Physical Examination of the Respiratory System

Clinical assessment of respiratory function remains invaluable in the diagnosis and management of patients with respiratory failure despite all the technologic advances that have occurred in monitoring. Initial evaluation begins with assessment of the child’s comfort and activity level. Coexisting nonpulmonary conditions such as pain and anxiety may make this assessment difficult. Observation of the child’s body position, respiratory pattern, and body habitus provides important information as to his or her level of respiratory distress.

Often children in respiratory distress assume a body position that helps them breathe more comfortably. This position may be splinting of the chest in a patient with pneumonia or assuming the “sniffing position” to maintain an open airway in a child with an upper airway obstruction. An infant is less independent and therefore often is held in position by his or her caregivers, irrespective of comfort.

The respiratory pattern provides information regarding the work of breathing in a distressed child. The respiratory rate varies with age, but an early sign of distress is tachypnea. Additional signs of increased work of breathing include grunting or irregular respirations, nasal flaring, use of accessory muscles of respiration (strap muscles of the neck), and retractions.

Inspection of the shape of the chest wall may reveal abnormalities that affect pulmonary function. Increased anteroposterior diameter can be seen in conditions associated with hyperinflation (e.g., asthma or cystic fibrosis). Scoliosis in severe cases can cause a reduction in lung volume. Neuromuscular disorders may be associated with an “A-shaped” chest and lung hypoplasia/dysplasia. Thoracic asymmetry may be associated with neuromuscular or skeletal deformities, pneumothorax, or a paralyzed diaphragm. Fingers and toes should be examined for evidence of clubbing (painless enlargement of the connective tissues of the distal phalanges), which is nonspecific but may be indicative of chronic hypoxemia. Growth parameters and neurodevelopment should be obtained and compared with age-appropriate normal subjects to assist in evaluation of long-standing or associated diseases.

Detection of cyanosis centrally (lips, tongue) or peripherally (nail beds) may be difficult. Arterial oxygen tension must drop below 80 mm Hg before cyanosis can be detected clinically. Cyanosis may be absent in patients with severe anemia or missed when lighting is poor. It also may be intermittent and be seen only with exercise or change in position. Cyanosis that does not resolve with oxygen therapy may indicate right-to-left shunting of blood in the lungs or heart or the formation of methemoglobin or sulfhemoglobin following the ingestion of certain drugs.

Evaluation of breath sounds provides assessment of airflow through the tracheobronchial tree, the presence of fluid in or obstruction of the airways, and conditions outside the lung and pleural space. The child’s chest wall is thinner than that of the adult, which allows better access to breath sounds but impedes localization of the lesion because the breath sounds can be referred. Upper airway abnormalities may present with stridor or muffling of the voice. Bronchial breath sounds suggest consolidation, whereas wheezes result from narrowed airways. Crackles may be fine or coarse. They represent air bubbling through secretions and the reopening of closed airways. A friction rub may be heard when the inflamed surfaces of the pleura move against each other through the respiratory cycle. Breath sounds may be absent if there is a significant pleural effusion or complete lobar collapse resulting from a mucus plug or pneumothorax. Heart tones often are shifted away from the pneumothorax and toward the atelectasis because of complete airway obstruction. Assessment for pulsus paradoxus (i.e., exaggerated decrease in the pulse or systolic blood pressure with inspiration) should be made during the evaluation of severe airway obstruction or pulmonary embolus.

Although each assessment should include observation and a limited physical examination, it is recognized that interobserver repeatability of physical signs is poor and independent of the experience of the observer.² The assessments of clubbing, wheezes, friction rub, and crackles are the most reliable and reproducible.^2,3 The lack of accuracy of repeated physical examinations and the complexity of critically ill pediatric patients require adjunctive tests/assessments to monitor these patients.

