Respiratory Care Management of COPD Exacerbations

Abstract

A COPD exacerbation is characterized by an increase in symptoms such as dyspnea, cough, and sputum production that worsens over a period of 2 weeks. Exacerbations are common. Respiratory therapists and physicians in an acute care setting often treat these patients. Targeted O₂ therapy improves outcomes and should be titrated to an S_pO2 of 88–92%. Arterial blood gases remain the standard approach to assessing gas exchange in patients with COPD exacerbation. The limitations of arterial blood gas surrogates (pulse oximetry, capnography, transcutaneous monitoring, peripheral venous blood gases) should be appreciated so that they can be used wisely. Inhaled short-acting bronchodilators can be provided by nebulizer (jet or mesh), pressurized metered-dose inhaler (pMDI), pMDI with spacer or valved holding chamber, soft mist inhaler, or dry powder inhaler. The available evidence for the use of heliox for COPD exacerbation is weak. Noninvasive ventilation (NIV) is standard therapy for patients who present with COPD exacerbation and is supported by clinical practice guidelines. Robust high-level evidence with patient important outcomes is lacking for the use of high-flow nasal cannula in patients with COPD exacerbation. Management of auto-PEEP is the priority in mechanically ventilated patients with COPD. This is achieved by reducing airway resistance and decreasing minute ventilation. Trigger asynchrony and cycle asynchrony are addressed to improve patient-ventilator interaction. Patients with COPD should be extubated to NIV. Additional high-level evidence is needed before widespread use of extracorporeal CO₂ removal. Care coordination can improve the effectiveness of care for patients with COPD exacerbation. Evidence-based practices improve outcomes in patients with COPD exacerbation.

Introduction

A COPD exacerbation is characterized by an increase in symptoms such as dyspnea, cough, and sputum production that worsens over a period of 2 weeks. Exacerbation history is a marker of disease severity. Recognizing the importance of exacerbations, the 2023 update of the Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines collapses the previous C and D groups into a single E group (Fig. 1).¹ A frequent exacerbator phenotype is defined as those patients with COPD who have 2 or more exacerbations per year.^2-4 Increasing frequency of exacerbations worsens the rate of decline in lung function and health-related quality of life, and an increased rate of hospitalized exacerbations is associated with increased risk of death.⁴ Doers et al⁵ identified 3 risk factors that were significantly associated with 30-d serious adverse events among patients in the emergency department (ED) with COPD exacerbations: triage venous P_CO2, Charlson comorbidity index, and hospitalization within the previous year.

Fig. 1.

Categorization of the severity of COPD based on exacerbation history and symptoms according to the 2023 Global Initiative for Chronic Obstructive Lung Disease report. mMRC = Modified Medical Research Council dyspnea scale; CAT = COPD Assessment Test.

In 2018, the incidence of COPD exacerbations for patients presenting to the ED in the United States was 6.2/10,000, which is about twice the rate in 2010.⁶ Interestingly, the mortality rate in this same cohort dropped nearly in half over the same time. There was a significant reduction in the rate of COPD exacerbations during the COVID-19 pandemic, likely related to mask wearing, social distancing, and hand hygiene.^7,8 Whether the rates of COPD exacerbations increase post pandemic to pre-pandemic levels remains to be seen. In this paper, I review the respiratory care management of severe COPD exacerbation requiring admission to the ED of hospital.

Who Was Dr Petty?

The Thomas L Petty Memorial Lecture is named for Dr Tom Petty (1932–2009), a legend of respiratory care.⁹ Along with being a co-author on the first paper to define ARDS,¹⁰ Dr Petty is considered the father of home O₂ therapy as the result of his work on the Nocturnal Oxygen Therapy Trial.¹¹ He spent a long and productive career at the University of Colorado, where he trained some of the future leaders in pulmonary medicine and respiratory care, including a former editor of this Journal, David J Pierson, who presented the first Petty Lecture in 2013.¹² I did not know Dr Petty well, but I had several opportunities to collaborate with him. I remember him not only for his knowledge but for his gentle affect dealing with a young respiratory therapist (RT) like me. It is my pleasure to have been invited to present the tenth Thomas L Petty Memorial Lecture.

O₂ Therapy

A randomized controlled trial (RCT) by Austin et al¹³ informs the practice of O₂ administration for patients with COPD exacerbation. They compared un-titrated O₂ therapy to titrated O₂ therapy for subjects with an acute COPD in the prehospital setting. In the un-titrated group, O₂ was administered by non–rebreather mask at a flow of 8–10 L/min. In the titrated group, O₂ was administered by nasal cannula at a flow sufficient for an S_pO2 of 88–92%. In subjects with confirmed COPD, mortality was significantly lower (2%) in the titrated O₂ group than in the un-titrated group (9%) (P = .02). This is a number needed harm of 14, meaning that for every 14 subjects treated with un-titrated O₂ therapy one will die.

A large observational study by Echevarria et al¹⁴ examined 1,027 subjects with COPD exacerbation who were receiving O₂ therapy at admission. Subjects were divided into groups according to admission S_pO2 < 88%, S_pO2 88–92%, S_pO2 93–96%, and S_pO2 97–100%. The lowest mortality was in the S_pO2 88–92% group (8.7%). Mortality was 17.1% for the S_pO2 97–100% group, 11.7% for the S_pO2 93–96%, and 17.1% for the S_pO2 < 88% group. After adjusting for baseline risk, the odds ratio for mortality was 2.97 in the S_pO2 97–100% group, 1.98 in the S_pO2 93–96% group, and 1.36 in the S_pO20 < 88% group. These results suggest that either too little O₂ (S_pO2 < 88%) or too much O₂ (S_pO2 > 92%) is associated with increased risk of death.

