Journal of Postgraduate Medicine
 Open access journal indexed with Index Medicus & EMBASE  
     Home | Subscribe | Feedback  

[Download PDF
Year : 2000  |  Volume : 46  |  Issue : 2  |  Page : 149-52  

Non-invasive respiratory monitoring in paediatric intensive care unit.

UB Nadkarni, AM Shah, CT Deshmukh 
 Department of Paediatrics, Division of Intensive Paediatric Care, Seth G. S. Medical College and K. E. M. Hospital, Parel, Mumbai - 400 012, India., India

Correspondence Address:
U B Nadkarni
Department of Paediatrics, Division of Intensive Paediatric Care, Seth G. S. Medical College and K. E. M. Hospital, Parel, Mumbai - 400 012, India.


Monitoring respiratory function is important in a Paediatrics Intensive Care Unit (PICU), as majority of patients have cardio-respiratory problems. Non-invasive monitoring is convenient, accurate, and has minimal complications. Along with clinical monitoring, oxygen saturation using pulse oximetry, transcutaneous oxygenation (PtcO2) and transcutaneous PCO2 (PtcCO2) using transcutaneous monitors and end-tidal CO2 using capnography are important and routine measurements done in most PICUs. Considering the financial and maintenance constraints pulse oximetry with end tidal CO2 monitoring can be considered as most feasible.

How to cite this article:
Nadkarni U B, Shah A M, Deshmukh C T. Non-invasive respiratory monitoring in paediatric intensive care unit. J Postgrad Med 2000;46:149-52

How to cite this URL:
Nadkarni U B, Shah A M, Deshmukh C T. Non-invasive respiratory monitoring in paediatric intensive care unit. J Postgrad Med [serial online] 2000 [cited 2022 Dec 6 ];46:149-52
Available from:

Full Text

The ability to monitor variables related to respiratory function is important because a large proportion of critically ill patients admitted to paediatric intensive care unit (PICU) have respiratory disorders. The ideal respiratory monitoring system should be easy to use, and interpret and have high degree of technical accuracy, specificity and sensitivity for index to be measured. Further, such system should be cost effective and low risk to patient[1],[2],[3]. Arterial blood gas and derived indices have been most commonly used and form integral part of monitoring in PICU. However, it is neither sensitive nor specific and is affected by changes in ventilation and non-pulmonary factors that alter the mixed venous tensions of O2 and CO2. It is also invasive and intermittent which may not reflect the true status of constantly changing system[1],[3].

[Table:1] shows the various parameters for respiratory monitoring. The main modalities of non-invasive respiratory monitoring are summarised below:

  ::   Physical examinationTop

Cyanosis of tongue and oral mucosa indicate oxygen saturation (SaO2) of less than 80 percent. However, there is significant inter-observer variability and difficulty in reproducibility.

Serial measurement of respiratory rate is easy and has good accuracy. Presence of increased work of breathing is suggested by suprasternal and intercostal retractions, use of accessory muscles of respiration and paradoxical breathing.

  ::   Monitoring of o2 levelTop

Clinical uses of monitoring O2 level is summarised in [Table:2]. There are two non-invasive methods of monitoring O2 level.

Pulse oximetry

This utilises the principle of different light absorption spectra for saturated oxyhaemoglobin compared with reduced haemoglobin. The device consists of a disposable or a reusable probe and a display unit. A varying amount of light is absorbed by arterial blood with more absorption of light during systole. The changing ratio of adsorbed to transmitted light is measured and fed to a microprocessor which calculates SaO2 that is continuously displayed [1],[7],[8],[9].

Pulse oximeter measures “functional saturation” defined as the percentage of oxyhaemoglobin compared to the sum of oxyhaemoglobin and reduced haemoglobin. It does not measure other types of haemoglobins. As regards accuracy, pulse oximeters measure SaO2 within 95% confidence limits of ? 4% when the saturation is greater than 70%. Pulse oximeters measure oxygen saturation, which is related to PaO2 by the oxyhaemoglobin dissociation curve. Any factor that shifts the oxyhaemoglobin dissociation curve (temperature, pH, PCO2) will affect the relationship between the saturation and the PaO2. Because of the shape of the curve, measurements of oxygen saturations are relatively insensitive at detecting significant changes in PaO2 at high levels of oxygenations; for instance, a change in PaO2 from 100 mm Hg to 75 mm Hg will have only minimal effect on the saturation values. Once haemoglobin is fully saturated (saturation of 100 %; PaO2 approximately 120 mm Hg), no change in saturation readings will be noted.

When hyperoxia is a clinical concern to prevent oxygen toxicity, as in neonates, the measurement of PaO2 is more appropriate than measurement of oxygen saturation[1],[7],[8],[10],[11].

