Asthma and COPD Exacerbations

In depth

The one-page summary above is what the PDF prints. These sections go further.

Obstructive physiology: flow limitation, hyperinflation, V/Q

In asthma, smooth-muscle spasm, mucosal edema and mucus plugs narrow the airways, and most of the narrowing reverses with treatment. In COPD, emphysema destroys the elastic recoil that holds small airways open, so they collapse during exhalation (expiratory flow limitation) and extra effort no longer speeds emptying. Either way the expiratory time constant (resistance × compliance) lengthens, and the next breath starts before the lung has emptied.

The trapped gas raises end-expiratory lung volume (dynamic hyperinflation) and leaves alveolar pressure positive at end-expiration (auto-PEEP). The flattened diaphragm works at a mechanical disadvantage, and auto-PEEP is a threshold load: the patient must lower alveolar pressure by that amount before air flows in, seen as efforts that fail to trigger the ventilator. High intrathoracic pressure also cuts venous return, so positive pressure, sedation or fast bagging can drop the blood pressure abruptly, and overdistended alveoli can rupture.

Hypoxemia comes from low V/Q units, not shunt, and usually corrects with modest oxygen; in asthma, a PaO₂ under 55 mm Hg or a need for FiO₂ above 0.3–0.5 should prompt a search for pneumonia, atelectasis or pneumothorax (Corbridge and Hall 1995). Early in an asthma attack PaCO₂ is low; hypercapnia usually means FEV₁ below about 25% of predicted or a tiring patient (McFadden 1968). In COPD, PaCO₂ rises when resistance, hyperinflation and dead space outrun weak respiratory muscles.

Assessing severity and risk

Judge an asthma attack at the bedside first (see the one-page table). Measure PEF or FEV₁ before treatment and 1 h after: GINA 2025 recommends admission for a pre-treatment value under 25% of predicted or best, or a post-treatment value under 40%. An ABG is not routine; GINA suggests one when PEF or FEV₁ is under 50% or the response is poor, and reads PaO₂ under 60 mm Hg with a normal or rising PaCO₂ as respiratory failure.

GINA 2025 risk factors for asthma death include past near-fatal asthma needing intubation, a hospital or ED visit in the past year, recent oral steroids, no inhaled steroid, more than one albuterol canister a month, psychosocial problems and poor adherence.

GOLD 2025 ICU criteria for COPD: dyspnea unresponsive to emergency treatment, confusion or coma, PaO₂ under 40 mm Hg or pH under 7.25 despite oxygen and NIV, need for intubation, or need for vasopressors.

Bronchodilators, steroids, magnesium and parenteral β-agonists

Albuterol relaxes airway smooth muscle within minutes. Side effects scale with dose: tachycardia, tremor, hypokalemia, hyperglycemia and a lactic acidosis that raises the respiratory rate and can be mistaken for worsening asthma. Adding ipratropium in the ED cut admissions (RR 0.72; Kirkland, Cochrane 2017), mostly in severe attacks.

Steroids act over hours, so give them early: within 1 h of ED arrival they cut asthma admissions (OR 0.40, NNT 8; Rowe, Cochrane 2001), and more than 80 mg/day of methylprednisolone added nothing in hospitalized asthma (Manser, Cochrane 2001). In COPD, REDUCE 2013 (n = 314) found 5 days of prednisone 40 mg non-inferior to 14 (re-exacerbation 35.9% vs 36.8%). Trials in ventilated COPD conflict (Alía 2011: less NIV failure; Abroug 2014: no benefit, more hyperglycemia), and methylprednisolone above 240 mg/day brought more insulin use and fungal infection without lower mortality (Kiser 2014, n = 17,239, observational).

GINA 2025 advises against routine IV magnesium but notes fewer admissions in adults with FEV₁ under 25–30% predicted at presentation or persistent hypoxemia after initial treatment. Kew (Cochrane 2014) found 7 fewer admissions per 100 adults treated (OR 0.75), but the 3Mg trial (Goodacre 2013, n = 1,109) missed significance (admission 72% vs 78%), and nebulized magnesium did nothing.

If anaphylaxis is possible, IM epinephrine comes first (GINA 2025); IV terbutaline and aminophylline are not recommended in adults (Mein 2025 review).

