-invasive ventilation. Indian Pediatr 2015;52:74-5.  Back to cited text no. 3
Hiremath GM, Mukhopadhyay K, Narang A. Clinical risk factors associated with extubation failure in ventilated neonates. Indian Pediatr 2009;46:887-90.  Back to cited text no. 4
Yee K, Shetty AL, Lai K. ABG needle study: A randomised control study comparing 23G versus 25G needle success and pain scores. Emerg Med J 2015;32:343-7.  Back to cited text no. 5
Bedford RF. Wrist circumference predicts the risk of radial-arterial occlusion after cannulation. Anesthesiology 1978;48:377-8.  Back to cited text no. 6
Soderstrom CA, Wasserman DH, Dunham CM, Caplan ES, Cowley RA. Superiority of the femoral artery of monitoring. A prospective study. Am J Surg 1982;144:309-12.  Back to cited text no. 7
Rawat M, Chandrasekharan PK, Williams A, Gugino S, Koenigsknecht C, Swartz D, et al. Oxygen saturation index and severity of hypoxic respiratory failure. Neonatology 2015;107:161-6.  Back to cited text no. 8
Doreswamy SM, Chakkarapani AA, Murthy P. Oxygen saturation index – A noninvasive tool for monitoring hypoxemic respiratory failure in newborns. Indian Pediatr 2016;53:432-3.  Back to cited text no. 9
Iyer NP, Mhanna MJ. Non-invasively derived respiratory severity score and oxygenation index in ventilated newborn infants. Pediatr Pulmonol 2013;48:364-9.  Back to cited text no. 10
Nicolini A, Ferraioli G, Ferrari-Bravo M, Barlascini C, Santo M, Ferrera L. Early non-invasive ventilation treatment for respiratory failure due to severe community-acquired pneumonia. Clin Respir J 2016;10:98-103.  Back to cited text no. 11
Nicolini A, Piroddi IM, Barlascini C, Senarega R. Predictors of non-invasive ventilation failure in severe respiratory failure due to community acquired pneumonia. Tanaffos 2014;13:20-8.  Back to cited text no. 12


  [Figure 1], [Figure 2], [Figure 3]

  [Table 1]

This article has been cited by
1 Saturation oxygenation pressure index: a non-invasive bedside measure for severity of respiratory disease in neonates on CPAP
Deepti Thandaveshwara,Ashok Huduguru Chandrashekar Reddy,Manjunath Vaddamabal Gopalakrishna,Srinivasa Murthy Doreswamy
European Journal of Pediatrics. 2020;
[Pubmed] | [DOI]


Similar in PUBMED
  Search Pubmed for
  Search in Google Scholar for
Related articles
Access Statistics
Email Alert *
Add to My List *
* Registration required (free)

  In this article
Materials and Me...
Article Figures
Article Tables

 Article Access Statistics
    PDF Downloaded311    
    Comments [Add]    
    Cited by others 1    

Recommend this journal