LAUSR

Introduction and Objectives COPD is a clinical diagnosis comprising symptoms, risk factors and evidence of post-bronchodilator airflow obstruction (AFO). While spirometry is fundamental to the diagnosis, the National COPD Audit Programme reported in 2016 that one quarter of spirometry values were not consistent with COPD. Our objective was to explore this further, using patient-anonymised data in Hampshire Health Record Analytical database, an NHS database for ≥1.4 million patients, to compare characteristics, comorbidities and respiratory medication in patients diagnosed with COPD with and without evidence of AFO. Methods Read codes in primary care records were used to identify a prevalent COPD cohort as at 01/01/2011 and define and describe three patient categories based on consistency of FEV1/FVC% since diagnosis: all <70% (persistent AFO), some <70% (variable AFO), all ≥70% (absent AFO). Results 16 479 patients were identified with diagnosed COPD of whom 13 653 (82.9%) had FEV1/FVC% data: 7609 (55.7%) showed persistent AFO, 4413 (32.3%) variable AFO and 1631 (11.9%) absent AFO (Table 1). In the 2826 patients without recorded FEV1/FVC%, half had no evidence of any spirometry. In patients without AFO, mean FEV1/FVC% was high (80.5%), though mean FEV1 was <80% predicted (77.6%). There was no clinically relevant age difference across the groups, but patients without AFO were more often women, had higher mean BMI and contained proportionally fewer active smokers than those with persistent AFO. Never smokers were rare in all groups. Mean number of comorbidities was highest in patients without AFO. Overall, 47.7% of this cohort also had a recorded diagnosis of asthma, slightly more among those with variable AFO unsurprisingly, but with very similar proportions in those with persistent and absent AFO. Among patients without AFO, 58% were receiving some form of respiratory medication (compared to 70% of those with persistent AFO). Many were receiving expensive treatments only recommended for confirmed moderate or severe COPD (long-acting bronchodilators or inhaled corticosteroids). Conclusion Twelve percent of patients with a primary care COPD diagnosis did not have obstructive spirometry. If COPD diagnosis is incorrect, there is potential overuse of harmful or ineffective treatments and other causes of patients’ symptoms may be missed. Abstract P31 Table 1 Baseline demographic and clinical characteristics by consistency of airflow obstruction Consistency of airflow obstruction in patients with FEV1/FVC% (n=13653, 82.9% of total cohort) Persistent (all values FEV1/FVC% <70%) (n=7609, 55.7%) Variable (some values FEV1/FVC% <70%) (n=4413, 32.3%) Absent (all values FEV1/FVC% ≥70%) (n=1631, 11.9%) p-value* Gender male n (%) 4400 (57.8) 2254 (51.1) 752 (46.1) <0.001 Age, years Mean (SD) 69.9 (10.5) 70.8 (10.4) 68.4 (11.7) <0.001 BMI, kg/m2 mean (SD) 26.3 (5.7) 28.1 (5.9) 29.2 (6.7) <0.001 Smoking status n (%) <0.001 Active smoker 2971 (41.0) 1490 (34.5) 534 (35.4) Ex-smoker 4187 (57.7) 2742 (63.5) 915 (60.7) Never smoker 94 (1.3) 83 (1.9) 58 (3.8) FEV1%predicted mean (SD) 54.4 (18.1) 65.2 (18.5) 77.6 (19.3) <0.001 FEV1/FVC% mean (SD) 51.1 (10.6) 66.0 (12.5) 80.5 (7.7) <0.001 MRC Dyspnoea score (1–5) median (IQR) 2 (2,3) 2 (2,3) 2 (1,3) <0.001 Number of comorbidities mean (SD) 2.4 (1.7) 2.9 (1.8) 3.0 (1.9) <0.001 On treatment at baseline n (%) Short-acting bronchodilator 4231 (55.6) 2370 (53.7) 695 (42.6) <0.001 Long-acting bronchodilator 4172 (54.8) 2238 (50.7) 611 (37.5) <0.001 Theophyllines 364 (4.8) 159 (3.6) 36 (2.2) <0.001 Inhaled corticosteroids 3731 (49.0) 2130 (48.3) 631 (38.7) <0.001 Mucolytics 436 (5.7) 205 (4.6) 56 (3.4) <0.001 On none of the above drugs 2266 (29.8) 1283 (29.1) 692 (42.4) <0.001 BMI, body mass index; FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity; MRC, Medical Research Council. *Test for difference depends on variable type / summary measures stated: ANOVA (for mean, SD), Kruskal-Wallis (for median, IQR) and Chi-squared test (for n, %).