Respiratory Monitoring during Mechanical Ventilation

Todd W. Sarge

Ray Ritz

Daniel Talmor

Respiratory function may be simply classified into ventilation and oxygenation, where ventilation and oxygenation are quantified by the ability of the respiratory system to eliminate carbon dioxide and form oxyhemoglobin, respectively. The goal of respiratory monitoring in any setting is to allow the clinician to ascertain the status of the patient’s ventilation and oxygenation. The clinician must then use the data appropriately to correct the patient’s abnormal respiratory physiology. As with all data, it is imperative to remember that interpretation and appropriate intervention are still the onus of the clinician, who must integrate these data with other pieces of information (i.e., history and physical examination) to make a final intervention. In the acutely ill patient, the principal intervention with regard to respiratory function and monitoring usually involves the initiation, modification, or withdrawal of mechanical ventilatory support. This chapter focuses on respiratory monitoring for the mechanically ventilated patient.

Mechanical ventilation entails the unloading of the respiratory system by the application of positive pressure to achieve the goal of lung insufflation (i.e., inspiration) followed by the release of pressure to allow deflation (i.e., expiration). These simplified goals of mechanical ventilation are achieved in spite of complex and dynamic interactions of mechanical pressure with the physical properties of the respiratory system, namely elastance (E_rs) and resistance (R_rs). Furthermore, the patient’s neurologic and muscular conditions can also affect the goals of respiration, and they need to be monitored and evaluated as well. This chapter focuses on three specific areas in monitoring the mechanically ventilated patient: (a) the evaluation of gas exchange, (b) respiratory mechanics, and (c) respiratory neuromuscular function.

Gas Exchange

Basic Physics of Gas Exchange

As mentioned earlier, the primary function of the respiratory system is gas exchange (i.e., elimination of carbon dioxide while instilling oxygen to form oxyhemoglobin). Inadequate ventilation and oxygenation within the intensive care setting are typically caused by hypoventilation, diffusion impairment, or shunt and ventilation–perfusion (V–Q) mismatch.

Hypoventilation is defined as inadequate alveolar ventilation, and it is commonly caused by drugs, neurologic impairment, or muscle weakness/fatigue, which results in hypercarbia, according to the following equation:

where P_aCO₂ is the arterial partial pressure of carbon dioxide, [V with dot above]_CO₂ the production of carbon dioxide in the tissues, [V with dot above]_A the alveolar ventilation, and k the constant. Fortunately, the institution of mechanical ventilatory support readily corrects hypoventilation while the underlying cause is determined and corrected.

Diffusion impairment is a result of inadequate time for the exchange of oxygen across the capillary–alveolar membrane. This may occur due to pathologic thickening of the membrane or high-output cardiac states such as sepsis. However, the relative clinical significance of diffusion impairment
in the intensive care unit (ICU) is debatable. This is because the hypoxemia that results from the acute exacerbation of diffusion impairment is usually corrected by supplemental oxygen therapy. Furthermore, P_aCO₂ is rarely affected by diffusion impairments because it is highly soluble and is eliminated in multiple forms, such as bicarbonate.

The most common cause of hypoxemia in the ICU is ventilation–perfusion ([V with dot above]/[Q with dot above]) mismatch. One manifestation of [V with dot above]/[Q with dot above] mismatch is shunting. The true shunt fraction is the amount of cardiac output that results in venous blood mixing with end-arterial blood without participating in gas exchange. This has little effect on carbon dioxide tension; however, increases in shunt can lead to hypoxemia. The true shunt is expressed by the shunt equation as follows:

where Q_s and Q_t are the shunt and total blood flows, and C_c, C_a, and C_v represent the oxygen contents of pulmonary end-capillary, arterial, and mixed venous blood, respectively. The absolute oxygen content of arterial and mixed venous blood is calculated according to the oxygen content equation:

where C_x, S_xO₂, and P_xO₂ are the oxygen content, saturation, and partial pressure of oxygen within arterial and mixed venous blood, respectively. The oxygen content of end-capillary blood is estimated by the alveolar gas equation as follows:

where P_atm and P_H₂O are the partial pressures of the atmosphere and water (typically 760 and 47 at sea level), respectively; while F_iO₂ is the concentration of inspired oxygen; P_aCO₂ the arterial partial pressure of carbon dioxide; and RQ the respiratory quotient. The significance of true shunt is the fact that it is not amenable to supplemental oxygen therapy. Shunted blood reenters the circulation and dilutes oxygenated blood, resulting in a lower partial pressure of oxygen (P_aO₂) in the arterial system. Increasing the F_iO₂ will not improve oxygenation since the shunted fraction of blood does not meet alveolar gas.

