Which Of The Following Changes Does Not Cause Decreased Efficiency Of The Respiratory System?

• Loading metrics

Open Admission

Peer-reviewed

Inquiry Commodity

Analyzing changes in respiratory charge per unit to predict the hazard of COVID-19 infection

• Dean J. Miller,
• John 5. Capodilupo,
• Michele Lastella,
• Charli Sargent,
• Gregory D. Roach,
• Victoria H. Lee,
• Emily R. Capodilupo

x

• Published: Dec 10, 2020
• https://doi.org/10.1371/journal.pone.0243693

Figures

Abstract

COVID-nineteen, the illness caused past the SARS-CoV-ii virus, can cause shortness of breath, lung impairment, and dumb respiratory part. Containing the virus has proven difficult, in big role due to its high transmissibility during the pre-symptomatic incubation. The study's aim was to determine if changes in respiratory rate could serve as a leading indicator of SARS-CoV-2 infections. A full of 271 individuals (age = 37.3 ± ix.5, 190 male, 81 female) who experienced symptoms consequent with COVID-xix were included– 81 tested positive for SARS-CoV-ii and 190 tested negative; these 271 individuals collectively contributed 2672 samples (days) of data (1856 healthy days, 231 while infected with COVID-xix and 585 while negative for COVID-nineteen but experiencing symptoms). To train a novel algorithm, individuals were segmented every bit follows; (1) a preparation dataset of individuals who tested positive for COVID-nineteen (due north = 57 people, 537 samples); (2) a validation dataset of individuals who tested positive for COVID-19 (n = 24 people, 320 samples); (3) a validation dataset of individuals who tested negative for COVID-xix (n = 190 people, 1815 samples). All information was extracted from the WHOOP organisation, which uses data from a wrist-worn strap to produce validated estimates of respiratory charge per unit and other physiological measures. Using the preparation dataset, a model was developed to estimate the probability of SARS-CoV-2 infection based on changes in respiratory rate during dark-time sleep. The model's power to identify COVID-positive individuals not used in training and robustness against COVID-negative individuals with similar symptoms were examined for a critical half dozen-mean solar day period spanning the onset of symptoms. The model identified twenty% of COVID-nineteen positive individuals in the validation dataset in the two days prior to symptom onset, and lxxx% of COVID-19 positive cases by the tertiary twenty-four hours of symptoms.

Citation: Miller DJ, Capodilupo JV, Lastella M, Sargent C, Roach GD, Lee VH, et al. (2020) Analyzing changes in respiratory rate to predict the chance of COVID-xix infection. PLoS Ane fifteen(12): e0243693. https://doi.org/10.1371/journal.pone.0243693

Editor: Enzo Pasquale Scilingo, Universita degli Studi di Pisa, Italian republic

Received: June 23, 2020; Accepted: Nov 26, 2020; Published: December 10, 2020

Copyright: © 2020 Miller et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted utilize, distribution, and reproduction in whatsoever medium, provided the original author and source are credited.

Data Availability: Public data sharing for these purposes is ethically and legally restricted. The data is collected and used with the consent of individuals who purchase a WHOOP membership (collectively, "WHOOP Members") and hold to the WHOOP terms of utilize (https://www.whoop.com/termsofuse/) and privacy policy (https://world wide web.whoop.com/privacy/full-privacy-policy/). While WHOOP terms of employ and privacy policy allow WHOOP to use nerveless data that has been aggregated or de-identified in a manner that no longer identifies an individual for advisory purposes, analytics or WHOOP's ain research purposes, the consent obtained from WHOOP Members does not extend to making the data publicly available for a 3rd party to use for its own purposes. As such, WHOOP's legal department volition not permit the data to be shared for these purposes. Information access queries can be directed to the WHOOP Executive squad via the Vice President of Performance at WHOOP – Kristen Holmes (holmes@whoop.com).

Funding: There was no specific funding provided for this projection. Still, ERC, JVC and VHL are employees of Whoop Inc. In improver, Whoop Inc is sponsoring the employment of DJM's position at CQUniversity. The specific roles of these authors are articulated in the 'writer contributions' section. The funders had no role in study blueprint, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors of this manuscript have read the periodical's policy and have the following competing interests: ERC and JVC are shareholders and employees of Whoop, Inc; VHL is an employee of Whoop, Inc; DJM's position at CQUniversity is sponsored by Whoop Inc. There are no other competing interests for the authors. There are no patents, products in evolution or marketed products associated with this research to declare. This does not alter our adherence to PLOS One policies on sharing information and materials.

