Primary care use of a C-Reactive Protein (CRP) Point of Care Test (POCT) to help target antibiotic prescribing to patients with Acute Exacerbations of Chronic Obstructive Pulmonary Disease (AECOPD) who are most likely to benefit (The PACE Study)

Design: Individually Randomised Controlled Trial Setting: 60 General Practices in Wales, Thames Valley & South London Target Population: Adults with confirmed COPD (GOLD Grade 1 to 3, FEV/FVC 30%) consulting with acute exacerbation (AECOPD) Health technology assessed: C-reactive protein point-of-care test (CRP-POCT) with training in test interpretation used in addition to current best practice to enhance antibiotic prescribing decisions. Comparator: Current best practice Measurement of cost and outcomes: Co-primary outcomes will evaluate whether addition of the CRP-POCT reduces antibiotic consumption (taking one or more doses of an antibiotic course) within 4 weeks of randomisation without negatively impacting on condition-specific Health-related Quality of Life (HRQoL) at 2 weeks using the Chronic Respiratory Disease Questionnaire Self-Administered Standardised (CRQ-SAS). Secondary outcomes: Prevalence of antimicrobial-resistant respiratory bacteria isolated at 4 weeks; condition specific and generic HRQoL, medication use, adverse effects, primary & secondary care consultations, pneumonia requiring hospitalisation, and costs within 4 weeks. A qualitative process evaluation will examine implementation and acceptability. Sample size: 650 patients. We need 434 participants (inflating to 544 to account for 20% loss to follow-up) to enable us to detect whether there is a reduction (superiority) in the intervention group from a current estimated antibiotic prescribing rate for AECOPD of 70% to 55% within 4 weeks from randomisation at the 5% significance level, with 90% power, and allowing for possible clustering effects. We need 490 participants (inflating to 600 to account for 20% loss to follow-up) to assess whether the intervention group is no worse (non-inferiority) in condition specific HRQoL at 2 weeks, assuming; expected difference between groups of 0, non-inferiority margin of 0.3 (lower than the lower limit of the 95% CI for the minimal clinically important difference), common standard deviation of 1.132, at the 5% significance level, with 90% power. Analysis: Co-primary outcomes; logistic (antibiotic consumption) and linear (CRQ-SAS) regression analysis, based on the intention-to-treat principle. Clustering effects (practice/GP) will be tested for, but not included in final models unless significant. Microbiology: Prevalence of antibiotic resistant bacteria in sputum at 4 weeks by trial arm and +/- antibiotics, adjusting for baseline prevalence. Cost-effectiveness: Analyses from a health service perspective, assessing the total health service costs against the primary co-outcomes and cost utility analysis (CUA) of total costs against generic HRQoL (EQ5D) to generate Quality Adjusted Life Years (QALY). Bootstraping will be used in estimating 95% confidence intervals. Incremental Cost Effectiveness Ratios (ICERs) will be reported. Qualitative data will be analysed thematically. Timetable: 33 months in total. Internal pilot (15 practices) in Year 1 (month 5 - Month 11), with clear stop-go criteria for progression to full trial (60 practices) in Year 2 (Month 15 - month 23). Expertise: Multi-disciplinary team with expertise in microbiology, clinical trials, general practice, respiratory medicine, social science, and health economics, with experience of conducting an RCT of CRP-POCT for respiratory infections in primary care.