V–Q mismatch is the result of inequality of the normal ventilation/perfusion ratio within the lung. V–Q mismatch is a spectrum of abnormal ratios signifying inadequate gas exchange at the alveolar level. It is possible with supplemental oxygen to overcome hypoxemia that is caused by an abnormal ratio of ventilation and perfusion, which differentiates this form of hypoxemia from true shunt. However, in the extreme, as the V–Q ratio in any alveolus approaches zero (i.e., ventilation approaches zero), it approaches true shunt as described above. At the other end of the spectrum, as the ratio in any alveolus approaches infinity (i.e., as perfusion approaches zero), it becomes physiologic “dead space,” which denotes alveoli that are ventilated but not perfused. Dead space is described in greater detail later in this chapter.

Direct Blood Gas Analysis

Monitors of gas exchange in the mechanically ventilated patient are typically directed at measurements of gas content and their gradients from the ventilator circuit to the alveolus and from the alveolus to the end-artery. As with most monitors, sources of error abound at many points as gases flow down their concentration gradients. The most accurate assessment of gas exchange is direct measurement from an arterial blood sample. This provides the partial pressures of carbon dioxide (P_aCO₂) and oxygen (P_aO₂) in the blood as well as the pH, base deficit, and co-oximetry of other substances such as carboxyhemoglobin and methemoglobin. Advantages of arterial blood gas (ABG) analysis include the fact that it is a fairly exact representation of the current state of the patient with regard to acid–base status, oxygenation, and ventilation. However, the limitations of blood gas analysis as a tool for monitoring gas exchange are numerous, including the fact that it is invasive, wasteful (blood), and noncontinuous (i.e., it is only a snapshot of the patient’s condition at the time the ABG is drawn).

Central and peripheral venous blood gas sampling has been proposed as an acceptable surrogate to arterial blood for monitoring pH, P_aCO₂, and base deficit [1]. The obvious advantage is mitigation of the invasiveness (i.e., patients are not required to have arterial access or punctures), while the disadvantages are the need for correlation and inability to assess oxygenation. With the exception of patient’s undergoing cardiopulmonary resuscitation [2], good correlation has been observed between arterial and venous pH and P_aCO₂ in patients with acute respiratory disease, with one study noting an average difference of 0.03 for pH and 5.8 for P_aCO₂ [1]. Another study in mechanically ventilated trauma patients also demonstrated good correlation between arterial and central venous pH, P_aCO₂, and base deficit; however, the authors concluded that the limits of agreement (-0.09 to 0.03 for pH and – 2.2 to 10.9 for P_aCO₂) represented clinically significant ranges that could affect management and therefore should not be used in initial resuscitation efforts of trauma patients [3].

Pulse Oximetry

Without question, pulse oximetry has been the most significant advance in respiratory monitoring in the past three decades. On the basis of established oxyhemoglobin dissociation curve (Fig. 31.1), pulse oximetry allows for the continuous, noninvasive estimate of a patient’s oxyhemoglobin and is expressed as a percentage of total hemoglobin. A detailed explanation of pulse oximetry including the physics and limitations is provided in Chapter 26.

Expired Carbon Dioxide Measurements

Capnometry is the quantification of the carbon dioxide concentration in a sample of gas. Capnography is the continuous
plotting of carbon dioxide over time to create a waveform (Fig. 31.2). When capnography is performed on continuous samples of gas from the airway circuit, a waveform is created whereby the plateau is reported as the maximum pressure in millimeters Hg and termed end-tidal carbon dioxide, or P_etCO₂. Although continuous capnography has limited usefulness in the ICU, capnometry has many clinical uses such as early detection of esophageal intubation. For a detailed explanation of capnography and its uses, please refer to Chapter 26.

Figure 31.1. This is a schematic demonstrating a normal hemoglobin dissociation curve with 50% saturation at P_aO₂ of 27 mm Hg and approaching 100% saturation at a P_aO₂ of 80 mm Hg.

Figure 31.2. This is a schematic representation of a capnograph waveform with the expiratory plateau delineating the end-tidal CO₂ between 30 and 40 mm Hg.