Clinical practice guidelines (CPGs) by the British Thoracic Society recommend targeting an S_pO2 of 88–92% in patients with COPD exacerbation.¹⁵ However, these guidelines allow for an S_pO2 of 94–98% if there is arterial blood gas confirmation of a normal pH and P_aCO2. In patients with COPD requiring O₂ therapy, CPGs of the American Association for Respiratory Care recommend a target S_pO2 of 88–92% or a P_aO2 of 55–75 mm Hg.¹⁶

A concern related to O₂ administration in patients with COPD is the potential for worsening hypercapnia. Traditional teaching has been that, in patients with chronic hypercapnia, the usual high P_CO2 stimulus to breathe is replaced by a low P_aO2 stimulus to drive—the hypoxemic (hypoxic) respiratory drive. It follows that an increase in P_aO2 due to O₂ administration results in suppression of the hypoxemic drive to breathe and an increase in P_aCO2. However, this is not supported by available evidence.¹⁷ Patients with COPD exacerbation have a high respiratory drive as measured by mouth occlusion pressure.¹⁸ Respiratory drive decreases with O₂ administration but remains higher than normal, and the change in drive is not sufficient to explain the increase in P_aCO2.¹⁸ With O₂ administration, the relationship between ventilation and perfusion is affected such that dead-space ventilation increases.^19-23 The increase in dead space results in an increase in P_aCO2, unrelated to respiratory drive. Another factor related to an increase in P_aCO2 with O₂ administration is the Haldane effect. The Haldane effect describes the interaction between hemoglobin binding of O₂ and CO₂. With the administration of O₂, the hemoglobin O₂ saturation increases, resulting in off-loading of CO₂ from hemoglobin with a resultant increase in P_aCO2. Thus, the mechanism for an increase in P_aCO2 with O₂ administration is unlikely related to a depression of hypoxemic (hypoxic) drive and more likely related to an increase in dead space and the Haldane effect.

Assessing Gas Exchange

Pulse oximetry is commonly used to assess arterial oxygenation in patients with COPD exacerbation. Indeed, pulse oximetry is used to target an S_pO2 of 88–92%, as recommended in CPGs.^15,16 But the accuracy limits of S_pO2 are not commonly appreciated. In critically ill patients, inaccuracies in pulse oximetry may have an important impact on the detection of hypoxemia and management of O₂ therapy.²⁴ In subjects with COPD exacerbation, Kelly et al²⁵ reported a bias of −0.8% but with limits of agreement of −8.2 to 6.7%. Amalakanti and Pentakota²⁶ reported that S_pO2 overestimated S_aO2 in subjects with COPD exacerbation. Despite the reported imprecision of pulse oximetry, it does seem useful to target an S_pO2 of 88–92% in many patients with COPD exacerbation. However, it is important that clinicians appreciate the imprecision of S_pO2 when using it for clinical decision making. Moreover, pulse oximetry cannot assess pH and P_aCO2. Despite the utility of S_pO2, pulse oximetry should not completely replace the need for arterial blood gas measurements.¹⁵

End-tidal P_CO2 (P_ETCO2) can be measured during spontaneous breathing, noninvasive ventilation (NIV), and invasive ventilation. Kartal et al²⁷ evaluated P_ETCO2 in 118 spontaneously breathing subjects with COPD in an ED setting. Agreement between P_aCO2 and P_ETCO2 was 8 mm Hg with a precision of 11 mm Hg (limits of agreement −13 mm Hg to 30 mm Hg). Fujimoto et al²⁸ compared P_aCO2 and P_ETCO2 in 30 samples from spontaneously breathing subjects with hypoxemic respiratory failure and 30 samples from subjects with hypercarbic respiratory failure. They reported a bias of 6 mm Hg for the difference between P_aCO2 and P_ETCO2 and a precision of 6 mm Hg (limits of agreement −12 mm Hg to 18 mm Hg). Others have reported imprecision in P_ETCO2 used for noninvasively²⁹ and invasively ventilated subjects with acute COPD.³⁰ The available evidence suggests that P_ETCO2 might be an imprecise estimate of P_aCO2 and thus should be used with caution.

Ruiz et al³¹ evaluated transcutaneous P_CO2 (P_tcCO2) in 81 subjects (34 with COPD exacerbation) with acute respiratory failure and severe hypercapnia. The bias between P_aCO2 and P_tcCO2 increased with the level of hypercapnia. For P_aCO2 > 60 mm Hg, the bias was 7 ± 5 mm Hg with limits of agreement of −3 to 16 mm Hg. Fujimoto et al²⁸ reported a bias of 1 mm Hg with limits of agreement of −10 mm Hg to 12 mm Hg for P_tcCO2. Sørensen et al³² evaluated the agreement between P_tcCO2 and P_aCO2 during COPD exacerbation, 57 transcutaneous measurements in 20 subjects. The bias between P_tcCO2 and P_aCO2 was 3 mm Hg, with limits of agreement of −2 mm Hg to 16 mm Hg. The bias for ΔP_tcCO2 was 2 mm Hg with limits of agreement −4 mm Hg to 8 mm Hg. The authors concluded that P_tcCO2 did not accurately reflect P_aCO2 but that assessment of P_tcCO2 for P_aCO2 changes was acceptable, and thus transcutaneous monitoring may be useful for continuous monitoring in conjunction with periodic arterial blood gas analysis.

