Substance + claimed effects

COPD — chronic obstructive pulmonary disease — is a progressive airflow-limitation syndrome caused chiefly by tobacco smoke, biomass smoke, and (rarely) inherited alpha-1 antitrypsin deficiency. A new diagnosis is the only point in the disease course where a single ninety-day window changes the entire trajectory: spirometric confirmation, alpha-1 antitrypsin screening, symptom and exacerbation phenotyping, smoking cessation pharmacotherapy, first inhaler, the vaccines, and pulmonary rehabilitation either get launched here or get delayed by years. Each delayed by months loses lung function that does not return (Anthonisen et al., JAMA 1994). The entry covers the first ninety days end-to-end: workup, classification (GOLD ABE), the action list itself, and the consequences that follow — symptom control, exacerbation-risk reduction, smoking cessation, and AATD detection — across the meta dimensions of short-term wellness, longevity, energy, mood, sleep, and the burden of doing it all. Out of entry: the long-term management of established disease, exacerbation-response protocols (a separate respond-type entry), surgical / lung-volume-reduction interventions, and end-stage care.

Evidence by addressing question

mechanism

COPD is small-airway disease plus parenchymal destruction (emphysema). The proteinase / antiproteinase imbalance theory — cigarette smoke recruits neutrophils whose elastase digests lung elastin faster than antiproteases neutralise it — was the founding model and remains the best mechanistic frame; AATD is the experiment-of-nature that proves it (Stoller and Aboussouan, Lancet 2005). Two physiologic signatures dominate the felt experience: expiratory flow limitation (the FEV₁/FVC drop that defines diagnosis) and dynamic hyperinflation (air trapped on exhalation; the next breath starts from a higher resting volume, the diaphragm is mechanically disadvantaged, dyspnea on exertion follows). Once below an FEV₁ of about 50% predicted, the relationship between symptom intensity and lung function flattens — exacerbation frequency and comorbidities become the dominant prognostic levers, not the spirometry number itself (Suissa et al., Thorax 2012). The Copenhagen and Framingham cohort work also makes clear that a substantial fraction of COPD cases arise from failure to attain peak adult lung function, not accelerated decline from a normal peak — meaning the diagnosis at 50 may reflect a thirty-year head start, not a recent injury (Lange et al., NEJM 2015).

evidence — diagnosis and workup

The diagnostic anchor is post-bronchodilator spirometry showing FEV₁/FVC < 0.7, retained by the 2025 GOLD report despite long-standing dispute over the fixed cut-off (which over-diagnoses elderly normal lungs and under-diagnoses some young adults) (GOLD 2025). The 2025 update endorses a two-step approach: pre-bronchodilator spirometry first to rule out obstruction (if normal, stop); confirm with post-bronchodilator if the pre-BD ratio is below 0.7 or clinical suspicion is high; for borderline ratios (0.6–0.8), repeat on a separate day. Three additional first-90-day workup elements have hard guideline backing: (1) a one-time serum alpha-1 antitrypsin level on every patient with COPD regardless of age, ethnicity, or smoking history — recommended by GOLD, ATS/ERS, and the Canadian Thoracic Society, despite real-world testing rates of only ~5% (ATS/ERS 2003); (2) symptom quantification with the CAT score (8-item, 0–40) or mMRC dyspnea scale (0–4), with CAT ≥ 10 the standard threshold for the symptomatic phenotype (Jones et al., ERJ 2009); and (3) capture of exacerbation history in the prior 12 months (any moderate-or-worse event, any hospitalisation) — these define GOLD's ABE group and dictate initial inhaler choice (GOLD 2025). Blood eosinophil count is now routinely added: a baseline above ~300 cells/μL identifies patients whose exacerbation risk responds to inhaled corticosteroids, with progressively greater ICS benefit as the count rises (Pascoe et al., Lancet Respir Med 2019). Chest CT is not required for diagnosis but is endorsed for any patient meeting lung-cancer-screening criteria or with persistent symptoms — emphysema severity on visual CT classification independently predicts mortality in COPDGene (Lynch et al., Radiology 2018). Pulse oximetry on every visit; arterial blood gas if SpO₂ ≤ 92%.