Dead Space Measurements

Dead space is defined as any space in the respiratory system that is ventilated but not perfused, such that no gas exchange can occur. Measurement of dead space is a marker of respiratory efficiency with regard to carbon dioxide elimination. Dead space can be subdivided into several categories including alveolar and anatomic. Anatomic dead space is the sum of the inspiratory volume that does not reach the alveoli and, therefore, participate in gas exchange. For mechanically ventilated patients, the anatomic dead space includes the proximal airways, trachea, endotracheal tube, and breathing circuit up to the Y-adapter. In normal human subjects, anatomic dead space in cubic centimeters is approximately two to three times the ideal body weight in kilograms, or 150 to 200 cm³. Alveolar dead space is the conceptual sum of all alveoli that are ventilated but not participating in gas exchange, otherwise described as “West Zone 1” [4]. Physiologic dead space (V_d) is the sum of anatomic and alveolar dead space and is usually expressed as a ratio of the total tidal volume (V_t) and can be calculated at the bedside using the modified Bohr equation:

where P_aCO₂ is the partial pressure of carbon dioxide and P_expCO₂ the partial pressure of carbon dioxide in the expired tidal volume of gas. The P_expCO₂ is difficult to measure, often requiring metabolic monitoring systems. However, volume capnography is a novel and simple approach to estimating P_expCO₂, involving measurements of carbon dioxide at the Y-adapter, and has been shown to correlate with more complex methods of metabolic monitoring [5]. The P_aCO₂ is often estimated as end-tidal carbon dioxide, P_etCO₂, however this is known to be inaccurate in disease states. Therefore, determination of the P_aCO₂ is most often measured directly by an ABG.

Physiologic dead space, V_d/V_t, is often increased in critical illnesses that cause respiratory failure, such as acute respiratory distress syndrome (ARDS) and chronic obstructive pulmonary disease (COPD). V_d/V_t can also increase with dynamic hyperinflation or auto-PEEP, as well as with overaggressive application of extrinsic positive end-expiratory pressure (PEEP) due to overinflation of alveoli impeding pulmonary artery blood flow—effectively increasing the West Zone 1 volume. Serial measurements of V_d/V_t have been shown to correlate with outcome in ARDS [6] and have been used to monitor the degree of respiratory compromise in critically ill patients [7]. However, these data have not translated into changes in treatment. Furthermore, Mohr et al. [8] found no appreciable difference in V_d/V_t while studying a series of posttracheostomy patients successfully weaned from mechanical ventilation compared with those who had failed weaning.

Pulmonary Mechanics

Basic Pulmonary Variables

Modern ventilators allow manipulation and measurement of the airway pressures (P_aw), including peak, plateau, mean and end-expiratory; volumes (V); and flows ([V with dot above]). Integration of these measurements allows assessment of the mechanical components of the respiratory system. The mechanical components are influenced by various disease states, and understanding these relationships may allow delivery of more appropriate ventilator support. The airway pressure (P_aw) is described by the equation of motion and must be equal to all opposing forces. For the relaxed respiratory system ventilating at normal frequencies, the major forces that oppose P_aw are the elastive and resistive properties of the respiratory system as they relate to the tidal volume (V_t) and flow ([V with dot above]), respectively:

where E_rs and R_rs are the elastance and resistance of the respiratory system, respectively.

Constant flow inflation in a relaxed, ventilator-dependent patient produces a typical picture as depicted in Figure 31.3 [9]. The rapid airway occlusion method at end inflation results in zero flow and a drop in P_aw from the peak value (peak inspiratory pressure, PIP) to a lower initial value and then a gradual decrease over the rest of the inspiratory period until a plateau pressure (P_plat) is observed. The P_plat measured at the airway represents the static end-inspiratory recoil of the entire respiratory system [10].

Measurement of the pleural pressures would allow further partitioning of these pressures into the lung (i.e., transpulmonary pressure, P_L) and chest wall (i.e., pleural pressure, P_pl)
components using the equation:

Figure 31.3. Schematic drawing of an airway pressure waveform delineating PEEP, auto-PEEP, peak inspiratory pressure (PIP), plateau pressure (P_plat), resistance, and compliance.

Unfortunately, direct measurements of pleural pressure are not practical in the intensive care setting. Therefore, pleural pressures have often been estimated by esophageal balloon catheters measuring the pressure in the esophagus (P_es), which lies in the proximity of the pleura at mid-lung height. This alters the earlier equation as follows:

where P_es is esophageal pressure.

These partitioned pressures are presented graphically in Figure 31.4.

Figure 31.4. Esophageal pressure tracing (P_eso) can be seen superimposed on the airway pressure tracing (P_air) during pressure control ventilation (PCV). Transpulmonary pressure has been estimated as the difference between these pressures with specific assumptions.