Introduction

The novel coronavirus affliction (COVID-nineteen) is caused by the astringent acute respiratory syndrome coronavirus two (SARS-CoV-two) virus [1] and predominantly presents as a lower respiratory tract infection. Severe cases of the disease can effect in alveolar impairment and progressive respiratory failure [2]. Containing the virus has proven difficult due to its high transmissibility during the pre-symptomatic incubation phase [3] and widespread shortages of testing.

Exterior of traditional laboratory testing, few applied COVID-19 monitoring systems have been proposed. Some businesses looking to reopen post-obit physical distancing mandates have implemented daily monitoring of temperature to identify and isolate potentially infected individuals. While this method would be constructive in identifying and isolating workers that were experiencing fevers–infectious individuals that exercise not present with a fever may be exposed to colleagues during the ii to xiv-day pre-symptomatic incubation flow [4,5]. This limitation of fever-based screening is pregnant given that infectiousness is known to height 0–two days prior to symptom onset [6].

Respiratory rate is a mutual screening tool to identify lower respiratory tract infections in clinical settings [7]; guidelines define tachypnea as a respiratory rate greater than 20 respirations per minute (rpm), and advise further tests (e.g., chest radiography) when present [7]. While such thresholds are useful in clinical settings, they are only implemented one time symptoms have emerged and are non sensitive to intraindividual differences in normal respiratory function. Given that COVID-nineteen impairs and damages the respiratory organisation [2], information technology is reasonable to advise that changes in respiratory efficiency–and therefore resting respiratory charge per unit–might occur in the early stages of infection. In this context, noninvasive daily monitoring of respiratory rate may be used to detect subclinical intraindividual deviations and identify potential infections that would otherwise be disregarded by clinical thresholds [7].

If deviations in respiratory rate are found to exist an accurate indicator of COVID-19 infection, respiratory rate monitoring could form part of the protocol used by medical professionals and organizations to enforce self-isolation and target testing. The aim of this written report was to appraise the ability of a novel algorithm to classify changes in respiratory rate as indicative of COVID-19 infection immediately prior to and during the first days of symptoms and to evaluate the model'southward robustness to instances of like clinical presentations with differing etiology.

Materials and methods

Respiratory rate, resting heart charge per unit (RHR) and center rate variability (HRV) were measured using the WHOOP strap; the algorithms used to derive these metrics from the wearable'due south photoplethysmography sensor are across the scope of this newspaper, merely have been validated in third party analysis and shown to have high levels of agreement with gold standard methodology [8]. The WHOOP strap is a small, waterproof, and rechargeable device containing a photoplethysmogram, accelerometer, thermometer, capacitive bear on sensor, and gyroscope that tin can be worn comfortably 24-hours per day and lasts 5 days between charges. The wrist-worn strap wirelessly transfers data to mobile devices running the associated WHOOP app; from at that place, data is transferred to a secure cloud-based data storage and processing server, collectively known as the WHOOP system. This study was canonical by the Central Queensland Academy Human Enquiry Ethics Commission. Data were collected with the written consent of individuals via WHOOP Inc'southward terms of service.

The following physiological data were obtained from the WHOOP arrangement for this report:

• Respiratory rate–median value of respirations per infinitesimal, derived each night during the main sleep period via photoplethysmography.
• RHR–average beats per minute sampled during the concluding v minutes of the terminal episode of deadening wave slumber each night.
• HRV–sampled during the last five minutes of the last episode of slow wave sleep each night using the root mean square of successive RR interval differences (rMSSD) method in units of milliseconds.

In addition to automated tracking of physiological information, the WHOOP app supports tracking of manually reported contextual factors. In response to the COVID-19 pandemic, on fourteen March 2020, WHOOP added the ability to runway COVID-nineteen symptoms and test results. Member-reported incidences of COVID-19 symptoms and test results were extracted through 06 June 2020. Respiratory rate, RHR, and HRV were extracted betwixt 01 Nov 2019 and 30 November 2019; respiratory charge per unit was additionally sampled between 01 Jan 2020 and 06 June 2020 for individuals that reported test results for COVID-nineteen.