evidence — smoking cessation

The Lung Health Study is the foundational trial: 5887 smokers with mild-to-moderate COPD randomised to special intervention (group cessation + nicotine gum) vs. usual care, followed for 5 years for lung function and 14.5 years for mortality. FEV₁ declined at 28 mL/year in sustained quitters, 48 mL/year in intermittent quitters, and 62 mL/year in continuing smokers (Anthonisen et al., JAMA 1994). The long-term follow-up showed all-cause mortality 8.83 per 1000 person-years in the special-intervention group vs. 10.38 in usual care, a 15% relative reduction driven by cardiovascular and respiratory deaths (Anthonisen et al., Ann Intern Med 2005). Cessation also produces a one-year FEV₁ bump (a partial reversal of smoking's acute inflammatory drag), after which the decline rate parallels never-smokers' (Scanlon et al., AJRCCM 2000). Importantly, halving cigarettes per day produced no measurable FEV₁ protection — reduction is not cessation. Population cohorts confirm the cessation effect persists at age of quitting up to and beyond 60: Swedish data shows premature-mortality risk reduction even when cessation occurs after diagnosis at older age (Lindberg et al., ERJ 2015). Pharmacotherapy: a Cochrane review covering varenicline, NRT, and bupropion all increase quit rates in COPD smokers vs. counselling alone; varenicline + behavioural support is the highest-yield combination, with no excess neuropsychiatric signal in the EAGLES post-marketing era (van Eerd et al., Cochrane 2016) (Tashkin et al., Chest 2011). The 2020 ATS pharmacologic-treatment guideline endorses varenicline over NRT alone and over bupropion, recommends varenicline over placebo even in unwilling-to-quit smokers (a paradigm shift), and supports the controller-style 1–4 week pretreatment-before-quit-day model (Leone et al., AJRCCM 2020).

evidence — initial inhaler choice and exacerbation prevention

GOLD 2025 routes initial pharmacotherapy through the ABE classifier built from CAT/mMRC + exacerbation history. Group A (low symptoms, ≤1 moderate exacerbation, no hospitalisation): any single long-acting bronchodilator. Group B (high symptoms, ≤1 moderate exacerbation): LABA + LAMA combination from the outset. Group E (≥2 moderate exacerbations OR ≥1 hospitalisation in the past year): LABA + LAMA, with triple therapy LABA + LAMA + ICS reserved for patients with blood eosinophils ≥ 300 cells/μL (GOLD 2025). The IMPACT trial randomised 10,355 symptomatic patients with exacerbation history to once-daily fluticasone furoate/umeclidinium/vilanterol vs. either dual component, finding a 25% reduction in annual exacerbation rate with triple over LAMA/LABA and a mortality signal favouring triple (Lipson et al., NEJM 2018). ETHOS replicated the exacerbation and mortality reduction at glucocorticoid dose tiers with budesonide/glycopyrrolate/formoterol (Rabe et al., NEJM 2020). Critical caveat: in both trials, the mortality and exacerbation advantage was concentrated in the first 90 days and largely disappears thereafter, suggesting much of the apparent triple-therapy benefit reflects ICS withdrawal in the dual-arm comparator rather than additive triple effect. Real-world cohort data is consistent with no exacerbation benefit and increased pneumonia risk from blanket triple therapy in unselected COPD populations (Suissa et al., Chest 2018). Pneumonia risk under ICS is the durable downside — TORCH first quantified it for fluticasone/salmeterol (Calverley et al., NEJM 2007). Net practical inference: ICS in COPD is for the high-eosinophil or frequent-exacerbator phenotype, not the default.