Although pulse oximeter represents a significant advance it may give inaccurate information in the presence of dyshaemoglobinaemias, jaundice, anaemia, low perfusion state and use of intravascular dye such as methylene blue.

Motion artefact and excessive light from fluorescent and xenon arc surgical lights may cause false low values[1],[2],[7],[8],[10],[12].

The advantages of oximeters in short are pulse-by-pulse detection of rapid changes, no equilibration time, no calibration and substantially lower maintenance. The main disadvantages of this system are inability to use in cases of severe hypotension or marked oedema, and risk of hyperoxia at saturations between 90-100% [12].

In our experience pulse oximeters have been found very useful in managing emergency out-patients (asthmatics), admitted critically ill children (pulmonary oedema, cardiopulmonary resuscitation), initiation and weaning of ventilator patients and intra-operative monitoring.

Transcutaneous oxygenation (PtcO2)

A Clark type of sensor is applied to the skin over anterior chest or abdomen on an occlusive contact medium and held in place by an adhesive ring. The sensor is heated to 43 to 440C to produce localised hyperaemia, which maximises capillary blood flow. Tissue oxygen that diffuses across the epidermis is reduced, producing a current proportional to the PaO2 which has arterial blood gas (ABG) correlation of 0.90 to 0.95[1],[12],[13]. A 10-15 minute equilibration time is needed for sensor, which has to be calibrated at 43 to 440C 2.

Inaccurate measurements may result from sensor dislodgement, improper calibration, skin oedema, severe acidosis, hypothermia, shock, and severe anaemia. However radiant warmers, bilirubin lights have no effects on PtcO2 measurements[1],[4],[12],[14],[15],[16].

Determination of PaO2 provides an earlier indication of deteriorating pulmonary function. It also provides estimates of hyperoxaemic states and helps in evaluation of shunting in persistent pulmonary hypertension [3],[12],[17]. The main disadvantages being thermal injury, long equilibration time, and frequent calibrations [1],[3].

  ::   Monitoring of co2 levelTop

Clinical settings were CO2 monitoring may be useful is summarised in [Table:3]. The two modalities available are transcutaneous PCO2 (PtcCO2) and capnography to estimate end-tidal PCO2 (PetCO2).

Transcutaneous PCO2 (PtcCO2)

PtcCO2 monitors use a modified Stow-Severinghaus pH sensitive glass electrode that is applied to skin occlusively and heat it to 440C, which causes vasodilatation. PCO2 equilibrates across epidermis where it combines with the internal electrode buffer and an electric current is generated by changes in pH that is proportional to PaCO2[1],[12],[18],[19]. Reported correlations are 0.90-0.93 agreeing with PCO2 of ? 4 mm of Hg. There is a significant correlation between PtcCO2 and PaCO2 in a haemodynamically stable patients, it is not affected by patient’s gestational age, postnatal age, weight and blood pressure. Before use the sensor must be calibrated which requires about 10-15 minutes for equilibration after that it provides stable readings[1],[4],[12],[18].

PtcCO2 monitoring is used where continuous monitoring is indicated including non-intubated patients and especially after extubation. Difference in PtcCO2 and arterial CO2 can be used as shock parameter[12]. The main disadvantage being difficult calibration procedure, costly and fragile electrodes, which requires replacement of membranes periodically. Slow response time makes it less useful in detecting acute rises in PCO2. Sensor site has to be rotated every 2-6 hours to avoid burns. Calibration drift of PtcCO2 sensors and frequency of membrane change can be minimised by keeping the sensor in a moist environment [1],[2],[3],[4],[12],[20].

Local experience with PtcO2 and PtcCO2 suggest that it is found to be useful for initiation and weaning of ventilator patients and maintaining controlled hyperventilation.

Capnography (End tidal CO2: PetCO2)

Capnographs uses infrared spectroscopy or mass spectrometry readings of expired gas to analyse CO2 content. A device that measures and displays the breath to breath numeric values of CO2 is referred to as capnometer whereas a device that also displays the waveform of CO2 during the respiratory cycle is called a capnograph[1],[12],[21].

In healthy subjects, the end tidal CO2 (PetCO2) value is usually 1 mm less than PaCO2 (up to 5 mm of Hg). Correlation coefficients with PaCO2 are 0.69 to 0.92[1],[3],[12],[21]. In patients with stable cardiorespiratory status and a regular breathing pattern, in the PetCO2 is helpful in monitoring changes in patient’s ventilatory status or in guiding and assessing alterations in ventilator settings [1].

Monitors are of two types, sidestream or mainstream. Sidestream monitors aspirate a small sample of expired gas and transport it by way of a capillary tube to an infrared absorption chamber for measurement of CO2. Mainstream monitors are commonly used in the PICUs because they can be incorporated at the proximal end of an endotracheal tube[6].

