Aims In the United Kingdom, around 9 00 000 people have a formal diagnosis of COPD, however, it is believed that over 2 million more may be living with the disease. The ASSIST study (REC:16/5C/0629) aimed to investigate ‘case finding’ strategies in undiagnosed patients with histories and patterns of GP attendance suggestive of COPD.

Methods Possible cases were identified by applying a ‘Read code’ based electronic search algorithm based on previously reported medical record entries suggestive of COPD (Jordan et al, 2016), to GP practice records. The following algorithm variables were used: age, smoking status (ex and current smoker), history of dyspnoea (recorded in the previous 3 years), evidence of a prescription for salbutamol (and number of prescriptions) and evidence of a prescription for antibiotics.

These patients were then invited to their GP clinic for spirometry.

Results From the 2213 patients identified, 611 were excluded as ineligible by the GP, leaving 1601 invited to attend, from which 288 patients provided informed consent and attended. Of those, 76 (26.4%) had airflow obstruction (FEV₁/FVC<0.7) plus typical respiratory symptoms, indicating COPD. 156 (54.1%) had unobstructed lung function (FEV₁/FVC≥0.7) but reported respiratory symptoms similar to those with newly identified COPD: dyspnoea (FEV1/FVC≥0.7 vs FEV1/FVC<0.7)(62.8% vs 55.2%, p= 0.34) and cough (70.5% vs 68.4%, p= 0.85). Patterns of co-morbidity between symptomatic patients with and without airflow obstruction were not significantly different for gastroesophageal reflux disease (GORD)(18.5% vs 10.5%, p= 0.14), depression (26.2% vs 19.7%, p= 0.34) or reported respiratory tract infections in the previous 12 months (27.5% vs 23.6%, p= 0.15),but were for osteoarthritis (18.5% vs 6.5%, p= 0.02), and obesity (BMI >29.9) (30.0 vs 28.5, p= 0.03)

Conclusions Most patients identified by our electronic screening algorithm had significant respiratory symptoms, with approximately one quarter fulfilling COPD diagnostic criteria of post-bronchodilator FEV1/FVC<0.70, and a further half reporting a significant burden of respiratory symptoms and chest infections despite lacking persistent airflow obstruction. Possible causes for respiratory symptoms in the unobstructed group include deconditioning, obesity, or early signs of airways disease. Further clinical characterisations and long term follow up would be recommended for this group of patients.