Peripheral venous P_CO2 (P_vCO2) has been used as a surrogate for P_aCO2. In a systematic review of 6 studies, Sheng et al³³ reported a bias of 5 mm Hg between P_vCO2 and P_aCO2. This is consistent with the normal physiologic difference between P_vCO2 and P_aCO2. However, they did not report the limits of agreement. In a study by McKeever et al,³⁴ not included in the systematic review of Sheng et al,³³ the bias between P_aCO2 and P_vCO2 was 6 mm Hg—like Sheng et al.³³ However, the limits of agreement were wide, from −22 mm Hg to 11 mm Hg. These wide limits of agreement suggest that, in individual patients, there might be substantial inaccuracy for the use of P_vCO2 as a proxy for P_aCO2.

Arterial blood gases remain the standard approach to assessing gas exchange in patients with COPD exacerbation. An arterial blood sample can be analyzed for P_aO2, P_aCO2, pH, and S_aO2 (measured by CO-oximetry). However, with arterial blood gas analysis, there is a delay between sample acquisition and analysis. Moreover, arterial blood samples are intermittent and painful for the patient. This makes noninvasive alternatives like pulse oximetry attractive. It is important to appreciate the clinical implications and limitations when using arterial blood gas surrogates like S_pO2, P_ETCO2, P_tcCO2, and P_vO2.

Aerosol Therapy

According to the GOLD 2023 report,¹ short-acting inhaled β₂ agonists, with or without short-acting anticholinergics, are recommended as the initial bronchodilators to treat a COPD exacerbation. These inhaled drugs can be provided by nebulizer (jet or mesh), pressurized metered-dose inhaler (pMDI), pMDI with spacer or valved holding chamber, soft mist inhaler, or dry powder inhaler.³⁵ There are advantages and disadvantages for each device (Table 1). van Geffen et al³⁶ conducted a Cochrane review comparing bronchodilators delivered by nebulizer versus pMDI with spacer or dry powder inhaler for exacerbations of COPD. They concluded that there is a lack of evidence in favor of any delivery device over another for bronchodilators during exacerbations of COPD. In CPGs related to device selection and outcomes of aerosol therapy,³⁷ the authors concluded that devices used for the delivery of bronchodilators and steroids can be equally efficacious. Moreover, they state that the following questions should be addressed when selecting an aerosol delivery device:

1. In what devices is the desired drug available?

2. What device is the patient likely to be able to use properly?

3. Which devices are the least costly?

4. Can all types of inhaled drugs (eg, short-acting beta₂ agonist, anticholinergic) be delivered with the same type of device?

5. Which devices are the most convenient for the patient or medical staff to use?

6. How durable is the device?

7. Does the patient or clinician have any specific device preferences?

Proper technique is likely the most important consideration in device selection provided that the drug can be delivered with the selected device.

Table 1.

Advantages and Disadvantages of Various Aerosol Delivery Devices That Can Be Used During a COPD Exacerbation

A jet nebulizer can be powered with O₂ or air flow. Bardsley et al³⁸ evaluated O₂ versus air-driven nebulizers for exacerbations of COPD. This was a parallel group double-blind RCT in 90 hospitalized subjects with COPD exacerbation. Subjects were randomized to receive aerosolized albuterol with the nebulizer driven by air or O₂ at 8 L/min. Hypercapnia was evaluated by P_tcCO2. The mean ± SD change in P_tcCO2 was 3 ± 2 mm Hg in the O₂ group and 0 ± 1 mm Hg in the air group. The proportion of subjects with a P_tcCO2 change ≥ 4 mm Hg was 18/45 (40%) and 0/44 (0%) for O₂ and air groups, respectively. The increase in P_tcCO2 with the use of O₂-powered nebulizers suggests that the nebulizer should be powered with air in patients with COPD exacerbation.

A concern with the use of nebulizers that arose during the COVID pandemic was the generation of fugitive aerosols. In many hospitals in the United States, the use of nebulizers was banned due to concern for caregiver protection. In retrospect, this was an overreaction.³⁹ There are several strategies that can be used to minimize the release of fugitive aerosols. A mouthpiece should be used rather than a face mask. Scavenging devices and exhalation filters can be used. A breath-actuated nebulizer might release less aerosol into the room. Also, fugitive aerosol release may be less with a mesh nebulizer than with a jet nebulizer.⁴⁰ The use of personal protective equipment by caregivers is the most important defense against fugitive aerosols.⁴¹

Heliox

The low density of helium makes it attractive as a therapeutic agent in the setting of airways obstruction.⁴² However, COPD is characterized by disease of small airways, where flow is laminar. Laminar flow is density independent and viscosity dependent. Thus, it can be predicted that changing the gas density, such as the use of heliox (a gas mixture of helium and O₂), is not likely to affect flow in patients with COPD.⁴³ Use of heliox as a driving gas for jet nebulization of bronchodilators during the first 2 h of treatment of a COPD exacerbation failed to improve FEV₁ faster than the use of air.⁴⁴ Jolliet et al⁴⁵ conducted a multi-center RCT assessing the efficacy of heliox in severe COPD exacerbations. They reported that heliox improved respiratory acidosis, encephalopathy, and the breathing frequency (f) more quickly than air/O₂ but did not prevent NIV failure. A systematic review and meta-analysis⁴⁶ reported that, compared to air/O₂, heliox did not reduce the rate of NIV failure in hypercapnic COPD exacerbation. But it is associated with a lower incidence of NIV-related adverse events and a shortening of ICU stay with no increase in hospital costs. Another systematic review and meta-analysis evaluated the use of heliox during mechanical ventilation.⁴⁷ The authors found that there was no conclusive evidence indicating the beneficial effect supporting the use of heliox during mechanical ventilation. The available evidence for the use of heliox for COPD exacerbation is weak, and its use is not generally recommended.

