Management of chronic obstructive pulmonary disease (COPD) with inhaled corticosteroid/long-acting β₂-agonist (ICS/LABA) improves lung function and health status and reduces COPD exacerbation risk versus monotherapy. This study described treatment use, healthcare resource utilisation (HCRU), healthcare costs and outcomes following initiation of single-device ICS/LABA as initial maintenance therapy (IMT).

Retrospective cohort study.

Primary care, England.

Linked data from the Clinical Practice Research Datalink Aurum and Hospital Episode Statistics datasets.

Patients with COPD and ≥1 single-device ICS/LABA prescription between July 2015 and December 2018 were included.

Treatment pathways, COPD-related HCRU and healthcare costs, COPD exacerbations, time to triple therapy, medication adherence (proportion of days covered ≥80%) and indexed treatment time to discontinuation. Data for patients without prior maintenance therapy history (IMT users) and non-triple users were assessed over a 12-month follow-up period.

Of 13 451 new ICS/LABA users, 5162 were IMT users (budesonide/formoterol, n=1056; beclomethasone dipropionate/formoterol, n=2427; other ICS/LABA, n=1679), for whom at 3 and 12 months post-index, 45.6% and 39.4% were still receiving any ICS/LABA. At >6 to ≤12 months, the proportion of IMT users with ≥1 outpatient visit (10.1%) and proportion with ≥1 inpatient stay (12.6%) had increased from those at 3 months (9.0% and 7.4%, respectively). Inpatient stays contributed most to total COPD-related healthcare costs. For non-triple IMT users, at 3 and 12 months post-index, 4.5% and 13.7% had ≥1 moderate-to-severe COPD exacerbation. Time to triple therapy initiation and time to discontinuation of index medication ranged from 45.9 to 50.2 months and 2.3 to 2.8 months between treatments. Adherence was low across all time points (21.5–27.6%). Results were similar across indexed therapies.

In the year following treatment initiation, ICS/LABA adherence was poor and many patients discontinued or switched therapies, suggesting that more consideration and optimisation of treatment is required in England for patients initiating single-device ICS/LABA therapy.

Infections in primary care are often treated with non-steroidal anti-inflammatory drugs (NSAIDs). This study evaluates whether NSAID prescribing is associated with adverse outcomes for respiratory (RTIs) or urinary track (UTI) infections.

To determine whether there is an association between NSAID prescribing and the rate of adverse outcomes for infections for individual consulting in primary care.

Cohort study of electronic health records.

87 general practices in the UK Clinical Practice Research Datalink GOLD.

142 925 patients consulting with RTI or UTI.

Repeat consultations, hospitalisation or death within 30 days of the initial consultation for RTI or UTI. Poisson models estimated the associations between NSAID exposure and outcome. Rate ratios were adjusted for gender, age, ethnicity, deprivation, antibiotic use, seasonal influenza vaccination status, comorbidities and general practice. Since prescribing variations by practice are not explained by case mix—hence, less impacted by confounding by indication—both individual-level and practice-level analyses are included.

There was an increase in hospital admission/death for acute NSAID prescriptions (RR 2.73, 95% CI 2.10 to 3.56) and repeated NSAID prescriptions (6.47, 4.46–9.39) in RTI patients, and for acute NSAID prescriptions for UTI (RR 3.03; 1.92 to 4.76). Practice-level analysis, controlling for practice population characteristics, found that for each percentage point increase in NSAID prescription, the percentages of hospital admission/death within 30 days increased by 0.32 percentage points (95% CI 0.16 to 0.47).

In this non-randomised study, prescription of NSAIDs at consultations for RTI or UTIs in primary care is infrequent but may be associated with increased risk of hospital admission. This supports other observational and limited trial data that NSAID prescribing might be associated with worse outcomes following acute infection and should be prescribed with caution.