Radiography

Portable chest x-ray films are the most common films taken in the PICU. The technical quality of the chest radiograph affects the interpretation; therefore, it is important to assess the film for adequacy of penetration, degree and symmetry of lung inflation, and degree of chest rotation. Chest x-ray films can be used to evaluate for cardiac, vascular, bone, and lung abnormalities, to assess for device placement (e.g., an endotracheal tube), and to determine the need for intervention.⁴ Lateral decubitus films help identify and quantify pleural effusions, pneumothorax, and the position of chest tubes/lines. Cross-table lateral films also may be used for these purposes but are harder to interpret. A chest computed tomography (CT) scan is used when details on the plain film are obscured by the superimposition of structures or an opaque hemithorax. A chest CT scan also can be used to guide drainage of fluid collections and assist in obtaining biopsy specimens. A high-resolution chest CT scan examines 1- to 1.5-mm slices at 10-mm intervals; therefore it can be used to illustrate lung parenchymal details better than a conventional CT scan, which examines 7- to 10-mm slices at 10-mm intervals.⁵ This detail can be helpful in distinguishing the pathologic process causing diffuse lung diseases that appear as diffuse lung shadowing on chest x-ray films. Chest ultrasound, ventilation/perfusion scanning, spiral CT, and magnetic resonance imaging may be useful adjutants depending on the disease process.

Evaluation of Gas Exchange

Episodic hypoxemia is common in critically ill adults and has been associated with increased mortality. The frequency and impact of episodic hypoxemia in pediatrics has not been studied. Additionally, whether monitoring and early treatment of episodic hypoxemia will improve patient outcome remains unanswered, but continuous monitoring of oxygen saturation in the PICU has become the standard of care.

Noninvasive Respiratory Monitoring

Pulse Oximetry

Pulse oximetry is considered a significant technologic advance that has improved patient safety.^1,6–9 Its ease of application and accuracy have resulted in widespread use. Pulse oximetry is commonly used to detect hypoxemia and to wean the oxygen concentration in patients undergoing mechanical ventilation.

Pulse oximetry is based on the principles that (1) the pulsatile absorbance detected is arterial blood and (2) oxyhemoglobin and reduced hemoglobin have different absorption spectra.⁶ Red (660-nm) and infrared (940-nm) wavelengths of light are used to determine the ratio of oxygenated to deoxygenated blood. Deoxygenated blood absorbs more red light, whereas oxygenated blood absorbs more infrared light. The two wavelengths are passed through an arterial bed, and the ratio of infrared and red light transmitted to the photodetector is determined. The ratio is calibrated against measurements of arterial oxygen saturations from human volunteers and their absorbance ratios.

Several factors may affect the accuracy of pulse oximetry. Pulse oximetry measures oxygen saturation (SaO₂). SaO₂ and PaO₂ are not linearly related; the oxyhemoglobin dissociation curve is sigmoid in shape (Figure 39-1). Large changes in PaO₂ at high levels of oxygen, the upper flat portion of the oxyhemoglobin dissociation curve, may occur with little change in saturation. Additionally, a reduction in oxygenation on the steep portion of the curve may not be appreciated as significant because only a small change in saturations will have occurred.^6,7 The accuracy of pulse oximetry falls with arterial oxygen saturations less than 70%.^1,6 At arterial oxygen saturations below 70%, pulse oximetry may be more appropriate for showing trends.⁶

Figure 39–1 Hemoglobin-oxygen (Hb-O₂) dissociation curve shows the percentage saturation of hemoglobin at each PO₂. When the hemoglobin concentration is known, the content of oxygen can be calculated. The total content includes the small additional content of oxygen in solution, which becomes significant at high levels of PO₂. The saturation scale on the left applies only to the Hb-O₂ line. The scale on the right shows content values for a normal hemoglobin level of 15 g/100 mL blood.

(Modified from Hlastala MP: Blood gas transport. In Culver BH, editor: The respiratory system, Seattle, 1997, ASUW Publications. Redrawn in Albert RK, Spiro SG, Jett R, editors: Clinical respiratory medicine, ed 2, St. Louis, 2004, Mosby Elsevier.)

Abnormal hemoglobin levels (carboxyhemoglobin, methemoglobin) that have similar absorbance spectra can lead to overestimation of the true SaO₂.⁶ Intravenous dyes and certain colors of nail polish may falsely lower pulse oximetry readings.