Noninvasive Respiratory Support

Noninvasive Ventilation

NIV is standard therapy for patients who present with COPD exacerbation and is supported by CPGs. CPGs by the American Thoracic Society and the European Respiratory Society state:⁴⁸

1. We suggest NIV not be used in patients with hypercapnia who are not acidotic in the setting of a COPD exacerbation. (Conditional recommendation, low certainty of evidence.)

2. We recommend NIV for patients with acute respiratory failure leading to acute or acute-on-chronic respiratory acidosis (pH ≤ 7.35) due to COPD exacerbation. (Strong recommendation, high certainty of evidence.)

3. We recommend a trial of NIV in patients considered to require endotracheal intubation and mechanical ventilation unless the patient is immediately deteriorating. (Strong recommendation, moderate certainty of evidence.)

Thus, NIV should be considered with respiratory acidosis and an f > 24 breaths/min despite standard medical therapy. NIV is the preferred choice for patients with COPD who develop acute respiratory acidosis during hospital admission. With severe respiratory acidosis, patients should be closely monitored with access to endotracheal intubation and invasive ventilation if not improving.⁴⁸ Note that NIV is reserved for patients with respiratory acidosis; it is not indicated for those who do not present with respiratory acidosis.

Mosher et al⁴⁹ performed a retrospective cohort study of 427 subjects with COPD exacerbation to evaluate the treatment course of NIV and factors associated with NIV treatment failure. Their NIV success rate was 78%; 81% of failures occurred within 48 h of initiation. The first 8 h following NIV were the critical time where the subject was at high risk for failure. They found a median time to NIV treatment failure and success was 8 h and 16 h, respectively. Increasing age, body mass index, bicarbonate level, and creatinine level were identified as risk factors associated with treatment failure or persistent treatment.

Fisher et al⁵⁰ conducted in-depth interviews with key stakeholders from a sample of high-performing hospitals regarding the successful use of NIV for COPD exacerbation. Interviews were conducted with 32 participants from 7 hospitals (15 RTs, 10 physicians, and 7 nurses). The authors identified 3 domains that characterized effective NIV use. Key processes included timely identification of appropriate patients, early initiation of NIV, frequent reassessment of patients, and attention to patient comfort. Necessary structural elements included adequate equipment, enough qualified RTs, and flexibility in staffing. Important contextual factors included provider buy-in, RT autonomy, interdisciplinary teamwork, and staff education. Of note, RT autonomy facilitated essential processes of NIV use.

A safety issue with the use of NIV is facial skin breakdown. Common sense strategies can help, such as avoidance of fitting the interface too tightly, rotating interfaces to distribute pressure points on the face, and providing breaks in the use of NIV as tolerated. During NIV breaks, high-flow nasal cannula (HFNC) can be used, a strategy called sequential use of NIV and HFNC.^51,52 A total face mask might reduce the risk of skin breakdown when compared to an oronasal mask.^53,54 NIV with humidification has a potential disrupting effect on the barrier function of facial skin, increasing the risk of skin breakdown. This is in conflict with the usual practice of providing humidification during NIV to prevent airway drying and to improve patient comfort.⁵⁵ The risk of breakdown can be decreased using cushioning materials between the interface and the skin. Orlov and Gefen⁵⁶ compared the protection of 3 foam-based wound dressings to a hydrocolloid dressing when applied to between the interface and the skin. Using a face model, they found that foam-based dressings performed substantially better than the hydrocolloid but also that foam dressings differed in their protective performance.

Understanding the experience of patients with COPD exacerbation treated with NIV might inform strategies that can improve their experience and tolerance with NIV. McCormick et al⁵⁷ used a human-centered design perspective to study the subject experience when using NIV (emotions, experience, thinking). Themes supporting NIV tolerance were provider trust, a favorable impression of the hospital and staff, understanding why NIV was needed, how NIV helps and how long it is needed, immediate relief of the suffocating sensation, familiarity with similar treatments, use of meditation and mindfulness, and the realization that treatment was beneficial. Themes that discouraged NIV tolerance were physical and psychological discomfort with NIV, feeling of loss of control, and being misinformed. Appreciation for the patient experience can be useful to improve patient tolerance of NIV.

Early in the COVID-19 pandemic, NIV was classified as an aerosol-generating procedure, which restricted its use. But it became clear that the risk of disease transmission is likely low with NIV.^58,59 Several mitigating procedures have been proposed such as use of a non-vented mask or helmet interface, using a dual-limb ventilator circuit with a filter placed at expiratory port, or use of a single-limb ventilator with filter placed at exhalation port or between the mask and exhalation port.³⁹ However, it should be appreciated that placing filters in the circuit has the potential to interfere with the performance of the ventilator.^60,61 With the use of personal protective equipment, the use of filters and other mitigating procedures might not be necessary.

Patients with COPD exacerbation typically receive inhaled short-acting bronchodilators as part of their care. These can be delivered effectively with a pMDI and spacer, jet nebulizer, or mesh nebulizer.^62,63 The aerosol-generating device should be placed directly at the mask,⁶⁴ or in the case of the mesh nebulizer, it can be incorporated directly into the mask (Fig. 2).⁶⁵ For a patient receiving benefit from the therapy, NIV should not be interrupted for the administration of inhaled bronchodilators.