Increase in the PetCO2 may occur with increase in cardiac output, injection of bicarbonate solution and hypoventilation. Decrease in the PetCO2 is known to occur with hyperventilation, decrease in cardiac output, pulmonary embolism, obstruction of endotracheal or leakage in ventilator circuit[2],[21]. Capnographs are useful as they may provide intermittent trend for larger intubated patients with chronic lung disease. Further, they can also be used qualitatively at the nostrils of non-intubated infants to detect airflow obstruction and apnoea[12]. As regards the limitations PetCO2 cannot be employed during high frequency ventilation and the additional dead space introduced by the airway adapter can cause CO2 retention[12],[21].

In authors’ experience capnographs are mainly useful for early detection of complications on ventilator (tube block, pneumothorax) besides initiation and weaning of ventilator patients. Further, it has also been found useful for instituting controlled hyperventilation for management of raised intracranial tension.

In a developing country like ours, where institutions have financial constraints, we would like to recommend combined pulse oximetry with end-tidal CO2 monitoring as the most feasible. Though PtcO2 and PtcCO2 monitors give an idea about tissue oxygenation, the exorbitant cost and the consumables required form the major limitations for its routine use in PICU. Last but not the least, any monitoring system should always be interpreted in the light of good and through clinical examination, which forms a backbone of any PICU.


1 Westenkirchner DF, Eigen H. Respiratory monitoring. In: Holbrook PR, editor. Textbook of Paediatric Critical Care. 1st ed. Philadelphia: WB Saunders; pp 423-429 .
2Tobin MJ. Respiratory monitoring. JAMA 1990; 264:244-251.
3Tobin MJ. Respiratory monitoring in the intensive care unit. Am Rev Resp Dis 1988; 138:1625-1642.
4Ravindranath T. Non-invasive monitoring in the paediatric ICU, Part I: Transcutaneous oxygen monitoring (PtcO2). Indian J Paediatr 1990; 57:169-173.
5Narang VP. The utility of the pulse oxymeter during cardiopulmonary resuscitation. Anaesthesiology 1986; 65:239-240.
6Cote CJ, Goldstein EA, Cote MA, Hoaglin DC, Ryan JF. A single blind study of pulse oxymetry in children. Anaesthesiology 1988; 68:184-188.
7Alexander CM, Teller LE, Gross JB. Principles of pulse oxymetry: theoretical and practical considerations. Anesth Analg 1989; 68:368-376.
8Schnapp LM, Cohen NH. Pulse oxymetry. Uses and abuses. Chest 1990; 98:1244-1250.
9Ravindranath T. Non-invasive monitoring in the paediatric ICU, Part III: Pulse oximeter. Indian J Paediatr 1990; 57:179-182.
10Nickerson BG, Sarkision C, Tremper K. Bias and precision of pulse oximeters and arterial oximeters. Chest 1988; 93:515-517.
11Severinghaus JW, Naifeh KH. Accuracy of response of six pulse oximeters to profound hypoxia. Anaesthesiology 1987; 67:551-558.
12Richardson D, Stark AR. Blood gas monitoring. In: Cloherty JP, Stark AR, editors. Manual of neonatal care. 3rd edn. Boston: Little, Brown and Co; 1992. pp 209-214.
13Huch R, Lubbers W, Huch A. Reliability of transcutaneous monitoring of arterial PO2 in newborn infants. Arch Dis Child 1974; 49:213-218.
14American Academy of Oediatrics. Task force on transcutaneous oxygen monitors. Paediatrics 1989; 83:122-126.
15Rome ES, Stork EK, Carlo WA, Martin RJ. Limitations of transcutaneous pO2 and pCO2 monitoring in infants with bronchopulmonary dysplasia.Pediatrics 1984; 74:217-220.
16Peabody JL, Gregory GA, Willis MM, Tooley WH. Transcutaneous oxygen tension in sick infants. Am Rev Respir Dis 1978; 118:83-87.
17Barker SJ, Tremper KK, Gamel DM. A clinical comparison of transcutaneous pO2 and pulse oxymetry in the operating room. Anaesth Analg 1986; 65:805-808.
18Hansen TN, Tooley WH. Skin surface carbon dioxide tension in sick infants. Paediatrics 1979; 64:942-945.
19Ravindranath T. Non-invasive monitoring in the paediatric ICU, Part II: Transcutaneous carbon dioxide Monitoring (PtcCO2). Indian J Paediatr 1990; 57:175-178.
20Cohen A. Blood gas monitoring. In: Cloherty JP, Stark AR editors. Manual of neonatal care. 2nd ed. Boston: Little, Brown and Co; 1985. pp180-184.
21Swedlow DB. Capnometry and Capnography: The anaesthesia disaster early warning system. Seminars in Anaesthesia 1986, 5:194-205.

Tuesday, December 6, 2022
 Site Map | Home | Contact Us | Feedback | Copyright  and disclaimer