COPD specifics: oxygen, CO₂ retention, antibiotics, eosinophils

Titrate oxygen to SpO₂ 88–92% with a Venturi mask (24–28%) or low-flow cannula, drive nebulizers with air, and recheck a blood gas after changes (GOLD 2025; BTS 2017). In Austin 2010 (n = 405, prehospital), titrated oxygen cut mortality in confirmed COPD from 9% to 2% compared with high-flow oxygen. Excess oxygen raises PaCO₂ mainly by releasing hypoxic vasoconstriction, which sends blood to poorly ventilated units, and through the Haldane effect; in Aubier 1980, 100% oxygen raised PaCO₂ by 23 mm Hg although minute ventilation recovered to 93% of baseline.

Chronic retainers compensate: bicarbonate rises about 3.5 mEq/L per 10 mm Hg of chronic PaCO₂ rise (1 mEq/L acutely), so pH falls about 0.03 per 10 mm Hg instead of 0.08; a lower pH means acute-on-chronic failure. On the ventilator, aim for the patient's usual PaCO₂, not 40: overcorrection causes a post-hypercapnic metabolic alkalosis, and the kidney then excretes bicarbonate the patient needs once breathing alone.

Give antibiotics for 5–7 days when purulent sputum comes with more dyspnea or sputum volume, or when NIV or intubation is needed: an aminopenicillin with clavulanate, a macrolide or a tetracycline, with sputum culture and antipseudomonal cover considered in frequent exacerbators, severe airflow obstruction or ventilated patients (GOLD 2025). In ventilated COPD, ofloxacin cut hospital mortality from 22% to 4% (Nouira 2001, n = 93).

Low blood eosinophils predict a weak steroid response. CORTICO-COP 2019 (n = 318, hospitalized) gave prednisolone after day 1 only on days with eosinophils ≥300/µL: days alive and out of hospital were similar (8.9 vs 9.3 of 14) and median steroid days fell from 5 to 2. GOLD 2025 notes the weaker benefit but does not advise withholding steroids, and the guided approach is untested in critically ill patients (Mein 2025).

Rule out mimics: pneumonia, heart failure, pneumothorax, arrhythmia and pulmonary embolism, found in 5.9% of hospitalized exacerbations (Couturaud 2021) and 16.1% of unexplained ones (Aleva 2017). Give VTE prophylaxis (GOLD 2025).

NIV and high-flow nasal oxygen

Bilevel NIV unloads tired muscles: EPAP offsets auto-PEEP so efforts trigger sooner, and pressure support raises Vt. Across 17 trials (Osadnik, Cochrane 2017; n = 1,264), NIV for hypercapnic COPD cut mortality (RR 0.54, NNT 12) and intubation (RR 0.36, NNT 5). ERS/ATS 2017 recommends it for pH ≤7.35 with hypercapnia and as a trial before intubation unless the patient is deteriorating fast, and suggests against it for hypercapnia without acidosis.

• Oronasal mask; typical starting settings are EPAP 4–5 and IPAP 10–15 cm H₂O; raise pressure support toward a Vt of 6–8 mL/kg (Mein 2025) and a falling RR; oxygen to SpO₂ 88–92%
• Blood gas at 1–2 h; a pH still under 7.25 after 2 h sharply raises the risk of failure (Confalonieri 2005, n = 1,033)
• Maximize hours on NIV in the first 24 h, then taper over 2–3 days (BTS/ICS 2016)
• One nurse per two NIV patients for the first 24 h; check leak, skin under the mask, gastric distension, secretions and mental status (BTS/ICS 2016)

HFNC (up to 60 L/min) flushes CO₂ from upper-airway dead space and adds a little PEEP. In moderate hypercapnia (pH 7.25–7.35), Cortegiani 2020 (n = 79) found PaCO₂ fell almost as much as on NIV, but 32% switched to NIV within 6 h; Tan 2024 (n = 225) found more failure (25.7% vs 14.3%) and intubation (14.2% vs 5.4%) with HFNC. ERS 2022 suggests trying NIV first in hypercapnic COPD; HFNC suits NIV breaks, NIV intolerance and post-extubation support.

In asthma, NIV is unproven: ERS/ATS 2017 makes no recommendation, BTS/ICS 2016 advises against it in hypercapnic asthma, and GINA 2025 says to watch the patient closely and never sedate to make the mask tolerable. Observational data are favorable (Althoff 2020, 53,654 ICU admissions: NIV linked to less intubation), yet 22% of NIV patients still needed intubation. A short supervised trial in an alert, cooperative patient is reasonable if it does not delay intubation (Mein 2025).