Data assay

Stability of metrics

Respiratory rate was evaluated as a potentially sensitive indicator of infection due to anecdotal observations of depression internight variation in WHOOP data. A search of Pubmed showed no extant longitudinal studies reporting on variability of nightly respiratory charge per unit in healthy adults. Therefore, to support the use of this metric in the model, a supplementary dataset from Nov 2019 was generated for analysis. A engagement range of 1 November 2019 through 30 November 2019 was chosen to avoid confounding factors related to the COVID-xix pandemic. A total of 25,000 WHOOP members were randomly selected (n = 750,000 nights); the only inclusion criteria was having respiratory rate recorded on all xxx consecutive nights. Resting heart rate and resting heart rate variability over this period were included for comparison. The following variables were calculated from the November dataset for each of the physiological metrics:

• Mean intraindividual mean: hateful within-member ways.
• Standard deviation of intraindividual means: standard difference of within-member means.
• Mean intraindividual standard deviation: mean within-fellow member standard deviation.
• Standard departure of intraindividual standard deviations: standard deviation of within-member standard deviations.
• Coefficient of variation: intraindividual standard deviation divided by the intraindividual mean.

Predictive model

Data extraction.

A total of 271 adults (age = 37.three ± nine.v, 190 male, 81 female) were included in the report; inclusion criteria were (ane) self-reporting symptoms consistent with COVID-19 (i.e., cough, fever and/or fatigue) and (ii) having been tested for the SARS-CoV-2 virus. These individuals were separated into three groups:

• grooming dataset: COVID-19 positive individuals who began experiencing COVID-19 symptoms betwixt 14 March 2020 and 14 April 2020 (northward = 57);
• validation dataset i: COVID-19 positive individuals who began experiencing COVID-19 symptoms betwixt fourteen Apr and half-dozen June 2020 (n = 24);
• validation dataset ii: individuals who experienced COVID-nineteen symptoms but reported a negative examination result (n = 190).

In social club to develop the algorithm, data was categorized by twenty-four hours relative to symptom onset (solar day 0) into:

• healthy days: data extracted from 30 to 14 days prior to symptom onset;
• infected days: data extracted between two days prior to symptom onset and 3 days post symptom onset.

All 271 individuals contributed to both categories, with a maximum of xv salubrious days per person and vi infected days per person. For the training dataset, 146 infected days and 391 healthy days were included. Due to the form imbalance betwixt infected days and healthy days, synthetic samples (i.e., days) were generated for the positive grade (i.due east., infected days) by adding uniformly distributed random dissonance on the interval [0, 1) to each infected day, bringing the number of infected days to 292. Generation of synthetic samples was done but for training and was not repeated for the validation datasets. Synthetic samples were only used for training the model and were excluded from the analysis of the training fix presented throughout. For validation dataset 1, 85 infected days and 235 healthy days were included. For validation dataset ii, 585 infected days and 1230 healthy days were included.

Data transformation.

The daily respiratory rate value (herein, current value) for each private was transformed into features based on how it compared to the values taken on each of the 21 days prior. In all datasets, just current values for which the prior 21 consecutive nights' respiratory rates were available were included. These features capture the dynamics of deviation from recent trends along a multifariousness of time scales. In generating the classifier's features, the following metrics were used:

• RR ₀ : current value (a respiratory rate)
• : median of the respiratory rates in the fourteen day period between 21 and vii nights prior to the current value.
• σ: standard deviation of the respiratory rates in the 14 day period between 21 and 7 nights prior to the electric current value.
• μ ₂ : mean of the electric current value and immediately prior night'due south respiratory rate.
• μ ₃ : mean of the current value and immediately prior two nights' respiratory rates.
• μ _half-dozen : median of the immediately prior 6 nights of respiratory rates, excluding RR ₀ .
• chiliad ₆ : slope of the linear regression of the collection of the respiratory rates of the electric current day to 6 days prior, excluding RR ₀ .