evidence — vaccines

Influenza vaccination reduces COPD exacerbations: Cochrane meta-analysis (six RCTs, 2,469 participants) shows a significant reduction in the number of exacerbations per vaccinated person, with the protective effect most marked for exacerbations occurring three or more weeks post-vaccination — i.e., the influenza-specific causal pathway (Kopsaftis et al., Cochrane 2018). PPV23 / PCV13 / PCV15 / PCV20 pneumococcal vaccination reduces community-acquired pneumonia incidence in COPD patients (number-needed-to-vaccinate ~21 over 3 years in subgroup analysis) but the trial evidence for exacerbation reduction is weaker than influenza (Walters et al., Cochrane 2017). GOLD 2025 also endorses RSV (single dose, age ≥ 60 or chronic lung/heart disease), COVID-19 (per CDC schedule), pertussis (Tdap), and zoster (Shingrix, age ≥ 50) (GOLD 2025). The first 90 days is when these are scheduled — none are urgent in the sense of needing to wait, all are preventable losses if forgotten.

evidence — pulmonary rehabilitation

Pulmonary rehabilitation — a structured 6–12 week supervised programme of exercise training, education, breathing technique, and self-management coaching — is the second highest-yield intervention in COPD after smoking cessation. A 2015 Cochrane meta-analysis of 65 RCTs (3,822 patients) showed clinically significant improvements in dyspnea, fatigue, emotional function, and disease mastery, with effect sizes substantially exceeding the minimum clinically important difference on every domain (McCarthy et al., Cochrane 2015). Post-exacerbation rehab (begun during admission or within 4 weeks of discharge) reduces hospital readmission and mortality (Puhan et al., Cochrane 2016). A Medicare-cohort analysis of 197,376 patients hospitalised for COPD found a 37% relative reduction in 1-year mortality among those who initiated pulmonary rehabilitation within 90 days of discharge vs. those who did not (Lindenauer et al., JAMA 2020). The official ATS/ERS statement positions pulmonary rehabilitation as standard-of-care for symptomatic COPD; effect on dyspnea exceeds anything inhaler therapy produces, but uptake is catastrophically low — only 3–16% of eligible patients are even referred, and < 4% complete (Spruit et al., AJRCCM 2013).

evidence — alpha-1 antitrypsin deficiency screening

AATD is a co-dominantly inherited deficiency of the SERPINA1-encoded protease inhibitor; the severe PI*ZZ genotype produces serum AAT levels < 15% of normal and accelerates emphysema (typically panlobular, basal-predominant) by 1–3 decades in smokers and ~10–20 years in non-smokers (Stoller and Aboussouan, Lancet 2005). Pooled global estimates place PI*ZZ frequency at roughly 1 in 2,000–7,000 in populations of northern/western European descent, with proportionally lower frequencies in other ancestral backgrounds; the worldwide PI*ZZ pool is ~250,000 patients and severe AATD overall ~1.1 million if heterozygous severe combinations are included (Blanco et al., Int J COPD 2017). Despite GOLD and ATS/ERS recommending one-time testing for every COPD patient, real-world uptake remains around 5% — the dominant failure of the workup. Detection matters because (a) the AATD-COPD patient and family benefit from intensified cessation support, (b) siblings and children warrant cascade testing, (c) IV augmentation therapy (purified pooled-plasma AAT, given weekly) slows emphysema progression on CT density in PI*ZZ patients with established emphysema and is the only disease-modifying therapy in any COPD subgroup (ATS/ERS 2003).

