Chapter 16: Pulmonary Function Testing, Plethysmography, and Pulmonary Diffusing Capacity

This chapter examines the many practical applications of pulmonary function testing (PFT) for evaluating lung function and disease progression in individual patients, as well as identifying broader public health concerns across populations. The full range of PFT procedures can assess airway resistance, functional residual capacity (FRC) and residual volume, pulmonary diffusing capacity, arterial blood gases (ABGs), exercise capacity, and even energy expenditure. While spirometry was introduced in Chaps. 4 and 6, the more detailed descriptions here include specific diagnostic criteria for grading the severities of obstructive and restrictive lung disease types, as developed by the American Thoracic Society and European Respiratory Society (ATS/ERS) (Table 16.1).

Spirometry is non-invasive, dynamic pulmonary function testing that indirectly assesses airway resistance, notably in medium-sized bronchi. Among the lung volumes estimated by spirometry to distinguish obstructive and restrictive disorders are: vital capacity (VC), forced vital capacity (FVC), forced expiratory volume in one second (FEV₁), peak expiratory flow (PEF) rate, and forced expiratory volume over the middle half of expiration (FEF_25-75). There are three distinct steps to obtaining a reliable FVC. First, maximal inspiration is measured by the subject inhaling as deeply as possible toward TLC before pausing for 1 second at that volume. Second, the subject exhales maximally as quickly as possible toward RV, continuing for at least 6 seconds, yielding an expiratory flow volume curve or loop [Fig. 16.1 (a)]. Third, the subject immediately inhales again toward TLC to obtain an inspiratory flow volume loop. At least three such efforts, but not more than eight, are performed in a given PFT session, with brief resting intervals using a normal V_T between efforts if needed.

With these data, the subject's largest FVC and FEV₁ from any of the attempts are recorded, even if the highest values for FVC and FEV₁ do not occur in the same expiratory effort. However, the highest two estimates of FVC and FEV₁ should not differ by more than 150 mL to assure an appropriate degree of repeatability. If the patient's best results differ by more than 150 mL, it is usually recommended that another full test be conducted. Similarly, once the expiratory flow-volume loops have been plotted, the patient's PEF results should be evaluated to ensure that peak rates (usually near the start of expiration) do not vary by more than 5% between the highest two values. If patient variance exceeds this benchmark, another full test should be completed.

The shape of the flow-volume loop during expiration [Fig. 16.1 (b)] is modified by dynamic airway compression that is effort-independent during expiration (Chap. 6). A "choke point" may exist throughout expiratory flow, so that whenever P_IP exceeds P_AW, airway collapse can occur (Fig. 16.2), making expiratory flow-volume loops concave. Dynamic compression of airways does not occur during inspiration because P_IP < P_AW throughout that phase of breathing, stenting the airways and alveoli open.

Comparing a subject's measured spirometry results and subsequent PFT data (eg, lung volumes and DL_CO described below) to appropriate reference values is essential to proper interpretation. Reference values are derived from studies of large, well-defined healthy populations of children and nonsmoking adults, expressed as normal intervals or frequency distributions representing 95% of the sample population. The most common method of analysis is to compare the observed data to the mean reference value, with results expressed both as absolute numeric values and as the percent of predicted value. The lower limit of normal (LLN) describes the lowest 5th percentile according to reference equations based on age, gender, and height, and sometimes ethnicity. Any value below the LLN is deemed clinically abnormal. To determine the severity of obstruction, the FEV₁ expressed as percentage predicted is often used (Table 16.2).

Some additional ATS/ERS guidelines for the use and interpretation of basic spirometry data include the following points:

1. Interpretation of spirometric results should include FVC, FEV₁, and FEV₁/FVC as well as PEF, FEF_25-75, and TLC to reduce the number of false positives (false alarms) in healthy adult or pediatric populations.

2. Spirometric results vary with time and testing conditions, but ≥15% yearly changes in FEV₁ or FVC are generally considered clinically significant. Regularly monitoring a patient's FVC may provide the most useful data for following the course of a restrictive chest disease (Chap. 24).

3. Low lung function levels by spirometry predict poor prognoses in patients with heart and lung disease, even among never-smokers. In patients with cystic fibrosis (CF) (Chap. 38), an FEV₁ <30% of predicted has been associated with a two-year mortality rate of 50%. Nevertheless, symptom severity or mortality risk for an individual patient cannot be predicted by FEV₁ alone.

