Podcast

QCast Episode 73: Respiratory Clinical Trials

Written by Marketing Quanticate | Nov 20, 2026, 4:41:55 PM

Respiratory clinical trials are often treated as one therapeutic area, but the conditions grouped under that label behave very differently. Asthma, COPD, idiopathic pulmonary fibrosis, pulmonary hypertension, chronic cough, respiratory infections and Long COVID share an organ system and little else in terms of biology, progression or treatment goals. Some studies, such as those involving pulmonary nodules, are diagnostic or risk-stratification studies rather than tests of a therapy. In this QCast episode, co-hosts Jullia and Tom look at what that diversity means for protocol design, endpoint choice and data handling, for sponsors who need results they can interpret with confidence.

🎧 Listen to the Episode:

 

 

Key Takeaways

Why respiratory endpoints answer different questions

FEV1, the volume forced out in the first second, is the usual lung function measure in asthma and COPD. FVC tracks the progressive loss of lung volume in fibrosis and other interstitial lung diseases. Physiological change doesn’t always match how patients feel, so patient-reported outcomes and functional tests such as the six-minute walk test show whether a spirometry result translates into something meaningful. Any biomarker used as an outcome needs validating for that context first.

How exacerbation definitions and estimands protect the analysis

An annualised exacerbation rate loses meaning if one site records an event when oral steroids start and another only after an unscheduled hospital visit. Tight protocol definitions and consistent capture at every visit prevent that drift. Respiratory measures also vary within and between patients with season, infection, adherence and environmental exposure, which affects sample size assumptions and model choice. A clearly specified estimand decides in advance how events like rescue medication or a change in background therapy are reflected in the results, so endpoint derivation isn’t argued over at database lock.

Where respiratory data quality tends to slip

Spirometry depends on patient effort, coaching and equipment calibration, which is why ATS/ERS acceptability and repeatability criteria matter for the primary endpoint. Central over-reading of spirometry and imaging can identify a site producing poor-quality tests while there is still time to retrain staff, in line with current guidance on risk-based oversight of critical data. Home devices and electronic diaries remove the site technician from the process but add their own risk. Validated systems, audit trails and routine review of incoming data are needed so that missing entries or a device that has stopped syncing are caught early.

Full Transcript

Jullia
Welcome to QCast, the show where biometric expertise meets data-driven dialogue. I’m Jullia.

Tom
I’m Tom, and in each episode, we dive into the methodologies, case studies, regulatory shifts, and industry trends shaping modern drug development.

Jullia
Whether you’re in biotech, pharma or life sciences, we’re here to bring you practical insights straight from a leading biometrics CRO. Let’s get started.

Tom
So today we’re talking respiratory clinical trials. I’ll be honest, when I hear “respiratory” I picture a fairly uniform area. Spirometry, an inhaler, maybe a symptom diary. Is that roughly accurate?

Jullia
Well, it’s a picture a lot of people have, but it doesn’t really hold up once you look at what’s under that label. You’ve got asthma, COPD, idiopathic pulmonary fibrosis, pulmonary hypertension, chronic cough, respiratory infections, Long COVID. It’s the same organ system, yes, but the biology’s different, the way they progress is different, and what a treatment’s actually trying to do can be completely different.

Tom
Give me two that pull in really different directions.

Jullia
Asthma and IPF are a good pair. With asthma you’ve got different phenotypes, different inflammatory patterns, so a study might be going after fewer exacerbations, better symptom control, or getting people off oral steroids, often in a biomarker-defined group. IPF’s progressive, so the question is really whether you can slow the decline over a long stretch, and whether someone feels a bit better in the short term doesn’t tell you much.

Tom
And some of these aren’t even treatment trials, are they? I’m thinking of pulmonary nodules.

Jullia
Yeah, nodule work is mostly about diagnosis or risk stratification. So you might be looking at an imaging strategy, a molecular test, or some kind of decision-support tool. Long COVID’s almost the opposite problem. You’ve got breathlessness, but also fatigue, cognitive symptoms, reduced exercise tolerance, and that makes it genuinely hard to work out who to enrol and what you should be measuring.

Tom
Okay, that gets us onto endpoints. I think there’s a bit of a temptation to treat the list of respiratory measures like a menu. Pick a lung function test, add a questionnaire, done. What goes wrong there?

Jullia
Well, each one’s answering a different question, so if you swap one for another, you change what the trial can actually tell you. FEV1, that’s the volume you can force out in the first second, is the go-to lung function measure in asthma and COPD. But in fibrosis and the other interstitial lung diseases it’s FVC, the total volume you can breathe out, because that picks up the progressive loss of lung volume much better.

Tom
So if a COPD study shows FEV1 going up, can a sponsor assume patients actually feel better?

Jullia
I wouldn’t, no. The physiology and the symptoms don’t always move together, and that’s why patient-reported outcomes carry so much weight. They tell you about breathlessness, how someone’s managing day to day, their quality of life, from the patient’s side. And then something like the six-minute walk test shows you whether a change on the spirometer actually turns into something the person can do. You see that a lot in pulmonary hypertension and fibrosis studies.