Fig. 2.

A: Jet nebulizer to mask for noninvasive ventilation (NIV). B: Mesh nebulizer incorporated into mask for NIV. C: Pressurized metered-dose inhaler with spacer to mask for NIV. B from reference 65.

High-Flow Nasal Cannula

There is much academic and clinical interest in the use of HFNC for acute respiratory failure, including in patients with COPD exacerbation. The mechanisms of action for HFNC make it attractive for use in this patient population.⁶⁶ It reduces upper-airway resistance and washes out upper-airway dead space, both of which should increase the efficiency of ventilation and lower P_aCO2. It delivers warm and humidified gas, which might facilitate airway clearance. The precise F_IO2 is beneficial, and the small amount of positive pressure applied might counterbalance auto-PEEP. It is comfortable for the patient and easy to use for the care provider. But the important question is, when compared to NIV, does it reduce the risk for intubation and mortality?

Robust high-level evidence with patient important outcomes (eg, intubation, mortality) is lacking for the use of HFNC in patients with COPD exacerbation. Plotnikov et al,⁶⁷ in a multi-center prospective observational study of 138 subjects, reported a reduction in P_aCO2 (57 to 52 mm Hg) and f (29 breaths/min vs 21 breaths/min) compared with baseline. Cong et al,⁶⁸ in an RCT of 168 male subjects, reported that both NIV and HFNC were effective in improving gas exchange. Braunlich et al⁶⁹ reported the use of HFNC in 38 subjects with COPD exacerbation who did not tolerate NIV, showing a reduction in P_aCO2. A multi-center non-inferiority RCT enrolled 80 subjects with P_aCO2 ≥ 55 mm Hg.⁷⁰ HFNC reduced P_aCO2 and was statistically non-inferior to NIV in terms of P_aCO2. However, 32% of subjects receiving HFNC required NIV by 6 h and almost 50% needed NIV during hospitalization, and subjects who switched to NIV remained longer on ventilation. The authors stated that, despite the statistical non-inferiority criteria being met, these results question the clinical effectiveness of HFNC when compared to NIV. Of note, subject important outcomes such as intubation rate and mortality rate were not reported in these studies. In a multi-center RCT, Xia et al⁷¹ compared HFNC to conventional O₂ therapy and found that HFNC did not reduce the need for intubation among subjects with COPD exacerbation and mild hypercapnia; secondary analyses suggested HFNC increased the length of hospital stay and hospital costs.

Clinical and bench studies⁷² have described the delivery of aerosols with HFNC using mesh nebulizers,⁷³ jet nebulizers,⁷⁴ and pMDIs with spacer.⁷⁵ Beuvon et al⁷³ conducted a physiological crossover study of 15 subjects with severe COPD exacerbation. Pulmonary function tests were performed while breathing through HFNC at 30 L/min alone and after albuterol delivered by vibrating mesh nebulizer through HFNC. Mean FEV₁ significantly increased after albuterol by a mean of 87 mL (95% CI 30–145 mL). Similar results were reported by others in subjects with stable COPD.^76,77

At the time of this writing, the benefit of HFNC in patients with COPD exacerbation is unclear. RCTs with subject important outcomes are needed. In a systematic review and meta-analysis, Yang et al⁷⁸ concluded that HFNC might be used as an alternative to NIV for COPD exacerbation with mild to moderate hypercapnia. However, this is not a robust finding and was based on only 2 RCTs enrolling 122 total subjects; the non-inferiority of HFNC reported is not definitive and might be the result of an underpowered meta-analysis. Given the established benefit of NIV for COPD exacerbation, future trials should use NIV as the control group. HFNC might be used during NIV breaks, and short-acting β agonists can be delivered by HFNC during these breaks.

Invasive Mechanical Ventilation

Some patients with COPD exacerbation accept invasive mechanical ventilation. This is offered to the patient if the exacerbation worsens despite aggressive less invasive therapies such as NIV. Even in the hands of skilled physicians and RTs, NIV failure rates are about 20%.⁴⁹ Hajizadeh et al⁷⁹ conducted a retrospective cohort study of 4,791 subjects with COPD exacerbation on long-term O₂ treatment who received invasive mechanical ventilation for COPD exacerbation. Mortality was high; 23% died in the hospital, and 45% died within 12 months. Readmissions were also high, 67% in the subsequent 12 months. Of those discharged, 27% were to a nursing home or skilled nursing facility. The authors suggest that these outcomes have important implications for physician-patient discussions about prognoses if a COPD exacerbation occurs and whether to accept invasive mechanical ventilation.

As with all mechanically ventilated patients, lung-protective ventilation is important. However, the strategies and priorities are different than that for patients with ARDS. With ARDS, there is a focus on alveolar recruitment. But with COPD, the lungs are hyperinflated, and the focus is on air trapping and auto-PEEP. PEEP is used for alveolar recruitment in ARDS but to counterbalance auto-PEEP in COPD. Respiratory system compliance (C_RS) is low in patients with ARDS but preserved or increased in COPD. As in patients with ARDS, driving pressure and plateau pressure (P_plat) should be monitored in patients with COPD. But in patients with COPD, a high P_plat (> 28 cm H₂O) or driving pressure (> 15 cm H₂O) is due to dynamic hyperinflation and auto-PEEP. Note that, in the presence of auto-PEEP, driving pressure is the difference between P_plat and total PEEP, including any measured auto-PEEP.