4. Lack of acute improvement in a patient's FEV₁ with bronchodilators does not preclude a clinical response to long-term bronchodilator therapy, and the patient's reported volumes may change with medication due to reduction in dynamic gas trapping and chest hyperinflation. More on this subject is presented in later discussions of chronic obstructive lung disease and asthma (Chaps. 21 and 22) in the context of irreversible airway remodeling.

With these basic spirometric results in hand, clinicians then can decide whether follow-up procedures are warranted. The assessment and treatment algorithm (Fig. 16.3) developed by the ATS/ERS is a useful guide. However, patients may not present with every feature of a classic obstructive or restrictive pattern, depending on how immediate their illness is and its severity are affected by baseline lung function that existed before the onset of new symptoms or an acute exacerbation of preexisting disease.

Direct measurement of TLC is necessary to confirm a restrictive disease that may be suggested by a low spirometric estimate of VC. Gas dilution techniques (Chap. 4) and plethysmography can directly measure FRC and thus TLC, core PFT values not obtainable by spirometry alone. Unlike the helium (He) dilution and N₂ washout techniques described in Chap. 4, however, whole-body plethysmography estimates all intra-thoracic air volume, even gas-filled spaces that may be isolated from the major airways. Such noncommunicating air spaces are common in the form of the subpleural blebs or bullae in diseases like cystic fibrosis or congenital bullous emphysema (Chaps. 37 and 38). Thus, the use of plethysmography to measure FRC (and thus TLC) is preferred over gas dilution techniques in those patients with obstructive conditions where air trapping may occur, or in patients suspected of having coexistent restrictive and obstructive diseases.

Plethysmography is based on the physics of Boyle's law (Chap. 1), which states that isothermally changing the volume of a known amount of gas changes the gas pressure inversely, so that: P₁ · V₁ = P₂ · V₂. A whole-body plethysmograph is an airtight Plexiglas chamber (a "body box" in PFT lexicon) about the size of a telephone booth. According to current ATS/ERS criteria, a subject should be sitting upright in the box while wearing nose clips and breathing through a mouthpiece fitted with a pneumotachometer to measure airflow (Fig. 16.4). The chamber should be large enough to accommodate the subject without having to bend or hyperextend the neck to reach the mouthpiece. Multiple transducers are arrayed within the chamber to measure pressures within the box, in the airway, and across the pneumotachometer itself.

A subject first breathes quietly in the plethysmograph for 8-10 breaths, as system temperatures stabilize and a reproducible end-expiratory volume (ie, FRC) is achieved. Then as the subject next pauses at FRC, a shutter closes to occlude the airway for perhaps 2-3 seconds. With the shutter closed, the subject is asked to perform mild panting (f = 0.5-1 Hz) with enough effort to create oscillations in their P_AW of about ±8 mm Hg. Ideally, 3-5 technically correct panting maneuvers can be recorded before the shutter is released. Once the airway is reopened, the subject exhales completely to RV, then slowly inhales to TLC, and then slowly exhales back to RV. This sequence of steps is considered one complete FRC maneuver. The subject is then instructed to resume quiet normal respiration until the shutter is closed and another trial is attempted.

Although no air moves into the lungs during the attempted panting, PAW fluctuates by a measurable amount, ΔP_AW, while the fixed amount of gas in the lungs (= FRC) expands and contracts by an unknown amount, ΔV_L. During the same panting efforts, the volume of air surrounding the subject in the box also fluctuates by an unknown amount, ΔV_BOX, even as its pressure oscillates by a measurable amount, ΔP_BOX. Periodic air injections are made by a calibrated push-pull syringe operating at typical respiratory frequencies while the subject is in the box. These injections provide precise values for the ratio ΔV_BOX/ΔP_BOX with the subject inside. Those ratios are then used in sequential iterations of Boyle's law to solve first for the unknown ΔV_BOX (while the subject pants), then the subject's unknown ΔV_L (while panting), and finally the subject's actual FRC. To achieve an appropriate level of repeatability, at least three such panting maneuvers should be performed that yield estimates of the subject's actual FRC within 5% of each other.

