Respiratory rate is often measured inaccurately, so there is an urgent need for reliable technology that can consistently monitor this early indicator of declining health for patients in hospital and at home. Here, Isis Terrington describes the development of a novel medical-grade, wearable monitor, which does not require direct skin contact, and its impending clinical evaluation at University Hospital Southampton NHS Foundation Trust.
Changes in respiratory rate are often the initial marker of a patient’s clinical deterioration and therefore the first opportunity to identify and treat acute issues.
It is well established that the early detection of critical illness leads to improved morbidity and mortality rates and reduces the need for services such as high dependency or critical care.1–4 Therefore, it is crucial that these high-risk patients are identified promptly to enhance their outcomes.
Respiratory rate constitutes a vital component of the standard physiological assessment for all hospitalised patients, particularly those with severe conditions such as sepsis or pneumonia, or in cases of haemorrhaging. In addition to its role in detecting acute illness, monitoring respiratory rate is essential for patients with chronic conditions such as asthma, chronic obstructive pulmonary disease and heart failure.
Current approaches to respiratory rate monitoring and assessment
Current methods of monitoring respiratory rate are limited. The most common method of assessment is intermittent manual breath counting. On a general ward, this is completed every four to six hours, or more frequently in patients at higher risk of deterioration, such as every 10 minutes in the post-operative recovery unit.
However, in addition to this being labour-intensive, studies have revealed significant inaccuracies associated with manual counting, including the effects of personal bias and environmental pressures.5,6
Unlike all other vital signs – such as heart rate, blood pressure and temperature – respiratory rate is the only parameter lacking a standard electronic monitoring device for general ward patients. In recent years, an increasing number of continuous monitors have entered the market. None, however, have gained widespread acceptance due to a combination of device inaccuracy, invasiveness leading to patient discomfort, and cost.
Many of these devices determine respiratory rate by detecting subtle changes in heart rate that occur throughout the respiratory cycle, sensed by either photoplethysmography or electrocardiogram. While affordable and simple to use, this method of measuring respiratory rate is unreliable in certain patient subgroups, particularly those with abnormal heart rhythms, such as atrial fibrillation, where heart rate irregularities lead to desynchrony in respiratory rate calculations.
As critical illness can further induce abnormal heart rhythms, heart rate-based systems, although suitable for use in lifestyle devices such as smartwatches, present a risk of inaccuracy in an acute medical setting. Consequently, there is an increasing need for a medical-grade respiratory rate monitor that is non-invasive, affordable and accurate for all patients, regardless of any potential cardiac or respiratory conditions.
Addressing the unmet need in measuring respiratory rate
Our research team at the University of Southampton, in partnership with researchers at the University of Nottingham and part-funded by the National Institute for Health Research, has developed a continuous monitor based on capacitance technology.
The PneumoRator is a small, wearable device (see main image) that is attached to the chest via a simple, non-irritant adhesive patch, which allows for wireless monitoring.
The PneumoRator measures the capacitance variation of a contact sensor caused by breathing, enabling accurate detection of subtle changes due to chest expansion or variations in lung composition. Weighing only 20 g, it is affordable and capable of significantly reducing movement artifacts using gyroscopic technology, making it a novel respiratory rate monitor.
Extensive testing has demonstrated that it achieves accuracy within two breaths per minute, making it highly reliable in comparison to manual respiratory counts. The device’s capacity to filter out movement artefacts enables patients to mobilise normally while still providing precise respiratory rate data, which presents a challenge for many other devices on the market. This feature is particularly crucial for post-operative patients, for whom early mobilisation is a vital aspect of their long-term recovery.
Ensuring accurate validation
Alongside the practical difficulties of creating an accurate monitoring device, meticulous validation of these devices presents its own challenges.
The gold standard for measuring respiratory rate is capnography, which assesses levels of CO2 being inhaled and exhaled. This is particularly relevant during invasive mechanical ventilation where the respiratory rate is closely monitored and CO2 levels can be exported as raw data.
Accessing raw data from a proposed gold standard is essential for the validation process and yet most clinical devices do not facilitate this, making capnography exceptionally valuable. Capnography serves as a true gold standard for measuring the actual respiratory rate rather than providing a derived metric, and it also enables access to its raw data.
Data from the PneumoRator and the capnograph will be processed using the same algorithm, and their concordance will be evaluated using Bland-Altman plots. Once we can demonstrate the high accuracy of the PneumoRator as a clinical sensor, we will evaluate the algorithm we develop to display the measured respiratory rate.