The features derived from these metrics were:

1. μ ₂ /
2. (μ ₂ - ) / σ
3. RR ₀ - μ _iii
4. m ₆
5. μ ₂ - μ ₆

Collectively, these internally derived and novel features capture dynamics of the changes in respiratory rate over fourth dimension. Utilizing a modified z-score (i.eastward. utilizing a median value rather than mean), creates a baseline that is robust to outlier values and more stable over the short fourth dimension periods explored in this study. Using a lagged baseline, as in in features 1 and 2, allows data to increment during an incubation period without artificially elevating the baseline and masking the bear on of the SARS-CoV-2 infection.

A gradient additional classifier [9] was trained using Python Language Software (version 3.six.2) on the derived features to return a probability of SARS-CoV-2 infection on salubrious and infected days.

Model functioning.

In order to evaluate the model'southward performance for classifying salubrious and infected days, a threshold value was assigned to the probability output of the model such that meeting or exceeding that threshold was equivalent to classifying salubrious or infected days as COVID-19 positive (C+); while declining to exceed the threshold was equivalent to classifying good for you or infected days as COVID-nineteen negative (C-). The threshold value was strategically set at 0.3 to maximize the utility of the model by reducing the chance of false negatives at the expense of increasing fake positives, in recognition that false negatives may take higher costs to society than simulated positives. The model'due south operation for classifying healthy days and infected days for each dataset was also evaluated at that threshold past calculating sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV).

Results

Stability of metrics

Thirty-day intraindividual variability of metrics are presented in Table one. Respiratory charge per unit was constitute to have a lower coefficient of variation than both centre rate variability and resting heart rate.

Predictive model

The model returned a continuous probability that a given sample is indicative of a SARS-CoV-2 infection (Fig 1). Tabular array 2 summarizes the performance of this model afterwards mapping the model's continuous probability output into C+ and C- classifications, bifurcated on the threshold of 0.3.

Fig 1.

Top panel: Distribution of classifier output from the training dataset, separating infected days and healthy days. Middle panel: Distribution of the classifier output from the validation dataset separating COVID-19 positive infected days and healthy days. Lesser panel: Distribution of the classifier output from the validation dataset separating COVID-19 negative infected days and healthy days. For all panels, black confined denote infected days (summit and heart: COVID-19 positive, bottom: COVID-19 negative), greyness bars denote healthy days; y-axes show the relative frequency of the classifier's output values, binned with widths of 0.04, with infected days and healthy bars plotted side by side. Dashed blackness lines testify the threshold value of 0.iii which is used to map the continuous distribution of the model's output to a binary classification.

https://doi.org/10.1371/journal.pone.0243693.g001

Fig 2 quantifies the percentage of individuals in each dataset to whom the model would accept assigned a positive or negative COVID-19 classification relative to symptom onset. Note that simply the subset of users who had available samples on each of the 6 evaluated days were included. In the training dataset (north = 23), 73.nine% of individuals had at least ane correct C+ classification over the six 24-hour interval period. In validation dataset 1 (north = x), 80.0% of the individual subjects had at least i correct C+ classification; during that same relative fourth dimension period, while 34.2% of the individuals in validation dataset 2 (northward = 79) had ane or more than C+ classification.

Give-and-take

The aim of this study was to assess the power of a novel algorithm to classify changes in respiratory rate, equally indicative of COVID-19 infection immediately prior to and during the first days of symptoms. The major findings of this study are (1) the stability of nightly respiratory charge per unit measurements within healthy individuals makes it a useful metric for tracking changes in health; (2) the model is capable of distinguishing between healthy days and infected days for individuals who tested positive to COVID-19 as well as those who had symptoms just tested negative; (3) the model identified 20% of individuals of COVID-xix positive prior to the onset of symptoms, and correctly identified 80% of COVID-nineteen positive individuals past the third day of symptoms.

Stability of metrics

This is the showtime study to written report on nightly changes in resting middle rate, centre rate variability, and respiratory rate in healthy individuals. Our findings show that while interindividual variation in nightly respiratory rate can be large, intraindividual variability across 30 nights is typically quite pocket-size, with mean intraindividual standard departure of 0.51 ± 0.xx rpm. The finding that nighttime median respiratory rate in healthy individuals has low internight variability is a novel finding of this paper and suggests that deviations in respiratory rate may be a useful indicator of acute changes in lower respiratory tract health.

