For chronic disease research and treatment, objective symptom monitoring is key to provide clinical benchmarks that evaluate the effectiveness of therapies and the patient experience. In the world of chronic cough, a new report in Lancet Digital Health identifies the crucial role of objective monitoring.
After analyzing 77 studies evaluating digital cough monitoring tools, the paper sheds light that not all monitors are created equal. Here Alexandra Zimmer, PhD, the lead researcher from McGill University behind the report discusses the use cases in which these tools excel, the role of both subjective and objective data, and future opportunities for broader public health use.
-
From a digital health perspective, what gap does objective cough counting fill that traditional symptom reporting has struggled to address?
Objective cough counting addresses a fundamental gap in traditional cough reporting, which relies on intermittent, patient-biased symptom reporting. Traditional patient-reported outcomes (PROs) and outcome diaries are both inherently episodic and retrospective. They rely on a patient’s memory, are vulnerable to recall bias, and often will only reflect a patient’s perception of symptom burden rather than their physiological reality. Objective cough monitoring, by contrast, provides continuous, high-resolution data. This transforms cough from a subjective, unquantified complaint into a quantifiable digital signal. This is an important shift for clinical trials, where objective endpoints are essential for evaluating the true efficacy of new anti-tussive therapeutics.
-
Your review highlights cough as a potential digital biomarker. What characteristics make cough particularly well suited for continuous, connected monitoring?
Cough is uniquely suited for continuous monitoring because it is a naturally occurring, passive signal that does not require active effort from the patient, unlike a spirometry test or a blood draw. However, as our review highlights, cough frequency alone is rarely a definitive diagnostic indicator. Different respiratory diseases often share very similar cough counts, which makes differential diagnosis based on cough frequency alone quite challenging.
Its true value as a digital biomarker emerges when we look at it holistically and longitudinally. By combining frequency data with the acoustic signature, such as wet versus dry sounds, and temporal patterns, such as nocturnal versus daytime rhythms, we can build a much clearer picture of disease activity. Ultimately, cough should not be used as a standalone marker for clinical decision-making. Instead, it should be used as a vital component in a broader suite of biomarkers.
-
How do automated cough monitoring technologies compare technically and clinically with legacy methods like patient diaries or short-duration recordings?
Automated systems offer continuous, passive data capture, algorithmic classification, and scalable analytics. Legacy methods like patient diaries are labor intensive, not all-encompassing, and subject to patient recall bias and error. Portable, research-grade monitors like the Leicester Cough Monitor (LCM) and VitaloJAK are highly validated, but often involve external microphones that can be cumbersome and affect long-term patient adherence. Modern solutions like Hyfe’s CoughMonitor Suite leverage smartphones and mics in wearable devices to reduce patient burden and enable AI-powered monitoring in devices patients already carry.
Automated systems also provide objective, quantifiable endpoints that are more clinically significant than perception-based severity scores. While diaries capture how patients feel about their condition, digital tools capture frequency and severity. Both are certainly important metrics, but they measure differing aspects. With digital monitoring, clinicians and researchers can objectively evaluate patterns, treatment response trajectories, and early deterioration signals.
-
What did your findings reveal about the reliability of subjective patient-reported outcomes versus sensor-derived cough data?
The review revealed only moderate correlations between objective cough frequency and patient-reported severity or quality-of-life measures. This demonstrates how subjective tools are vulnerable to recall bias, anxiety-related amplification, end-of-scale effects (where responses cluster at the maximum values on a subjective scale, preventing the scale from capturing further increase in symptom severity), and even weak parent-child agreement (which occurs when a caregiver’s subjective assessment of a child’s symptoms differs significantly from the child’s actual experience). In contrast, objective sensors provide stable, reproducible frequency counts that aren’t influenced by psychological or situational factors.
This moderate correlation, however, underscores the impact of that patient experience. Even a small increase in cough frequency may produce a large quality-of-life impact depending on the patient’s baseline and context. This reinforces an important principle that sensor-driven data should augment, rather than replace, patient-reported data, together creating a multidimensional view of disease burden.
-
For developers of connected medical devices, what design considerations are most critical when building cough monitoring tools for real-world use?