Decisions regarding the development of the algorithm will include factors such as defining the duration for averaging the respiratory rate and the refresh rate, as well as identifying and managing noisy signals caused by actions such as talking and coughing.
Different algorithms will alter the displayed respiratory rate, hence the importance of validating against raw data rather than post-processed data to determine the accuracy of the device. Most commercially available devices have not undergone this level of scrutiny, opting instead to validate the sensor and the algorithm simultaneously.
There are also clinically used devices that are CE marked yet exhibit remarkably wide limits of agreement on their Bland-Altman plots7 when compared against their selected gold standards. For these devices, users cannot discern whether the lack of concordance results from the sensor or the algorithm.
Advancing to clinical evaluation of the respiratory rate device
Our team will soon begin the first clinical trial of the PneumoRatoron patients having major surgery at University Hospital Southampton NHS Foundation Trust. We will monitor their respiratory rate while they are anaesthetised and continue this for up to the first 48 hours of their recovery.
This will act as a validation trial against raw capnography data from the theatre and in the recovery room, followed by manual respiratory rate counts once the patients are on the ward.
The trial will assess the accuracy of the PneumoRator compared with the current standard respiratory rate methods most commonly used in UK hospitals. Additionally, we will evaluate the patient experience of the device to determine its durability and comfort, enabling us to explore the potential for longer-term wear.
This trial represents a significant step forward in formulating a device that addresses current clinical needs. We will subsequently advance our technology to enable the respiratory rate to be transmitted to an electronic patient record system via the Health Level 7 protocol, while also developing a patient-facing app for use at home.
Integrated algorithms will identify concerning trends, potentially alerting the relevant clinicians through various modalities about the possible need for intervention and escalation of care. Further clinical trials assessing markers of patient deterioration and mortality will evaluate the impact of using a PneumoRator compared with standard care.
The implementation of continuous monitoring and the management of potential alarm fatigue will also require collaboration with clinical teams to ensure that both the sensitivity and specificity of the alerts are clinically appropriate.
Conclusion
With ever-increasing technological advancements, respiratory rate monitoring remains an important issue for healthcare. We are excited to start the clinical trial of our novel device, which will bring us one step closer to improving the early detection of deteriorating patients and strengthening clinical outcomes.
Author
Isis Terrington MBChB BSc
Clinical research fellow (critical care), University Hospital Southampton NHS Foundation Trust
Acknowledgements
With thanks to co-researchers:
Neil White PhD DSc CEng CPhys FIET FInstP SMIEEE
Professor of intelligent sensor systems and director of the ECS Centre for Healthcare, School of Electronics and Computer Science, University of Southampton
Harry Akerman MBBS FRCA
Alexander Jackson MBChB
Both of University Hospital Southampton NHS Foundation Trust
Mahdi Shaban PhD
School of Electronics and Computer Science, University of Southampton
Rod Lane PhD CEng MIET
Zelemiq Ltd, Salisbury
Yang Wei PhD
Amjad Ali PhD
Both of the Department of Engineering, Nottingham Trent University
References
- Whebell SF et al. Increased time from physiological derangement to critical care admission associates with mortality. Crit Care 2021;25(1):226.
- Akel MA et al. Less is more: Detecting clinical deterioration in the hospital with machine learning using only age, heart rate, and respiratory rate. Resuscitation 2021;168:6–10.
- Buist M et al. Association between clinically abnormal observations and subsequent in-hospital mortality: a prospective study. Resuscitation 2004;62(2):137–41.
- Areia C et al. The impact of wearable continuous vital sign monitoring on deterioration detection and clinical outcomes in hospitalised patients: a systematic review and meta-analysis. Crit Care 2021;25(1):351.
- Latten GHP et al. Accuracy and interobserver-agreement of respiratory rate measurements by healthcare professionals, and its effect on the outcomes of clinical prediction/diagnostic rules. PLoS One 2019;14(10):e0223155.
- Churpek MM et al. Accuracy Comparisons between Manual and Automated Respiratory Rate for Detecting Clinical Deterioration in Ward Patients. J Hosp Med 2018;13(7):486–7.
- Hayward N et al. A capaciflector provides continuous and accurate respiratory rate monitoring for patients at rest and during exercise. J Clin Monitor Comput 2022;36(5):1535–46.