Predictive model

This is the first study to examine the potential for continuously monitored respiratory rate to identify early stages of COVID-19 infections. A predictive algorithm was formulated to leverage individual baseline data and determine if nightly respiratory rate when contextualized by 21-24-hour interval trends tin can predict COVID-19 infections. A pregnant finding was that twenty% of COVID-19 positive individuals were identified prior to the onset of symptoms and 80% of COVID-xix positive individuals were correctly identified by the 3rd day of symptoms (Fig 2). This suggests that the final stages of incubation and early on stages of the infection may have a detectable signature that tin identify individuals who should self-isolate and seek testing. This novel approach may be particularly advantageous for individuals with low resting respiratory rates, who despite experiencing significantly elevated respiratory rates relative to their personal baseline, might not exist medically classified equally tachypneic according to globally divers norms [seven].

There are a number of practical applications for the electric current model'due south power to analyze daily changes in respiratory rate, including aiding testing protocols and monitoring essential workers. The limited availability of testing kits and the fourth dimension-intensive nature of most laboratory tests makes repeated screening for an private both plush and impractical. Despite strict testing criteria, 3–12% of laboratory COVID-19 tests return a positive result [10]. Given the operation of the current model at discriminating between COVID-19 and other illnesses with like symptomatology, information technology could potentially be used to streamline testing protocols in areas that may accept testing kit shortages. In addition, this algorithm may exist peculiarly useful in situations where physical distancing is impractical (e.g., industry, aristocracy sport, healthcare), merely where a positive COVID-xix case could have major implications. Along with recommended hygiene and physical distancing protocols, wearable technology could exist used as a bespeak of care measure out to monitor employees and/or athletes during the transition back to piece of work and contest.

Some purlieus conditions should exist considered when interpreting the results of this written report. Firstly, COVID-19 examination results and date of symptom onset were reported by WHOOP members direct in the WHOOP app and were non verified past medical professionals. All COVID-19 positive individuals included in the analyses experienced symptomatic COVID-19 disease, thus the model has not been evaluated for its operation in fully asymptomatic cases; given that asymptomatic COVID-xix cases are contagious [eleven], further analysis is required to decide the utility of the algorithm in those cases. Last diagnoses from individuals who tested negative to COVID-19 were not collected in the present study, therefore the COVID-19 negative cohort may represent individuals with a multifariousness of illnesses; fFurther research beyond the scope of this study is warranted to segment model performance by not-COVID-19 diagnosis, particularly for weather condition with similar initial clinical presentations. The number of unique individuals included in the analyses could be seen every bit a limitation, however the model was trained using information extracted from multiple days from each user.

When interpreting the model's performance, information technology should exist noted that the sensitivity and specificity of the model are determined both past the discriminatory ability of the features and past the threshold selected to discriminate between C+ and C- designations. As illustrated in Fig 1, healthy days tend to be assigned lower probabilities of beingness COVID-19 positive while infected days tend to be assigned higher probabilities. For the aforementioned probability distributions, a higher threshold would effect in higher specificity just lower sensitivity, while a lower threshold makes the opposite tradeoff increasing sensitivity while decreasing specificity. The optimal threshold for a given model is dependent on its intended awarding; while a threshold of 0.5 would maximize accurateness, this is often non the metric most associated with applied utility. For the algorithm presented in this written report, a false positive–indicating that COVID-xix negative individual may be COVID-nineteen positive–ways that an private self-isolates unnecessarily, while a false negative–indicating that a COVID-nineteen positive private is COVID-19 negative–could result in the individual interacting with and potentially infecting others. Therefore, the reduced threshold value of 0.3 was called in recognition that imitation negatives have higher costs to lodge than false positives. We annotation that the threshold selection process was non particularly rigorous and that the optimal tradeoff between imitation positives and faux negatives would be dependent on a number of unknown factors beyond the scope of this analysis. Finally, information technology should exist noted that the WHOOP strap is not a medical device and should not be used as a substitute for professional medical advice, diagnosis or treatment.

Conclusions

This report presents a novel, non-invasive method for detecting SARS-CoV-2 infection prior to and during the first days of symptoms. The findings signal that the early stages of the infection may have a detectable signature that could help identify individuals who should self-isolate and seek testing. Hereafter investigations should examine the performance of respiratory rate based algorithms to classify infection amongst larger and more than various cohorts.

