Early access data, honestly reported
We present the full results from our 12-week NHS-affiliated cohort, alongside its limits and unanswered questions. The data remains preliminary and pending peer review.
Study design and conduct
The respiratory outpatient clinic at an NHS-affiliated teaching hospital in England enrolled n=47 COPD patients with Stage II-IV disease by GOLD criteria. For 12 weeks, participants used the RespiroScreen app for a 90-second acoustic breath check each morning. During monitoring, the AI was blinded to clinical exacerbation status.
Detection of clinically confirmed COPD exacerbations within a 7-day forward window after the first RespiroScreen alert was the primary endpoint. GOLD criteria defined exacerbations, which the attending respiratory physician adjudicated without knowing RespiroScreen alert status.
Participants' usual care pathway stayed unchanged. This monitoring-only study kept clinicians unaware of RespiroScreen alerts throughout observation, avoiding endpoint confounding.
The planned expansion will test the early-access findings across three respiratory centres and a broader population, including people with COPD and comorbid asthma. The ethics application is submitted, with enrolment targeted for Q4 2026.
Study outcomes
12-week early access results from n=47 patients. Preliminary, pending peer review.
Detection within the 7-day window
RespiroScreen monitoring compared with usual care
Comparator: same clinic's historical records from the preceding 12-month period. Preliminary, pending peer review.
Publication and study limits
The complete results are under review at a respiratory medicine journal. Once the manuscript is posted, its preprint will appear here. Current status: under review.
This single-centre cohort was small, at n=47, and used a historical rather than concurrent comparator. Participants were predominantly male (74%) and Stage III-IV (68%), which may limit generalisability.
Further work must assess larger multisite cohorts with concurrent controls, asthma and mixed disease, adherence beyond 12 weeks, and generalisability across ages and comorbidity profiles. We are not making claims beyond what the data supports.
RespiroScreen supports clinical decision-making as a monitoring and alerting tool. It is not a diagnostic device and does not replace clinical assessment. Regulatory pathway discussions remain ongoing.
Take part in the next early access wave
Help build the evidence while protecting patients. The pilot is free and requires no EHR integration.