protocol

The 90-day action set falls out of the evidence above and is the operative content the article translates to felt experience. Compressed: spirometry to confirm, AAT level once, CBC for eosinophils, CAT and mMRC and prior-year exacerbation count, chest imaging when indicated, SpO₂; smoking cessation pathway started day one with varenicline (or NRT + bupropion if varenicline contraindicated) plus behavioural support; first inhaler matched to ABE group; influenza + pneumococcal (PCV20 or PCV15→PPSV23) + COVID + RSV (if ≥ 60 or chronic disease) brought up to date; pulmonary rehabilitation referral made before week 12 with a target completion date; LTOT only if PaO₂ ≤ 55 mmHg (or 56–59 with cor pulmonale / polycythemia) on stable, optimised therapy and at least 30 days post-exacerbation; a written action plan for exacerbation recognition and home antibiotic / prednisone supply when warranted. Re-spirometry at 90 days to verify response and document baseline. Inhaler-technique check at every visit — a non-trivial fraction of "non-responders" are actually non-deposits.

contraindications

The 90-day plan itself has no contraindications — the only thing to actively contraindicate is the wholesale ICS-in-everyone reflex. ICS in low-eosinophil (< 100 cells/μL), low-exacerbation patients adds pneumonia risk without benefit (Suissa et al., Chest 2018) (Calverley et al., NEJM 2007). Varenicline is contraindicated in pregnancy and breastfeeding; the historical FDA black-box for neuropsychiatric events was removed in 2016 after the EAGLES trial. NRT and bupropion both have pregnancy-relative-risk considerations. Long-term oxygen given to non-hypoxaemic patients does not improve outcomes and the LOTT trial of moderate desaturators showed no benefit, so prescribing LTOT outside the NOTT/MRC indications is harm-neutral at best (LOTT, NEJM 2016).

misconceptions

"Cutting back is almost as good as quitting." Lung Health Study explicitly tested 50% reduction; no FEV₁ benefit at all (Anthonisen et al., JAMA 1994). "I'm too old / damaged for cessation to matter." All-cause mortality risk falls within years even when cessation begins in the seventh decade (Lindberg et al., ERJ 2015). "ICS is the cornerstone of COPD treatment." It is not — bronchodilators are; ICS is a phenotypic add-on. "Pulmonary rehab is for severe disease only." The largest functional gains accrue to mild-moderate patients who are still active enough to train; waiting for severity is waiting for irreversibility (Spruit et al., AJRCCM 2013). "AATD is a paediatric / rare-disease curiosity." It is a 1-in-2000-to-7000 condition in patients of European descent, vastly underdiagnosed in adult COPD clinics (Blanco et al., Int J COPD 2017).

failure-modes

Where the 90-day plan actually breaks: (1) the diagnosis was made off symptoms or chest X-ray without spirometric confirmation — common in primary care; the patient may have asthma, heart failure, bronchiectasis, or simply normal lungs (GOLD 2025). (2) AAT testing is omitted — the modal failure, with ~95% of patients never tested (ATS/ERS 2003). (3) Cessation is recommended without pharmacological support; quit rates with advice alone are < 5% sustained (van Eerd et al., Cochrane 2016). (4) Inhaler chosen by the prescriber's habit rather than the ABE group + eosinophil count; either ICS is added unnecessarily (pneumonia risk) or omitted when indicated (preventable exacerbations). (5) Pulmonary rehab is mentioned but never referred — the rehabilitation-uptake gap is the single largest unmet need in COPD care (Spruit et al., AJRCCM 2013). (6) Inhaler technique is never observed; up to 50% of patients use their device wrong, and the prescription is functionally a placebo. (7) Comorbidities (cardiovascular disease, anxiety/depression, osteoporosis, OSA) are not screened — and these drive much of COPD's all-cause mortality.

practicalities

Costs in the US first 90 days: spirometry $50–$300 covered by insurance; AAT serum level $30–$100; CBC trivial; chest CT $300–$1,500 if indicated. Inhalers vary wildly — generic tiotropium dry-powder $30–$100/month, branded triple therapies $500–$700/month list, frequently covered after deductible. Varenicline (now generic) ~$100/month for 12 weeks; NRT patches ~$30–$60/month. Pulmonary rehabilitation covered by Medicare and most private plans; out-of-pocket for self-pay $30–$80 per session × 16–36 sessions. The single highest-impact action — cessation pharmacotherapy + behavioural counselling — is generally covered (quitline services free in all US states and many countries). Time burden is real: pulmonary rehab is 2–3 sessions/week × 8–12 weeks; daily inhaler use 1–3 doses; weekly check-ins early in cessation. Insurance navigation is non-trivial.