Acknowledgments

We thank Dr. Douglas Johnston and Dr. Aaron Weiss from the Cleveland Clinic for their help with this project.

References

1. 1. Coronaviridae Study Group of the International Committee on Taxonomy of V. The species Severe acute respiratory syndrome-related coronavirus: classifying 2019-nCoV and naming it SARS-CoV-ii. Nat Microbiol. 2020;5(4):536–44. Epub 2020/03/04. pmid:32123347; PubMed Central PMCID: PMC7095448.
   • View Article
   • PubMed/NCBI
   • Google Scholar
2. 2. Huang C, Wang Y, Li X, Ren Fifty, Zhao J, Hu Y, et al. Clinical features of patients infected with 2019 novel coronavirus in Wuhan, Mainland china. Lancet Gob Health. 2020;395(10223):497–506. pmid:31986264
   • View Commodity
   • PubMed/NCBI
   • Google Scholar
3. 3. Liu Y, Gayle AA, Wilder-Smith A, Rocklöv J. The reproductive number of COVID-19 is higher compared to SARS coronavirus. J Travel Med. 2020;27(ii). pmid:32052846
   • View Commodity
   • PubMed/NCBI
   • Google Scholar
4. four. Lai C-C, Shih T-P, Ko W-C, Tang H-J, Hsueh P-R. Severe acute respiratory syndrome coronavirus two (SARS-CoV-2) and coronavirus illness-2019 (COVID-19): The epidemic and the challenges. Int J Antimicrob Agents. 2020;55(3):105924. pmid:32081636
   • View Article
   • PubMed/NCBI
   • Google Scholar
5. v. Lauer SA, Grantz KH, Bi Q, Jones FK, Zheng Q, Meredith 60 minutes, et al. The Incubation Period of Coronavirus Disease 2019 (COVID-19) From Publicly Reported Confirmed Cases: Estimation and Application. Ann Intern Med. 2020;172(9):577–82. Epub 2020/03/x. pmid:32150748; PubMed Central PMCID: PMC7081172.
   • View Commodity
   • PubMed/NCBI
   • Google Scholar
6. half dozen. He X, Lau EHY, Wu P, Deng X, Wang J, Hao X, et al. Temporal dynamics in viral shedding and transmissibility of COVID-xix. Nat Med. 2020;26(5):672–five. pmid:32296168
   • View Article
   • PubMed/NCBI
   • Google Scholar
7. 7. Ebell MH. Predicting pneumonia in adults with respiratory illness. Am Fam Physician. 2007;76(4):560–two. Epub 2007/09/15. pmid:17853631.
   • View Article
   • PubMed/NCBI
   • Google Scholar
8. 8. Berryhill S, Morton CJ, Dean A, Berryhill A, Provencio-Dean N, Patel SI, et al. Effect of Wearables on Slumber in Good for you Individuals: A Randomized Cross-Over Trial and Validation Study. J Clin Sleep Med. 2020;xvi(five):775–783. pmid:32043961
   • View Commodity
   • PubMed/NCBI
   • Google Scholar
9. nine. Pedregosa F, Varoquaux G, Gramfort A, Michel 5, Thirion B, Grisel O, et al. Scikit-learn: Machine Learning in Python. J Mach Learn Res. 2011;12:2825–30.
   • View Article
   • Google Scholar
10. 10. World Health Organization [Cyberspace]: COVID-19—virtual printing briefing; c2020; [cited 2020 June 19]. Bachelor from: https://world wide web.who.int/docs/default-source/coronaviruse/transcripts/who-audio-emergencies-coronavirus-press-conference-full-30mar2020.pdf?sfvrsn=6b68bc4a_2.
11. 11. Rothe C, Schunk Yard, Sothmann P, Bretzel Thousand, Froeschl G, Wallrauch C, et al. Manual of 2019-nCoV Infection from an Asymptomatic Contact in Germany. N Engl J Med. 2020;382(10):970–one. pmid:32003551
    • View Article
    • PubMed/NCBI
    • Google Scholar

Source: https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0243693

Posted by: lippertagetaing.blogspot.com

Share this post