Intubating severe asthma

Intubate for arrest, falling consciousness, exhaustion or worsening acidosis despite full treatment; no single PaCO₂ decides it. In Williams 1992, hypotension (20%), barotrauma (14%) and arrhythmias (10%) among ICU asthma admissions occurred only in ventilated patients. These are physiologically difficult airways: hypoxemic, hyperinflated, often dehydrated, and held up by catecholamines that vanish at induction.

1. Most experienced operator; preoxygenate with NIV or HFNC; give a fluid bolus and draw up a vasopressor
2. Rapid sequence: ketamine 1–2 mg/kg (bronchodilates, less hypotension) or propofol, with rocuronium 1–1.2 mg/kg or succinylcholine
3. Largest tube that fits, ideally 8.0 mm: adjusted mortality was 26.7% with 7.0 mm or smaller vs 11.0% with 8.0 mm or larger (Kashiouris 2022, retrospective)
4. Bag slowly with small breaths and long pauses; confirm with waveform capnography; set the ventilator for obstruction before handing over

In cardiac arrest: disconnect, let the chest empty, ventilate slowly, and decompress either side if tension pneumothorax cannot be excluded.

Ventilating obstructed lungs

Ventilate to let the lungs empty, not to normalize PaCO₂: Darioli and Perret 1984 accepted hypercapnia and had no deaths in 34 episodes, and Tuxen 1987 showed that trapping and hypotension rose with rate, Vt and slower inspiratory flow, while peak pressure tracked flow, not lung volume.

Accept a pH down to about 7.20 unless intracranial pressure is raised; sources differ on the floor (BTS/ICS 2016 aims for 7.20–7.25, Mein 2025 for 7.25–7.30). Lowering the rate buys more exhalation time than raising flow: with Vt 500 mL at 60 L/min, inspiration takes 0.5 s, leaving 4.5 s to exhale at 12 breaths/min. Raising flow to 100 L/min adds 0.2 s; dropping the rate to 10 adds 1 s. Gains shrink below about 8–12 breaths/min because most trapped gas leaves in the first seconds (Mein 2025).

• Plateau: inspiratory hold of about 0.5–1 s in volume control; keep under 30 cm H₂O. It tracks end-inspiratory lung volume, which above 1.4 L (about 20 mL/kg) predicted hypotension and barotrauma (Williams 1992)
• Auto-PEEP: end-expiratory hold until pressure levels off; total PEEP minus set PEEP. Patient effort falsifies both holds
• Flow waveform: expiratory flow not back to zero when the next breath starts means trapping
• Peak − plateau: resistive pressure; a falling gap means bronchodilation is working
• Hidden auto-PEEP: closed airways can make auto-PEEP read low while plateau stays high (Leatherman and Ravenscraft 1996); believe the plateau

In a patient who triggers breaths, auto-PEEP causes missed triggers. With flow limitation, typical of COPD, external PEEP up to about 80% of measured auto-PEEP narrows that gap without adding hyperinflation (the waterfall effect). In passive patients the response varies: Caramez 2005 saw PEEP raise, leave unchanged or paradoxically lower plateau and total PEEP, so add it in 2 cm H₂O steps and keep it only if plateau does not rise; BTS/ICS 2016 advises not exceeding 12 cm H₂O.

Give bronchodilators by MDI with a spacer in the inspiratory limb or by mesh nebulizer, with the HME out during treatment. Each shift record peak, plateau, auto-PEEP and Vt, and check potassium, magnesium, phosphate and lactate.

Sedation, paralysis and rescue therapies

Deep sedation early suppresses the air hunger that drives a fast rate and dyssynchrony. Propofol (bronchodilating, but hypotensive) with fentanyl is common; ketamine adds bronchodilation with less hypotension but more secretions; avoid morphine, which releases histamine. Lighten sedation as obstruction resolves.

Use neuromuscular blockade only when deep sedation fails to control dyssynchrony or hyperinflation, preferably as boluses and briefly. With steroids, acute myopathy is common: weakness in 20 of 69 paralyzed asthma patients vs 0 of 38 given steroids alone (Leatherman 1996, 107 episodes), with odds rising 2.1-fold per extra day of paralysis (Behbehani 1999); atracurium was no safer than aminosteroids. Monitor train-of-four and protect the eyes.