stakes

Untreated, COPD is the third leading cause of death globally (Boers et al., JAMA Network Open 2023). The decade after a missed-and-progressed COPD diagnosis is shorter walks, more breathlessness on stairs, higher cardiovascular event rate (the smoker's metabolic / inflammatory load is shared), recurrent winter exacerbations, accelerating hospitalisations, eventual disability, then the survival horizon contracts. Continuing smokers lose FEV₁ at roughly twice the rate of quitters (Anthonisen et al., JAMA 1994); their 14.5-year all-cause mortality runs ~15% higher than otherwise-similar cessation-arm patients (Anthonisen et al., Ann Intern Med 2005). Each severe exacerbation roughly doubles the risk of the next, and post-hospitalisation 1-year mortality is ~22% without pulmonary rehab uptake (Suissa et al., Thorax 2012) (Lindenauer et al., JAMA 2020). The social signal — being the one who has to stop on the kerb to catch breath; being unable to keep up with grandchildren; the rasping cough strangers notice — arrives years before the prognostic numbers do.

payoff

The first 90 days, done right, buy back lung function (the cessation-bump within a year), drop the exacerbation rate by half or more (combination inhaler + rehab + vaccines), and shift the survival curve toward never-smokers' over the following decade (Anthonisen et al., Ann Intern Med 2005). Felt: morning cough subsides over weeks of cessation; stairs become available again after rehab; the dread of winter recedes once exacerbation frequency drops. For the AATD-positive minority, family members get screened — a 90-day workup that prevents another generation's disease.

out-of-scope

Long-term inhaler titration past month 3, exacerbation-response protocol (the at-home prednisone/antibiotic decision tree), lung volume reduction (surgical or endobronchial valve), transplant evaluation, AATD IV augmentation therapy details, ventilation in end-stage disease.

Credibility range

Optimist case

The first 90 days after a COPD diagnosis are the single highest-leverage window in the disease. Cessation in this window (especially with pharmacotherapy + structured behavioural support) shifts the patient onto the never-smoker mortality curve over the following decade; pulmonary rehabilitation cuts dyspnea and exacerbations by clinically dominant margins; AAT screening salvages the 1–2% who can benefit from augmentation and who would otherwise progress unchecked; properly matched inhaler therapy halves exacerbation risk in the symptomatic and exacerbation-prone phenotype. All of this is evidence-rich (multiple Cochrane reviews, large RCTs, guideline-backed). It is one of the most clearly settled longevity interventions in chronic disease medicine. The under-delivery problem is operational, not scientific.

Skeptic case

The FEV₁/FVC < 0.7 cut-off systematically over-diagnoses healthy older adults and under-diagnoses some young patients, inflating the prevalence numbers. Much of the triple-inhaler "mortality benefit" in IMPACT and ETHOS reflected ICS withdrawal in the comparator arm, not true triple superiority; real-world cohort data is null. Pulmonary-rehab effect sizes are real in trials but uptake is so low (< 4%) that the population effect is tiny. The much-cited 14.5-year Lung Health Study mortality benefit was a 15% relative reduction; the absolute numbers were modest. AAT augmentation evidence is CT-density and observational; mortality RCTs do not exist. And outside the cessation pillar, much of "COPD management" is symptom-modifying, not disease-modifying — the lungs do not regrow.