The primary goals when setting the ventilator for a patient with COPD are (1) minimize air trapping, (2) avoid overdistention, (3) promote patient-ventilator synchrony, (4) provide adequate ventilation, and (5) provide adequate oxygenation. Note that adequate ventilation and oxygenation do not mean normal arterial blood gases. Normalization of P_aCO2 might increase the degree of air trapping and auto-PEEP. Moreover, the increase in auto-PEEP might drive up the P_plat and result in additional overdistention. Generally, ventilation should only be sufficient to achieve the baseline P_aCO2.

Modes and Initial Settings

The choice of volume control ventilation versus pressure control ventilation is based on individual bias or institutional preference. Evidence is lacking to support that either is superior to the other. A concern with volume control is the high peak inspiratory pressure (PIP) that can occur with volume control ventilation. However, the high PIP is the result of elevated airway resistance (R_aw) (Fig. 3). It can also be the result of a high P_plat resulting from air trapping and auto-PEEP. Ventilation and P_aCO2 are maintained with volume control ventilation if auto-PEEP increases but with an increase in P_plat and PIP (overdistention). With pressure control ventilation, there is less risk for overdistention if auto-PEEP increases but with a decrease in ventilation and an increase in P_aCO2. Indeed, if auto-PEEP increases to the pressure control setting (+ PEEP), there is no ventilation.

Fig. 3.

Airway pressure and flow waveforms. Note the large difference in peak inspiratory pressure and plateau pressure (P_plat) in the presence of high airways resistance. Also note the P_plat is affected by the level of auto-PEEP. Adapted from reference 81. PIP = peak inspiratory pressure. P_plat = plateau pressure.

Initial ventilator settings should be selected to minimize the development of auto-PEEP. These include a tidal volume (V_T) of 6–8 mL/predicted body weight, an inspiratory time (T_I) of 0.8–1.2 s, and an f of 12–15 breaths/min. This is to maximize expiratory time (T_E) and, in that way, to minimize air trapping.

Auto-PEEP

Incomplete emptying of the lungs occurs if the expiratory phase is terminated prematurely. The pressure resulting from this trapped gas is called auto-PEEP, intrinsic PEEP, or occult PEEP.⁸⁰ Auto-PEEP increases end-expiratory lung volume and thus causes dynamic hyperinflation.⁸¹ Auto-PEEP is measured by applying an end-expiratory pause (Fig. 4). The pressure measured at the end of this maneuver more than the PEEP set on the ventilator is defined as auto-PEEP. For a valid measurement, the patient must be relaxed and breathing in synchrony with the ventilator, as active breathing invalidates the measurement (Fig. 4). The end-expiratory pause method can underestimate auto-PEEP when some airways close during exhalation (Fig. 5). In spontaneously breathing patients, measurement of esophageal pressure can be used to determine auto-PEEP (Fig. 6).

Fig. 4.

Auto-PEEP is measured using an end-expiratory pause. If the patient is actively exhaling, pressure increases during the breath-hold, as indicated by the dashed line; this invalidates the measurement of auto-PEEP. Adapted from reference 81. PIP = peak inspiratory pressure.

Fig. 5.

The auto-PEEP measured at the proximal airway is less than the auto-PEEP in some lung regions if airways collapse during exhalation. Adapted from reference 81.

Fig. 6.

Airway pressure, flow, volume, and esophageal pressure (P_es) waveforms in a patient with auto-PEEP. Note the decrease in P_es required to trigger the ventilator, which represents the amount of auto-PEEP. Also note that flow does not return to zero at the end of exhalation, and the inspiratory effort does not trigger the ventilator. Adapted from reference 81. P_aw = airway pressure; P_es = esophageal pressure.

Auto-PEEP is a function of ventilator settings, V_T and T_E, and lung function, R_aw and C_RS:⁸¹

$$\text{auto}-\text{PEEP} = {\mathrm{V}}_{\text{T}} / {\text{C}}_{\mathrm{R}\text{S}} \times ({\text{e}}^{({\mathrm{K}}_{\text{X}} \times {\mathrm{T}}_{\text{E}})} - 1)$$

where K_x is the inverse of the T_E constant (τ; resistance × compliance), 1/τ_E. Auto- PEEP is increased with increased resistance and compliance. The administration of bronchodilators and airway clearance decreases R_aw and thus decreases auto-PEEP. Increased f or increased T_I increases the amount of auto-PEEP, as both decrease T_E. An increase in V_T also increases the potential of auto-PEEP. A clinician question arises related to whether it is better to use a higher V_T and lower f or a lower V_T and higher f in terms of auto-PEEP. Mathematically, it can be shown that, for the same minute ventilation (V̇_E), the auto-PEEP is lower with a higher V_T and slower f (Table 2).⁸² However, the greater effect on auto-PEEP is to lower the V̇_E and allow permissive hypercapnia. Clinically, auto-PEEP can be decreased by decreasing V̇_E, increasing T_E (decreasing rate or T_I), or decreasing R_aw (eg, bronchodilator administration). End-expiratory flow is present if R_aw is high and T_E is not sufficient, indicating the presence of air trapping (auto-PEEP). This can be examined by assessing the ventilator graphs for the presence of end-expiratory flow⁸¹ or by clinical examination (inspection, palpation, auscultation).⁸³

Table 2.