The pulmonary diffusing capacity for carbon monoxide (DL_CO) is a standard lung function test that estimates alveolar-capillary diffusion for gases such as O₂ and CO₂. However, measuring rate-limiting O₂ transfer across the alveolar membranes presents technical difficulties that are caused by high background levels of O₂. Consequently, CO has been used extensively as a surrogate index of O₂ transfer, due primarily to its high affinity for Hb that results in CO being considered a diffusion-limited gas (Chaps. 3 and 9).

The most widely used DL_CO protocol is the single-breath method (Fig. 16.5). After a few normal breaths at rest, subjects exhale to RV and then inhale a standard mixture of gases (F_Ico = 0.003; F_IHe = 0.100; F_Io₂ = 0.207; F_In₂ = 0.690) to TLC. To complete the single-breath test properly, >85% of a subject's VC should consist of this standard gas mixture just inhaled to TLC. While holding at TLC for 10 seconds, subjects should avoid every positive airway pressures that decrease pulmonary blood flow and capillary volume, V_c (see Valsalva maneuver) or extremely negative airway pressures that artifactually increase Q̇ and V_c (see Müller maneuver). After the 10-second breathhold at TLC, the subject exhales completely to RV in a smooth, unforced, and uninterrupted manner. Owing to undiluted gases within the dead space volume, the first liter of exhaled air is discarded. The second liter of gas exhaled by the subject is collected and analyzed to compute both the equilibrated alveolar concentration of CO, F_Aco and its partial pressure, P_Aco (Chap. 8) at the end of the breathhold. DL_CO is then calculated as the total amount of CO absorbed by the subject from the amount inspired (F_Ico – F_Aco) divided by the P_Aco that served as the driving pressure gradient into the patient's pulmonary capillary blood. Simultaneous measurement of the equilibrium [He] allows an accurate estimate of lung volume by the dilution principles discussed in Chap. 4.

The DL_CO estimated for the entire lung by this test consists of the sequential diffusive capacities of the alveolar-capillary membrane and of pulmonary capillary blood (Chap. 9). Indeed, DL_CO is calculated using the equation for resistors in series:

where DM_CO = alveolar membrane diffusing capacity for CO, V_c = pulmonary capillary blood volume (Chap. 7), and Θ_CO is the blood transfer conductance coefficient for CO. The Θ_CO is the standard rate at which 1 mL of whole blood will take up CO in mL STPD per minute per milliliter of mercury of partial pressure. The unit for DL_CO is mL·(min^–1·mm Hg^–1) or mL·min^–1·mm Hg^–1, or it can be written as mL/(min·mm Hg). The units for DL_CO in most textbooks and research papers report DL_CO as mL/min/mmHg (mL/min/kPa in Europe). The most commonly used approximation for Θ_CO was derived by Roughton and Forster (1957) who showed that:

where [Hb] is expressed in g/dL of arterial or venous blood. Membrane diffusive resistance, 1/DM_CO and red cell resistance, 1/(Θ_CO · V_c) usually contribute about equally to the overall diffusive resistance of the lung, 1/DL_CO. To estimate the individual contributions made by DM_CO and V_c to DL_CO, the single-breath method is done twice, at an F_Io₂ = 0.207 and then at 0.897 to achieve two widely separated values for P_Ao₂ of about 100 mm Hg and 600 mm Hg, respectively. Using the subject's P_Ao₂ measured during each single-breath test, the value of 1/DL_CO is plotted on the y-axis and 1/Θ_CO is plotted on the x-axis. A straight line drawn through these two points yields the y-intercept (1/DM_CO) and the slope (1/V_c) (Fig. 16.6). Sufficient accuracy can usually be achieved with just four to six single-breath tests, usually two to three at low P_Ao₂ and two to three at high P_Ao₂.

No more than five DL_CO tests should be performed per session to avoid accumulation of HbCO, since HbCO increases by ~0.7% per single-breath test. For each 1% increase in HbCO, there is an approximate 1% decrease in DL_CO due to the reduction in 1/Θ_CO that is caused by such dysfunctional hemoglobins. Regular smokers may have an HbCO of 5%-10%, much higher than in nonsmokers, whose HbCO = 0.7%-1.0%. When two DL_CO tests fall within 3 mL·min^–1·mm Hg^–1 of each other, the average of both tests is reported. Normal month-to-month variations in DL_CO are about 5 mL·min^–1·mm Hg^–1. Thus larger variations are clinically meaningful. A classification scheme for severity of a decrease in DL_CO compared to normative values is shown in Table 16.3.