Author's call

The 90-day workup-and-launch is the highest-confidence longevity intervention in this corner of the catalogue: every component is guideline-endorsed, the dominant intervention (cessation) has fifty years of replication, and the failure modes are all under-delivery, not over-claim. The skeptic critiques are mostly about the marginal additions (triple therapy, augmentation), not about the core (confirm diagnosis, quit smoking, vaccinate, rehab, screen for AATD). Evidence rates 5; controversy rates 2 — quiet pushback on triple therapy and on the spirometric cut-off, but no foundational disagreement about what to actually do in the first 90 days.

Stakeholder + incentive map

• Inhaler manufacturers (GSK, AstraZeneca, Boehringer Ingelheim, Chiesi) — sponsor the large triple-therapy trials; push for guideline language that broadens eligibility. The 90-day windowing in IMPACT/ETHOS analyses has been used in marketing.
• Pulmonologists and primary-care professional bodies (ATS, ERS, ACP, NICE, Canadian Thoracic Society) — converge on cessation + bronchodilator + rehab + AATD screening as the evidence-driven core.
• Tobacco industry — historical counter-incentive; current incentive shifted toward harm-reduction product (vapes, heated tobacco). Cessation pharmacotherapy is a competing product.
• Pulmonary rehabilitation programmes — under-resourced, under-reimbursed in most systems; the access bottleneck is administrative, not scientific.
• AATD-focused charities (Alpha-1 Foundation, EU registries) — sponsor cascade testing; push for universal screening of COPD; receive partial industry support from augmentation-therapy makers (Grifols, Takeda, CSL Behring).
• Insurance / payers — fund spirometry and inhalers reliably, fund pulmonary rehab inconsistently, fund AATD testing because it is cheap; high-cost inhalers create coverage friction that displaces effort onto patient and prescriber.
• Quitline / behavioural-cessation programmes (public-health-funded in most high-income countries) — free at point of care; under-referred to from new-COPD encounters.

Population variability

Age — most new COPD diagnoses are 50+; AATD-related cases skew younger (40s in smokers, 50s in non-smokers). Cessation benefit attenuates but does not disappear at older ages of quitting (Lindberg et al., ERJ 2015). Sex — historically male-predominant but the female fraction has been climbing for decades; women diagnosed with COPD tend to have lower smoking exposure for equivalent damage and higher prevalence of comorbid anxiety / depression. Ancestry — the FEV₁/FVC fixed cut-off was derived chiefly from white populations; reference equations and AATD allele frequencies differ across ancestral backgrounds, but the workup recommendations apply universally. Baseline severity — Group A (mild) patients see proportionally larger gains from cessation + rehab and small gains from inhalers; Group E (frequent exacerbators) see the inverse. Comorbidities — cardiovascular disease, OSA, anxiety / depression, osteoporosis, and lung cancer are massively over-represented; the 90-day plan needs to detect them. Eosinophil phenotype — ~30% of COPD patients have blood eosinophils ≥ 300 cells/μL and constitute the ICS-responsive subgroup; the other ~70% derive no exacerbation benefit and accept pneumonia risk. AATD genotype — PI*ZZ patients (the ~1 in 2000–7000) and PI*SZ heterozygotes warrant the augmentation conversation; MZ heterozygotes have intermediate risk but no augmentation indication.

Knowledge gaps

• Whether triple-inhaler therapy has any meaningful benefit beyond LAMA/LABA + targeted ICS in the high-eosinophil phenotype, once ICS-withdrawal-confound is controlled. The ICS-step-down question.
• Whether AAT augmentation produces a mortality benefit (only CT-density and exacerbation evidence exist; mortality RCTs are infeasible at the relevant sample size).
• The optimal cessation pharmacotherapy sequence for the unmotivated COPD smoker — varenicline pre-loading, cytisine, e-cigarettes-as-cessation-tool all have open questions.
• Whether earlier pulmonary rehabilitation (at the time of diagnosis, not after first exacerbation) prevents the first exacerbation in the first place.
• The role of biologics (anti-IL-5, anti-TSLP) in the eosinophilic COPD phenotype — early trials are mixed; reliable subgroup identification is unresolved.
• Whether the spirometric cut-off should be revised to the lower-limit-of-normal, reducing over-diagnosis in older adults — adoption stalled because the fixed cut-off is operationally simpler and the over-diagnosed cohort is largely asymptomatic.