Effect of Minute Ventilation and Combinations of Breathing Frequency and Tidal Volume on auto-PEEP

There are several complications related to the dynamic hyperinflation resulting from auto-PEEP, which is accompanied by an increase in P_plat.⁸⁴ Due to the pressure-volume relationship of the respiratory system, an increase in lung volume increases the work of breathing. The increase in intrathoracic pressure can lead to hemodynamic compromise. A quick test to determine whether hemodynamic compromise is due to auto-PEEP is to briefly disconnect the patient from the ventilator. With ventilator disconnection, there is a prolonged exhalation, leading to a decrease in auto-PEEP, a decrease in lung volume, a decrease in intrathoracic pressure, and an improvement in hemodynamics (eg, an increase in blood pressure).⁸⁵ This is a diagnostic test, not a therapeutic procedure as the effect is short lived; when the patient is reconnected to the ventilator, the dynamic hyperinflation and hemodynamic compromise return unless the V̇_E is reduced. Overdistention can lead to air leak and pneumothorax. The hyperinflation also increases physiologic dead space, which contributes to hypercapnia.

Trigger asynchrony can result from auto-PEEP. Auto-PEEP is a threshold load that must be overcome by inspiratory effort to change pressure or flow at the proximal airway to trigger the ventilator. Ineffective efforts occur if the patient cannot generate enough effort to overcome the threshold load of the auto-PEEP. Notching in the expiratory flow waveform suggests the presence of ineffective efforts (Figure 7). The presence of ineffective efforts and flow limitation suggests that PEEP might effectively counterbalance auto-PEEP (Figure 8),⁸⁶ and an increase in PEEP might not affect P_plat and overdistention. If pushing on the abdomen, which increases intrathoracic pressure, results in no additional expiratory flow, flow limitation is present (Figure 9). Neurally adjusted ventilatory assist (NAVA) is a ventilator mode that might improve patient-ventilator interaction in the setting of ineffective efforts.⁸⁷

Fig. 7.

Flow, airway pressure, and esophageal pressure in a patient with severe COPD. The arrows represent ineffective efforts. Adapted from reference 81.

Fig. 8.

Effect of PEEP, auto-PEEP, and trigger effort in the setting of flow limitation. Adapted from reference 81.

Fig. 9.

Effect of increased intra-abdominal pressure on expiratory flow in a patient with expiratory flow limitation. Note that there is no change in expiratory flow as intra-abdominal (and hence intrathoracic) pressure increases. Adapted from reference 81.

Caramez et al⁸⁸ described 3 possible effects of PEEP in the setting of auto-PEEP. In patients with pure expiratory flow limitation, there is no change in hyperinflation (P_plat) with an increase in PEEP until about 80% of the auto-PEEP (Figure 10).⁸⁶ Without flow limitation, an increase in PEEP increases P_plat. In some cases, a paradoxical response occurs in which increases in PEEP decreases P_plat. This paradoxical response might occur in patients with expiratory flow limitation and highly heterogeneous lungs. The effect of PEEP on hyperinflation is unpredictable, and thus P_plat should be assessed when increasing PEEP.⁸⁹ PEEP that results in an increase in P_plat should be avoided.

Fig. 10.

In the top graphic, a patient with COPD has an auto-PEEP of 10 cm H₂O. When the PEEP is increased to 8 cm H₂O, there is no change in peak inspiratory pressure and plateau pressure, consistent with flow limitation.

A concern with ineffective efforts is the potential for lung injury related to pendelluft, although this has not been studied in the setting of auto-PEEP. With pendelluft, a vigorous inspiratory effort results in flow from non-dependent to dependent alveoli before the breath is triggered, which results in overdistention of dependent alveoli. This occurs with no flow from the ventilator and no change in P_plat. This concern is theoretical but might occur with ineffective efforts associated with auto-PEEP and COPD as has been described in the setting of ARDS.⁹⁰

Aerosol Delivery

Aerosols can be delivered effectively during invasive mechanical ventilation.^63,91 Either mesh nebulizers, jet nebulizers, or a pMDI with spacer can be used. At the time of this writing, dry powder inhalers and soft mist inhalers cannot be used in intubated patients. Aerosol therapy is commonly used in mechanically ventilated patients with COPD exacerbation despite that no formulations are specifically cleared by the United States FDA for use in this clinical setting, and thus the use of such medications in mechanically ventilated patients represents an off-label application. Goals of aerosol therapy in mechanically ventilated subjects have been proposed by Dhand⁹¹ (Table 3).

Table 3.

Goals of Aerosol Therapy in Mechanically Ventilated Patients

Liberation

After a period of stabilization and full ventilatory support, patients are typically transitioned to a spontaneous mode of ventilation such as pressure support. But pressure support ventilation can be problematic in patients with COPD. Due to their high R_aw, flow decreases slowly when a fixed pressure is applied to the airway. Because pressure support is flow cycled, this can result in a prolonged inspiratory phase and activation of expiratory muscles to terminate the inspiratory phase.⁹² This is a form of cycle asynchrony in which the T_I provided by the ventilator exceeds neural T_I. This can be observed at the bedside by palpating expiratory abdominal muscles and observing the pressure graphic for an increase in pressure at the end of inspiration. This can be addressed by adjusting the flow cycle criteria (usually increasing it, effectively shortening T_I) (Figure 11),⁹³ using a lower level of pressure support, or using pressure control ventilation with a fixed T_I. Alternatively, a proportional mode like NAVA or proportional assist ventilation can be used, but additional investigation is needed to establish the role of proportional modes.⁹⁴ With any spontaneous breathing mode, it is important that PEEP is set sufficiently to counterbalance auto-PEEP and reduce the risk of trigger asynchrony.

Fig. 11.

Pressure support ventilation with the cycle criteria set a 10, 25, and 50%. With a higher cycle criterion, note that the inspiratory time is shorter. The higher cycle criteria are appropriate for patients with COPD. Adapted from reference 81.

