Risk of exacerbations, hospitalisation, and mortality in adults with physician-diagnosed chronic obstructive pulmonary disease with normal spirometry and adults with preserved ratio impaired spirometry in Sweden: retrospective analysis of data from a nationwide cohort study

Abstract

Background

Physician diagnosed COPD with normal spirometry (dnsCOPD) (sometimes labeled pre-COPD) and Preserved Ratio Impaired Spirometry (PRISm) has been studied in population-based cohorts, but not in physician diagnosed COPD (dCOPD) patients from routine clinical practice. The Swedish National Airway Register (SNAR) is a large nationwide register including data from dCOPD patients from over 1000 clinics across all regions of Sweden and is representative of the COPD care in Sweden. We aimed to identify and characterize patients with dnsCOPD, PRISm and spirometrically confirmed COPD (sCOPD) from dCOPD patients in SNAR, stratify them further according to symptoms and exacerbations risk using the Global Initiative for Chronic Obstructive Lung Disease (GOLD) A/B/E classification, and assess differences in risk for exacerbations, cause-specific hospitalisations and mortality.

Methods

We enrolled patients aged ≥30 years with dCOPD in the SNAR from 1 January 2014 to 30 June 2022 with complete spirometry i.e., postbronchodilator values for both forced expiratory volume in 1 s (FEV₁) and forced vital capacity (FVC) (index date). Patients with concomitant asthma were excluded. Patients were stratified into dnsCOPD (FEV₁/FVC ≥0.7 and FEV₁ ≥80% predicted), PRISm (FEV₁/FVC ≥0.7 and FEV₁ <80% predicted) and sCOPD (FEV₁/FVC <0.7). Further substratification was based on GOLD A/B/E (A: COPD assessment test (CAT) score <10 points and <2 moderate, 0 severe exacerbations within 1 year before the index date, B: CAT-score ≥10 points and <2 moderate, 0 severe exacerbations, E: ≥2 moderate or ≥1 severe exacerbation(s)). Patients were followed until 31 November 2022. Competing risk regression was used to calculate subdistribution hazard ratios (SHR)s with 95% confidence intervals (CIs) for exacerbation, hospitalisation and mortality.

Findings

Of 45,653 patients with dCOPD, 5.4% had dnsCOPD, 11.4% had PRISm and 83.3% had sCOPD. Smoking history was similar between groups (ever smoker: dnsCOPD: 79% PRISm: 82% sCOPD: 86%) and inhalation therapy was common in all groups (any inhaler: 75%, 80% and 80%, triple combination: 22%, 28% and 35%). Patients with PRISm had a high prevalence of obesity (dnsCOPD: 30%, PRISm: 43%, COPD: 22%), cardiovascular disease (dnsCOPD: 39%, PRISm: 48%, COPD: 41%) and diabetes (dnsCOPD: 10%, PRISm: 17%, COPD: 9%). Baseline GOLD group B or E were highly prevalent in dnsCOPD (B: 54%, E: 11%), PRISm (B: 59%, E: 14%), as well as in COPD (B: 54%, E: 17%). DnsCOPD and PRISm patients had lower risk of exacerbations (SHR 0.69, 95%CI 0.64–0.74 and 0.85, 95%CI 0.81–0.89), respiratory hospitalisation (0.40, 95%CI 0.34–0.46 and 0.68, 95%CI 0.62–0.73), and respiratory mortality (0.22, 95%CI 0.13–0.37 and 0.60, 95%CI 0.48–0.75) compared to sCOPD. Cardiovascular mortality was lower in dnsCOPD (0.41, 95%CI 0.19–0.86), but similar in PRISm (0.73, 95%CI 0.49–1.08) compared to sCOPD. The A/B/E classification was predictive for all outcomes in dnsCOPD and PRISm. DnsCOPD and PRISm group E patients had higher risks for all outcomes than sCOPD group A or B.

Interpretation

DnsCOPD and PRISm are prevalent in a real-life cohort of patients with a physician diagnosis of COPD. These patients are symptomatic, might suffer from exacerbations and are commonly treated with inhaled therapy, equally to sCOPD. Patients with PRISm had a high prevalence of obesity, diabetes and cardiovascular disease. DnsCOPD and PRISm had generally lower overall risks of exacerbation or respiratory events, although PRISm patients showed similar cardiovascular risk to sCOPD. The A/B/E classification predicted future events, even in dnsCOPD and PRISm patients.

Funding

This study is performed with support from The Swedish Heart-Lung Foundation (20200150) and the Swedish government and country council ALF grant (ALFGBG-824371).