Heliox (80:20 is about one-third as dense as air) lowers the pressure needed for turbulent flow but helps only while FiO₂ stays under about 0.3–0.4. It showed no overall benefit in non-intubated asthma (Rodrigo, Cochrane 2006), and in COPD on NIV it sped improvement in pH without reducing NIV failure (Jolliet 2017, n = 445: 14.7% vs 14.5%). Use ventilators calibrated for heliox; an oxygen flowmeter under-reads it (multiply by 1.8 for 80:20, 1.6 for 70:30).

Low-dose ketamine (0.1–0.2 mg/kg, then 0.5 mg/kg/h) added nothing in non-intubated ED patients (Howton 1996, n = 53); higher-dose infusions in ventilated patients rest on case series, as do inhaled anesthetics (isoflurane, sevoflurane), which need an anesthesia ventilator or anesthetic-conserving device (Mein 2025).

VV-ECMO removes CO₂ so the ventilator can rest the lungs; ELSO 2021 lists severe asthma and suggests ECMO for pH under 7.25 with PaCO₂ ≥60 despite optimal ventilation. In the ELSO registry (Yeo 2017, n = 272) survival to discharge was 83.5%, with complications in 65% (bleeding 28%). Lower PaCO₂ gradually: a fall over 50% in the first 24 h was linked to more neurologic complications (Cavayas 2020: 9.8% vs 6.4%).

ECCO₂R removes CO₂ through smaller cannulas at lower blood flow. In COPD failing NIV it avoided intubation in 14 of 25 patients but caused major bleeding in 9 (ECLAIR, Braune 2016); VENT-AVOID (Duggal 2024, n = 113, stopped early) found no gain in ventilator-free days and higher mortality in the NIV stratum (22% vs 0%).

Liberation, discharge and follow-up

Start liberation when the peak–plateau gap, plateau and auto-PEEP have fallen and PaCO₂ drops at the same minute ventilation. Extubate hypercapnic COPD patients straight to NIV: ERS/ATS 2017 suggests NIV to wean hypercapnic patients and to prevent post-extubation failure in high-risk patients, with HFNC if NIV is not tolerated. Rescue NIV for established post-extubation failure raised mortality (Esteban 2004), so use it preventively. Post-extubation support in asthma is untested.

Before an asthma patient leaves hospital: start or step up inhaled-steroid treatment (GINA 2025 prefers ICS–formoterol as maintenance and reliever), finish 5–7 days of oral steroid, check inhaler technique, give a written action plan and arrange review within 2–7 days. Past near-fatal asthma is a lasting risk factor for death (GINA 2025); refer to a specialist.

Before a COPD patient leaves: long-acting bronchodilators started, inhaler technique checked, steroid and antibiotic stop dates clear, oxygen need reassessed, smoking treated, vaccines updated, rehabilitation arranged, and follow-up within 1 month and again at about 3 months (GOLD 2025).

Persistent hypercapnia (PaCO₂ over 53 mm Hg) 2–4 weeks after the acidosis resolves identifies candidates for home NIV: in HOT-HMV 2017 (n = 116), home NIV plus oxygen cut 12-month readmission or death from 80.4% to 63.4%. About half of patients die within 5 years of a COPD admission (GOLD 2025), so discuss goals of care before discharge.

Exam pearls

• A 'normal' PaCO₂ in a wheezing, tachypneic asthmatic means fatigue
• SpO₂ targets: COPD 88–92%; asthma 93–95% (GINA 2025)
• Oxygen-induced hypercapnia is mostly V/Q worsening and the Haldane effect, not loss of hypoxic drive
• NIV first for COPD with pH ≤7.35 and PaCO₂ >45; it cuts intubation and death
• COPD steroid: prednisone 40 mg for 5 days (REDUCE); oral works as well as IV
• Auto-PEEP: measure with an end-expiratory hold; lowering the rate is the strongest fix
• High peak with normal plateau = resistance; high plateau = hyperinflation, pneumothorax or stiff lung
• Hypotension after intubating an asthmatic: disconnect first, then think tension pneumothorax
• COPD antibiotics: purulent sputum plus more dyspnea or volume, or any ventilation; 5–7 days