Once there is improvement in the exacerbation, a spontaneous breathing trial (SBT) is initiated. Whether this is done with a low level of pressure support, low level of PEEP, or without support is an area of controversy. Sklar et al⁹⁵ suggest that approaches that apply no pressure more accurately reflect the physiologic conditions after extubation. However, this might be moot if the patient is extubated to NIV. In this case, the SBT can be conducted with a low level of pressure support and PEEP, after which the patient is extubated to the same settings on NIV if the SBT is successful. Invasively ventilated patients with COPD exacerbation are at high risk for extubation failure and thus, in most cases, should be extubated to NIV. NIV is superior to HFNC in this setting,⁹⁶ but sequential use of NIV alternating with HFNC might be beneficial.⁹⁷ Using a composite end point of re-intubation and switching between NIV and HFNC, Tan et al⁹⁸ concluded that use of HFNC after extubation did not result in increased rates of treatment failure compared with NIV in subjects with COPD and severe hypercapnic respiratory failure. However, this conclusion was challenged by others.⁹⁸

NIV has also been used as a method to reduce the exposure to invasive mechanical ventilation. With this approach, patients who do not meet conventional extubation criteria are extubated directly to NIV, eg, fail an SBT or are not ready to undergo an SBT. NIV is then reduced over time, minimizing exposure to invasive ventilation. Burns et al⁹⁹ performed a systematic review and meta-analysis to evaluate this approach. In 14 RCTs enrolling 922 subjects, mortality was significantly lower in subjects extubated to NIV (risk ratio [RR] 0.36 [95% CI 0.25–0.51]). There was also a reduction in ventilator-associated pneumonia rate (RR 0.22 [95% CI 0.15–0.33]) and duration of mechanical ventilation (−1.32 d [95% CI −2.21 to −0.43]). These data support this approach to earlier extubation and deserve greater uptake in clinical practice.

Extracorporeal CO₂ Removal

Extracorporeal CO₂ removal is a form of extracorporeal life support (ECLS) intended to remove CO₂ in patients with acute hypoxemic or acute hypercapnic respiratory failure so as to treat respiratory acidosis.¹⁰⁰ Extracorporeal CO₂ removal is a subset of ECLS in which lower extracorporeal blood flow (200–1,500 mL/min) is used to remove CO₂. The flow needed to reduce hypercapnia is much lower than required to achieve adequate oxygenation in hypoxemic patients. This therapy is attractive for the management of COPD exacerbation with hypercapnia and high levels of auto-PEEP. In a multi-center, retrospective, case-controlled study of 42 subjects with acute hypercapnic respiratory failure not responding to NIV, 90% of those who received extracorporeal CO₂ removal with a pumpless arteriovenous technique did not require intubation.¹⁰¹ In the prospective study by Braune et al,¹⁰² 25 subjects with acute hypercapnic respiratory failure refractory to NIV were treated with venovenous extracorporeal CO₂ removal. Endotracheal intubation was avoided in 56% of subjects in the extracorporeal CO₂ removal group (mean extracorporeal blood flow of 1.3 L/min). But 36% of subjects suffered major bleeding complications, and there were no significant differences in length of stay or mortality in comparison with a matched historical control. In a matched cohort study with historical control, Del Sorbo et al¹⁰³ reported similar results in 25 subjects with COPD. Before widespread use of extracorporeal CO₂ removal in patients with COPD exacerbation, it is important that RCTs are conducted to establish indications and the risk-benefit relationship.¹⁰⁰

Care Coordination

Several studies have evaluated the role of a care coordinator—specifically the role of an RT care coordinator—in the care of subjects with COPD exacerbation. Silver et al¹⁰⁴ conducted a prospective, single-center, unblinded, randomized trial to determine whether an RT disease management program could reduce rehospitalization and ED visits for subjects with COPD. This was a large study that enrolled 428 subjects. The disease management program was associated with fewer readmissions, fewer ICU days, and shorter hospital stays due to COPD exacerbations. Zafar et al¹⁰⁵ reported that reliable adherence to a COPD care bundle reduced 30-d ED revisits among those treated in the ED observation unit. The bundle was designed to address 5 local care delivery failures: (1) inadequate medication regimen, (2) lack of follow-up appointments, (3) late follow-up appointments, (4) lack of standard emergency plan, and (5) confusion about inhalers due to different brands and devices leading to inappropriate inhaler use.

Atwood et al¹⁰⁶ evaluated the effectiveness of a COPD transition bundle, with and without a care coordinator, on rehospitalizations and ED revisits. The care coordinator was either an RT or a registered nurse. They reported that the use of a COPD transition bundle was effective at reducing 7-d and 30-d hospital readmissions. Truumees et al¹⁰⁷ evaluated the impact of a home RT to reduce 30-d readmission rates for exacerbation of COPD. This was a retrospective before and after study comparing hospital readmissions for subjects with COPD exacerbation that received standard of care at home versus an RT-led COPD program. They reported a 10% absolute reduction in 30-d readmission when utilizing the COPD disease management program.

Summary

RTs and physicians working in an acute care setting see patients with COPD exacerbation on a regular basis. Evidence-based practices such as controlled O₂ administration, delivery of short-acting bronchodilators, and use of NIV improve patient outcomes. Management of auto-PEEP is a priority in mechanically ventilated patients with COPD exacerbation. The evidence related to HFNC and extracorporeal CO₂ removal for this patient population is immature, and RCTs are needed. Care coordination is effective in the care of patients with COPD exacerbation.