Pulse oximetry sensors may be unable to distinguish a true signal from background in low perfusion states that result in diminished pulsations (e.g., vasoconstriction, low cardiac output, and hypothermia). This situation usually is displayed as an inadequate pulse message.¹ Low peripheral perfusion and motion artifact are the most common causes of inaccurate pulse oximetry readings.^9–11 Newer designs of pulse oximeters with signal-progressing algorithms that detect and ignore motion and pulse rate interferences may overcome these limitations.^10,11

The complications of pulse oximetry are rare. They include skin burns and pressure necrosis in newborns.^1,6 Limited understanding of pulse oximetry by health care providers may be an underrecognized problem, along with time spent determining whether alarms are false.^9,12

Capnography

End-tidal carbon dioxide monitoring is the noninvasive measurement of exhaled carbon dioxide at the plateau of the carbon dioxide waveform (Figure 39-2). End-tidal carbon dioxide concentration reflects PaCO_2, cardiac output, percentage of dead space, and airway time constants. In healthy subjects, the end-tidal carbon dioxide concentration is 1 to 5 mm Hg less than the PaCO₂.¹³ End-tidal carbon dioxide concentration represents the PCO₂ of all ventilated alveoli, whether or not they are perfused. Therefore any condition that reduces pulmonary perfusion of ventilated alveoli increases the difference between PaCO₂ and end-tidal carbon dioxide. Comparison between end-tidal carbon dioxide concentration and PaCO₂ helps to differentiate between change in alveolar ventilation, CO₂ production, or pulmonary perfusion as a cause of the change in end-tidal carbon dioxide concentration. Additionally, end-tidal carbon dioxide monitoring can be used to verify tracheal intubation, detect complete airway obstruction, and monitor ventilation during sedation.^13–16

Figure 39–2 Normal capnograph. A to B is dead space, B to C is start of expiration (mix of dead space and alveolar gas), C to D is the alveolar plateau with point D representing the end-tidal value, and D to E is the start of inhalation.

Sampling of exhaled carbon dioxide can be at the patient-ventilator interface (mainstream), diverted to a monitor (sidestream), or an intermediate connection.¹⁶ The exhaled gas sample is exposed to various wavelengths of infrared light. The relative amount of light absorbed by the exhaled sample is compared with the amount of light absorbed by a sample that does not contain carbon dioxide. The difference between the two samples is the concentration of carbon dioxide. Carbon dioxide also can be measured semiquantitatively using a pH-sensitive indicator that changes from purple to yellow when exposed to carbon dioxide.¹⁴

Respiratory Mechanics

Institution of mechanical ventilation in a patient with acute respiratory failure may result in improvement and even normalization of blood gases and measures of ventilation without improving the underlying disease process. Assessment of respiratory mechanics can provide information on the status and progression of the disease process and information that is useful for minimizing ventilator-induced lung injury.

The development of microprocessor-equipped ventilators has simplified the measurement of respiratory mechanics.²¹ Simple observation of the displays of an intubated patient who is undergoing ventilation with a microprocessor-equipped ventilator provides significant information about the respiratory mechanics of the patient. Failure of expiratory flow to fall to zero (on the flow-time trace) before the onset of inspiration is the hallmark of breath-stacking, also termed autoPEEP. Inability of the patient to trigger the ventilator can be seen as “bumps” during expiration.

The simple maneuver of a rapid airway occlusion using the ventilator buttons can provide further information. The end-expiratory occlusion gives a direct measure of autoPEEP once a plateau in the airway pressure is reached.²¹ The occlusion of the airway at end-inspiration for 5 seconds will measure the inspiratory plateau pressure, which approximates alveolar pressure. The risk of barotrauma is thought to increase at inspiratory plateau pressures greater than 35 cm H₂O. The inspiratory plateau pressure also can be used to calculate the static compliance. Static compliance is the change in volume over the pressure change when flow has stopped in the airways and lung, measured as plateau pressure. Therefore the calculation of static compliance is

where V_t is tidal volume and PEEP is positive end-expiratory pressure. Dynamic compliance is the change in volume divided by the change in pressure when flow in the upper airway is zero but flow continues in the lung. It can be calculated at the bedside as

where PIP is peak inspiratory pressure. Dynamic and static compliance should be similar. Compliance may be markedly reduced in patients with ALI/ARDS because of the presence of pulmonary edema and alveolar flooding.

