Improving data quality in respiratory clinical trials

Why respiratory data quality is so challenging

Spirometry data quality depends on a complex interplay of factors: proper technique, effective coaching, cooperative participants, and well‑trained site staff. Subtle breakdowns at any point in this process such as incomplete inhalation, hesitation at the start of exhalation, early termination, or inconsistent effort can introduce significant error. Variability in patient understanding, language barriers, fatigue, or disease severity can further affect performance and reproducibility, even when instructions are clearly provided.

Even when tests appear to meet ATS/ERS acceptability criteria, data may still be inaccurate or biologically implausible. Passing basic quality checks does not guarantee that the captured values truly reflect lung function; errors such as submaximal inspiration or poor blast effort may go undetected without deeper review. In multi-site clinical trials, these risks are amplified by differences in equipment calibration, staff experience, and adherence to standardized procedures.

Unrecognized inhalation problems and protocol violations can increase variability in study results, making it harder to detect true treatment effects. Over time, this added “noise” can dilute statistical power, mask meaningful changes in lung function, and complicate endpoint interpretation. Ultimately, these issues may lead to incorrect conclusions about efficacy or safety, potentially impacting regulatory outcomes, trial timelines, and decision-making around promising therapies.

Centralized data collection and risk‑based monitoring

Regulatory guidance from the FDA encourages sponsors to adopt an approach that focuses oversight on the data and processes most critical to participant safety and study integrity. Centralized data collection plays a key role in this strategy by enabling routine, cross‑site data reviews enabling early identification of anomalies, trends, and inconsistencies.

Sponsors that have actively managed data quality using a centralized approach report meaningful improvements in both data quality and operational efficiency. Statistical analyses, variability assessments, and plausibility reviews allow teams to move beyond surface‑level compliance and identify hidden risks with data inconsistency.

Looking beyond ATS/ERS acceptability

ATS/ERS acceptability standards alone are not enough. While these guidelines are essential, they do not detect issues such as submaximal inhalation, medication washout failures, or implausible longitudinal changes in lung function.

Clario data shows a significant percentage of spirometry measurements in clinical trials may be implausible, even when graded acceptable per ATS/ERS guidelines. High variability patients, particularly those with an elevated coefficient of variation, are far more likely to show exaggerated or misleading treatment responses that may not be representative of a true drug effect.

This is where secondary and tertiary data reviews, including plausibility analysis, become critical. Studies using these methods have demonstrated that up to 85% of the lowest quality data points can be accurately identified and addressed, protecting study outcomes and data quality.

The hidden impact of Submaximal Inhalation Error (SIE)

Submaximal Inhalation Error is a common but often overlooked cause of respiratory data variability. Even spirometry tests that appear technically flawless can be compromised by slow or incomplete inhalation prior to forced exhalation.

Analysis across multiple asthma trials showed that SIE:

• Is common across sites and operators
• Can exceed one liter in volume discrepancy
• Is a major driver of spirometry longitudinal variability

Addressing SIE through proper coaching and practice can dramatically reduce variability and increase confidence in spirometry results. 

Training: The single biggest lever for better data

Targeted, high‑quality training can be an effective tool for improving respiratory endpoint quality. Clinical trials that prioritized in‑person, standardized, and protocol‑specific training experienced:

• Fewer support calls and site issues
• Higher proficiency test pass rates
• Significantly improved spirometry acceptability

Training benefits both sites and participants by reducing repeat visits, minimizing burden, and ensuring accurate data capture the first time. Clario’s approach combines expert trainers, flexible delivery models, and ongoing proficiency monitoring to maintain quality throughout the trial lifecycle.

As respiratory trials continue to grow in complexity, investing in comprehensive data quality strategies is essential for confident decision‑making and successful drug development. Effective quality review needs to go beyond traditional approaches which just focus on ATS/ERS compliance checks to consider the plausibility of the data and changes over time.