Compliance and Elastance

The static compliance (C_st,rs) of the respiratory system and its reciprocal, elastance (E_st,rs), are easily measured at the bedside using the aforementioned end-inspiratory airway occlusion method to produce zero flow and thus negate the resistive forces within the system. The elastance of the respiratory system (E_st,rs) is simply the pressure gradient between the total PEEP (PEEP_t) and the plateau pressure (P_plat) divided by the tidal volume (V_t) to yield the following equation:

E_st,rs may also be separated into its lung (E_L) and chest wall (E_cw) components by applying this equation to the P_L and P_es tracings obtained using P_es tracings (see Fig. 31.4) and by the equation:

The relative contributions of the lung and chest wall to the total elastance may be dependent on the etiology of respiratory failure. By way of example, pulmonary edema, either cardiogenic or as a result of ARDS, will lead to an elevated lung E_st and reduced compliance. ARDS of a nonpulmonary origin, for example, sepsis, may also lead to edema of the chest wall and abdominal distension. Both of these will lead to an additional increase in the E_st,rs as a result of an increase in the elastance of the chest wall.

Resistance

According to Ohm’s law, resistance is a function of the airway pressure gradient (Δ P_aw) divided by flow ([V with dot above]). Airway resistance can be measured in ventilator-dependent patients by using the technique of rapid airway occlusion during constant flow inflation. The maximum resistance (R_max) of the respiratory system is calculated by

And the minimum resistance (R_min) of the respiratory system can be computed by dividing

R_min reflects ohmic airway resistance, while the difference between R_max and R_min (Δ R) reflects both the viscoelastic properties (stress relaxation) and the time–constant variability within the respiratory tissues (pendelluft effect).

Pressure–Volume Curves

Static Measurements of the Pressure–Volume Curve

The gold standard of pressure–volume (P–V) curve measurement is the super-syringe method. Using a large calibrated syringe, increments of volume of 50 ± 100 mL gas are used to inflate the lung up to a total volume of 1,000 ± 2,000 mL. After each increment, the static airway pressure is measured during a pause lasting a few seconds during which there is no flow, and the pressure is the same in the entire system from the super-syringe to the alveoli. The lung is then deflated in the same manner and the pressure at each decrement of gas is recorded and the inspiratory and expiratory P–V curves are plotted. Continued oxygen uptake from the blood during this slow inflation–deflation cycle, coupled with equalization of the partial pressure of CO₂ in the blood and alveoli, will lead to a decrease in the deflation volume as compared with the inflation volume of gas. This artifact may appear to contribute to the phenomena of hysteresis. The more important mechanical cause of hysteresis is based on the slow inflation of the lung during the P–V curve maneuver. This slow inflation recruits or opens up areas of the lung with slow time constants and collapsed alveoli. This again will lead to a decreased expiratory volume and hysteresis.

Semistatic Measurements of the Pressure–Volume Curve

There are two methods for obtaining semistatic measurements of the P–V curve. These methods do not require the specialized skill and equipment needed for the super-syringe technique. The multiple occlusion technique uses a sequence of different-sized volume-controlled inflations with an end-inspiratory pauses [11,12]. Pressure and volume are plotted for each end-inspiratory pause to form a static P–V curve. If expiratory interruptions are also done, the deflation limb of the P–V curve may also be plotted. This process may take several minutes to complete, but yields results close to those obtained by static measurements. The second method is the low-flow inflation technique. This technique uses a very small constant inspiratory flow to generate a large total volume. The slope (compliance) of the curve is parallel with a static P–V curve only if airway resistance is constant throughout the inspiration. This is likely not the case as the low flow lessens airway resistance. The low flow also causes a minimal but recognizable pressure decrease over the endotracheal tube, which means that the dynamic inspiratory pressure–volume curve will be shifted to the right [13,14]. The long duration of the inspiration produces the same artifacts as the super-syringe technique, which is represented as hysteresis. Another drawback of static and semistatic methods is that they require stopping therapeutic ventilation while the maneuver is performed. The question has been raised, therefore, if these maneuvers are relevant in predicting the mechanical behavior of the lung under dynamic conditions, where resistance and compliance depend on volume, flow, and respiratory frequency.

Dynamic Measurements of the Pressure–Volume Curve

Dynamic measurement of the P–V curve allows continuous monitoring of the respiratory mechanics and in particular of the response to ventilator changes. These measurements are done with the patient on therapeutic ventilator settings and therefore may reflect more accurately the complex interaction of patient, endotracheal tube, and ventilator. A continuous display of pressure may be obtained either proximal to the endotracheal tube (at the patient connector or from the ventilator itself) or distal to the endotracheal tube. This pressure may then be plotted against tidal volume to produce a dynamic P–V curve. Each of these methods has advantages; however, the more commonly used proximal method suffers from the disadvantage of being heavily influenced by the resistance of the endotracheal tube. Neither the peak pressure nor the end-expiratory pressures are accurately recorded, and this will lead to an underestimation of compliance [12].

Only gold members can continue reading. Log In or Register to continue