Precision Medicine Meets AI: Advances in Cough Monitoring
Cough is a common and, at times, problematic medical condition that is ripe for disruption. It is one of the main reasons why patients seek medical care. In Europe and the U.S., between 9% and 33% of the population reports cough at any given time; and up to 5% experience chronic cough, defined as cough lasting eight or more weeks in the previous 12 months. Currently, there are limited therapies for cough, but potent new antitussives will soon be available. The high prevalence of cough and limited therapeutic options results in a remarkable financial burden on health systems and large opportunity costs at a societal level.
What is Chronic Cough?
Cough can be both a symptom (something a patient notices, e.g. pain) or a sign (something a provider tracks, e.g. blood pressure). In both cases, it can stem from diseases of the respiratory tract, such as pneumonia, diseases from other systems, like gastroesophageal reflux (GERD), or even be the result of an adverse reaction to a drug such as angiotensin-converting-enzyme (ACE) inhibitors. When it becomes chronic and no underlying disease can be found, cough can also be considered its own disease.
When cough becomes chronic, it can have an enormous impact on a patient’s quality of life. It can cause physical problems such as fatigue and sleep disruption, urinary incontinence, vomiting and even fractured ribs. From a psychological point of view, three out of four patients with chronic cough fear they have a severe disease such as cancer or tuberculosis and often need assurance that this is not the case. From a social point of view, coughing can negatively impact a couple’s relationship (“My spouse cannot tolerate my cough”) and, particularly in the COVID era, causes social embarrassment and limits certain activities (“I can no longer go to the movies”).
Quantifying Cough and the Need for Precision Medicine
Despite living in the era of “precision medicine” in which health care is improved by objective measurements, until recently it was simply not possible to measure cough. Currently, cough is managed entirely based on subjective assessment. For example, the physical, psychological and social dimensions are assessed by healthcare providers using questionnaires and other patient reported outcome measures such as the Leicester Cough Questionnaire and the visual analog scale.
While several of these questionnaires and outcome measures have been validated in patients with chronic cough, they remain subject to perception bias. It is no secret that one’s mood can influence the perception of certain symptoms, such as pain. Chronic cough is no exception as it can cause anxiety and depression via the dimensions described above. These adverse mental states can negatively influence a patient’s perception of his or her cough as well. In other words, chronic cough and anxiety/depression can be intertwined in a vicious cycle that makes patients feel worse than a simple tallying of symptoms could predict. This cycle can greatly increase the cost of care, mostly via diagnostic tests. One way to address the cost issue is to assess not only the impact of cough on quality of life but also the objective quantity of cough (“Is your cough bothering you?” vs “Are you coughing less?”)
Cough is highly variable; it can vary seasonally, throughout the course of a disease and its treatment, and even throughout the day, posing additional complexity in its diagnosis and management. And yet, cough as a quantifiable sign is, still today, only assessed objectively in the few minutes doctors have in a medical consultation. This is akin to exploring a large mural painting through a keyhole; only a small portion of the picture can be seen at any time, and it becomes difficult to understand the general situation or its trend. This large difference between a patient’s experience and the small window the healthcare provider can observe directly, leads to misunderstandings and discussions about the real versus the perceived impact of cough. Patients feel unheard and like their cough is not taken seriously.
The diagnosis and management of cough is largely based on serial empiric trials of medication—and yet, the impact of these drug trials are not actually measured. And providers grow frustrated that chronic cough patients return again and again to the clinic. The result is a damaged relationship between provider and patient, as well as increased costs via serial consultations with pulmonologists, otolaryngologists, gastroenterologists and allergists.
Consider the lack of objective assessment of cough compared with the situation for another respiratory symptom/sign: inability to exercise. Patients with COPD or pulmonary hypertension often report shortness of breath in response to milder physical effort than their healthy peers of the same age and weight. Healthcare providers have developed a clever method to measure this highly variable sensation, the six-minute walk test. On this highly standardized test, a patient is asked to walk for six minutes and the total distance achieved as well as the shortness of breath are assessed by the provider. The distance achieved can be used as an objective indicator of disease progression, response to treatment or even prognosis. Shorter distances or faster decline are associated with higher mortality in patients with COPD. Associations such as this are possible today because the six-minute walking test is easy to implement, leading to a broad clinical use and the collection of a large amount of patient data.
Looking Ahead: AI and Remote Cough Monitoring
Recent developments, specifically artificial intelligence and the ubiquity of smart devices, now enable us to monitor cough unobtrusively and continuously for periods of time that can span months. The importance of tracking cough over longer periods of time is twofold: 1) it enables underlying patterns to emerge from the hour-to-hour and day-to-day noisiness inherent in cough; 2) it enables healthcare providers to measure how cough is varying over time in response to lifestyle, medication and other changes. Objective cough quantification can be combined with patients’ perceptions to better determine diagnosis, treatment response and prognosis. We now can bring cough into the era of digital and precision medicine, rather than continuing to squeeze its observation into the too-narrow confines of the patient clinic visit.
Today, the average smart watch can measure, store, transmit and even send alerts related to physical activity, heart rate, ECG and sleep. Acoustic signals such as cough pose a slightly harder challenge, but tools are arriving on the market and in the clinic. What stands in the way of widespread adoption of objective cough monitoring? Only hesitance in the face of technological innovation and a lack of familiarity with novel cough counting tools.
Original post: https://medtechintelligence.com/column/precision-medicine-meets-ai-advances-in-cough-monitoring/
Post publish date: