Skip to main content
PMC home page
PLOS ONE logoLink to PLOS ONE
. 2020 Dec 10;15(12):e0243122. doi: 10.1371/journal.pone.0243122

Can the detection dog alert on COVID-19 positive persons by sniffing axillary sweat samples? A proof-of-concept study

Dominique Grandjean 1,*,#, Riad Sarkis 2,#, Clothilde Lecoq-Julien 1,#, Aymeric Benard 3,#, Vinciane Roger 1,‡, Eric Levesque 4,‡, Eric Bernes-Luciani 3,‡, Bruno Maestracci 3, Pascal Morvan 5,‡, Eric Gully 6,‡, David Berceau-Falancourt 7,‡, Pierre Haufstater 8, Gregory Herin 9, Joaquin Cabrera 1,‡, Quentin Muzzin 1,‡, Capucine Gallet 1,‡, Hélène Bacqué 1, Jean-Marie Broc 10, Leo Thomas 10, Anthony Lichaa 2, Georges Moujaes 2, Michele Saliba 2, Aurore Kuhn 10, Mathilde Galey 10, Benoit Berthail 10, Lucien Lapeyre 4, Anthoni Capelli 5,‡, Steevens Renault 5,‡, Karim Bachir 5,‡, Anthony Kovinger 5,‡, Eric Comas 6, Aymeric Stainmesse 6,‡, Erwan Etienne 6,‡, Sébastien Voeltzel 6,‡, Sofiane Mansouri 6,‡, Marlène Berceau-Falancourt 11,‡, Aimé Dami 3, Lary Charlet 6, Eric Ruau 6, Mario Issa 2, Carine Grenet 12, Christophe Billy 13, Jean-Pierre Tourtier 4,#, Loïc Desquilbet 1,#
Editor: Nadine Ravel14
PMCID: PMC7728218  PMID: 33301539

Abstract

The aim of this proof-of-concept study was to evaluate if trained dogs could discriminate between sweat samples from symptomatic COVID-19 positive individuals (SARS-CoV-2 PCR positive) and those from asymptomatic COVID-19 negative individuals. The study was conducted at 2 sites (Paris, France, and Beirut, Lebanon), followed the same training and testing protocols, and involved six detection dogs (three explosive detection dogs, one search and rescue dog, and two colon cancer detection dogs). A total of 177 individuals were recruited for the study (95 symptomatic COVID-19 positive and 82 asymptomatic COVID-19 negative individuals) from five hospitals, and one underarm sweat sample per individual was collected. The dog training sessions lasted between one and three weeks. Once trained, the dog had to mark the COVID-19 positive sample randomly placed behind one of three or four olfactory cones (the other cones contained at least one COVID-19 negative sample and between zero and two mocks). During the testing session, a COVID-19 positive sample could be used up to a maximum of three times for one dog. The dog and its handler were both blinded to the COVID-positive sample location. The success rate per dog (i.e., the number of correct indications divided by the number of trials) ranged from 76% to 100%. The lower bound of the 95% confidence interval of the estimated success rate was most of the time higher than the success rate obtained by chance after removing the number of mocks from calculations. These results provide some evidence that detection dogs may be able to discriminate between sweat samples from symptomatic COVID-19 individuals and those from asymptomatic COVID-19 negative individuals. However, due to the limitations of this proof-of-concept study (including using some COVID-19 samples more than once and potential confounding biases), these results must be confirmed in validation studies.

Introduction

The “One Health—One Medicine” concept is currently more important than ever as it is bringing medical doctors, veterinary surgeons, epidemiologists, and dog handlers together to share their knowledge and experience in an attempt to combat the SARS-CoV-2 pandemic. Quick, reliable, and widespread testing is one of the key measures to control the pandemic. In June 2020, some European countries managed to control the COVID-19 epidemic by maintaining a low number of COVID-19 positive cases. In this context, the timing of testing is crucial. A high rate of testing is more effective at slowing further outbreaks if conducted early, when the disease prevalence is lower [1].

There is now some evidence that dogs may play a critical role in detecting infectious and parasite diseases as well as cancers [2]. Some studies involving detection dogs suggested that detection of non-infectious diseases by dogs was comparable with standard diagnostic methods [3], especially for cancer detection [4, 5]. In 1989, Williams and Pembroke suggested that dogs may be able to detect malignant tumours based on their specific odour [6]. The first clinical investigation of lung and breast cancers was published by McCulloch et al. in 2006 [7]. In 2010, Willis et al. published a clinical investigation on bladder cancer [8], after having published a proof-of-principle study in 2004 [9]. Subsequent studies have been performed to investigate the ability of dogs to detect colorectal cancer [10], lung [11–13], prostate [14–16], and liver [17] cancers, and melanoma [18, 19]. Regarding other non-infectious diseases, several studies have suggested that dogs can be used as “alert dogs” for diabetic [20–22] and epileptic patients [23], to improve patient quality of life.

Several studies have used dogs to differentiate a range of target insect and parasite odours. For example, Wallner and Ellis [24] trained dogs to locate gypsy moth eggs with a 95% success rate. Richards et al. assessed the dogs’ ability to differentiate between nematodes infected and uninfected sheep faeces [25]. Guest et al. showed that trained dogs can use olfaction to identify people with malaria [26]. Bacteriological diseases can also be detected by dogs [27, 28]. Dogs seem able to differentiate cellular cultures infected by the virus causing bovine mucosal disease from non-infected cultures, or from those infected by other viruses (bovine herpes 1 or bovine parainfluenza 3) [29].

As pointed out by Edwards et al. in 2017, “one of the largest threats to the validity of the results from [olfactory detection dog] studies is the possibility that systematic differences, other than disease status, between positive and control samples in both training and testing phases were responsible for the obtained results.” [4]. Therefore, the current amount of evidence suggesting that dogs can be used as a complementary tool to detect infectious and non-infectious diseases may not be as high as the number of published studies on the subject suggests. In order to accurately demonstrate that dogs can detect the studied disease, a rigorous protocol must be followed to prevent biases and over-interpretation of the results. Protocol recommendations include ensuring the dog handler is unaware of the individual’s disease status when presenting to the dog (i.e., single-blind criterion; the double-blind criterion where all other staff present during the testing procedure are also blinded strengthens the evidence but may add difficulties when rewarding the dog), ensuring the dog is presented one sample no more than once during the training and the testing sessions (the dog may memorise just the odour, instead of generalising it; this criterion includes the use of different samples between training and testing sessions), ensuring control samples are comparable to positive samples except for disease status (to avoid confounding bias), and randomising sample positions in the line-up (if line-ups are used) [4, 30, 31].

The Nosaïs project, conducted by the UMES (Unité de Médecine de l’Elevage et du Sport) at the National Veterinary School of Alfort (Maisons-Alfort, France), has been established to develop the scientific approach of medical detection dogs. In May 2020, due to the potential occurrence of a second COVID-19 wave in many countries, the Nosaïs team initiated a multicentre proof-of-concept study on the olfactive detection of COVID-19 positive individuals by dogs. This proof-of-concept study is based on the preliminary assumption that dogs could be trained to discriminate between COVID-19 positive and negative individuals due to their strong olfactory acuity. This assumption has been reinforced by the results from a recent pilot study on the olfactory identification of COVID-19 positive individuals by detection dogs [32]. Our assumption was based on the potential excretion of specific catabolites in the sweat, induced by SARS-CoV-2 cellular actions or replications in the organism’s cells, through the apocrine sweat glands, generating Volatile Organic Compounds (VOCs) that the dogs can detect.

VOCs are volatile at an ambient temperature, and may be detectable by the dogs, in breath, urine, tears, saliva, faeces and sweat. Most studies on volatile biomarkers have been conducted on breath samples. VOCs emanating from the skin contribute to an individual’s body odour, and may convey important information about metabolic processes [33, 34]. They are produced by eccrine, apocrine and sebaceous gland secretions, and are the major source of underarm odorants, playing a role in chemical signalling [35]. Several studies have been performed to better characterise VOCs produced by axillary apocrine glands [36–39]. Sweat from the palms of the hands, soles of the feet and the forehead mainly comes from eccrine glands and sebum. Most of these sweat compounds are organic acids ranging in carbon size from C2 to C20, the most abundant being saturated, monounsaturated and di-unsaturated C16 and C18, which are not volatile at body temperature [40]. More recent studies confirmed that human sweat compounds differ according to the anatomic site [41], and that the secretory capacity of eccrine sweat glands appears larger on the trunk (including underarms) compared to other parts of the body [42].

Material and methods

This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Animals edicted by french law (articles R214-87 to R214-137 of the rural code) updated by decree 2013–118 and 5 decrees edited on february 1st 2013. The protocol was approved by the Committee on the Ethics of Animal Experiments of the Ecole Nationale Veterinaire d’Alfort and by the Protection of Persons Committee of Assistance Publique Hopitaux de Paris (Protocol Covidog of global study Covidef). No invasive study was performed on the dogs participating to the study.

Institutional review board: APHP (Assistance Publique Hopitaux de Paris) Biosafety committee approval: ENVA (Ecole Nationale Veterinaire d’Alfort) Ethic committee approval: ENVA (Ecole Nationale Veterinaire d’Alfort) Scientific committee approval: APHP (Assistance Publique Hopitaux de Paris) and Hopital d’Instruction des Armees Begin Approval of the research and form of consent obtained: name "COVIDOG" by APHP (Assistance Publique Hopitaux de Paris global project "COVIDeF").

Inclusion criteria of COVID-19 negative and positive individuals

This is a two-site study (National Veterinary School of Alfort, Maisons-Alfort, France, and French-Lebanese University Saint Joseph, College of Medicine, Beirut, Lebanon) which recruited COVID-19 negative and positive individuals from four hospitals in Paris, France (Hôpital d’Instruction des Armées Bégin, Saint-Mandé; Centre Hospitalier François Quesnay, Mantes-la-Jolie; Grand Hôpital de l’Est Francilien, Jossigny; Hôpital Henri-Mondor, Créteil) and one hospital in Beirut, Lebanon (Hôpital Hôtel Dieu de France, Beirut).

Inclusion criteria for COVID-19 positive individuals included showing COVID-19 clinical symptoms and being COVID-19 positive on RT-PCR or PCR test for SARS-CoV-2. To avoid potential interferences with long-term medical treatments in the sweat VOCs, COVID-19 positive individuals were not included in the study if they had received a medical treatment for more than 36 hours prior to the PCR test.

To avoid potential confounding due to “hospital odour” [5, 43], when one COVID-19 positive individual was recruited into the study, a COVID-19 negative individual was recruited from the same hospital who showed no clinical symptoms related to COVID-19 and was COVID-19 negative on the SARS-CoV-2 PCR test. These COVID-19 negative individuals were patients hospitalised for reasons other than COVID-19 symptoms or nurses working in the same hospital.

Samples

Sweat samples (whether collected from COVID-19 positive or negative individuals) were collected from the underarm region by doctors, interns or nurses, with helpers, who were trained not to contaminate the samples with their own odours. A training video and the necessary equipment were provided to the clinical staff collecting sweat samples. The sampler wore two pairs of new gloves and full COVID-19 safety protection when collecting COVID-19 positive samples. COVID-19 negative samples did not require particular safety protection but the sampler wore a new pair of surgical gloves to prevent contaminating samples with her/his own odours. The same brands of gloves were used for COVID-19 positive and negative individuals from the same hospital. Before being sampled, each individual was told about the protocol and signed an informed consent form (Fig 1).

Fig 1. Individual informed consent form.

Fig 1

Open in a new tab

Axillary sweat samples were collected because it seems a promising substrate for canine detection [21, 44], is the key odour for search and rescue or tracking dogs [45], and the axillary region is easily accessible (Fig 2). Furthermore, this site is unlikely to be contaminated by the saliva of a COVID-19 positive patient. However, since it is not known whether sweat is a SARS-CoV-2 transmission route via skin-to-object-to-mucosa contact [46], careful manipulation of axillary sweat samples is required.

Fig 2. Underarm sampling.

Fig 2

Open in a new tab

The sampling material used were 2x2-inch sterile gauze swabs used by the hospitals or sterile gauze filters used for explosive detection (provided by DiagNose comp.), and inert polymer tubes used for explosives, drugs or criminology detection (provided by Gextent comp.) (Fig 3). The polymer tubes can adsorb both polar and apolar molecules to try to identify potential SARS-CoV-2 sweat biomarkers. Two sampling materials (i.e., gauze or polymer tubes) were used because at the beginning of this proof-of-concept study, it was not known which of the two was the most effective. The sampling material remained in contact with the skin for 20 minutes. The average amount of sweat obtained is around 75 mg for both the gauze swab and the cellulosic filter.

Fig 3. Sampling materials.

Fig 3

Open in a new tab

Samples were stored in hospital anti-UV sterile containers, disinfected by the sampler’s helper, coded (including left or right underarm), then placed into a second plastic envelope. Individual anonymous data were registered on a form for each coded sample (Fig 4). All samples were transferred from the sampling site to the testing site in separated coolers (of the same brand) for COVID-19 positive and negative samples. Coolers were cleaned and disinfected with a 10% aqueous acetone solution after each use. All samples were then stored in the same place at the two sites (but positives and negatives were not mixed) at constant temperature (+18°C for Alfort site and +6°C for Beirut site) and humidity (45%).

Fig 4. Anonymous form for each coded sample.

Fig 4

Open in a new tab

Canine resources

Dogs trained for this proof-of-concept study were explosives detection dogs, search and rescue dogs, and colon cancer detection dogs. Explosives detection dogs are trained to detect 30 to 40 different types of explosives and work on a line of samples that they have to sniff individually. For such dogs, if COVID-19 positive samples have a specific odour, they only have to memorise this additional odour (and subsequently generalise it). Search and rescue dogs are trained to perform area searches, mainly using the victim’s scent. Colon cancer detection dogs are trained on rectal gases. Drug detection dogs were not selected as candidate dogs since it cannot be ruled out that COVID-19 positive or negative individuals have used prohibited substances, the catabolites of which would be excreted into the sweat and subsequently detected by such dogs.

Safety of the dogs regarding SARS-CoV-2 infection

There have been very few reports on dogs being passive carriers of SARS-CoV-2. The most synthetic and documented case concerned two dogs living in a SARS-CoV-2 infected household, out of a total of fifteen households [47]. The dogs were quarantined and remained asymptomatic. The authors concluded that it was unclear whether infected dogs could transmit the virus back to humans. In the USA, Idexx Laboratories tested more than 3,500 dogs, cats and horses from places where community transmission of SARS-CoV-2 was occurring in humans and found no positive animals [48]. A recent study, which is not yet published, provided some evidence of the absence of SARS-CoV-2 infection in dogs in close contact with a cluster of COVID-19 patients [49]. Finally, the CDC (Centre for Disease Control and Prevention, USA) attests that there is no evidence that pet animals, and especially dogs, play any significant role in SARS-CoV-2 transmission or spread [50].

It is assumed that like SARS-CoV, SARS-CoV-2 does not survive longer than a few hours on cotton [51]. So, for safety reasons, samples were not used for training or testing sessions within 24 hours of collection. A more recent study concludes that absorbent materials like cotton are safer than unabsorbent materials for protection from SARS-CoV infection [52].

Training protocol

Fourteen dogs were trained to work on a line-up of cones, used as sample carriers (Figs 5 and 6), and to mark the cone containing the COVID-19 positive sample by sitting in front of it (Fig 7). The training method was based on positive reinforcement (the dog gets its toy for each correct marking). The whole training session was carried out according to the following 4-step procedure: (1) learning line-up work (all cones in the line-up are empty and the dog is rewarded with its toy when he sniffs each of the empty olfaction cones); (2) memorising the COVID-19 sample odour (the dog is rewarded when he marks the cone with the COVID-19 positive sample and all the remaining cones in the line-up are empty); (3) empty cones are replaced by mocks (i.e., cones containing the sample material only; the dog is rewarded when he marks the COVID-19 positive sample); (4) COVID-19 negative samples are introduced into the line-up which contains one COVID-19 positive sample, one or two mock(s), and COVID-19 negative sample(s). For each dog, memorising the specific COVID-19 positive sample odour required less than one day, with the dogs sniffing between four and ten COVID-19 positive samples. The handler judged when the dog was trained and ready for the testing session, based on the dog’s behaviour.

Fig 5. Testing equipment.

Fig 5

Open in a new tab

Fig 6. 4-olfactory cone line-up.

Fig 6

Open in a new tab

Fig 7. A dog marking a cone on a 4-cone lineup.

Fig 7

Open in a new tab

Testing protocol

Once the dog was considered trained, the testing sessions started with new COVID-19 positive and negative samples (i.e., samples which were not used for the training sessions). During the testing sessions, each “trial” was conducted in a dedicated room containing 3 or 4 cones located at the back of the room (see Fig 6). The data recorder randomly placed one COVID-19 positive sample behind one cone, and at least one COVID-19 negative sample, and between zero and two mocks behind the remaining cones in the line-up. All the samples in the line-up were from the same hospital and were made of the same material (gauze or polymer tubes). The presence or absence of mocks depended on the way the dog usually worked with its handler for explosive detection, search and rescue, or for colon cancer detection. All the trials for one dog used the same total number of cones in the line-up (four cones for five dogs, and three cones for one dog). One sample (either COVID-19 positive or negative) could be used in two or three separate trials for one dog. As recommended by Johnen et al. [31], the location of the COVID-19 positive and negative samples in the line-up was randomly assigned by a dedicated website (http://www.randomization.com/), and was blinded to the dog and its handler. During the testing session, the data recorder knew the COVID-19 positive sample location but there was no visual contact between the data recorder and either the dog or the dog handler. Once the samples were placed behind the cones, the dog entered the room with its handler who asked the dog to sniff each cone in the line-up. Since the dog had been trained to mark the position of one, and only one, COVID-19 positive sample behind a cone, the dog was left free to sniff the cones before marking one cone. Once a cone was (correctly or incorrectly) marked, the trial was stopped. The data recorder indicated to the dog handler whether the dog correctly marked the cone or not; if correct, the dog handler rewarded the dog. The trial was considered a “success” if the dog correctly marked the cone with the COVID-19 positive sample, otherwise it was considered a “failure”. After each trial, the dog and its handler left the room, and the data recorder placed new samples behind the cones in the line-up. The line-up in this protocol is “simultaneous” rather than “sequential” (i.e., each cone is sniffed by the dog and if the dog marks a cone, the handler asks the dog to resume the task for the remaining cones in the line-up).

The testing period lasted 21 days because most dogs could not work every day. During the training and the testing sessions, the standard environmental conditions for canine olfactory detection were observed (temperature of +18°C and 50% humidity). No disease, symptoms or clinical abnormalities have been reported for any of the dogs involved in the study.

Both samples and cones were manipulated by the same person (the data recorder), wearing sanitary barrier protections and a pair of new surgical gloves at each trial, to avoid any olfactory contamination or interaction. All cones were cleaned with clean water after each trial, and twice daily were cleaned and disinfected with 10% aqueous acetone solution, and then dried.

Statistical analysis

Since this is a proof-of-concept study, the initial objective was to provide evidence that a trained dog can detect one COVID-19 positive sample in one olfactory cone from a line-up of three to four cones, containing either mocks or COVID-19 negative samples. It was not possible to calculate the sensitivity, specificity or false alert rates (as recommended by Johnen et al. [31]) because the line-up was simultaneous not sequential. The study outcome focused on the success rate, which was calculated by dividing the number of successful trials by the number of trials carried out by the dog. If, for one dog, the lower bound of the 95% confidence interval (calculated by using the Clopper-Pearson’s method [53]) of the estimated success rate was higher than the success rate obtained by chance alone (thereafter called “random choice proportion”), then this result would provide some evidence that this dog could actually detect COVID-19 positive samples. The dogs were initially trained to discriminate between sweat samples and mocks before being trained to discriminate between COVID-19 positive and negative samples (step 3 of the training). Therefore, the random choice proportion was calculated by subtracting the number of mocks present in the line-up, then dividing 1 by the number of COVID-19 positive or negative sweat samples in the line-up. For example, a 4-cone line-up containing one mock would produce a random choice proportion of 33% (i.e., 4–1 = 3, 1/3 = 33%). To address the impact on the success rates of the dogs being presented the same COVID-19 positive sample several times, success rates are also shown according to whether it was the first, second, or third time the sample had been presented.

Results

A total of 95 symptomatic COVID-19 positive individuals and 82 asymptomatic COVID-19 negative individuals were recruited for the study, producing 177 sweat samples for the use in the study trials. The Paris site study sample consisted of 27 symptomatic COVID-19 positive individuals and 34 asymptomatic COVID-19 negative individuals. The percentage of females was similar between the two groups (52% for COVID-19 positive and 56%, for COVID-19 negative individuals), but the mean age was higher in the COVID-19 positive group (70 years) than in the COVID-19 negative group (42 years). The Beirut site study sample consisted of 68 symptomatic COVID-19 positive individuals and 48 asymptomatic COVID-19 negative individuals. The percentage of females (43% and 46%, respectively) and the mean age (48 years and 42 years, respectively) were similar between the COVID-19 positive and COVID-19 negative groups.

Out of the 14 dogs which started training, eight dogs did not participate in testing because they were not considered ready at the time of the beginning of the testing session (which could not have been postponed). Six dogs participated in testing and their training period ranged from one to three weeks.

The characteristics of the six dogs are presented in Table 1. Five dogs were Belgian Malinois Shepherds because this is the most commonly used breed in French working dog organisations. Three dogs were explosive detection dogs (Guess, Maika, and Gun), two were colon cancer detection dogs (Bella and Jacky), and one was a search and rescue dog (Gun).

Table 1. Characteristics of the 6 dogs participating in the study testing sessions.

Name Gender Breed Age Organisation involved Specialty
Paris Guess F Belgian Malinois 8 years Cynopro detection Dog Explosives
Maika F Belgian Malinois 3 years Fire and Emerg. Dept. 77 Search and rescue
Gun M Belgian Malinois 16 months Cynopro detection Dog Explosives
Oslo M Belgian Malinois 18 months DiagNose Explosives
Beirut Bella F Belgian Malinois 6 years Mario-K9 Colon Cancer
Jacky M Jack Russell terrier 3 years Mario-K9 Colon Cancer
Open in a new tab

For three dogs (Oslo, Bella, and Jacky), the 4-cone line-ups contained one COVID-19 positive sample and three COVID-19 negative samples without any mocks. For one dog (Gun), most of the trials (44/47) were performed on 3-cone line-ups containing one COVID-19 positive sample and two COVID-19 negative samples without any mocks. For the last two dogs (Guess and Maika), the 4-cone line-ups contained one COVID-19 positive sample, at least one COVID-19 negative sample, and either one or two mocks.

Table 2 presents the results obtained from the six dogs, according to the number of mocks in the line-ups. The success rates ranged from 76% (for Maika completing 17 trials with 4-cone line-ups containing two mocks) and 100% (for Bella and Jacky completing 4-cone line-ups containing no mocks). There was one other 100% success rate but it was calculated from just three trials containing one mock in 3-cone line-ups (for Gun). In all but one situation, the lower bound of the 95% confidence interval (CI) of the success rates was higher than or equal to the random choice proportion. The exception was the 100% success rate calculated from the three trials with one mock completed by Gun, where the lower bound of the 95% CI was 29% and the random choice proportion was 50%.

Table 2. Results obtained from the 6 dogs participating in the study testing sessions.

Number of mocks / number of cones in the line-ups* Number of trials Correct identifications Success rate [95% CI] Random choice proportion
Guess 1/4 34 26 76% [59%-89%] 33%
2/4 18 17 94% [73%-100%] 50%
Maika 1/4 38 32 84% [69%-94%] 33%
2/4 17 13 76% [50%-93%] 50%
Gun 0/3 44 39 89% [75%-96%] 33%
1/3 3 3 100% [29%-100%] 50%
Oslo 0/4 31 29 94% [79%-99%] 25%
Bella 0/4 68 68 100% [95%-100%] 25%
Jacky 0/4 68 68 100% [95%-100%] 25%
Open in a new tab

CI: Confidence interval;

* All line-ups contained one COVID-19 positive sample, and at least one COVID-19 negative sample.

Table 3 presents the success rates according to whether the COVID-19 positive samples were being presented for the first, second, or third time to the four Paris site dogs (the data were not available for the Beirut site). For Guess and Maika, the success rates were higher for trials involving COVID-19 positive samples presented to the dog for the first time (92% and 87%, respectively), compared to trials where the positive samples were presented for the second (75% and 77%, respectively) and the third (75% and 83%, respectively) times. Conversely, the success rates for Gun and Oslo were higher for trials involving COVID-19 positive samples presented for the second time (96% and 100%, respectively) compared to trials where positive samples were presented for the first time (83% and 90%). However, for the trials when Gun and Oslo were presented the COVID-19 positive samples for the first time, the lower bound of the 95% CI of the success rates was higher than the maximum value of the random choice proportion reached in our study (50%, see Table 2).

Table 3. Results obtained from the Paris site dogs according to first, second, or third presentation of the COVID-19 positive samples.

First, second, or third presentation of the COVID-19 positive sample Number of trials Correct identifications Success rate [95% CI]
Guess 1st 24 22 92% [73%-99%]
2nd 24 18 75% [53%-90%]
3rd 4 3 75% [19%-99%]
Maika 1st 23 20 87% [66%-97%]
2nd 26 20 77% [56%-91%]
3rd 6 5 83% [36%-100%]
Gun 1st 24 20 83% [63%-95%]
2nd 23 22 96% [78%-100%]
Oslo 1st 21 19 90% [70%-99%]
2nd 10 10 100% [69%-100%]
Open in a new tab

CI: Confidence interval.

Discussion

In a context where there is a lack of diagnostic tests available to perform mass detection of SARS-CoV-2 in many countries worldwide, exploring the possibility of using dog olfactory detection as a rapid, reliable and cheap “tool” to pre-screen people or perform rapid checking in certain circumstances is important. The first step for such an approach was to determinate if trained dogs can discriminate between axillary sweat samples from symptomatic COVID-19 positive patients and asymptomatic COVID-19 negative individuals.

The results of this proof-of-concept study are a promising first step providing some evidence that dogs may be able to detect COVID-19 positive samples collected from the axillary sweat of individuals showing clinical COVID-19 symptoms and who are positive on SARS-CoV-2 RT-PCR or PCR tests.

During the testing sessions, two samples collected from a priori COVID-19 negative individuals (i.e., individuals who met the inclusion criteria to be recruited as controls) were repeatedly marked by two dogs. The information was immediately sent to the concerned hospital and PCRs were performed again on these two individuals who turned out to be positive. These individuals and their corresponding samples, as well as the trials which included these two samples, were excluded from the data since, in these cases, the line-up contained more than one COVID-19 positive sample. False negative RT-PCR results are not uncommon [54–56], but these two events support our working hypothesis. Overall, our results are in accordance with a recent pilot study which investigated the ability of trained dogs to detect saliva or tracheobronchial secretion samples from symptomatic COVID-19 positive patients [32].

By filming each trial for each dog, we were able to understand why some trials were unsuccessful (i.e., a dog marked a COVID-19 negative sample or a mock). For example, a horse walking close to the testing room (on the site at the Alfort veterinary school), or people not respecting the protocol and making too much noise close to the testing room. However, there were other unsuccessful trials which could not be explained. This observation is not as negative as it first appears. Even if trained dogs are able to correctly discriminate symptomatic COVID-19 positive individuals from asymptomatic COVID-19 negative ones, they should not be considered a perfect diagnostic method, but rather a complementary tool in settings where diagnostic tests are not readily available or while diagnostic tests do not have high levels of accuracy [57].

Mocks were included in the line-ups for the study dogs who are accustomed to working with mocks for explosive detection or search and rescue. However, because the trial may have been easier for the dog when a line-up contained at least one mock compared to a line-up not containing any mocks, it was necessary to remove these mocks when calculating the random choice proportion. The lower bound of the 95% CI of the estimated success rates were all higher than or equal to the random choice proportion, except for one success rate. This success rate was 100% but was calculated from just three trials. We therefore cannot rule out that we may have lacked some statistical power. Although mocks are necessary to train the dogs, and can be taken into account in statistical analyses, future validation study testing protocols should not include mocks in their line-ups.

Our proof-of-concept study protocol used a simultaneous line-up rather than a sequential one. This therefore meant it was not possible to calculate the true positive and false negative rates, as recommended by Concha et al. in 2014 [58], or sensitivity and specificity, as recommended by Johnen et al. [31]. Success rates were instead provided in this present study because each trial used a line-up containing just one COVID-19 positive sample, the rest being either mocks or COVID-19 negative samples. Providing a success rate for detection dogs is however considered an acceptable option by Johnen et al. [31]. This study was a necessary first step before conducting subsequent studies with sequential line-ups in which sensitivities and specificities will be calculated.

One other limitation of our study is the repeated use of some samples during the testing sessions. In this situation, we cannot rule out that the dog memorised the odour of a COVID-positive sample and marked it the second time because it had been presented before (dogs can memorise at least 10 odors [59]). However, in our study, the success rates did not seem to depend on whether the dog was being presented the COVID-19 positive sample for the first, second, or third time. For two out of the four dogs at the Paris site, the success rates were higher when the trials involved COVID-19 positive samples presented for the first time. For the other two dogs, the success rates were higher for the second presentation but were still above 80% for the trials when the COVID-19 positive samples were presented for the first time. These results support the hypothesis that olfactory memory did not play a major role, if any, in the discrimination task in our study.

The comparability criterion (i.e., positive samples are comparable to negative samples, except for the disease status) is recommended for detection dog studies [4]. If this criterion is not met, confounding bias may occur [5]. The criterion was met for the “hospital” factor, since a COVID-19 positive sample was collected in the same hospital as a COVID-19 negative sample but it may be possible that COVID-19 positive patient rooms had a different odour to COVID-19 negative patient rooms. However, as Walczak et al. wrote in their paper published in 2012, “hospital rooms may have a hospital odour, derived mainly from disinfectants, which may be a common component of all samples taken from donors who are inside a hospital.” [43]. Therefore, in our data, there should not be confounding bias due the hospital odour.

Although the proportions of females were similar between COVID-19 positive individuals and COVID-19 negative individuals, both at the Paris and Beirut sites, the comparability criterion was not met for age at the Paris site (COVID-19 positive individuals were older than the negative ones). This lack of comparability is a strong limitation in this study. Body odour differs with age [60, 61] and humans seem to be able to discriminate age based on body odour alone [62]. This lack of age comparability at the Paris site may have introduced some noise that possibly altered the dog’s performance, since the French dogs had lower success rates than the Lebanese dogs. Furthermore, it is possible that the French detection dogs detected odours excreted by the elderly because of their age, and not because of the COVID-19 positive status, so age then becomes a confounding factor [30]. However, the observed results from the Lebanese dogs (a success rate of 100% for the two dogs), where mean age was similar between COVID-19 positive and negative groups, do not support this latter hypothesis.

Information about medical conditions or medication use for recruited individuals was not systematically collected, so it was impossible to check the comparability criterion was met for these characteristics. COVID-19 negative individuals in our study were hospitalised for another reason and were probably on medication which differed from the ones used by the COVID-19 positive individuals. This situation may have impaired the dogs’ performance or introduced some confounding bias. However, when some conditions (including diabetes, arterial hypertension, respiratory disorders, and hypothyroidism) were recorded for some included individuals, and for which the comparability criterion was not met, the dogs rarely marked COVID-19 negative individuals suffering from one of these conditions (data not shown). We cannot generalise this statement to other disease conditions which were not recorded in our data.

Besides medical conditions or medication use which may differ between COVID-19 positive and negative individuals, the amount of sweat collected in the samples was not individually recorded and we cannot exclude that the quantity of sweat collected from COVID-19 positive individuals was higher than that from COVID-19 negative individuals. This may have been a confounding factor. However, the sampling procedure described in our study led to an inter-individual variability in sweat quantities not related to sample status (COVID-19 positive or negative), but rather related to individual characteristics.

Two sampling materials (sterile gauze swabs or filters, and inert polymers tubes) were used because at the beginning of this proof-of-concept study, it was not known which sampling material was the most effective. The sampling material should however not play a role as a confounder because the same sampling material was used within each trial. Therefore, the dog cannot have been influenced by the sampling material odour when trying to detect the COVID-19 positive sample.

Due to the aforementioned limitations, more research in this field is needed with studies which meet all the recommended criteria [4, 30, 31]. However, despite these limitations, our study has some strengths in accordance with previous recommendation [4, 30, 31]. These strengths included the fact that samples used for the training sessions were different from those used for the testing sessions, six dogs were used in this study which is considered a large number when compared to other detection dog studies, only one sweat sample was collected per individual and a large number of individuals were recruited in our study (n = 177), the sample position was randomised, and the dog and its handler were blinded to the sample locations. Although the data recorder was not blinded to the sample location, he remained at the back of the room and could not be seen by either the dog handler or the dog during the session. This situation reduced the chance of the data recorder directly influencing the dog’s behaviour, but some unintentional influence on the choice of samples cannot be totally ruled out. Furthermore, this single blinding allowed the data recorder to tell the handler when the dog correctly marked a cone so the dog could be rewarded as soon as it corrected marked the cone. It should however be noted that the study should ideally be double-blind (i.e., both the handler and the data recorder are blinded to the sample positions), while being able to reward the dog using various methods [30].

The training period for the six study dogs (between one and three weeks) was much shorter than reported in some detection dogs studies. One reason for this is that eight out of the 14 dogs which initially participated in the training sessions were not considered ready by their handler at the time of the beginning of the testing sessions (which could not have been postponed). In operational settings, it is recommended to expect around six weeks to train a COVID-19 detection dog. Furthermore, this training time is in accordance with the study by Jendrny et al. on COVID-19 detection dogs [32]. The short training period can also be explained by the fact that the study dogs were already well-trained detection dogs. The explosives detection dogs are trained to detect between 30 and 40 different types of explosives and are accustomed to working on a line of samples that they have to sniff individually. For such dogs, if COVID-19 positive samples have a specific odour, they only have to memorise one additional odour. Search and rescue dogs are trained to perform area searches and work using the scent of the sweat. Although these dogs are trained to detect general human odour, without focusing on a person’s specific odour (e.g. odour(s) related to a disease), they can easily be retrained to work in a scent line-up and discriminate odours. We also used colon cancer detection dogs which are trained on rectal gases. These detection dogs are accustomed to working with samples on line-ups and they quickly memorise a new odour.

The next step would be to carry out a double-blind validation study involving as many COVID-19 positive and negative individuals as this proof-of-concept study, which will provide the sensitivity and specificity of the dogs’ ability to discriminate between COVID-19 positive and negative samples. Ideally, in this validation study, samples should be received from an independent source (e.g. from another organisation or even country) which have not been processed by the research team before being presented to the dogs. If the sensitivity and specificity are high enough, then this validation study will provide some evidence that national authorities could use trained COVID-19 detection dogs in settings where equipment and/or money are lacking to perform standard serology or RT-PCR tests, or as a complementary method in other settings.

Supporting information

S1 Data

(XLSX)

Click here for additional data file. (52.3KB, xlsx)

Data Availability

Data are provided as Supporting information file.

Funding Statement

DiagNose, Cynopro Detection Dogs volunteered their dogs to participate to the study without any financial request. DiagNose, Cynopro Detection Dogs provided support in the form of salaries for authors [PM, MBF, AC, SR, KB], but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. Additionally, ICTS Europe, Biodesiv SAS, and Mario K9 also provided support in the form of salaries for authors [PM, GH, MI], but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the ‘author contributions’ section.

References

  • 1.Hasell J. Testing early, testing late: four countries’ approaches to COVID-19 testing compared. 2020; https://ourworldindata.org/covid-testing-us-uk-korea-italy. [Google Scholar]
  • 2.Angle C, Waggoner LP, Ferrando A, Haney P, Passler T. Canine Detection of the Volatilome: A Review of Implications for Pathogen and Disease Detection. Front Vet Sci 2016,3:47 10.3389/fvets.2016.00047 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Bijland LR, Bomers MK, Smulders YM. Smelling the diagnosis: a review on the use of scent in diagnosing disease. Neth J Med 2013,71:300–307. [PubMed] [Google Scholar]
  • 4.Edwards TL, Brown CM, Schoon A, Cox C, Poling A. Animal olfactory detection of human diseases: Guidelines and systematic review. J Vet Behav 2017,20:59–73. [Google Scholar]
  • 5.Pirrone F, Albertini M. Olfactory detection of cancer by trained sniffer dogs: A systematic review of the literature. J Vet Behav 2017,19:105–117. [Google Scholar]
  • 6.Williams H, Pembroke A. Sniffer dogs in the melanoma clinic? Lancet 1989,1:734 10.1016/s0140-6736(89)92257-5 [DOI] [PubMed] [Google Scholar]
  • 7.McCulloch M, Jezierski T, Broffman M, Hubbard A, Turner K, Janecki T. Diagnostic accuracy of canine scent detection in early- and late-stage lung and breast cancers. Integr Cancer Ther 2006,5:30–39. 10.1177/1534735405285096 [DOI] [PubMed] [Google Scholar]
  • 8.Willis CM, Britton LE, Harris R, Wallace J, Guest CM. Volatile organic compounds as biomarkers of bladder cancer: Sensitivity and specificity using trained sniffer dogs. Cancer Biomark 2010,8:145–153. 10.3233/CBM-2011-0208 [DOI] [PubMed] [Google Scholar]
  • 9.Willis CM, Church SM, Guest CM, Cook WA, McCarthy N, Bransbury AJ, et al. Olfactory detection of human bladder cancer by dogs: proof of principle study. BMJ 2004,329:712 10.1136/bmj.329.7468.712 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Sonoda H, Kohnoe S, Yamazato T, Satoh Y, Morizono G, Shikata K, et al. Colorectal cancer screening with odour material by canine scent detection. Gut 2011,60:814–819. 10.1136/gut.2010.218305 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Boedeker E, Friedel G, Walles T. Sniffer dogs as part of a bimodal bionic research approach to develop a lung cancer screening. Interact Cardiovasc Thorac Surg 2012,14:511–515. 10.1093/icvts/ivr070 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Buszewski B, Ligor T, Jezierski T, Wenda-Piesik A, Walczak M, Rudnicka J. Identification of volatile lung cancer markers by gas chromatography-mass spectrometry: comparison with discrimination by canines. Anal Bioanal Chem 2012,404:141–146. 10.1007/s00216-012-6102-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Ehmann R, Boedeker E, Friedrich U, Sagert J, Dippon J, Friedel G, et al. Canine scent detection in the diagnosis of lung cancer: revisiting a puzzling phenomenon. Eur Respir J 2012,39:669–676. 10.1183/09031936.00051711 [DOI] [PubMed] [Google Scholar]
  • 14.Bjartell AS. Dogs sniffing urine: a future diagnostic tool or a way to identify new prostate cancer markers? Eur Urol 2011,59:202–203. 10.1016/j.eururo.2010.10.033 [DOI] [PubMed] [Google Scholar]
  • 15.Cornu JN, Cancel-Tassin G, Ondet V, Girardet C, Cussenot O. Olfactory detection of prostate cancer by dogs sniffing urine: a step forward in early diagnosis. Eur Urol 2011,59:197–201. 10.1016/j.eururo.2010.10.006 [DOI] [PubMed] [Google Scholar]
  • 16.Taverna G, Tidu L, Grizzi F, Torri V, Mandressi A, Sardella P, et al. Olfactory system of highly trained dogs detects prostate cancer in urine samples. J Urol 2015,193:1382–1387. 10.1016/j.juro.2014.09.099 [DOI] [PubMed] [Google Scholar]
  • 17.Kitiyakara T, Redmond S, Unwanatham N, Rattanasiri S, Thakkinstian A, Tangtawee P, et al. The detection of hepatocellular carcinoma (HCC) from patients’ breath using canine scent detection: a proof-of-concept study. J Breath Res 2017,11:046002 10.1088/1752-7163/aa7b8e [DOI] [PubMed] [Google Scholar]
  • 18.Pickel D, Manucy GP, Walker DB, Hall SB, Walker JC. Evidence for canine olfactory detection of melanoma. Appl Anim Behav Sci 2004,89:107–116. [Google Scholar]
  • 19.Campbell LF, Farmery L, George SM, Farrant PB. Canine olfactory detection of malignant melanoma. BMJ Case Rep 2013,2013 10.1136/bcr-2013-008566 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Rooney NJ, Guest CM, Swanson LCM, Morant SV. How effective are trained dogs at alerting their owners to changes in blood glycaemic levels?: Variations in performance of glycaemia alert dogs. PLoS One 2019,14:e0210092 10.1371/journal.pone.0210092 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Rooney NJ, Morant S, Guest C. Investigation into the value of trained glycaemia alert dogs to clients with type I diabetes. PLoS One 2013,8:e69921 10.1371/journal.pone.0069921 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Wilson C, Morant S, Kane S, Pesterfield C, Guest C, Rooney NJ. An Owner-Independent Investigation of Diabetes Alert Dog Performance. Front Vet Sci 2019,6:91 10.3389/fvets.2019.00091 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Kirton A, Winter A, Wirrell E, Snead OC. Seizure response dogs: evaluation of a formal training program. Epilepsy Behav 2008,13:499–504. 10.1016/j.yebeh.2008.05.011 [DOI] [PubMed] [Google Scholar]
  • 24.Wallner WE, Ellis TL. Olfactory detection of gypsy moth pheromone and egg masses by domestic canines. Environ Entomol 1976,5:183–186. [Google Scholar]
  • 25.Richards KM, Cotton SJ, Sandeman RM. The use of detector dogs in the diagnosis of nematode infections in sheep feces. J Vet Behav 2008,3:25–31. [Google Scholar]
  • 26.Guest C, Pinder M, Doggett M, Squires C, Affara M, Kandeh B, et al. Trained dogs identify people with malaria parasites by their odour. Lancet Infect Dis 2019,19:578–580. 10.1016/S1473-3099(19)30220-8 [DOI] [PubMed] [Google Scholar]
  • 27.Bomers MK, van Agtmael MA, Luik H, Vandenbroucke-Grauls CM, Smulders YM. A detection dog to identify patients with Clostridium difficile infection during a hospital outbreak. J Infect 2014,69:456–461. 10.1016/j.jinf.2014.05.017 [DOI] [PubMed] [Google Scholar]
  • 28.Maurer M, McCulloch M, Willey AM, Hirsch W, Dewey D. Detection of Bacteriuria by Canine Olfaction. Open Forum Infect Dis 2016,3:ofw051. 10.1093/ofid/ofw051 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Angle TC, Passler T, Waggoner PL, Fischer TD, Rogers B, Galik PK, et al. Real-Time Detection of a Virus Using Detection Dogs. Front Vet Sci 2016,2:1–6. 10.3389/fvets.2015.00079 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Elliker KR, Sommerville BA, Broom DM, Neal DE, Armstrong S, Williams HC. Key considerations for the experimental training and evaluation of cancer odour detection dogs: lessons learnt from a double-blind, controlled trial of prostate cancer detection. BMC Urol 2014,14:22 10.1186/1471-2490-14-22 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Johnen D, Heuwieser W, Fischer-Tenhagen C. An approach to identify bias in scent detection dog testing. Appl Anim Behav Sci 2017,189:1–12. [Google Scholar]
  • 32.Jendrny P, Schulz C, Twele F, Meller S, von Kockritz-Blickwede M, Osterhaus A, et al. Scent dog identification of samples from COVID-19 patients—a pilot study. BMC Infect Dis 2020,20:536 10.1186/s12879-020-05281-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Amann A, Costello Bde L, Miekisch W, Schubert J, Buszewski B, Pleil J, et al. The human volatilome: volatile organic compounds (VOCs) in exhaled breath, skin emanations, urine, feces and saliva. J Breath Res 2014,8:034001 10.1088/1752-7155/8/3/034001 [DOI] [PubMed] [Google Scholar]
  • 34.Grabowska-Polanowska B, Miarka P, Skowron M, Sulowicz J, Wojtyna K, Moskal K, et al. Development of sampling method and chromatographic analysis of volatile organic compounds emitted from human skin. Bioanalysis 2017,9:1465–1475. 10.4155/bio-2017-0128 [DOI] [PubMed] [Google Scholar]
  • 35.Wysocki CJ, Preti G. Facts, fallacies, fears, and frustrations with human pheromones. Anat Rec A Discov Mol Cell Evol Biol 2004,281:1201–1211. 10.1002/ar.a.20125 [DOI] [PubMed] [Google Scholar]
  • 36.Troccaz M, Starkenmann C, Niclass Y, van de Waal M, Clark AJ. 3-Methyl-3-sulfanylhexan-1-ol as a major descriptor for the human axilla-sweat odour profile. Chem Biodivers 2004,1:1022–1035. 10.1002/cbdv.200490077 [DOI] [PubMed] [Google Scholar]
  • 37.Hasegawa Y, Yabuki M, Matsukane M. Identification of new odoriferous compounds in human axillary sweat. Chem Biodivers 2004,1:2042–2050. 10.1002/cbdv.200490157 [DOI] [PubMed] [Google Scholar]
  • 38.Zeng XN, Leyden JJ, Lawley HJ, Sawano K, Nohara I, Preti G. Analysis of characteristic odors from human male axillae. J Chem Ecol 1991,17:1469–1492. 10.1007/BF00983777 [DOI] [PubMed] [Google Scholar]
  • 39.Zeng XN, Leyden JJ, Spielman AI, Preti G. Analysis of characteristic human female axillary odors: Qualitative comparison to males. J Chem Ecol 1996,22:237–257. 10.1007/BF02055096 [DOI] [PubMed] [Google Scholar]
  • 40.Bernier UR, Kline DL, Barnard DR, Schreck CE, Yost RA. Analysis of human skin emanations by gas chromatography/mass spectrometry. 2. Identification of volatile compounds that are candidate attractants for the yellow fever mosquito (Aedes aegypti). Anal Chem 2000,72:747–756. 10.1021/ac990963k [DOI] [PubMed] [Google Scholar]
  • 41.Taylor NA, Machado-Moreira CA. Regional variations in transepidermal water loss, eccrine sweat gland density, sweat secretion rates and electrolyte composition in resting and exercising humans. Extrem Physiol Med 2013,2:4 10.1186/2046-7648-2-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Murota H, Matsui S, Ono E, Kijima A, Kikuta J, Ishii M, et al. Sweat, the driving force behind normal skin: an emerging perspective on functional biology and regulatory mechanisms. J Dermatol Sci 2015,77:3–10. 10.1016/j.jdermsci.2014.08.011 [DOI] [PubMed] [Google Scholar]
  • 43.Walczak M, Jezierski T, Górecka-Bruzda A, Sobczyńska M, Ensminger J. Impact of individual training parameters and manner of taking breath odor samples on the reliability of canines as cancer screeners. J Vet Behav 2012,7:283–294. [Google Scholar]
  • 44.Lippi G, Cervellin G. Canine olfactory detection of cancer versus laboratory testing: myth or opportunity? Clin Chem Lab Med 2012,50:435–439. 10.1515/CCLM.2011.672 [DOI] [PubMed] [Google Scholar]
  • 45.Woidtke L, Dressler J, Babian C. Individual human scent as a forensic identifier using mantrailing. Forensic Sci Int 2018,282:111–121. 10.1016/j.forsciint.2017.11.021 [DOI] [PubMed] [Google Scholar]
  • 46.Propper RE. Is sweat a possible route of transmission of SARS-CoV-2? Exp Biol Med (Maywood) 2020,245:997–998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Sit THC, Brackman CJ, Ip SM, Tam KWS, Law PYT, To EMW, et al. Infection of dogs with SARS-CoV-2. Nature 2020. 10.1038/s41586-020-2334-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.IDEXX. SARS-CoV-2 (COVID-19) RealPCR Test. https://www.idexx.com/en/veterinary/reference-laboratories/overview-idexx-sars-cov-2-covid-19-realpcr-test/.
  • 49.Temmam S, Barbarino A, Maso D, Behillil S, Enouf V, Huon C, et al. Absence of SARS-CoV-2 infection in cats and dogs in close contact with a cluster of COVID-19 patients in a veterinary campus. BioRxiv 2020:2020.2004.2007.029090. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Centers for Disease Control and Prevention. COVID-19 and Animals. 2020 [updated June 22, 2020]; https://www.cdc.gov/coronavirus/2019-ncov/daily-life-coping/animals.html.
  • 51.Lai MY, Cheng PK, Lim WW. Survival of severe acute respiratory syndrome coronavirus. Clin Infect Dis 2005,41:e67–71. 10.1086/433186 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Ren SY, Wang WB, Hao YG, Zhang HR, Wang ZC, Chen YL, et al. Stability and infectivity of coronaviruses in inanimate environments. World J Clin Cases 2020,8:1391–1399. 10.12998/wjcc.v8.i8.1391 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Brown LD, Cai TT, DasGupta A. Interval Estimation for a Binomial Proportion. Stat Sci 2001,16:101–117. [Google Scholar]
  • 54.Xiao AT, Tong YX, Zhang S. False negative of RT-PCR and prolonged nucleic acid conversion in COVID-19: Rather than recurrence. J Med Virol 2020. 10.1002/jmv.25855 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Li Y, Yao L, Li J, Chen L, Song Y, Cai Z, et al. Stability issues of RT-PCR testing of SARS-CoV-2 for hospitalized patients clinically diagnosed with COVID-19. J Med Virol 2020,92:903–908. 10.1002/jmv.25786 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Tahamtan A, Ardebili A. Real-time RT-PCR in COVID-19 detection: issues affecting the results. Expert Rev Mol Diagn 2020,20:453–454. 10.1080/14737159.2020.1757437 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Venter M, Richter K. Towards effective diagnostic assays for COVID-19: a review. J Clin Pathol 2020,73:370–377. 10.1136/jclinpath-2020-206685 [DOI] [PubMed] [Google Scholar]
  • 58.Concha A, Mills DS, Feugier A, Zulch H, Guest C, Harris R, et al. Using sniffing behavior to differentiate true negative from false negative responses in trained scent-detection dogs. Chem Senses 2014,39:749–754. 10.1093/chemse/bju045 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Williams M, Johnston JM. Training and maintaining the performance of dogs (Canis familiaris) on an increasing number of odor discriminations in a controlled setting. Appl Anim Behav Sci 2002,78:55–65. [Google Scholar]
  • 60.Haze S, Gozu Y, Nakamura S, Kohno Y, Sawano K, Ohta H, et al. 2-Nonenal newly found in human body odor tends to increase with aging. J Invest Dermatol 2001,116:520–524. 10.1046/j.0022-202x.2001.01287.x [DOI] [PubMed] [Google Scholar]
  • 61.Chen D, Haviland-Jones J. Human olfactory communication of emotion. Percept Mot Skills 2000,91:771–781. 10.2466/pms.2000.91.3.771 [DOI] [PubMed] [Google Scholar]
  • 62.Mitro S, Gordon AR, Olsson MJ, Lundstrom JN. The smell of age: perception and discrimination of body odors of different ages. PLoS One 2012,7:e38110 10.1371/journal.pone.0038110 [DOI] [PMC free article] [PubMed] [Google Scholar]

Decision Letter 0

Nadine Ravel

13 Jul 2020

PONE-D-20-17256

Detection dogs as a help in the detection of COVID-19

Can the dog alert on COVID-19 positive persons by sniffing axillary sweat samples ?

Proof-of concept study

PLOS ONE

Dear Dr. grandjean,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

As you will read the 3 treviewers raised relatively similar methodological options and my opinion is converging in this direction.

To summarize as suggested by one of the reviewer you have two options:

- Additional data is added, providing the results of a double-blind test using samples from new people that have not previously been screened by the dogs, or

- The current abstract, discussion and conclusions are revised to provide a more critical appraisal of the results, covering possible limitations of the study and whether there are any other possible explanations for the dogs' detection performance and possible sources of unintentional bias.

It seems also clear that your manuscript would benefit from being corrected by a native english speaker.

Please submit your revised manuscript by Aug 27 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

  • A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

  • A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

  • An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Nadine Ravel

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2.  Thank you for stating the following in the Financial Disclosure section:

"The authors received no specific funding for this work"

We note that one or more of the authors are employed by a commercial company: "DiagNose, Cynopro Detection Dogs and Biodesive SAS"

a) Please provide an amended Funding Statement declaring this commercial affiliation, as well as a statement regarding the Role of Funders in your study. If the funding organization did not play a role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript and only provided financial support in the form of authors' salaries and/or research materials, please review your statements relating to the author contributions, and ensure you have specifically and accurately indicated the role(s) that these authors had in your study. You can update author roles in the Author Contributions section of the online submission form.

Please also include the following statement within your amended Funding Statement.

“The funder provided support in the form of salaries for authors [insert relevant initials], but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the ‘author contributions’ section.”

If your commercial affiliation did play a role in your study, please state and explain this role within your updated Funding Statement.

b) Please also provide an updated Competing Interests Statement declaring this commercial affiliation along with any other relevant declarations relating to employment, consultancy, patents, products in development, or marketed products, etc. 

Within your Competing Interests Statement, please confirm that this commercial affiliation does not alter your adherence to all PLOS ONE policies on sharing data and materials by including the following statement: "This does not alter our adherence to  PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests) . If this adherence statement is not accurate and  there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

Please include both an updated Funding Statement and Competing Interests Statement in your cover letter. We will change the online submission form on your behalf.

Please know it is PLOS ONE policy for corresponding authors to declare, on behalf of all authors, all potential competing interests for the purposes of transparency. PLOS defines a competing interest as anything that interferes with, or could reasonably be perceived as interfering with, the full and objective presentation, peer review, editorial decision-making, or publication of research or non-research articles submitted to one of the journals. Competing interests can be financial or non-financial, professional, or personal. Competing interests can arise in relationship to an organization or another person. Please follow this link to our website for more details on competing interests: http://journals.plos.org/plosone/s/competing-interests

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Partly

Reviewer #3: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: No

Reviewer #3: I Don't Know

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This article investigates whether dogs can differentiate the axillary odor of individuals infected by SARS-CoV-2 from individuals that are not infected. The study is very well justified and potential applications for the management of the current pandemic (especially in case of a feared second wave) are highly relevant.

I however have some major issues with this study, as detailed below.

1) Design. With the current information provided in the method, the reader cannot be convinced that the authors controlled for a number of confounding factors. Namely, it should be made clear that the negative and positive samples cannot be different from each other based on other factors than infection to SARS-CoV-2. It is known that body odor changes according to medication intake, gender, age, health status, hormonal status etc. With this in mind, a number of questions can be raised here. Who are the COVID-negative patients: do they have other pathologies? If yes, which ones? Do they all have the same pathology, or different ones? Are they on medication? Are the COVID+ on medication and if yes, is it the same one for everybody? How are the negative samples chosen with regards to gender and age, according to the target (COVID+ sample) characteristics? Two odor sampling methods are mentioned (gauze swabs or filters and polymer tubes): Are they both used for negative and positive samples, and is the same sampling support used in the whole line during a trial? In sum, more information is expected in the method section on these aspects; the limits related to the possible confounding factors must be discussed in more details than it currently is; and the conclusion will probably need to be more cautious (as a consequence).

2) Statistical analyses. The authors mention accuracy of the dog’s responses in introduction and conclusion, in terms of sensitivity and specificity. It should be clearly defined, especially with regards of the tests performed in the present article. The signal detection theory is evoked only in discussion, and false positives are discussed in that same section. It is therefore expected to see the current data analyzed using these criteria, which is not the case.

3) Paragraph about human pheromones in the discussion (lines 359-382). I suggest to simply remove this paragraph, which is 1) erroneous (no pheromones have ever been identified in humans, and the androgenous compounds you are referring to are male – not female – compounds supposed to attract females), 2) very speculative (how did you estimate the fertile phase of the donor menstrual cycle?), 3) disproportionally long compared to other points to discuss that are more crucial and more directly related to the main question.

Minor points.

1) Please describe what the mocks are, and how many are there in the line.

2) Please better explain how the training was conducted with regards to sample type: are there negatives at this stage? Mocks?

3) Please explain body odor sample management. From what I understand, they are used fresh (minimum 24 hrs between odor collection and training). Are they thrown away after 3 trials? Are they frozen at any time in the process? Refrozen?

Finally, on a more formal aspect, the article should be proofread for English. Also, improvement of the writing would be welcome. Some sentences are difficult to understand (i.e., lines 237-240 / 395-398) and some others do not integrate well in the chain of arguments (i.e. lines 62-64 / 343-345). Most importantly, in its current form the article is very fractionated (“bullet points” feeling in the whole document): It would very much gain in fluidity and articulation of the ideas if the authors work on this aspect.

Reviewer #2: Review of the manuscript PONE-D-20-17256

The idea of using trained dogs for COVID-19 screening was boosted by mass media in many countries e.g. by the BBC in UK, and by announcing by many researchers and dog trainers, that they intend to train dogs to detect COVID-19. It is obvious that first published experimental works in this field, will have a great potential for being broadly cited or comercialized. However, it could be presumed that such experiments and manuscripts are probably accomplished under a strong pressure of time, to be published as the first. Therefore the papers like the present one, which, to my knowledge is the first paper on this topic, should be reviewed and evaluated particularly causciously and critically. It is important not to disseminate facts that are not based on sound scientific experiments. This is particularly important since the reviewed paper is already available online at https://www.biorxiv.org/content/10.1101/2020.06.03.132134v1.full.pdf as not certified by peer review.

For justification of their study the authors cited publications on cancer detecting dogs and dogs alerting hypoglycaemia and seisure in human patients. However, it should be mentioned that despite of about 20 publications within the last 17 years, the medical detection dogs have not been applied for practical clinical screening so far. This was evident on the latest international conference on cancer detection by dogs, organized by the Curie-Institute in Paris in October 2019. Also, despite of numerous chemical studies within more than 25 years, using gas chromatography / mass spectrometry, which found some differences in volatile organic compounds (VOCs) between cancer positive and cancer negative patients, no single VOC or a simple combination of VOCs were identified, that could be reliable cancer marker(s), mature for practical cancer screening. The reason for problematic application of methods based on identivication of VOCs for disease screening, either by chemical analyses or by trained animals, is a great variability of emited VOCs due to day-to-day differences in diet, medication and numerous other confounding factors.

The aim of the reviewed study was to assess if dogs could be trained to discriminate body odor of patients infected with SARS-Cov-2, from body odor of healthy controls.

Such aim is justified by the current and probably future situation with COVID-19, untill an effective and commonly accessible COVID-19 vaccine is available.

The experiments were conducted at 3 different locations, with a total of 18 dogs which were previously trained either for explosives detection or for search and rescue or for cancer detection.

The Authors collected altogether 101 sweat samples from patients showing clinical symptoms of COVID-19, being also COVID-19 positive on RT-PCR or PCR test for SARS-CoV-2. Negative samples (n=97) were collected the same way, by the same trained staffs, from patients who have no clinical symptoms related to COVID-19, and were negative on SARS-CoV-2 PCR test.

The main flaw of collecting sweat samples for this study was, that the samples were not controlled quantitatively for the amount of the sweat. It is known that one of the typical COVID symptoms is a high fever. Humans with fever sweat more intensively than healthy people, so the amount of the sweat on the sample may be a clear cue for dogs.

As an example, our early studies in the eighties, on dogs detecting so called „silent estrus” in cows, could be mentioned. We used cotton swabs for collection vaginal mucus from cows in estrus vs diestrus. Our dogs perfectly indicated estrus samples. Only later we noticed that cows in estrus usually secreted more mucus and this was a cue for the dogs. When the dogs became more „experienced” they perfectly indicated cotton swabs moistured even with water. This made me causcious and critical about dog’s reliability. For dogs, odors like cow’s in estrus, human cancer, drugs or explosives, or SARS-CoV-2, play no biological role and such odors are indicated by dogs only because an operant conditioned response has been trained by producing association between odors and a reward in form a treat or favorite object to be retrieved and play with. Therefore, the trained dogs, especially when they are more „experienced”, will seek for a cue or a simple solution, to earn a reward. Such strategy of dogs may produce frequent false alerts. At identification of human scents by dogs it should be remembered that samples of human scent contain both individual component and other components as those related to diet, disease(s), medication, hygiene, cosmetics, background odors of the location as well as own odors of the sampling materials (odor carriers).

In lines 346-349 the authors report on two samples that were negative according to inclusion criteria for negative samples, but were indicated by two dogs as positive and the PCR re-test showed that in fact they were positive. This, however, is not a convincing evidence that the dogs sniffed out specific molecules induced by the SARS-CoV-2 virus presence. These two persons were not controlled for the amount of the sweat or body temperature measured precisely at the time of sample collection. Without knowing if the odor samples differ quantitatively in the amount of the sweat, it is not clear whether the dogs indicated a hypothetic, characteristic odor emited by COVID positive patients, or rather indicated samples from people with fever, who sweat more.

The Authors mentioned that the odor samples were collected „ at the same location”. This is methodologically sound but it should be precised what does it means „the same location” in this context ? Was it the same room or the same buiding or the same district ? The authors must be more precise here, because it is known that the dogs might be conditioned on the characteristic „hospital odors”, as shown by Walczak et al.(2012) in cancer detecting dogs, when the cancer positive samples were collected mostly in hospitals.

Another problem with COVID-positive sweat samples is that positive patients are staying either in infectious hospitals or in isolation wards that are probably intensively desinfected.

In addition, the access of healthy COVID-negative healthy donors (controls) to the same room where sick patients are staying, is hazardeous and strictly limited. This may be a serious constraint for collecting samples for further studies by other researchers.

The description of the training methods only in 3 lines is very scarce. Were the dogs rewarded in every training and testing trial and who and how gave a signal that the dog’s indication was correct ?

Lines 219-220 - It is true that search and rescue dogs (SRD), trained to perform disaster and area search, mainly work through the scent of the sweat. However, it must be taken into consideration that SRD are trained for detection of a generalized human odor, without discriminating individual persons and persons’ specific odor e.g. odor(s) related to a disease. Such dogs are relatively easy to be re-trained to work in a scent lineup to discriminate odours of particular person, but a re-training, to be reliable, takes longer than 1-4 hours and 4-10 trials.

In lines 262-264 the Authors stated: „…For each involved dog, the acquisition of the specific odour of COVIS-19+ sweat samples required from one to four hours, with an amount of positive samples sniffing ranging from four to ten….”

I am skeptical about the ability of each dog to learn precisely the task within such a short time. For example for cancer detecting dogs, the training period according to 13 papers published in the period 2003-2014, ranged from several weeks (Pickel et al. 2004) to 16 months (Cornu et al.2010), and on average took 8 months.

Even if the Authors could observe that their dogs correctly indicated the target samples after one hour, performing only four trials from being confronted for the first time with COVID sweat samples, it seems inredible that dogs could acquire the ability to detect COVID odor so quickly. It can not be excluded that the dogs indicated samples in the lineup, that are merely different from the others e,g, by containing more sweat.

In lines 265-267 or in lines 294-296 the Authors did not precise what was a criterion (percentage of correct indications, false alerts, misses, regularity of sniffing the samples in the lineup ?) for considering the dog to be fully trained and ready for testing.

From the lines 294-296 it could be concluded that separate samples were used for the training and testing, which is methodically correct. However, there is no information how many positive and negative or mock samples, out of the totals given in Table 1, were used for the trainig and real testing.

The authors did not conduct so called „zero trial” with no target samples in the lineup. This is another flaw which could have a consequence by making the overall conclusion in lines 407-408 premature. Although negative and mock samples were used in the lineup (lines 269-270), but there is no explanation what kinds of mocks were used. Were they samples with more sweat ? or samples from donors with fever but not related to COVID-19 (for example from people with seasonal influenza) ? „Zero trials” usually provoke more false alerts, especially on samples that are simple different from the others.

Minor remarks:

The manuscript shoud be better structured. For example the major portion of the Conclusion is appropriate rather for the Discussion section, whereas the real conclusion is given as the last sentence or the Discussion (in lines 407-408)

Line 64 More original papers should be cited instead of a handbook.

Lines 76-78 „…Scent detection by animals has been published in a large number of diagnostic studies, which all suggest similar or even superior accuracy compared with standard diagnostic methods…” – nevertheless, the scent detection by animals is not used for medical diagnoses or screenings on a large scale (probably except for tuberculosis detecting Gambian pouched rats in Africa). The medical diagnosis of a disease should not be confused with screening. Scent detection by animals will never be a diagnosis.

Line 83 „…The first clinical investigation of cancer was published by Willis on bladder cancer in 2010, after having published a proof of principle study in 2004…” – this information is only partly true: McCulloch et al. published their study on canine detection of lung and breast cancer in 2006.

Lines 88-89 „….The dog’s nose is also now currently utilised in order to prevent the consequences of crisis (or even prevent them) for diabetic [30, 31, 32, 33] and epileptic people [70]….” - The accuracy and reliability of hypoglycaemia and seizure alerting dogs, when considering on a daily basis, is still questionable. The alerting dogs rather contribute to feeling more safe and thus to improving the quality of life of the patients, than to precise alerting the hypoglycaemia or seizure episodes.

Lines 115-117 „….In a second time……” – this assumption is not quite clearly formulated in English.

Lines 154-158 Why two types of sampling supports are used ?:

(1) - Sterile gauze swabs utilised by the hospitals or sterile gauze filters utilised for explosive detection, and (2)- Inert polymers tubes utilised for explosives, drugs or criminology detection)

Line 289 Why acetone was used for disinfection of sample carriers ? Acetone is an irritant, causing mild skin irritation and moderate to severe eye irritation which may be repelling for dogs.

Lines 292-297 „…A failure was defined if the dog alerts on any of the other boxes. Trials were registered from the moment dogs recognized positive sample among other samples (mock or negative ones) without having sniffed any of them previously to the test. If the dog marks first a negative sample, the trial is considered as failed, even if he marks the positive afterwards….”

- the authors did not precise whether the dogs have to sniff ALL samples prior to the indicatication. Have the dogs to sniff the remaining samples after the correct indication? If so, would the trial be declared as a failure, if the dog made a false alert AFTER prior correct indication in the same trial? . This makes that the calculation of correct indication by chance should take into account how many samples a dog actually has sniffed. Therefore, when the dog indicates correctly for example on the first sample in the lineup and does not sniff the remaining samples, the probability of correct indication by chance at 7 samples in a lineup is not 14% but 50%. This would mean that the probability of correct indication by chance is varying from trial to trial ?

Line 350 „….dogs have no reason to lie….” – this is too colloquially expressed. With regards to odor detecting dogs one should not speak about "lying" or "telling truth".

From the dog's perspective, detecting trained odors as for example the hypothetical COVID-19 odor, that have no biological relevance, is a rather a game to earn a reward. Dogs, even if tey are well trained, when they are getting more routined, will try to get a reward at a simple way, which means that they will try to find any additional cues to indicate the target odor in the lineup. Therefore the dogs' performance in the scent lineup has to be permamently controlled throughout the whole deployment period, because it may vary. This seems to be one of the reason why so called medical detection dogs have not been used in medical screening on a regular basis.

Table 3. The terms detection sensitivity and specificity, which derive from the point 1-4 in lines 385-388, should be given, or percentage of false alerts should be given instead of the column with the number of correct indications, which is redundant in this table

The English of the manuscript is for me as a non-native speaker understandable but suffers from many language and wording imperfections. It should be edited by a native English speaker, who also has some knowledge in the specific study field.

Line 48 „..State of play..” ? – rather „State-of-the-art.”

Line 51 What does it means „..for those of interest in dog’s olfaction…” in this context?

Line 148. „…In order to avoid any dog creancement (?) on the olfactive “background noise

In several places of the manuscript there are statements that are difficult to follow

In lines 115-117 the authors write: „…In a second time, we attempt to validate the method with the dogs checking samples one by one instead of comparatively, in order to fit with what could be an operational utilisation of such trained dogs...” – what does it means exactly

Line 372 „ In Sheakspeare’s day…” ? - rather is „ in Sheakspeare’s era” or „in Sheakspeare’s time”

Lines 372-374 the passage with the anecdote on peeled apple under woman’s arm should be deleted because it is too colloquial and not closely related to the reviewed study.

Line 404 „ …cynotechnicians…” ? or dog handlers ?

Line 455 „…by small…” ? – should be „ by smell”

In conclusion, this first report on dogs trained to detect COVID-19 is interesting but a flaw in the methodology of collecting scent samples (no control for the amount of the sweat) makes that the conclusion that „…there is a high evidence that dogs can detect a person infected by the virus responsible for COVID-19 disease …” seems to be premature.

In addition, some language imperfections make me to evaluate the manuscript as not acceptable for publication in PloS ONE in the present form and needing a major revision before resubmission.

First of all, in the major revision the issue of the amount of sweat in odor samples should be addressed, including replication of the trials.

Reviewer #3: General comments

Thank you for conducting and reporting this very interesting study. In my opinion, this study is a good first step towards testing whether dogs can detect odours associated with SARS-CoV-2 infection. You have used a reasonable sample size of dogs and SARS-CoV-2 samples and the methods are generally appropriate. A main difficulty with any study of this type, where it is not known whether the odour the dogs are being trained to detect exists (in this case SARS-CoV-2 related), is demonstrating beyond reasonable doubt that dogs are detecting the odour rather than incidental contamination, and demonstrating robustly that the research team is not unintentionally influencing/cuing the dogs. Sources of odour contamination may include, for example, medication used in SARS-CoV-2 patients, hospital odours, differences in the odour of the storage containers of positive and negative samples and different odours arising from how positive and negative samples are handled. You have taken some steps to try to control for these issues. However, in my opinion, it is not possible to conclusively say that the sweat of SARS-CoV-2 contaminated persons can be detected by trained dogs based on the methods and results you present so far. In places, the methodology and controls you used just need further clarification in order to rule out possible sources of bias and alternative explanations that the dogs may be detecting contaminating odours associated with the SARS-CoV-2 samples.

Most importantly, to support the conclusion that the sweat of SARS-CoV-2 contaminated persons can be detected by trained dogs, a more rigourous, double-blind test is required, in which the research team is provided with new samples from an independent source which they are asked to assess using the dogs. The dog handlers, sample handlers and data recorders (everyone involved in assessing the samples) should be blind to the sample identities (positive or negative) until the dogs have conducted the test. Methods for how to do this have been presented in a number of previous medical detection dog studies.

I would suggest two options for revising the paper:

- Additional data is added, providing the results of a double-blind test using samples from new people that have not previously been screened by the dogs, or

- The current abstract, discussion and conclusions are revised to provide a more critical appraisal of the results, covering possible limitations of the study and whether there are any other possible explanations for the dogs' detection performance and possible sources of unintentional bias.

Potential sources of bias/limitations in medical detection dogs studies have been summarised previously (e.g. Elliker et al 2014. Key considerations for the experimental training and evaluation of cancer odour detection dogs; Edwards et al 2017 Animal olfactory detection of human diseases: Guidelines and systematic review; Johnen D, 2017. An approach to identify bias in scent detection dog testing). I recommend you review these guidelines and report where your study meets these criteria, and/or discuss limitations where your study does not meet these criteria.

The entire manuscript, including references, needs proof reading and correction for typographical errors, spelling, grammar, and consistency of tense and language.

Detailed comments

Abstract

- For the acquisition phase it would be useful to clarify how many “runs” dogs required during the one to four hours of training. Was it just a single presentation of each sample or were they presented multiple times during?

- Please clarify the wording around how many dogs were used for data collection in the study (it currently confusing for the reader whether 8 or 18 dogs were used for data collection).

- As per my general comments, I feel the conclusion that the sweat of SARS-CoV-2 contaminated persons can be detected by trained dogs is too strong at present, as the study does not appear to provide a robust double blind assessment and possible sources of bias/limitations have not been discussed. The study appears to be a good first step, but there are other possible explanations for the results based on the data presented. I suggest the conclusion is caveated that this study is a promising first step, and that a double blind trials is needed to more robustly confirm the results (unless this data is available and can be presented).

Medical detection dogs: state of play

- To avoid bias towards only discussing positive studies and to provide a more balanced discussion, I suggest mentioning some of the difficulties of conducting these studies where it is unknown whether the target odour exists, and that not all studies to date have shown disease odours to be reliable detectable by dogs. Issues are summarised in papers such as Elliker et al 2014, Edwards et al 2017, Johnen D, 2017.

Material and Methods

Samples

- Please clarify whether these 198 samples were each from individual people (i.e. 198 different people), or whether multiple samples were taken from the same person. Please could you clarify which samples were used for training and which were used for the trial stage (if training samples were reused for the trial stage this is a limitation that needs mentioning).

Practical realisation and storage of samples

- Please clarify if there is any risk that positive and negative samples could have been contaminated with different odours due to the way they were collected. For example, you state positive samples were collected with the person wearing full COVID-19 safety equipment – could this have transferred odour to the positive samples and not the negatives? Were negative samples also disinfected - if not, it is possible that dogs were training on the odour of disinfectant on the positives?

Canine Ressources (note spelling mistake in title)

The dogs

- The photos of dogs are nice but are not required in a scientific paper – suggest delete.

Training and first testing of the dogs

- Lines 262 to 263. The statement that the acquisition of the specific odour of COVIS-19+ sweat samples required from one to four hours ideally needs further supporting details on the training methods used and statistical evidence that the discrimination had been learnt (e.g. what was the learning criterion, were the handlers blind, were blank/negative samples also placed out, how many presentations did each dog require?). Please clarify if the learning criterion was based on the subjective opinion of the handlers rather than a statistical criterion.

- This seems to be extremely rapid acquisition (some medical detection dog studies have required weeks or months of training) so a comment on this would be useful in the discussion section.

- Were the training samples different to those used for actual tests? If not, please include as a study limitation as the dogs could have learnt the specific odours of the training samples rather than SARS-CoV-2 generally.

Testing protocol

- Lines 282-283. Repeated use of the same sample could be considered "pseudoreplication". Please clarify if this was taken into account during the data analysis (i.e. the 3 samples should not be considered as independent trials if the dog had 3 attempts to detect the same sample - it’s 3 attempts at a single "diagnosis"). Ideally each sample, positive or negative, should be used only once in the test. If they were used more than once this need to be discussed as an important study limitation, as the dogs may have learnt individual samples rather than generalising based on a SARS-CoV-2 odour.

- Lines 290-291. Please comment on whether there is any risk that the person putting out the samples could have unintentionally influenced the results by handling or storing the positive and negative samples differently. The dogs may learn minor differences in packaging odour, storage location or handling (e.g. positives and negatives stored in different fridges may carry different environmental odours that the dogs learn).

- The ideal experimental design would be to run the trial double blind (the person putting out the samples does not know if they are positive or negative during the trial). If it was not double blind this needs to be mentioned as a study limitation.

- Lines 292-293. Please comment on whether the individuals collecting data were blind to the sample allocations and whether they could have biased the results (ideally they should be blind too).

Lines 295-296. This sentence suggests that dogs had not previously been exposed to the samples used in the trials – please clarify if this is correct. Does this mean new samples were from new individuals not previously assessed by the dogs?

Results and statistical analysis

- Line 313. As above, if samples have been reused several times for the same dog this may be considered psuedoreplication of data. Please clarify if this was taken into account during the statistical analysis. Reuse of samples/individuals may result in detection based on learning of individual odours rather than COVID odour generalisation (it is not impossible that dogs could potentially learn 33 different individual samples)

Table 3: Results obtained on the dogs that finished the training period (n=8)

- Please state how many different positive samples (from different people) each dog was presented. It is not clear whether samples collected from the same individuals were reused more than once per dog. If they were reused, this needs to be included as a study limitation.

- Ideally it would be best to provide your underlying data (breakdown of trials for each dog showing positive and negative samples put out and response of the dog) as a supplementary file/download.

Discussion

- As per my general comments above, I recommend that the discussion is revised to highlight that this study a promising first step, but you include a more critical discussion of the possible limitations and possible alternative explanations for the dogs’ performance. I think the current conclusion (lines 331-332) is too strong based on the methods and data presented, without a more robust double-blind and independently validated trial taking place.

- Lines 359-382. Although interesting, I feel that the discussion of female hormone detection is moving too far away from the focus of the study subject and would be better deleted from the discussion (the evidence for the dogs indicating these samples due to female pheromones is also weak/anecdotal).

Conclusion

- As per my general comments above, I think the current conclusion (lines 417-419) is too strong based on the methods and data presented, without a more robust double-blind and independently validated trial taking place – suggest revision of wording in line with my general comments above.

Figures

- Figures 3. 4 and 11 are not readable due to poor resolution.

- Suggest delete figures 5 and 6.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2020 Dec 10;15(12):e0243122. doi: 10.1371/journal.pone.0243122.r002

Author response to Decision Letter 0

1 Sep 2020

Point-by-point response to reviewers.

Reviewer #1

This article investigates whether dogs can differentiate the axillary odor of individuals infected by SARS-CoV-2 from individuals that are not infected. The study is very well justified and potential applications for the management of the current pandemic (especially in case of a feared second wave) are highly relevant.

I however have some major issues with this study, as detailed below.

1) Design.

With the current information provided in the method, the reader cannot be convinced that the authors controlled for a number of confounding factors. Namely, it should be made clear that the negative and positive samples cannot be different from each other based on other factors than infection to SARS-CoV-2. It is known that body odor changes according to medication intake, gender, age, health status, hormonal status etc. With this in mind, a number of questions can be raised here. Who are the COVID-negative patients: do they have other pathologies? If yes, which ones? Do they all have the same pathology, or different ones? Are they on medication? Are the COVID+ on medication and if yes, is it the same one for everybody? How are the negative samples chosen with regards to gender and age, according to the target (COVID+ sample) characteristics? Two odor sampling methods are mentioned (gauze swabs or filters and polymer tubes): Are they both used for negative and positive samples, and is the same sampling support used in the whole line during a trial? In sum, more information is expected in the method section on these aspects; the limits related to the possible confounding factors must be discussed in more details than it currently is; and the conclusion will probably need to be more cautious (as a consequence).

We thank the reviewer for this very relevant comment.

The COVID-19 positive and negative samples were drawn from patients or individuals working in one of five hospitals (four located in the Paris or in the Suburb of Paris, France, and one located in Beirut, Lebanon), as now indicated in the manuscript. In each hospital, once a sample from a COVID-19 positive patient was collected (who fulfilled the following inclusion criteria: showing clinical symptoms of COVID-19, who were COVID-19 positive on RT-PCR or PCR test for SARS-CoV-2), a sample from a COVID-19 negative individual from the same hospital (patients or nursing staff) was collected (who fulfilled the following inclusion criteria: no clinical symptoms related to COVID-19 and negative on SARS-CoV-2 PCR test). Therefore, the potential confounding factor of the “hospital odor” was taken into account. We added this point in the Discussion section.

Regarding the comment about the odor sampling supports, first of all, for the training sessions as well as for the testing sessions, the same sampling support was used on each of the 3 or 4 cones of the lineup. Therefore, the dogs could not have been influenced by the odor of the sampling support when it indicated a cone. For one dog, only the polymer tubes were used both for the training and testing sessions. For the five other dogs, they were trained and tested on each of the two sampling supports. We added this important point raised by the reviewer in the Material and Method section, as well as in the Discussion section.

Regarding the comparability criteria between COVID-19 positive and negative individuals, we added in the manuscript the percentage of female and the mean age between the two groups, for Paris and Beirut sites (the data from the Ajaccio site were excluded because there were not enough COVID-19 positive samples and these ones were used too many times for one dog). As the reviewer can see, the percentages of female are similar between the two groups (COVID-19 positives and COVID-19 negatives, both in Paris and Beirut). However, although similar between the two groups in Beirut site, the means of age were not similar in Paris site (48 years for the COVID-19 positive group versus 70 years for the COVID-19 negative group). We were able to collect data for some COVID-19 positive and negative individuals of the Parisian site on diabetes, overweight, arterial hypertension, respiratory disorders (such as asthma, or previous respiratory viral infections), and hypothyroidism.

The results are the following:

COVID-19 positive COVID-19 negative

Diabetes 4/14 (29%) 1/8 (13%)

Overweight 12/19 (63%) 4/11 (36%)

Arterial hypertension 9/17 (53%) 3/6 (50%)

Respiratory disorders 3/15 (20%) 3/4 (75%)

Hypothyroidism 4/14 (29%) 1/3 (33%)

Regarding the overweight, only one dog of the Parisian site (Gun) marked two (out of the four) COVID-19 negative but overweight individuals. The three other dogs did not mark these COVID-19 negative but overweight individuals.

Regarding the diabetes, the same dog (Gun) and only this dog (and not the other three) marked this COVID-19 negative individual having diabetes.

Although these data are very scarce, these results do not provide strong evidence that dogs marked diabetes or overweight condition rather than COVID-19 condition. However, the low number of available data prevent us from providing such number in the manuscript and claiming such statement. We mentioned this point in the discussion section, and the fact that confounding bias cannot be ruled out.

2) Statistical analyses.

The authors mention accuracy of the dog’s responses in introduction and conclusion, in terms of sensitivity and specificity. It should be clearly defined, especially with regards of the tests performed in the present article. The signal detection theory is evoked only in discussion, and false positives are discussed in that same section. It is therefore expected to see the current data analyzed using these criteria, which is not the case.

The word “accuracy” was mentioned in the Introduction section only when we referred to the paper by Maurer (Open Forum Infect Dis, 2016), inside which sensitivity and specificity terms were mentioned. This “accuracy” term has been removed from the manuscript to avoid any confusion by the readers. We now mentioned the sensitivity and specificity terms only when referring to subsequent studies.

Our data collection does not allow us to calculate any sensitivity or specificity since a trial consisted of a lineup containing 3 or 4 cones, with only one positive sample on the lineup, and the dog had only one “guess” to mark it out of the 3 or 4 samples of the lineup. (We removed the “false positive” expression in the manuscript in order to avoid any confusion.) The success rate (ie, the number of correct indications divided by the total number of trials) was rather calculated with its 95% confidence interval. Again, this study is a proof-of-concept study, and it therefore presents some limitations compared to a confirmatory study in which sensitivities and specificities can be calculated. To make clear that the readers should not expect sensitivities and specificities, we indicated this point in the Material and Method section.

3) Paragraph about human pheromones in the discussion (lines 359-382).

I suggest to simply remove this paragraph, which is 1) erroneous (no pheromones have ever been identified in humans, and the androgenous compounds you are referring to are male – not female – compounds supposed to attract females), 2) very speculative (how did you estimate the fertile phase of the donor menstrual cycle?), 3) disproportionally long compared to other points to discuss that are more crucial and more directly related to the main question.

As also suggested by another reviewer, these four paragraphs were removed from the manuscript.

Minor points.

1) Please describe what the mocks are, and how many are there in the line.

We added the description of a mock in the Material and Methods section.

Please find the requested information below for each of the 4 dogs of the Parisian site.

Gun (47 trials, lineup of 3 cones): 44 trials with 0 mock and 3 trials with 1 mock.

Guess (52 trials, lineup of 4 cones): 34 trials with 1 mock and 18 trials with 2 mocks.

Maika (55 trials, lineup of 4 cones): 38 trials with 1 mock and 17 trials with 2 mocks.

Oslo (31 trials, lineup of 4 cones): 31 trials with 0 mock.

2) Please better explain how the training was conducted with regards to sample type: are there negatives at this stage? Mocks?

We agree that the training session was too scarcely described. We detailed it, and we hope that it is better explained.

3) Please explain body odor sample management. From what I understand, they are used fresh (minimum 24 hrs between odor collection and training). Are they thrown away after 3 trials? Are they frozen at any time in the process? Refrozen?

Any sample, whatever COVID-19 positive or negative, was never frozen. The samples were used fresh and thrown away after all dogs carried out the trials. The time between sample collection at the hospital and the trial (either during the training period or during the testing period) was independent of the sample status (COVID+ or COVID-).

Finally, on a more formal aspect, the article should be proofread for English. Also, improvement of the writing would be welcome. Some sentences are difficult to understand (i.e., lines 237-240 / 395-398) and some others do not integrate well in the chain of arguments (i.e. lines 62-64 / 343-345). Most importantly, in its current form the article is very fractionated (“bullet points” feeling in the whole document): It would very much gain in fluidity and articulation of the ideas if the authors work on this aspect.

We deeply revised the writing of the manuscript, and we do hope that its current form is now acceptable.

 

Reviewer #2

The idea of using trained dogs for COVID-19 screening was boosted by mass media in many countries e.g. by the BBC in UK, and by announcing by many researchers and dog trainers, that they intend to train dogs to detect COVID-19. It is obvious that first published experimental works in this field, will have a great potential for being broadly cited or comercialized. However, it could be presumed that such experiments and manuscripts are probably accomplished under a strong pressure of time, to be published as the first. Therefore the papers like the present one, which, to my knowledge is the first paper on this topic, should be reviewed and evaluated particularly causciously and critically. It is important not to disseminate facts that are not based on sound scientific experiments. This is particularly important since the reviewed paper is already available online at https://www.biorxiv.org/content/10.1101/2020.06.03.132134v1.full.pdf as not certified by peer review.

For justification of their study the authors cited publications on cancer detecting dogs and dogs alerting hypoglycaemia and seisure in human patients. However, it should be mentioned that despite of about 20 publications within the last 17 years, the medical detection dogs have not been applied for practical clinical screening so far. This was evident on the latest international conference on cancer detection by dogs, organized by the Curie-Institute in Paris in October 2019. Also, despite of numerous chemical studies within more than 25 years, using gas chromatography / mass spectrometry, which found some differences in volatile organic compounds (VOCs) between cancer positive and cancer negative patients, no single VOC or a simple combination of VOCs were identified, that could be reliable cancer marker(s), mature for practical cancer screening. The reason for problematic application of methods based on identification of VOCs for disease screening, either by chemical analyses or by trained animals, is a great variability of emited VOCs due to day-to-day differences in diet, medication and numerous other confounding factors.

The aim of the reviewed study was to assess if dogs could be trained to discriminate body odor of patients infected with SARS-Cov-2, from body odor of healthy controls.

Such aim is justified by the current and probably future situation with COVID-19, untill an effective and commonly accessible COVID-19 vaccine is available.

The experiments were conducted at 3 different locations, with a total of 18 dogs which were previously trained either for explosives detection or for search and rescue or for cancer detection.

The Authors collected altogether 101 sweat samples from patients showing clinical symptoms of COVID-19, being also COVID-19 positive on RT-PCR or PCR test for SARS-CoV-2. Negative samples (n=97) were collected the same way, by the same trained staffs, from patients who have no clinical symptoms related to COVID-19, and were negative on SARS-CoV-2 PCR test.

The main flaw of collecting sweat samples for this study was, that the samples were not controlled quantitatively for the amount of the sweat. It is known that one of the typical COVID symptoms is a high fever. Humans with fever sweat more intensively than healthy people, so the amount of the sweat on the sample may be a clear cue for dogs.

As an example, our early studies in the eighties, on dogs detecting so called „silent estrus” in cows, could be mentioned. We used cotton swabs for collection vaginal mucus from cows in estrus vs diestrus. Our dogs perfectly indicated estrus samples. Only later we noticed that cows in estrus usually secreted more mucus and this was a cue for the dogs. When the dogs became more „experienced” they perfectly indicated cotton swabs moistured even with water. This made me causcious and critical about dog’s reliability. For dogs, odors like cow’s in estrus, human cancer, drugs or explosives, or SARS-CoV-2, play no biological role and such odors are indicated by dogs only because an operant conditioned response has been trained by producing association between odors and a reward in form a treat or favorite object to be retrieved and play with. Therefore, the trained dogs, especially when they are more „experienced”, will seek for a cue or a simple solution, to earn a reward. Such strategy of dogs may produce frequent false alerts. At identification of human scents by dogs it should be remembered that samples of human scent contain both individual component and other components as those related to diet, disease(s), medication, hygiene, cosmetics, background odors of the location as well as own odors of the sampling materials (odor carriers).

In lines 346-349 the authors report on two samples that were negative according to inclusion criteria for negative samples, but were indicated by two dogs as positive and the PCR re-test showed that in fact they were positive. This, however, is not a convincing evidence that the dogs sniffed out specific molecules induced by the SARS-CoV-2 virus presence. These two persons were not controlled for the amount of the sweat or body temperature measured precisely at the time of sample collection. Without knowing if the odor samples differ quantitatively in the amount of the sweat, it is not clear whether the dogs indicated a hypothetic, characteristic odor emited by COVID positive patients, or rather indicated samples from people with fever, who sweat more.

A recent study (Concha AR, Guest CM, Harris R, Pike TW, Feugier A, Zulch H, et al. Canine Olfactory Thresholds to Amyl Acetate in a Biomedical Detection Scenario. Frontiers in veterinary science. 2019;5:345) actually showed that the accuracy of detection can be very high even at low levels of concentrations. By analogy, we expect that small amounts of sweat in the sample would be as efficient (in dog’s detection) as large amounts of sweat. Furthermore, it is well known that large amounts of a substance that a dog must detect often provide poor results due to the saturation of its olfactory receptors. With the sampling procedure we used in our study, the average amount of sweat obtained was around 75 mg for a 20 minute-contact of a 2x2 inches gauze with the skin, and the same amount of sweat were obtained for cellulosic filters. The inter-individual variability in sweat quantities obtained after this 20 minute-contact were not related to the status of the sample (COVID-19 positive or negative), but rather on individual characteristics of the sampled persons.

This important point raised by the reviewer was added in the Discussion section.

The Authors mentioned that the odor samples were collected „at the same location”. This is methodologically sound but it should be precised what does it means „the same location” in this context ? Was it the same room or the same buiding or the same district ? The authors must be more precise here, because it is known that the dogs might be conditioned on the characteristic „hospital odors”, as shown by Walczak et al.(2012) in cancer detecting dogs, when the cancer positive samples were collected mostly in hospitals.

The COVID-19 positive and COVID-19 negative samples were collected from individuals from the same hospital (one COVID-19 positive sample was matched to one COVID-19 negative sample inside the same hospital). The COVID-19 negative samples were collected either from patients of the hospital or from the nursing staff of the hospital. We cannot rule out that inside the hospital, rooms of COVID-19 positive patients may have a different odor compared to rooms of COVID-19 negative patients. However, as Walczak et al. wrote in their 2012 paper, “hospital rooms may have a hospital odor, derived mainly from disinfectants, which may be a common component of all samples taken from donors who are inside a hospital.”. Therefore, in our data, there should not be confounding bias due the hospital odor. We added this point in the Material and Methods section, as well as in the Discussion section.

Another problem with COVID-positive sweat samples is that positive patients are staying either in infectious hospitals or in isolation wards that are probably intensively desinfected.

In addition, the access of healthy COVID-negative healthy donors (controls) to the same room where sick patients are staying, is hazardeous and strictly limited. This may be a serious constraint for collecting samples for further studies by other researchers.

The COVID-19 negative individuals never stayed for a long time in the same room as the COVID-19 positive patients (never for the COVID-19 negative patients, and during a short period of time for the COVID-19 negative nursing staff). Furthermore, as previously cited (Walczak et al. (2012)), the hospital disinfection protocol should be the same between rooms inside the hospital.

The description of the training methods only in 3 lines is very scarce. Were the dogs rewarded in every training and testing trial and who and how gave a signal that the dog’s indication was correct?

We agree that the training protocol was scarcely described, and we described it now more in details. We also modified the description of the testing protocol, where now it is written that the dog was rewarded once it correctly marked a COVID-19 positive cone (the data recorder, who was in the same room, knew the result of the indication of the dog, and told the result to the dog handler right after the indication by the dog, allowing the dog handler to reward the dog if the dog correctly marked the cone). We added in the manuscript that the data recorder, although not in a blinded situation, was located into the room such that there was no visual contact between the data recorder and either the dog or the dog handler.

Lines 219-220 - It is true that search and rescue dogs (SRD), trained to perform disaster and area search, mainly work through the scent of the sweat. However, it must be taken into consideration that SRD are trained for detection of a generalized human odor, without discriminating individual persons and persons’ specific odor e.g. odor(s) related to a disease. Such dogs are relatively easy to be re-trained to work in a scent lineup to discriminate odours of particular person, but a re-training, to be reliable, takes longer than 1-4 hours and 4-10 trials.

As written in our manuscript (but modified since the original submitted version), three types of detection dogs were used. (1) Three explosives detection dogs, which are trained to detect between 30 and 40 different types of explosives, and are used to work on a line of samples that they have to sniff individually. For such dogs, if COVID-19 positive samples have a specific odor, they only have to memorize only one additional odor in their “olfactory memory bank”. (2) One search and rescue dog, which is trained to perform disaster and area searches, and works through the scent of the sweat (Grandjean D., Haak R., Massey J., Pritchard C., Schuller P., Riviere S.; The search and recue dog, Aniwa ed., Paris, France, 2077; pp296). We must point out that such search and rescue dogs are trained for detection of a generalized human odor, without discriminating individual persons and persons’ specific odor (e.g. odor(s) related to a disease). However, such dogs can be easily re-trained to work in a scent lineup to discriminate odors of particular person, but a re-training, to be reliable, takes longer than for explosives detection dogs. (3) Two colon cancer detection dogs trained on rectal gases (Sarkis R, Khazen J, Issa M, Khazzaka A, Hilal G, Grandjean D. Dépistage du cancer colorectal par détection olfactive canine. J Chir Visc. 2017;154(Supp 1):22). These detection dogs are used to work with samples on lineups, and the acquisition of a new odor is fast.

In lines 262-264 the Authors stated: „…For each involved dog, the acquisition of the specific odour of COVIS-19+ sweat samples required from one to four hours, with an amount of positive samples sniffing ranging from four to ten….”

I am skeptical about the ability of each dog to learn precisely the task within such a short time. For example for cancer detecting dogs, the training period according to 13 papers published in the period 2003-2014, ranged from several weeks (Pickel et al. 2004) to 16 months (Cornu et al.2010), and on average took 8 months.

Even if the Authors could observe that their dogs correctly indicated the target samples after one hour, performing only four trials from being confronted for the first time with COVID sweat samples, it seems inredible that dogs could acquire the ability to detect COVID odor so quickly. It can not be excluded that the dogs indicated samples in the lineup, that are merely different from the others e,g, by containing more sweat.

A recent study on COVID-19 detection dogs published by Jendrny et al. (Jendrny P, Schulz C, Twele F, Meller S, von Kockritz-Blickwede M, Osterhaus A, et al. Scent dog identification of samples from COVID-19 patients - a pilot study. BMC infectious diseases. 2020;20(1):536.) mentioned a similar time to train COVID-19 detection dogs. Of course, several weeks (depending on dog's ability and behavior) are necessary but are enough (between one and three for our six dogs). From the experience we now have with our collaborators of 5 different countries (Emirates, Brazil, Argentina, Chile, Mexico), the training session for one dog already trained for olfactory detection takes also between one and three weeks, and take between four and eight weeks for working dogs without any detection experience.

Positive samples do not contain more sweat than negative. We saw that in Lebanese as well as in Emirians studies, where negative individuals do sweat as much (and sometimes even more) than COVID-19 positive hospitalized ones.

In lines 265-267 or in lines 294-296 the Authors did not precise what was a criterion (percentage of correct indications, false alerts, misses, regularity of sniffing the samples in the lineup ?) for considering the dog to be fully trained and ready for testing.

In the more detailed training protocol paragraph, we added the requested information: “The dog was considered as trained and ready for the testing session by the subjective opinion of the handler, based on the behavior of the dog”.

From the lines 294-296 it could be concluded that separate samples were used for the training and testing, which is methodically correct. However, there is no information how many positive and negative or mock samples, out of the totals given in Table 1, were used for the training and real testing.

We decided to remove Table 1 and to replace it by the text in the Results section describing the study sample of COVID-19 negative and positive individuals, both for the Parisian site and for the Beirut site. If such Table of individuals would have still been present in the manuscript, it would have been uneasy to read if we added data regarding the trials (how many positive and negative or mock samples were used per dog). However, we described more in details the training and testing protocols. Furthermore, here are the data requested by the reviewer for the 4 Parisian dogs:

Gun (47 trials, lineup of 3 cones): 44 trials with 0 mock and 3 trials with 1 mock.

Out of the 5 unsuccessful trials, Guess marked a COVID-19 negative sample 5 times.

Guess (52 trials, lineup of 4 cones): 34 trials with 1 mock and 18 trials with 2 mocks.

Out of the 9 unsuccessful trials, Guess marked the mock 3 times, and a COVID-19 negative sample 6 times.

Maika (55 trials, lineup of 4 cones): 38 trials with 1 mock and 17 trials with 2 mocks.

Out of the 10 unsuccessful trials, Maika marked the mock 3 times, and a COVID-19 negative sample 7 times.

Oslo (31 trials, lineup of 4 cones): 31 trials with 0 mock.

We do hope that this addresses the issue raised by the reviewer.

The authors did not conduct so called „zero trial” with no target samples in the lineup. This is another flaw which could have a consequence by making the overall conclusion in lines 407-408 premature. Although negative and mock samples were used in the lineup (lines 269-270), but there is no explanation what kinds of mocks were used. Were they samples with more sweat ? or samples from donors with fever but not related to COVID-19 (for example from people with seasonal influenza) ? „Zero trials” usually provoke more false alerts, especially on samples that are simple different from the others.

The lineup contained one (and only one) COVID-19 positive sample. All the other samples were negative samples or mock(s). The mock contained only the sample support (ie, the sterile gauze or the inert polymers tube). We added this information in the paragraph detailing the training protocol. As written in the manuscript, the inclusion criteria for COVID-19 negative individuals were the following: no clinical symptoms related to COVID-19, and having a negative PCR test for SARS-CoV-2. COVID-19 negative individuals may therefore not have fever. The absence of “zero trial” may prevent us from drawing the written conclusion (in lines 407-408 in the original version of the manuscript). We modified such conclusion accordingly.

Minor remarks:

The manuscript should be better structured. For example the major portion of the Conclusion is appropriate rather for the Discussion section, whereas the real conclusion is given as the last sentence or the Discussion (in lines 407-408)

We modified the structure of the manuscript, and especially the Discussion section. We removed the Conclusion section.

Line 64 More original papers should be cited instead of a handbook.

The paragraph where the handbook was cited has been removed.

Lines 76-78 „…Scent detection by animals has been published in a large number of diagnostic studies, which all suggest similar or even superior accuracy compared with standard diagnostic methods…” – nevertheless, the scent detection by animals is not used for medical diagnoses or screenings on a large scale (probably except for tuberculosis detecting Gambian pouched rats in Africa). The medical diagnosis of a disease should not be confused with screening. Scent detection by animals will never be a diagnosis.

We agree with the reviewer and modified this sentence accordingly.

Line 83 „…The first clinical investigation of cancer was published by Willis on bladder cancer in 2010, after having published a proof of principle study in 2004…” – this information is only partly true: McCulloch et al. published their study on canine detection of lung and breast cancer in 2006.

Actually, the sentence in the manuscript was about bladder cancer, not all cancers. However, we acknowledge that the sentence was not clear, and we modified it by adding the study of McCulloch suggested by the reviewer.

Lines 88-89 „….The dog’s nose is also now currently utilised in order to prevent the consequences of crisis (or even prevent them) for diabetic [30, 31, 32, 33] and epileptic people [70]….” - The accuracy and reliability of hypoglycaemia and seizure alerting dogs, when considering on a daily basis, is still questionable. The alerting dogs rather contribute to feeling more safe and thus to improving the quality of life of the patients, than to precise alerting the hypoglycaemia or seizure episodes.

We modified the sentence according to the comment of the reviewer.

Lines 115-117 „….In a second time……” – this assumption is not quite clearly formulated in English.

We modified the sentence.

Lines 154-158 Why two types of sampling supports are used ?:

(1) - Sterile gauze swabs utilised by the hospitals or sterile gauze filters utilised for explosive detection, and (2)- Inert polymers tubes utilised for explosives, drugs or criminology detection)

First of all, and most importantly, for the training period as well as for the testing period, the same sampling support was used on each of the 3 or 4 cones of the lineup. Therefore, the dogs could not have been influenced by the odor of the sampling support when it indicated a cone. For one dog, only the polymer tubes were used both for the training period as well as for the testing period. For the five other dogs, they were trained and carried out the testing session on each of the two sampling supports. Two sampling supports were used because at the time to the beginning of the proof-of-concept study, we did not know which one of the two sampling supports was the most effective. However, the main point is that the same sampling support was used on each of the 3 or 4 cones of the lineup.

We added this point raised by the reviewer in the Material and Method section, as well as in the Discussion section.

Line 289 Why acetone was used for disinfection of sample carriers ? Acetone is an irritant, causing mild skin irritation and moderate to severe eye irritation which may be repelling for dogs.

We modified the sentence since it originally did not well describe the used procedure. It is now: “All cones were cleaned with straight water after each trial, were twice a day cleaned and disinfected with 10p100 aqueous acetone solution, and then dried up.”

Lines 292-297 „…A failure was defined if the dog alerts on any of the other boxes. Trials were registered from the moment dogs recognized positive sample among other samples (mock or negative ones) without having sniffed any of them previously to the test. If the dog marks first a negative sample, the trial is considered as failed, even if he marks the positive afterwards….”

- the authors did not precise whether the dogs have to sniff ALL samples prior to the indication. Have the dogs to sniff the remaining samples after the correct indication? If so, would the trial be declared as a failure, if the dog made a false alert AFTER prior correct indication in the same trial? This makes that the calculation of correct indication by chance should take into account how many samples a dog actually has sniffed. Therefore, when the dog indicates correctly for example on the first sample in the lineup and does not sniff the remaining samples, the probability of correct indication by chance at 7 samples in a lineup is not 14% but 50%. This would mean that the probability of correct indication by chance is varying from trial to trial?

First of all, the dog was trained such that it knew that there was only one positive sample on the lineup. Once the dog was trained and ready for the testing period, it was presented into the room where the samples were already and randomly placed behind each cone. Only one cone out of the cones contained the positive sample. All the other cones contained negative samples or mock. The dog was first presented to the first cone (the one on the very left of the room), then to the second (the next one on the right), and so on. The dog could freely sniff all the cones, but the rule was the following: once the dog indicates a cone (by sitting in front of it), the trial stops. If the dog indicates a cone with the positive sample, the trial was considered as a success; it was considered as a failure in the other case. In such a protocol, the calculation of the probability of correct indication by chance is still one divided by the number of cones on the lineup. For instance, in a lineup of four cones, if the first cone (ie, the one on the very left into the room) contained the positive sample, and if the dog indicates this first cone, the probability of indicating this cone by chance only is still 25%, because the trial consists in only one sit by the dog (again, the trail stops once the dog sits). The game for the dog was: “pick the only one positive sample out of the 3 or 4 samples on the lineup, and you have only one trial”. However, we acknowledge that the description of the protocol was not clear enough, and we modified it for better understanding of the protocol. We do hope that it is clearer now.

Line 350 „….dogs have no reason to lie….” – this is too colloquially expressed. With regards to odor detecting dogs one should not speak about "lying" or "telling truth".

>From the dog's perspective, detecting trained odors as for example the hypothetical COVID-19 odor, that have no biological relevance, is a rather a game to earn a reward. Dogs, even if they are well trained, when they are getting more routined, will try to get a reward at a simple way, which means that they will try to find any additional cues to indicate the target odor in the lineup. Therefore the dogs' performance in the scent lineup has to be permanently controlled throughout the whole deployment period, because it may vary. This seems to be one of the reason why so called medical detection dogs have not been used in medical screening on a regular basis.

We agree with the reviewer and we modified the expression “dogs have no reason to lie”.

Table 3. The terms detection sensitivity and specificity, which derive from the point 1-4 in lines 385-388, should be given, or percentage of false alerts should be given instead of the column with the number of correct indications, which is redundant in this table

Neither Fjellanger et al. (ref 65) nor Concha et al. (ref 67) mentioned the sensitivity (Se) and specificity (Sp) terms in their study – we removed ref 66 in our manuscript, only because it was a handbook and not an original paper. In our study, we deliberately did not mention these terms of Se and Sp, simply because Se and Sp could not have been calculated with such our protocol. In order to calculate Se and Sp, we would have needed one (positive or negative) sample behind one cone only, and the dog would have had to sit, or not, in front of this specific cone (exactly like a diagnostic tool evaluating the presence of a disease from one sample of one patient). In our study, because the lineup contained 3 or 4 cones, all of them minus one containing negative samples or mocks, it was not possible to calculate such Se and Sp. In the same way, it is not a question of “false alert” (again, it would have been the case if there had been only one cone, either positive or negative). This study is a proof of concept study. It has not the strength of a study in which Se and Sp (or percentage of false alerts) can be calculated. Calculating such statistical indicators from our data would have led to an over-interpretation of our results. However, the 95% confidence intervals of the success rates we were able to calculate never included the success rate that would have been observed by chance, which provides some evidence that dogs seem a promising tool to detect COVID-19 positive people. We added this point in the Discussion section.

The English of the manuscript is for me as a non-native speaker understandable but suffers from many language and wording imperfections. It should be edited by a native English speaker, who also has some knowledge in the specific study field.

We modified the writing of the manuscript to avoid wording imperfections.

Line 48 „..State of play..” ? – rather „State-of-the-art.”

We removed this expression.

Line 51 What does it means „..for those of interest in dog’s olfaction…” in this context?

We modified the sentence.

Line 148. „…In order to avoid any dog creancement (?) on the olfactive “background noise

In several places of the manuscript there are statements that are difficult to follow.

We modified the sentence.

In lines 115-117 the authors write: „…In a second time, we attempt to validate the method with the dogs checking samples one by one instead of comparatively, in order to fit with what could be an operational utilisation of such trained dogs...” – what does it means exactly?

We removed this sentence because it finally appeared that is was not relevant.

Line 372 „ In Sheakspeare’s day…” ? - rather is „ in Sheakspeare’s era” or „in Sheakspeare’s time”

Lines 372-374 the passage with the anecdote on peeled apple under woman’s arm should be deleted because it is too colloquial and not closely related to the reviewed study.

As suggested by another reviewer, we removed the four paragraph regarding pheromones, including the one where the two expression cited above were written.

Line 404 „ …cynotechnicians…” ? or dog handlers ?

We modified the word “cynotechnicians” by dog handlers.

Line 455 „…by small…” ? – should be „ by smell”

The reviewer was right. We however removed the paragraph including the reference of the book since it was redundant with the following one in the manuscript.

In conclusion, this first report on dogs trained to detect COVID-19 is interesting but a flaw in the methodology of collecting scent samples (no control for the amount of the sweat) makes that the conclusion that „…there is a high evidence that dogs can detect a person infected by the virus responsible for COVID-19 disease …” seems to be premature.

We agree with the reviewer that the conclusion was too strong and premature regarding the protocol of the proof-of-concept study. We modified the discussion section accordingly.

In addition, some language imperfections make me to evaluate the manuscript as not acceptable for publication in PloS ONE in the present form and needing a major revision before resubmission.

First of all, in the major revision the issue of the amount of sweat in odor samples should be addressed, including replication of the trials.

Regarding the replication of the trials, we are conducting a second study which confirms the results of this proof-of-concept study. However, we needed to conduct and write the manuscript of this proof-of-concept study first, in order to implement the second one.

Regarding the issue about the sweat quantity, we answered the reviewer that, first, if the quantity of sweat has any impact of dog’s detection, it would be on the opposite way as the one suggested by the reviewer (ie, larger amounts of sweat would reduce the ability of the dog to detect a COVID-19 positive sample), and second, that the quantity of sweat was not related to the status of the sample but rather to the individual characteristics. We do hope that our explanations answered the issues raised by the reviewer.

We added in the Material and Method section that some samples could have been used in different trials for one dog. This is an important point we discussed in the Discussion section.

 

Reviewer #3

Thank you for conducting and reporting this very interesting study. In my opinion, this study is a good first step towards testing whether dogs can detect odours associated with SARS-CoV-2 infection. You have used a reasonable sample size of dogs and SARS-CoV-2 samples and the methods are generally appropriate. A main difficulty with any study of this type, where it is not known whether the odour the dogs are being trained to detect exists (in this case SARS-CoV-2 related), is demonstrating beyond reasonable doubt that dogs are detecting the odour rather than incidental contamination, and demonstrating robustly that the research team is not unintentionally influencing/cuing the dogs. Sources of odour contamination may include, for example, medication used in SARS-CoV-2 patients, hospital odours, differences in the odour of the storage containers of positive and negative samples and different odours arising from how positive and negative samples are handled. You have taken some steps to try to control for these issues. However, in my opinion, it is not possible to conclusively say that the sweat of SARS-CoV-2 contaminated persons can be detected by trained dogs based on the methods and results you present so far. In places, the methodology and controls you used just need further clarification in order to rule out possible sources of bias and alternative explanations that the dogs may be detecting contaminating odours associated with the SARS-CoV-2 samples.

Most importantly, to support the conclusion that the sweat of SARS-CoV-2 contaminated persons can be detected by trained dogs, a more rigourous, double-blind test is required, in which the research team is provided with new samples from an independent source which they are asked to assess using the dogs. The dog handlers, sample handlers and data recorders (everyone involved in assessing the samples) should be blind to the sample identities (positive or negative) until the dogs have conducted the test. Methods for how to do this have been presented in a number of previous medical detection dog studies.

I would suggest two options for revising the paper:

- Additional data is added, providing the results of a double-blind test using samples from new people that have not previously been screened by the dogs, or

- The current abstract, discussion and conclusions are revised to provide a more critical appraisal of the results, covering possible limitations of the study and whether there are any other possible explanations for the dogs' detection performance and possible sources of unintentional bias.

Potential sources of bias/limitations in medical detection dogs studies have been summarised previously (e.g. Elliker et al 2014. Key considerations for the experimental training and evaluation of cancer odour detection dogs; Edwards et al 2017 Animal olfactory detection of human diseases: Guidelines and systematic review; Johnen D, 2017. An approach to identify bias in scent detection dog testing). I recommend you review these guidelines and report where your study meets these criteria, and/or discuss limitations where your study does not meet these criteria.

We thank a lot the reviewer for her/his very relevant comments.

We chose the second option the reviewer suggested. At the mean time, we deeply modified the manuscript to clarify some points raised by the reviewer. The potential confounding bias due to “hospital odor” was already taken into account thanks to the matching of COVID-19 positive individuals to COVID-19 negative individuals from the same hospital (this important information has been added in the manuscript).

We also added in the manuscript that both the dog handler and the sample handler were blinded to the sample identity (COVID-19 positive or negative). The data recorder was not blinded but was located in the back of the room, where neither the dog handler nor the dog could see the data recorder when the dog was searching the positive sample.

Regarding the comparability criteria between COVID-19 positive and negative individuals, we added in the manuscript the percentage of female and the mean age between the two groups, for Paris and Beirut sites (the data from the Ajaccio site were excluded because there were not enough COVID-19 positive samples and these ones were used too many times for one dog). As the reviewer can see, the percentages of female are similar between the two groups, both in Paris and Beirut. However, although similar between the two groups in Beirut site, the means of age were not similar in Paris site (48 years for the COVID-19 positive group versus 70 years for the COVID-19 negative group). We were able to collect data for some COVID-19 positive and negative individuals of the Parisian site on diabetes, overweight, arterial hypertension, respiratory disorders (such as asthma, or previous respiratory viral infections), and hypothyroidism.

The results are the following:

COVID-19 positive COVID-19 negative

Diabetes 4/14 (29%) 1/8 (13%)

Overweight 12/19 (63%) 4/11 (36%)

Arterial hypertension 9/17 (53%) 3/6 (50%)

Respiratory disorders 3/15 (20%) 3/4 (75%)

Hypothyroidism 4/14 (29%) 1/3 (33%)

Regarding the overweight, only one dog of the Parisian site (Gun) marked two (out of the four) COVID-19 negative but overweight individuals. The three other dogs did not mark these COVID-19 negative but overweight individuals.

Regarding the diabetes, the same dog (Gun) and only this dog (and not the other three) marked this COVID-19 negative individual having diabetes.

Although these data are very scarce, these results are not in favor of that dogs mark diabetes or overweight condition rather than COVID-19 condition. However, the low number of available data prevent us from providing such number in the manuscript. We will mention this point in the discussion section, and the fact that, as the reviewer wrote, confounding bias cannot be ruled out.

We also thank the reviewer for the suggested references, which are now cited in the manuscript.

The entire manuscript, including references, needs proof reading and correction for typographical errors, spelling, grammar, and consistency of tense and language.

We modified the writing of the manuscript and we hope that it is now suitable for the reviewer.

Detailed comments

Abstract

- For the acquisition phase it would be useful to clarify how many “runs” dogs required during the one to four hours of training. Was it just a single presentation of each sample or were they presented multiple times during?

The training protocol is now described more in details in the Material and Methods section.

- Please clarify the wording around how many dogs were used for data collection in the study (it currently confusing for the reader whether 8 or 18 dogs were used for data collection).

After the removal of the data from the Ajaccio site, 14 dogs were trained and 6 dogs finally performed the trials for testing. We only mention now the six dogs used for the study; we did not mention anymore the number of dogs used for the training, since it added a non relevant information. The new Table 1 (ex Table 2) contained only the six dogs which performed the testing session.

- As per my general comments, I feel the conclusion that the sweat of SARS-CoV-2 contaminated persons can be detected by trained dogs is too strong at present, as the study does not appear to provide a robust double blind assessment and possible sources of bias/limitations have not been discussed. The study appears to be a good first step, but there are other possible explanations for the results based on the data presented. I suggest the conclusion is caveated that this study is a promising first step, and that a double blind trials is needed to more robustly confirm the results (unless this data is available and can be presented).

We totally agree with the reviewer, and we modified the Discussion section and the conclusion sentence accordingly.

Medical detection dogs: state of play

- To avoid bias towards only discussing positive studies and to provide a more balanced discussion, I suggest mentioning some of the difficulties of conducting these studies where it is unknown whether the target odour exists, and that not all studies to date have shown disease odours to be reliable detectable by dogs. Issues are summarised in papers such as Elliker et al 2014, Edwards et al 2017, Johnen D, 2017.

We thank the reviewer and added the points raised in the three mentioned papers in the manuscript.

Material and Methods

Samples

- Please clarify whether these 198 samples were each from individual people (i.e. 198 different people), or whether multiple samples were taken from the same person.

There were 177 samples collected from 177 different individuals (one sample per individual). We added the clarification into the manuscript.

Please could you clarify which samples were used for training and which were used for the trial stage (if training samples were reused for the trial stage this is a limitation that needs mentioning).

Actually, samples used for the training sessions were not used the testing sessions. This point has been clarified at the beginning of the testing protocol paragraph.

Practical realisation and storage of samples

- Please clarify if there is any risk that positive and negative samples could have been contaminated with different odours due to the way they were collected. For example, you state positive samples were collected with the person wearing full COVID-19 safety equipment – could this have transferred odour to the positive samples and not the negatives? Were negative samples also disinfected - if not, it is possible that dogs were training on the odour of disinfectant on the positives?

There was no risk of odors contamination between positive and negative samples. Worn equipment was the same for positives and negatives. However, we agree that the original text was not clear enough about this subject, and modified it accordingly (see “Samples” part of the manuscript).

Canine Ressources (note spelling mistake in title)

The dogs

- The photos of dogs are nice but are not required in a scientific paper – suggest delete.

We deleted the photos.

Training and first testing of the dogs

- Lines 262 to 263. The statement that the acquisition of the specific odour of COVIS-19+ sweat samples required from one to four hours ideally needs further supporting details on the training methods used and statistical evidence that the discrimination had been learnt (e.g. what was the learning criterion, were the handlers blind, were blank/negative samples also placed out, how many presentations did each dog require?). Please clarify if the learning criterion was based on the subjective opinion of the handlers rather than a statistical criterion.

We developed this training protocol of the manuscript, in order to provide more information for the reader. Regarding the learning criterion, we added: “The dog was considered as trained and ready for the testing session by the subjective opinion of the handler, based on the behavior of the dog.”

- This seems to be extremely rapid acquisition (some medical detection dog studies have required weeks or months of training) so a comment on this would be useful in the discussion section.

As written in our manuscript (but modified since the original submitted version), three types of detection dogs were used. (1) Three explosives detection dogs, which are trained to detect between 30 and 40 different types of explosives, and are used to work on a line of samples that they have to sniff individually. For such dogs, if COVID-19 positive samples have a specific odor, they only have to memorize only one additional odor in their “olfactory memory bank”. (2) One search and rescue dog, which is trained to perform disaster and area searches, and works through the scent of the sweat (Grandjean D., Haak R., Massey J., Pritchard C., Schuller P., Riviere S.; The search and recue dog, Aniwa ed., Paris, France, 2077; pp296). We must point out that such search and rescue dogs are trained for detection of a generalized human odor, without discriminating individual persons and persons’ specific odor (e.g. odor(s) related to a disease). However, such dogs can be easily re-trained to work in a scent lineup to discriminate odors of particular person, but a re-training, to be reliable, takes longer than for explosives detection dogs. (3) Two colon cancer detection dogs trained on rectal gases (Sarkis R, Khazen J, Issa M, Khazzaka A, Hilal G, Grandjean D. Dépistage du cancer colorectal par détection olfactive canine. J Chir Visc. 2017;154(Supp 1):22). These detection dogs are used to work with samples on lineups, and the acquisition of a new odor is fast.

From the experience we now have with our collaborators of 5 different countries (Emirates, Brazil, Argentina, Chile, Mexico), the training session for one dog already trained for olfactory detection takes also between one and three weeks, and take between four and eight weeks for working dogs without any detection experience.

A recent study published by Jendrny et al. on COVID-19 detection dogs (Jendrny P, Schulz C, Twele F, Meller S, von Kockritz-Blickwede M, Osterhaus A, et al. Scent dog identification of samples from COVID-19 patients - a pilot study. BMC infectious diseases. 2020;20(1):536.) mentioned a similar time to train COVID-19 detection dogs.

We added this point in the Discussion section.

- Were the training samples different to those used for actual tests? If not, please include as a study limitation as the dogs could have learnt the specific odours of the training samples rather than SARS-CoV-2 generally.

As previously mentioned, samples for training were not used for testing sessions.

Testing protocol

- Lines 282-283. Repeated use of the same sample could be considered "pseudoreplication". Please clarify if this was taken into account during the data analysis (i.e. the 3 samples should not be considered as independent trials if the dog had 3 attempts to detect the same sample - it’s 3 attempts at a single "diagnosis"). Ideally each sample, positive or negative, should be used only once in the test. If they were used more than once this need to be discussed as an important study limitation, as the dogs may have learnt individual samples rather than generalising based on a SARS-CoV-2 odour.

We thank the reviewer for the relevant comment. COVID-19 positive or negative samples could have been used up to three times for one dog, during the testing session. As the reviewer said, this is a limitation of the study that we discuss now the in Discussion section. Importantly, because Williams et al. showed that the dogs can learn at least 10 odors (Williams M, Johnston JM. Training and maintaining the performance of dogs (Canis familiaris) on an increasing number of odor discriminations in a controlled setting. Appl Anim Behav Sci. 2002;78(1):55-65), we removed the data from the Ajaccio site since there were only 6 COVID-19 positive samples (and 15 COVID-19 negative samples). In such a case, we cannot rule out, based on the study of Williams et al., that the dogs from Ajaccio used their olfactory memory to detect COVID-19 positive samples. However, for the remaining sites (Paris and Beirut), the dogs worked on at least 24 COVID-19 positive samples (and at least 33 COVID-19 negative samples), and we make the hypothesis according to which the olfactory memory of dog did not play a major role in the detection task, but may have been able to generalize the odor of COVID-19 positive samples used for training. We added this important point in the paragraph describing the testing session as well as in the Discussion section.

- Lines 290-291. Please comment on whether there is any risk that the person putting out the samples could have unintentionally influenced the results by handling or storing the positive and negative samples differently. The dogs may learn minor differences in packaging odour, storage location or handling (e.g. positives and negatives stored in different fridges may carry different environmental odours that the dogs learn).

Positive and negative samples were collected in the same way by the person. They were also stored in the same fridges. Therefore, the positive and negative samples coming from the same hospital (they were matched on hospital) were in the same environment from collection to sniffing by the dog. We clarified this point in the Material and Methods section.

- The ideal experimental design would be to run the trial double blind (the person putting out the samples does not know if they are positive or negative during the trial). If it was not double blind this needs to be mentioned as a study limitation.

Both the dog handler as well as the sample handler were blinded to the sample status (COVID-19 positive or negative). We added this point in the manuscript, in the testing protocol paragraph.

- Lines 292-293. Please comment on whether the individuals collecting data were blind to the sample allocations and whether they could have biased the results (ideally they should be blind too).

The data recorder was not blinded to the sample identity. However, he was located in the back of the room, where neither the dog handler nor the dog could see the data recorder when the dog was searching the positive sample. According to us, he could not have influenced the behavior of the dog. Furthermore, this absence of blindness allowed the data recorder to tell the handler whether the trial was successful or not, allowing to reward the dog each time it corrected marked the cone. This clarification was added in the testing protocol paragraph. We added this point in the Discussion section.

Lines 295-296. This sentence suggests that dogs had not previously been exposed to the samples used in the trials – please clarify if this is correct. Does this mean new samples were from new individuals not previously assessed by the dogs?

Yes, this is correct. We clarified this point in the testing protocol paragraph.

Results and statistical analysis

- Line 313. As above, if samples have been reused several times for the same dog this may be considered psuedoreplication of data. Please clarify if this was taken into account during the statistical analysis. Reuse of samples/individuals may result in detection based on learning of individual odours rather than COVID odour generalisation (it is not impossible that dogs could potentially learn 33 different individual samples)

We do agree that because one sample could have been used up to three times, it is not impossible, as the reviewer said, that the dogs used their olfactory memory to detect COVID-19 positive samples. However, we checked in the Parisian data the number of trials between two trials using the same COVID-19 positive sample. Here are the results for the 4 dogs regarding the COVID-positive samples:

Guess: 4 COVID+ samples were used once, 21 were used twice, and 2 were used 3 times.

Median number between two trials using the same COVID+ sample [p25-p75] (min-max): 7 [5-10] (3-22)

Maika: 3 COVID+ samples were used once, 20 were used twice, and 4 were used 3 times.

Median number between two trials using the same COVID+ sample [p25-p75] (min-max): 6 [4-10] (2-27)

Gun: 5 COVID+ samples were used once, 21 were used twice.

Median number between two trials using the same COVID+ sample [p25-p75] (min-max): 8 [4-12] (2-21)

Oslo: 11 COVID+ samples were used once, 10 were used twice.

Median number between two trials using the same COVID+ sample [p25-p75] (min-max): 3 [1-4] (1-9)

For three out of the four dogs (Guess, Maika, and Gun), the “time” interval between two identical COVID-19 positive samples seems enough to make the hypothesis that the olfactory memory played in minor role in the detection positive cones. For the last dog (Oslo), the high number of COVID+ samples used only once counterbalances the fact that the 1st quartile of the number between two trials using the same COVID+ sample is equal to 1.

Although we did not include these numbers in the manuscript, we added this important point about the repeated use of samples in the Discussion section.

Table 3: Results obtained on the dogs that finished the training period (n=8)

- Please state how many different positive samples (from different people) each dog was presented. It is not clear whether samples collected from the same individuals were reused more than once per dog. If they were reused, this needs to be included as a study limitation.

We indicate now the number of COVID-19 positive and negative individuals which were used for the study in the Results section (n=177). We also added in the testing protocol paragraph that one sample could have been used in up to three trials for one dog. We added this point in the Discussion section.

- Ideally it would be best to provide your underlying data (breakdown of trials for each dog showing positive and negative samples put out and response of the dog) as a supplementary file/download.

The data for Paris site are provided to the Editor. Because of the recent exploding event in Beirut, it was not possible to obtain the data from the testing sessions in Beirut.

Discussion

- As per my general comments above, I recommend that the discussion is revised to highlight that this study a promising first step, but you include a more critical discussion of the possible limitations and possible alternative explanations for the dogs’ performance. I think the current conclusion (lines 331-332) is too strong based on the methods and data presented, without a more robust double-blind and independently validated trial taking place.

We agree with the reviewer, and modified the Discussion section accordingly.

- Lines 359-382. Although interesting, I feel that the discussion of female hormone detection is moving too far away from the focus of the study subject and would be better deleted from the discussion (the evidence for the dogs indicating these samples due to female pheromones is also weak/anecdotal).

As suggested also by another reviewer, we deleted the four paragraphs regarding female hormone detection.

Conclusion

- As per my general comments above, I think the current conclusion (lines 417-419) is too strong based on the methods and data presented, without a more robust double-blind and independently validated trial taking place – suggest revision of wording in line with my general comments above.

We modified the Discussion section and the conclusion sentence.

Figures

- Figures 3. 4 and 11 are not readable due to poor resolution.

We improved the resolution of the figures mentioned above.

- Suggest delete figures 5 and 6.

We deleted these two figures.

Attachment

Submitted filename: Response to Reviewers v1.docx

Click here for additional data file. (70.4KB, docx)

Decision Letter 1

Nadine Ravel

29 Sep 2020

PONE-D-20-17256R1

Can the detection dog alert on COVID-19 positive persons by sniffing axillary sweat samples?

A proof-of concept study

PLOS ONE

Dear Dr. grandjean,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Two reviewers ar still requiring some clarification or some revision of your conclusions and importantly recognize some indications of the limits of your study.

Please submit your revised manuscript by Nov 13 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

  • A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

  • A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

  • An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Nadine Ravel

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

Reviewer #2: All comments have been addressed

Reviewer #3: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

Reviewer #3: Partly

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: I Don't Know

Reviewer #2: Yes

Reviewer #3: No

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The authors have considerably improved the manuscript, by providing significant information that was lacking in the previous version, and by improving the structure of the manuscript. For this kind of study, and especially due to the potential impact its results may have on the current management of the Covid-19 disease, it remains essential that the methodology is carefully discussed, so that the validity of the conclusions can be properly evaluated. In the current version, the additional details provided have elicited further questions about the validity of the conclusions:

1) The pairing between COVID-19 positive and COVID-19 negative odor donors represents a strong limitation in this study. Whereas the Lebanese groups are comparable, age differences between both groups in Paris is considerable: as body odor differs with age (to human noses as well as in terms of chemical composition: Mitro et al 2012 “The smell of age: perception and discrimination of body odors of different ages”, Chen & Haviland-Jones 2000 “Human olfactory communication of emotion”, Haze et al 2001 “2-Nonenal newly found in human body odor tends to increase with aging”), this may interfere with the dogs’ discrimination. The results suggest that it may introduce noise that alters the dog’s performance, since French dogs had lower success rates than Lebanese dogs. In future studies, targets and controls should at least be paired for sex and age. Another even more important limit is the fact that COVID-19 negative donors were partly people hospitalized for another reason than COVID-19. Meaning that they probably are on medication, and on medication that differs from the one taken by COVID-19 positive patients. How it may impair dogs’ performance or bias there responses towards the COVID-19 positive sample is currently difficult to predict in the absence of further information regarding patients’ treatment.

2) Mocks. The authors now explain that there were 3 or 4 cones in the lineups, including one COVID-19 positive sample and zero, one or two mocks. First, I assume this was zero or one for the 3-cones lineups, and zero to two mocks for the 4-cones lineups, i.e. there was always at least one COVID-19 negative sample in the lineup. Please confirm and write it explicitly in the text. Second, what was the rationale for using mocks? This is potentially a problem, since the test is much easier for the dog when they have 2 mocks plus 2 body odors (i.e., one positive and only one negative sample): they actually are discriminating between 2 and not 4 samples (the level of successfully answering by chance is much higher), since they were trained before to discriminate sweat samples from mocks (step 3 of the training).

3) Where was the dogs’ testing performed? Why did one dog have only 3 cones, and not 4 as the others?

4) Repetition of sample presentations during testing may also be an issue, as raised by the other reviewers. The authors response to this limit was not convincing: rather, they could test whether the performances increased on the second or third repetition compared to the first presentation. In fact, the results section is very short: the authors could very well test some of the possible bias arising from methodological flaws and present them here. For example, in addition to testing the effect of sample repetition, I suggest that they i) test whether lineups including only COVID-19 positive and healthy COVID-19 negative (no sick COVID-19 negative) elicit higher/lower performance, ot ii) test whether one sampling method provided better performances than the other. On the latter point, the fact that “two sampling supports were used because at the time to the beginning of the proof-of-concept study, we did not know which one of the two sampling supports was the most effective” should appear in the manuscript, otherwise no one will understand why you did so. This study being a proof-of-concept study, this is fine to mention that. And it would be nice to have some info regarding effectiveness of both methods.

5) Last paragraph of introduction: you mention the recommendations on how to proceed to rigorously test your research question but we expect to read how you study is positioned within these recommendations. This is currently not the case.

6) There are erroneous references to the bibliography. First, page 4 “sweat does not appear to be a route of excretion for SARS-CoV-2 virus [44, 45]”. In these papers, sweat is not mentioned. Instead, it appears that in the absence of scientific study on that topic, we don’t know whether sweat is able to convey the virus or not, see the following paper: Propper, R.E., 2020. Is sweat a possible route of transmission of SARS-CoV-2? Second, page 6: “SARS-CoV-2 virus does not survive more than a few hours on cotton and disposable gauze [52]. A more recent study concludes that absorbent materials like cotton and gauze are safer than no-absorptive materials for protection from SARS-CoV-2 infection [53].” References 52 and 53 are about SARS-CoV and not SARS-CoV-2, and about the comparison between virus survival on cotton gown versus disposable (polypropylene) gown. Reformulation is needed.

Overall, to make this study publishable, I would recommend to formulate the discussion and abstract even more cautiously. The conclusions are still too affirmative according to the numerous limitations. It should be explicitly stated that more research is need in the future, addressing a number of limitations present in this study – and then all limitations should be carefully listed. Also, the manuscript has clearly not been proofread by an English native speaker, as many language and wording imperfections remain.

Reviewer #2: Review of the Revision R1

The Authors have thoroughly addressed my remarks and comments, both in their responses to my review and in the revision. The plenty of changes to the manuscript makes the tracking of them hardly possible, thus the revision could be considered almost as a new manuscript. Obviously the revision substantially benefits from such considerable changes.

The Authors are aware that in order to avoid confounding factors, control samples must be in all aspects comparable to positive samples, except for the disease status. This was now evident from the lines 131-132, lines 142-145 and 350-365 of the revision.

In lines 371-373 of the revision the Authors properly address the problem of differences in the amount of sweat collected from COVID positive and negative donors, due to possible fever, which is a typical COVID-19 symptom. This issue was my main concern. In the revision the authors inform that the amount of sweat was not recorded and it can not be excluded that it could be a confounding factor. This statement is correct, however, further discussion in lines 374-378 including citation of Concha et al. (2019), is not quite pertinent to the problem I would like to point out. In the study of Concha et al. (2019) on detection threshold of amyl acetate, the dogs had to discriminate between amyl acetate diluted in mineral oil at different concentrations and pure mineral oil as control. There were defined qualitative chemical differences between target odour and control odour up to a dilution of amyl acetate which was for dogs not distinguishable from mineral oil. On the contrary, in the present study it is still not known which volatile organic compounds in the sweat of COVID-positive patiens make that the dogs are able to discriminate the sweat odour of COVID positive from negative humans. Also, the saturation of olfactory receptors by a great amout of the odorant (revision lines 377-378), is not pertinent to the problem of the possible differences in the amount of sweat in the samples. I wanted only to point out that potential differences in the amount of sweat could be a confounding factor and/or could be a cue for dogs. I suggest to delete the passage in revision lines 374-378.

An important information was given only in the response to my review, namely that the samples contained approximately 75 mg sweat and no differences were found between positive and negative samples e.g. due to more sweating in fever. In order to avoid doubts of the readers, this information should be given not only in responses to the review, but first of all in the revision, in the paragraph dedicated to sample collection,.

In the revision the Authors have resigned from the 4 Ajaccio dogs and 21 sweat samples collected in Ajaccio. This was convincingly justified in Authors’ response to the Reviewers #1 and #3. As to the number of dogs used, the Authors in their explanations to the Reviewers #3 inform that now they used six dogs (for the real tests) and they did not mention anymore the number of 14 dogs used for the training. The Authors now explain that „…it added a non relevant information…”. However, in my opinion it is just an importat information, because it shows that not all dogs are able to meet the training or qualification criteria for this task. An information on how many dogs out of all candidates that started the training, were qualified for the test, would be useful for other researchers or dog trainers who will intend to continue or to repeat such experiments.

As to the qualification criterion, the Authors’ information that „…The dog was considered as trained and ready for the testing session by the subjective opinion of the handler, based on the behavior of the dog….” – is acceptable, though for the sake of scientific accuracy in future experiments more precise quantitative criteria would be recommended.

It is not clear to me why the Authors think the calculation of the sensitivity and specificity was not possinle ? (revision lines 276-277).

It depends on how the outcome of a trial in the lineup is assessed. The authors apparently considered their lineup as a simultaneous lineup, where the dog makes a choice out of 3 or 4 samples. This approach seems to be more „convincing” because the theoretical probability of the correct indication by chance is relatively low (1/4 or 1/3). However, the actual probability of correct choices by chance depends on how many samples the dog had really sniffed. The not sniffed samples do not count for calculation of correct hits by chance. Therefore the Authors are not correct, claiming in their explanations that this probability is still 25% if the dog hits correctly on the first cone, without sniffing the remaining cones. Thus, the probabilities given in the last right-hand column of the Table 2 are not constant but variable from trial to trial.

The other approach is a sequential lineup where the dogs are considered to make separate and independent yes/no decisions to each sniffed sample, at the constant 50% probability of correct hits by chance. The sequential lineup is believed to be statistically more sound. If the Authors would consider their lineup as the sequential one, all sniffed but not indicated cones with negative samples are „true negative”, all correct hits on positive samples can be regarded as „true positive”, and the indications labelled in Authors’ protocols as “failures” - are „false positive”. Since in the present protocol „…the dog was previously trained to detect one and only one COVID-19 positive cone, the dog was left free to sniff the cones before marking one cone….” and „….Once a cone was correctly or incorrectly marked, the trial was stopped..(revision lines 258-260) – it means that in the protocol like this, there were no „false negatives” (f.n.=0). Assuming that f.n. =0, the calculated sensitivity would be 100% [Ʃ true positive / Ʃ (true positive + false negative). The specificity for some dogs would be <100% [= Ʃ true negative / Ʃ (false positive + true negative)]. However, from the statement in revision lines 258-260, one can suspect that there was a bias toward reducing „false negative” responses (a miss to indicate the target sample), since the handler knew, and the dogs were trained, that in each trial one of the samples must be the target to be indicated. A problem would be when it comes to testing samples of unknown health status in real scenario. The way to avoid this problem would be either to introduce from time to time a „zero” or „blank” trial (with all negative samples in the lineup), or to set a limit for how many times the dog is allowed to sniff positive samples without indicating, before it eventually hit them. For example if the dogs has sniffed the positive sample 2 or 3 times without indicating, the outcome can be classified as the „false negative”.

Since the aim of the study was only to confirm that dogs are able to discriminate COVID positive and negative sweat samples, the term „Success rate” used by the authors, is acceptable, but for the sake of scientific accurracy in future studies the Authors should be able to calculate the sensitivity and specificity.

It is worth underlining that the revision is better structured than the first submission. The concluding remarks are more cautious and the references are updated.

Minor remarks

The terms „odour detection” and „odour discrimination” are frequently used interchangeably, however, in the context of this work I suggest to use the wording „odour discrimination” rather than „odour detection”. Odour detection refers rather to searching for unknown location of odour source in a field setting. In the odour lineup like this, the dogs are trained to visit the odour locations (cones) which are visible, and their task is to compare and to discriminate between the target and control odours.

In lines 344 and 352 the year of the publication should be completed for more clarity. This is particularly important when two or more papers of the same first author are cited.

Linie 373 „COVID-19 negative” should be inserted for more clarity.

To sum up, only few amendments (minor revision), as listed above, should be made to the text to be fully acceptable for publication in the PLoS ONE.

Reviewer #3: Thank you to the authors for taking into account the previous comments from the three reviewers, providing very detailed responses to the points raised. The revised manuscript is much improved, provides a more balanced appraisal of the study and is much clearer to read. Some of the additional information that has been added has raised further questions and comments, detailed below, but I think the manuscript would be appropriate for publication if these can be addressed.

General comments

- The general use of English language and grammar has greatly improved in this revision. However, there are still a few typographical errors and language issues in places that need addressing with a thorough proof reading.

- Throughout the manuscript (title, abstract, discussion etc.) suggest you clarify that you were testing a discrimination of *symptomatic* SARS-CoV-2 positive vs asymptomatic SARS-CoV-2 negative. This is an important distinction, as symptomatic individuals may produce confounding odours for a wide variety of reasons.

Abstract

- Line 49 “The aim … was to evaluate if a sample of sweat produced by COVID-19 50 positive individuals (SARS-CoV-2 positive) can be detected by trained detection dogs among 51 sweat samples produced by COVID-19 negative individuals.” This wording suggests that dogs were evaluated using just a single (one) SARS-CoV-2 PCR positive sample. The line needs clarifying; I think you mean to say something like: “evaluate if dogs could detect individual samples of sweat produced by COVID-19 50 positive individuals (SARS-CoV-2 PCR positive) presented in a line up amongst sweat samples produced by COVID-19 negative individuals”?

- I think you need to state in the abstract that there are limitations to the study (e.g. processes for collecting positive and negative samples were not completely identical, the trial stage was not fully double blind, some samples were re-used several times during the trial stage) and that a more rigorous validation study is ideally needed to confirm the results.

Introduction

- For balance, I think you should mention that not all studies testing canine detection of disease have shown positive results. There is currently very little evidence relating to canine detection of virus odours, particularly using live subjects (you cite one example using cell cultures).

Materials and Methods

Samples

- Line 157. I think you need to briefly discuss in your discussion section/conclusions the possible limitation that SARS-CoV-2 positive samples were collected using full PPE whereas negatives were not. It could have produced a consistent difference in odour that confounded the results. Were the same brands of gloves used for collecting positive and negative samples? If different brands/types were consistently used for positives and negatives, it’s a major limitation as the dogs may have trained on different glove odours.

- Line 177. Were the two coolers identical (same brand, from the same source etc.?). Worth stating if this was the case. If not, it’s a limitation as dogs may have trained on the odour of two different coolers.

Safety of the dogs regarding SARS-CoV-2 virus infection

- Line 216. Should read “non-absorptive”

Testing protocol

- Line 248. Please clarify the minimum number of SARS-CoV-2 negative samples that were placed out each time. If zero negatives samples were placed out, that would not be valid discrimination data and should be excluded from the analysis/results. If 1, 2 or 3 samples were placed out, please clarify how this was accounted for in the statistical analysis.

Statistical analysis

- Line 275. As per the abstract, this wording suggests that dogs were evaluated using just a single (one) SARS-CoV-2 positive sample. The line needs clarifying.

- As mentioned in the testing protocol comment, some additional detail is needed on how the statistical analysis accounted for the different number of negative samples placed out. If some line-ups included zero negatives, these trials should be excluded from the analysis. It is vital that the analysis is based on the number of negatives placed out, not the number of cones. Clearly, dogs would have a greater chance of detecting the positive samples with lower numbers of negatives present. If not all the cones contained negatives the statistical analysis needs to be revised to account for this.

Results

- Line 292. The ages are the wrong way around?

Table 2, as stated above, the random choice proportion should be based on the number of negatives placed out, not the number of cones (if not all cones (expected the positive) contained SARS-CoV-2 negative samples).

Discussion

- Line 335. This line needs revising to reflect that only positive and negative samples (not “mocks”) should have been included in the statistical analysis.

- line 389 to 396. Recommend you clarify that the lack of double blinding remains a limitation that should ideally be addressed in a future validation study. Previous studies have used various techniques to reward dogs whilst remaining double blind (e.g. use a live mobile phone conversation with an independent referee who has a list of sample “identities” for each line up). The ideal validation study would be to receive samples from an independent source (e.g. from another organisation or even country) which have not been processed by your research team before being presented to the dogs.

- 411 to 416. Recommend key improvements for the validation trial should include double-blinding and samples prepared and delivered by an independent organisation/source.

- Line 412. Should read ‘sensitivity’ not ‘sensibility’.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2020 Dec 10;15(12):e0243122. doi: 10.1371/journal.pone.0243122.r004

Author response to Decision Letter 1

29 Oct 2020

Dear Editor,

Please find the revised version of our manuscript entitled “Can the detection dog alert on COVID-19 positive persons by sniffing axillary sweat samples? A proof-of-concept study”. It has been modified according to the comments of the three reviewers, and has been revised by a native English speaker.

We do hope that this version adequately answers the points raised by the reviewers.

Furthermore, I would like to modify the order of the co-authors, by placing Loic Desquilbet as the last author because of his significant investment in the writing of the manuscript as well as in the statistical analysis and interpretation of the results. I would like to place Jean-Pierre Tourtier as the co last-author if possible, or as the second author if it is not.

With my kind regards,

Prof. Dominique Grandjean

Attachment

Submitted filename: Response to Reviewers (Revision #2) v2.docx

Click here for additional data file. (78.2KB, docx)

Decision Letter 2

Nadine Ravel

17 Nov 2020

Can the detection dog alert on COVID-19 positive persons by sniffing axillary sweat samples?

A proof-of concept study

PONE-D-20-17256R2

Dear Dr. grandjean,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Nadine Ravel

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #3: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: In this second revision of their manuscript, the authors have answered all my comments in a satisfactory manner. To me, the paper is now acceptable providing some minor changes as detailed below:

1) Ethical statements (page 5) should be formulated as proper sentences (or as such but in a separate section of the manuscript before the references, according to the editor’s preference).

2) Figures 1 and 4 are not required within the manuscript, they may appear either as Appendices at the end of the manuscript or as supplementary materials, again according to the editor’s preference.

3) As early as in the method section, it should be stated explicitly that, within a trial in a line of cones, samples are from the same hospital (are they?) and are made of the same material (gauze swabs or gauze filters). Line 190: it may be useful to explicitly say that polymer tubes were not used in the present study but collected for chemical analysis not presented here.

4) There seem to be a contradiction between Line 193 stating an average quantity of sweat secreted and Line 441 stating that amount of sweat was not recorded. Please address this issue.

Reviewer #3: Thank you to the authors for revising the manuscript and taking into account the comments of the three referees. Overall, I feel the manuscript has been improved again and now provides a clear and balanced discussion of the study and the possible limitations. I have made a few minor comments below, but otherwise would be content for it to be accepted/published – many thanks.

Minor comments:

Introduction

Line 82: The wording “was good compared with standard diagnostic methods” should be clarified (e.g. to “was comparable with” or “was better than” or “was more accurate than” – whichever term explains the comparison more clearly).

Line 126: Suggest change to VOCs “_may_ be detectable by the dogs” (we cannot say for certain that they “are”).

Methods

Line 191: You say two sampling methods were used but three are discussed above (gauze swabs, gauze filters, polymer tubes)?

Discussion

Lines 459-461: “Although the data recorder was not blinded to the sample location, he remained at the back of the room and could not be seen by either the dog handler or the dog during the session. Therefore, he could not have influenced the dog’s behaviour.”

Because the data recorder put out the samples there is always a risk he/she unintentionally influenced the dogs’ choices. There are many theoretical ways this could happen (e.g. they accidentally touched the COVID positive samples more than the negatives so they smell more strongly of the data recorder; they always put the positive samples out first and they had more time to release odour than the negatives etc. etc.). Fundamentally, if the team knows the sample locations there always the risk they may influence the dogs in some way they do not realise and cannot possibly predict – this is the basis of the Clever Hans Effect. That is why double-blind testing is so important for validation of these studies. I suggest you change the text to clarify that standing at the back of the room reduced the chance of the data recorder directly influencing the dog’s behaviour, but the risk of them having some unintentional influence on the choice of samples cannot be completely ruled out (without the double-blind validation study).

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #3: No

Acceptance letter

Nadine Ravel

24 Nov 2020

PONE-D-20-17256R2

Can the detection dog alert on COVID-19 positive persons by sniffing axillary sweat samples? A proof-of-concept study

Dear Dr. Grandjean:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Nadine Ravel

Academic Editor

PLOS ONE

Associated Data

    This section collects any data citations, data availability statements, or supplementary materials included in this article.

    Supplementary Materials

    S1 Data

    (XLSX)

    Click here for additional data file. (52.3KB, xlsx)
    Attachment

    Submitted filename: Response to Reviewers v1.docx

    Click here for additional data file. (70.4KB, docx)
    Attachment

    Submitted filename: Response to Reviewers (Revision #2) v2.docx

    Click here for additional data file. (78.2KB, docx)

    Data Availability Statement

    Data are provided as Supporting information file.

    Articles from PLoS ONE are provided here courtesy of PLOS

    RESOURCES