AI in PRImary Care Spirometry Pathways for Diagnosis of Lung Disease (APRIL) (APRIL)

Real-world Evaluation of an Artificial-Intelligence Support Software (ArtiQ.Spiro) in Primary Care Spirometry Pathways for the Detection of Lung Disease

To investigate the feasibility of performing a future real-world randomised controlled trial to determine the clinical effectiveness of ArtiQ.Spiro in supporting diagnostic performance of primary care staff in the interpretation of spirometry

This is a feasibility study for a larger multicentre randomised controlled trial (RCT) assessing the impact of the ArtiQ.Spiro software on diagnostic accuracy, care processes, patient and health economic outcomes. The primary objective of this study is to assess the feasibility of a future randomised controlled trial designed to evaluate the real-world clinical effectiveness of an artificial intelligence enabled software to support the diagnostic interpretation and quality assessment of primary care spirometry. Endpoints of the current study will be focused on feasibility and acceptability outcomes. Secondary objectives of this study are to collect data on potential clinical and health economic endpoints for a future randomised controlled trial designed to evaluate the real-world clinical effectiveness of an artificial intelligence enabled software to support the diagnostic interpretation and quality assessment of primary care spirometry. This will help determine the primary endpoint for a future trial and provide data to support a sample size calculation to ensure that any future trial will have adequate power. This is a mixed methods randomised controlled feasibility trial. In PICO format: Population: Individuals with respiratory symptoms referred clinically for primary care spirometry in Hillingdon, Leicestershire, West Hampshire. Intervention: Local primary care spirometry pathway supported by additional artificial intelligence enabled software (ArtiQ.Spiro) Control: Local primary care spirometry alone. Outcomes: Feasibility outcomes with a particular focus on recruitment and retention, and acceptability of intervention and trial design. Clinical outcomes including referrer diagnostic performance and patient health status. Health economic outcomes including health care usage and economic modelling. All participants will undergo their locally agreed spirometry pathway.

In participants randomised to the control group, the primary care referrer will receive the spirometry report plus any additional information as per local spirometry pathway. The primary care referrer will not receive an interpretation and quality assessment report generated by ArtiQ.Spiro.

In participants randomised to the intervention group, the primary care referrer will receive the spirometry report plus any additional information as per local spirometry pathway plus an interpretation and quality assessment report generated by ArtiQ.Spiro.

ArtiQ.Spiro is a decision support software that combines two sub-components - one focussing on quality assessment (ArtiQ.QC), and one on spirometry interpretation (ArtiQ.PFT). It is intended to be used as an adjunct to spirometry to assist with the grading of spirometry quality and the interpretation of spirometry by providing the probability of six disease / or no disease categories.

Contributing data: Screening & recruitment log and trial consort diagram. Progression criteria: Green (≥80% screened eligible, ≥60% eligible recruited), Amber (60-79% screened eligible, 40 - 59% eligible recruited), Red (<60% screened eligible, <40% eligible recruited)

Contributing data: Participation data and Trial Consort diagram. Progression criteria: Green (≥75% retained at 3 months, ≥60% retained at 6 months), Amber (60-75% retained at 3 months, 40-59% retained at 6 months), Red (<60% retained at 3 months, <40% retained at 6 months).

Contributing data: Numbers randomised to control/intervention groups and baseline distribution of characteristics of intervention and control groups. Progression criteria: Green: <10% difference in sample size, proportion of women, proportion of non-white ethnicity, proportion of > 65 years between control/intervention for each site. Amber: <15% difference in above criteria; Red: >15% difference in above criteria.

Contributing data: Email Logs, and Primary Care Referrer Questionnaire. Progression Criteria: Green (≥90% primary care referrers of the participants in the intervention group sent ArtiQ.Spiro report, ≥90% primary care referrers of the participants in the intervention group accessed ArtiQ.Spiro report for interpretation). Amber (85-94% primary care referrers of the participants in the intervention group sent ArtiQ.Spiro report, 75-89% primary care referrers of the participants in the intervention group accessed ArtiQ.Spiro report for interpretation). Red (<85% primary care referrers of the participants in the intervention group sent ArtiQ.Spiro report, <75% primary care referrers of the participants in the intervention group accessed ArtiQ.Spiro report for interpretation).

Contributing data: Email Logs. Progression Criteria: Green (≤10% primary care referrers of the participants in the control group receiving ArtiQ.Spiro report). Amber (11-15% primary care referrers of the participants in the control group receiving ArtiQ.Spiro report) Red (>15% primary care referrers of the participants in the control group receiving ArtiQ.Spiro report)

Contributing data: Stakeholder focus groups. Progression criteria: Green (Reported as acceptable (or can be with minimal modification). Amber (Reported as acceptable with modification).Red (Intervention not acceptable).

Contributing data: Stakeholder focus groups, and Clinical Outcome Measure collection. Progression criteria: Green (Reported as acceptable or can be with minimal modification, Missing data of ≤10% for each measure).Amber (Reported as acceptable with modification, 11-25% missing data for each measure). Red: (Intervention not acceptable, >25% missing data for each measure).

Contributing data: Diagnostic accuracy of ArtiQ.Spiro compared with reference standard. Progression criteria: Green: ≥ 80%, Amber: 65-79%, Red: <65%.

Contributing data: Participation data. Progression criteria: Green (Blinding maintained for ≥85% participants), Amber (Blinding maintained for 84-70%), Red (Blinding maintained for <70%).

Referrers will be asked to rate the quality of the spirometry of the participant (Acceptable, usable, unusable/unacceptable) based on spirometry report (with or without ArtiQ.Spiro report). This will the be compared with Reference Standard(panel of specialists)

Referrers will be asked to rate the confidence in their quality assessment on a Likert scale (from 0: Very unsure to 10: Very confident).

Referrer will be asked to provide their Technical/Pattern Interpretation of the spirometry report received (Normal, Obstructive, Restrictive, Mixed), and their confidence in their technical/pattern interpretation.

Referrers will be asked to rate their confidence in their Technical/Pattern Interpretation on a Likert scale (from 0: Very unsure to 10: Very confident).

Referrers will be asked to provide a preferred diagnosis based on spirometry report (with/without ArtiQ.Spiro report) and information from primary care records.

Referrers will be asked to rate the confidence in their diagnosis on a continuous visual analogue scale (from 0: Very unsure to 10: Very confident).

This will be measured using a Patient Healthcare Resource Use Questionnaire at baseline, 3 and 6 months post spirometry.

The health economic analysis will compare health-related costs and benefits of ArtiQ.Spiro supported spirometry with standard spirometry pathways, as observed within the study.

Three focus groups will be conducted, comprising trial patient participants, primary care referrers and spirometry practitioners. These will be used to assess feasibility and acceptability of the trial procedures, intervention, trial outcomes.

Inclusion Criteria: Adults ≥18 years, irrespective of sex, ethnicity, disability, sexual orientation, marital status, educational level, Referral by GP or nominated representative for primary care spirometry during study period. Patients able to provide informed consent. Exclusion Criteria: Age <18yrs; Absolute contraindication to spirometry. Any locally defined exclusion to spirometry.