Studies that have used pressure-volume curves to set ventilator parameters and compared survival outcome have generated interest in the clinical application of pressure-volume curves. Three methods can be used to generate these curves.²² All methods require that the patient be deeply sedated and usually medically relaxed. This step is necessary to eliminate the influence of the respiratory muscles. The first method used was the super syringe method.²² In this method, the lungs are inflated in a stepwise manner using a calibrated syringe of a known volume (1.5 to 2 L) from the resting volume to a maximum of 40 to 50 cm H₂O. The limitations of the method are that it takes a long time, the patient must be repeatedly disconnected from the ventilator, and the results may be influenced by oxygen consumption, changes in gas temperature, and humidity. Because of these limitations, its use is mainly limited to research.

The second method is the multiple occlusion technique.²² The ventilator system is checked to ensure the absence of leaks. Multiple end-inspiratory occlusions are used to achieve different inflating volumes, each starting at the same lung volume. After several seconds of pause, the static pressure values are obtained and the exhaled volume is read off the ventilator after the occlusion is released. The volumes are plotted against the static pressures to make the pressure-volume curve. This procedure is fairly complex and may not be well tolerated by the patient. In addition, the sequential inflations at different tidal breaths may modify the lung volume history and influence the shape of the curve.

The third method is the low-flow technique.²² This method is based on the concept that the rate of airway pressure change is inversely related to the compliance of the respiratory system during a passive lung inflation with constant inspiratory flow (15 L/min). This method is rapid and reproducible and does not require that the patient be disconnected from the ventilator.

However generated, the pressure-volume curve is an S-shaped inspiratory curve of three segments (Figure 39-3). The lower inflection point is at the transition from the initial flat segment and the linear part of the curve. It is thought to reflect reopening of collapsed airways. The upper inflection point is at the transition of the linear part of the curve and the final portion of the curve. It is thought to reflect the start of overinflation. The portion of the curve between the upper and lower inflection points represents the target for ventilation in patients with ARDS to reduce the potential for ventilator-induced lung injury.²¹

Figure 39–3 Thepressure-volume curve is an S-shaped inspiratory curve that can be divided into three sections. The lower inflection point (P-flex) is at the transition from the initial flat segment and the linear part of the curve. It is thought to reflect reopening of collapsed airways. The upper inflection point is at the transition of the linear part of the curve and the final portion of the curve. It is thought to reflect the end of recruitment and the start of overinflation. The portion of the curve between the upper and lower inflection points represents the target for ventilation in patients with ARDS to reduce the potential for further lung injury.

(From Albert RK, Spiro SG, Jett JR, editors: Clinical respiratory medicine, ed 2, St. Louis, 2004, Mosby Elsevier.)

Endoscopy

Bronchoscopy and laryngoscopy are sometimes used in the PICU to visually examine the airway. Direct endoscopy can be performed with a flexible or rigid bronchoscope.^23,24 Flexible bronchoscopy offers the advantages of requiring little sedation and can be done through the endotracheal tube. Flexible bronchoscopy is used to evaluate the size, patency, and compression of the airway during respiration. It also is useful for evaluating the anatomy/structure of the airway and for removing or obtaining secretions from the lung. Rigid bronchoscopy continues to have a role for several reasons. Visualization of the posterior aspects of the larynx and cervical trachea is difficult using a flexible bronchoscope. Therefore better evaluation of vocal cord paralysis, laryngeoesophageal clefts, and H-type tracheoesophageal fistulas are obtained with a rigid bronchoscope. Removal of foreign bodies is performed more easily and safely with a rigid bronchoscope.

Summary

A variety of methods for assessing pulmonary status and function have been reviewed. Use of the derived information may allow ventilator settings to be altered and better matched to patient demands and comfort. Better understanding and interpretation of respiratory monitoring may result in improved patient care.

References are available online at http://www.expertconsult.com.