Real-world performance must be prioritized over laboratory accuracy alone. Controlled trial environments minimize background noise and ensure device compliance, while real-world use is more complicated, introducing acoustic clutter and adherence challenges. Ergonomics are equally critical, as wearables may achieve high fidelity but can lead to device fatigue in routine use. Solutions like Hyfe’s that incorporate smartphones or watches aim to reduce some of this friction by embedding monitoring tools into devices that are already accepted socially and practically. Ultimately, human factors like privacy perception, stigma, and comfort are just as important for patients as algorithmic accuracy is for developers. To ensure adoption beyond pilot phases, developers must also consider interoperability, battery efficiency, secure data handling, and clinical-facing summarization tools. Successful implementation requires the continuous collection of qualitative end-user data. By incorporating end-user feedback into the design process, developers can ensure these tools are not only accurate, but also equitable and practical for diverse populations.
-
How might continuous cough monitoring integrate with other digital health data streams, such as wearables, remote patient monitoring platforms, or AI-driven analytics?
Integration into broader ecosystems makes continuous monitoring even more powerful. Combination with metrics like heart rate, respiratory rate, oxygen saturation, activity levels, or sleep data can contribute to a multidimensional digital phenotype. AI analytics can establish personal baselines and detect deviations rather than relying on population averages, making cough a contextualized and predictive biomarker. For example, a rise in cough counts paired with declining activity and elevated resting heart rate may signal early exacerbation in COPD.
However, moving this data from a personal device to a clinical provider requires overcoming significant hurdles. We currently face a lack of standardized protocols for interoperability. Healthcare systems and digital interfaces vary wildly between countries and even between hospitals in the same city, making it difficult to relay a patient’s personal cough data into a clinician’s workflow seamlessly. This also raises vital questions about privacy. For this to work, we need transparent opt-in processes where patients have clear control over their data.
Stationary systems like Flu Sense offer a different way to use this technology for public health surveillance. These systems can be placed in hospital waiting rooms to count coughs in the environment. If these tools are built to be interoperable with local health offices while keeping people completely anonymous, they could help health officials track the spread of illness in a community.
-
What are the biggest barriers to clinical adoption of objective cough monitoring, and how can health technology companies help overcome them?
The largest barriers to adoption relate to workflow integration, reimbursement, and demonstrated clinical utility. Clinicians require synthesized, actionable summaries rather than raw data streams to reduce their burden and ease integration into their existing workflows – a gap Hyfe is bridging with its solutions. Health systems will seek alignment with existing reimbursement structures and value-based care initiatives to earn their adoption. Payors will look for incremental clinical value to demonstrate improved outcomes, while regulators will require data that backs device performance in real-world settings. Research-grade systems like LCM and VitaloJAK have strong validation histories, but their complexity may limit scalability. App- and wearable-based solutions like Hyfe, however, lower barriers to deployment but still must demonstrate incremental clinical value. Additionally, patients will need reassurance that data collected by always-on microphones is protected.
Finally, while many studies have focused on the technological development of these tools or their use in controlled clinical trials, there are not enough studies evaluating them in real-world clinical settings. We need more research that specifically measures how objective cough data transforms patient health and improves clinical decision-making.
-
Looking ahead, how do you see objective cough data influencing the future of digital endpoints, regulatory approval, and value-based care models?
Objective cough data has significant implications for digital endpoints and regulatory pathways. Continuous cough counts are already being incorporated into antitussive and respiratory trials because they provide quantifiable, statistically robust endpoints compared with subjective measures and with 24-hour snapshots. Regulators have shown interest in having a regulated cough counter available for clinical use.
As regulators increasingly evaluate digital biomarkers based on analytical and clinical validity, tools like Hyfe and established systems like LCM and VitaloJAK contribute to a growing evidence pool that supports cough as a measurable physiologic signal. For value-based care models, continuous monitoring that predicts deterioration or tracks recovery could shift respiratory management from reactive to proactive and from episodic in-clinic visits to real-world physiologic measurement. Ultimately, it is a matter of time before cough joins other endpoints among the set that are routinely monitored in clinical settings.