Tom
Exacerbations, though. That has to be the most concrete endpoint of the lot, surely?

Jullia
You’d think so, until two sites define an exacerbation differently. Say one site counts it when the patient starts oral steroids, and the other only counts it if there’s an unscheduled hospital visit. Now your annualised rate means one thing here and something else over there, and nobody’s necessarily noticed. So you need a tight definition in the protocol, and it has to be captured the same way at every visit. Biomarkers are similar, actually. If you’re using one as an outcome, it needs validating for that context before anyone leans on it.

Tom
I want to push a bit on variability. Respiratory patients have good weeks and bad weeks for all sorts of reasons. Doesn’t that make some of these trials really hard to power?

Jullia
It makes it harder, definitely, and that’s a big reason to get statisticians involved early. FEV1, symptom scores, exacerbation counts, they all vary, both within the same patient and between patients. Seasons play into it, infections, whether people are taking their medication, what they’re exposed to. So your sample size assumptions have to account for that, and so does the model you choose for all those repeated measurements.

Tom
The estimand framework comes up a lot here right, but people don’t often say what it actually changes on a study.

Jullia
So an estimand is basically a precise statement of the treatment effect you’re trying to estimate. Where it really earns its keep is that it makes you decide, upfront, what happens when something disrupts the picture after randomisation. Picture a patient who has an exacerbation in month four, goes onto rescue medication, and then later their background therapy gets changed. If the protocol and the analysis plan already say how those readings are handled, your endpoint derivation stays consistent, and nobody’s arguing about it at database lock.

Tom
Though you’d think tighter eligibility criteria would sort out a lot of that variability before it even starts.

Jullia
Well they do help, up to a point. Stricter criteria can cut down the noise, but they also push up screen failures, they slow recruitment, and they narrow who your results actually apply to. Confirming the diagnosis properly, with spirometry or imaging or biomarkers or documented prior exacerbations, does make the results easier to interpret. It also adds work for the sites. And with severe disease, you’re recruiting people who might be on oxygen, or have limited mobility, or be really fatigued, so feasibility work, specialist centres and disease registries make a real difference.

Tom
Okay, so say you’ve got the population and endpoints right. Where does data quality tend to slip?

Jullia
A lot of the time it’s in assessments that depend on how hard the patient tries and how well the test’s run, and spirometry’s the obvious example. If a patient gets coached differently from one visit to the next, or the equipment drifts out of calibration, your primary endpoint picks up noise that’s got nothing to do with the drug. That’s why spirometry should follow the ATS/ERS standards, with acceptability and repeatability criteria and a clear process for tests that don’t meet them.

Tom
And central review picks those up?

Jullia
It can, if it’s set up from the start. Central over-reading of spirometry and imaging might show you one site that keeps producing technically poor tests, or a run of missed visits, while there’s still time to retrain the staff. And that fits with the risk-based approach in current Good Clinical Practice expectations. You work out which data are critical to participant safety and to endpoint reliability, and you concentrate your oversight there.

Tom
We’re seeing more home devices and electronic diaries now. Does that actually fix the site variability, or does it just move it somewhere else?

Jullia
Honestly, it moves some of it. You’ve taken the site technician out of the loop, but now you need validated systems, audit trails, controlled access, and a proper process for reviewing the data as it comes in and raising queries. Otherwise you can have a fortnight of blank diary entries, or a device that’s quietly stopped syncing, and nobody spots it until the data can’t be recovered. And blinding feels close to impossible if it’s a device or a rehab programme.

Tom
How do you protect the result then?

Jullia
If you can’t fully blind the participant, you put the protection around the assessment instead. So independent endpoint assessors, adjudication of events like exacerbations, and objective measures all help reduce the room for bias. And I think it’s worth just stating the limitation plainly in the analysis, rather than hoping nobody brings it up.

Tom
If a sponsor team walked away remembering only a couple of things from this, what would you want them to be?

Jullia
I’d want them to design around the indication, because what works for asthma won’t carry over to fibrosis or chronic cough. I’d want endpoints picked for the question they actually answer. And I’d want them to decide early which assessments really are critical, and put their standardisation and review effort there.

With that, we’ve come to the end of today’s episode on respiratory clinical trials. If you found this discussion useful, don’t forget to subscribe to QCast so you never miss an episode and share it with a colleague. And if you’d like to learn more about how Quanticate supports data-driven solutions in clinical trials, head to our website or get in touch.

Tom
Thanks for tuning in, and we’ll see you in the next episode.

About QCast

QCast by Quanticate is the podcast for biotech, pharma, and life science leaders looking to deepen their understanding of biometrics and modern drug development. Join co-hosts Tom and Jullia as they explore methodologies, case studies, regulatory shifts, and industry trends shaping the future of clinical research. Where biometric expertise meets data-driven dialogue, QCast delivers practical insights and thought leadership to inform your next breakthrough.

Subscribe to QCast on Apple Podcasts or Spotify to never miss an episode.