Scope vs. brief. The brief named five threads: workup, symptom control, exacerbation risk, smoking cessation pathways, AATD detection. All five are covered end-to-end in the body. The entry stops cleanly at day 90 — long-term inhaler titration, exacerbation-response decision trees, lung-volume-reduction, transplant, and AATD augmentation-therapy mechanics are flagged in out-of-scope and belong to separate future entries (see below).

Action type call: respond, not do. The substance is triggered by a discrete event (diagnosis) and runs as a bounded course; respond + course match better than do + daily. The daily-inhaler component is a downstream artefact of the course, not the substance itself.

Audience scoping. Restricted to ages 40–59 and 60+ — the modal age band for new COPD diagnoses. AATD-related cases skew younger and would expand 18–39, but that subset is small enough that the over-inclusion (mistargeting young readers without COPD) would have outweighed under-inclusion. No gender restriction; women's COPD is rising and the workup is identical.

Focus dimension: scored 0, considered 1. Cessation produces real cognitive recovery after the acute withdrawal phase, and treating chronic hypoxaemia improves cognition in severe disease. Both effects are too small or too population-narrow to defend as "produces that effect" for the modal newly-diagnosed Group A/B reader. Setting to 0 keeps the high scores legible.

Controversy 2, not 3. The triple-therapy / ICS-withdrawal dispute (Suissa et al. vs. IMPACT/ETHOS authors) is a real but bounded methodological argument; the core 90-day actions (confirm spirometry, AATD test, cessation pharmacotherapy, vaccinate, pulmonary rehab) are universally endorsed. A 3 would over-state the disagreement on what the reader actually does in the first 90 days.

Effort burden 4 vs 3. Quitting smoking alone is among the hardest single behaviour changes a person makes. Adding twice-weekly pulmonary rehabilitation × 8–12 weeks, daily inhaler routine, and follow-up clinic schedule pushes this firmly into "major". Considered 3 if rehab uptake were assumed low (it usually is), but the meta scores the substance, not the realistic-real-world uptake. Reader is told honestly in highlights that this is one of the harder behaviour changes.

Triple therapy framing call. The article presents ICS as a phenotype-targeted add-on rather than as a default, because the real-world overprescription pattern (Suissa 2018) and the ICS-withdrawal-confound critique of IMPACT/ETHOS line up. Reasonable specialists could push back. The contraindications section is the place this critique lives in the reader-facing voice.

Inhaler step-down at 90 days. Considered including the GOLD recommendation that if no symptom improvement is seen at 1–3 months, step down. Cut for length / scope — fits better in a follow-on entry on long-term inhaler titration.

Future-link candidates (do not yet exist):

• copd-exacerbation-at-home — symptom-recognition and the antibiotic + prednisone home-action-plan decision tree.
• aatd-augmentation-decision — IV augmentation therapy mechanics, indications, and cascade testing for family.
• copd-osa-overlap — the overlap syndrome workup and CPAP / BiPAP question.
• smoking-cessation-pharmacotherapy — varenicline-vs-NRT-vs-bupropion deep dive that this entry references but does not own.
• pulmonary-rehabilitation — the programme itself, accessible to other respiratory conditions, not COPD-specific.
• inhaler-technique — short device-specific entry; the "non-deposit prescription" problem deserves its own home.

Citation note. The LOTT trial is stored under ref Sciurba2019 with year 2016 — ref-name mismatch carried from an early draft; the record's year field is correct at 2016. Worth a one-pass rename in the library if a future entry uses LOTT again.