SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

OBJECTIVES

When cats purr during examination it is difficult to perform auscultation. The objective of this study was to determine the prevalence of purring in cats during examination, and identify interventions that could be used to stop purring.

METHODS

Cats seen at a first opinion clinic were enrolled in the study and their purring status recorded. Thirty of the purring cats were exposed to up to three different interventions in an attempt to stop purring in a randomised controlled trial including blowing at the ear, use of an ethanol-based aerosol near the cat and proximity to a running tap.

RESULTS

The 30 cats in the trial were subjected to a total of 54 attempts to stop purring, proximity to a running tap caused 17 of 21 (81%) cats to stop purring, blowing at the cat's ears worked in 2 of 15 (13%) cats, spraying an aerosol close to the cat was effective in 9 of 18 (50%) cases. In 2 cats (7%), none of the interventions interrupted purring.

CLINICAL SIGNIFICANCE

This study provides evidence that placing a purring cat near a running tap and in proximity to the discharge of an ethanol-based aerosol are effective measures to stop purring in order to allow auscultation.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Purring is defined as a soft buzzing sound created by vibrations within the upper respiratory tract which varies rhythmically with breathing (Remmers & Gautier 1972, Frazer-Sissom et al. 1991). The precise behavioural role of purring has not been established although it is generally believed to be a form of communication. Several species of felids vocalise in this way (Stogdale & Delack 1985, Turner & Bateson 2000, Peters 2002). Purring is often cited as a sign of contentment in the domestic cat; on the other hand it has been observed that injured, sick or frightened cats may also purr, so the precise function of this phenomenon remains an enigma. Cats sometimes purr whilst undergoing physical examination at a veterinary practice, but no data have yet been published describing the prevalence of this behaviour. The sound of purring is a distraction when performing thoracic auscultation in this species, and important clinical abnormalities such as heart murmurs and cardiac gallop sounds (both of which are brief, low frequency sounds, often of quite low intensity) can be masked by purring. Purring may also hamper the measurement of heart rate and recognition of dysrhythmias during auscultation. Similarly, purring may prevent the effective auscultation of the respiratory system.

The purpose of this study was to establish the frequency with which cats purr during routine thoracic auscultation in veterinary clinical practice, and to look for possible associations with purring. A further aim was to identify an effective and safe intervention that could be used to halt purring of cats during a clinical examination. Such an intervention may be of use to clinicians in the course of their clinical work.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Pilot study

A number of veterinary clinicians were approached to describe any tactics, which they believed could be effective in interrupting or stopping a cat from purring during physical examination. A search of the Internet was also used to identify further potentially useful methods. A list of these methods is shown in Table 1 together with a description of each method. Informal pilot studies were carried out by the investigators during which purring cats were exposed to several of these tactics (up to a maximum of three tactics per cat).

Table 1. Tactics used to stop cats purring
TacticDescription
  1. A description of the tactics used to try to stop cats from purring used in the pilot study. Blowing at the ears, a running tap and the use of an aerosol were selected for evaluation in the randomised controlled trial

Blow at the earsGently blow at the ears of the cat with a force to cause the guard hairs at the cavum concha to bend slightly.
Run a tapHold the cat within 0 · 5 to 2 m of a running tap (faucet). The tap should be running at a rate which could fill a cup of water in less than 10 seconds but more than 2 seconds.
“Squeak” a toySqueak a high-pitched children's or dogs’ toy within 0 · 5 to 2 m of the cat.
Touch the noseTouch the bridge of the cat's nose with a finger using a similar pressure to that used on the keypad of a touch-screen mobile phone.
Expose to surgical spirit vapourHold a swab dipped in surgical spirit within 10 cm of the nose of the cat.
Aerosol sprayWith the nozzle of the spray directed away from the animal, spray an aerosol onto the table within 15 cm of the face of the cat.

The results indicated that neither touching the nose of the cat nor squeaking a squeaky toy close to the cat appeared to be effective as they interrupted or stopped the cat from purring in only 1 of 10 or 2 of 10 cases, respectively. No further investigations were performed using these tactics. When purring cats were exposed to a swab soaked in surgical spirit, 3 of 10 stopped purring. However in one of these cases, the cat immediately became quite excited and difficult to handle and, consequently, no further use was made of this tactic. In these pilot tests, the three other tactics (blowing at the cat's ears, running water from a tap in the vicinity of the cat and the use of an alcohol-based aerosol spray sprayed onto the examination table) appeared to be more effective ways to stop or interrupt cats from purring as they worked in 3 of 10, 7 of 8 and 5 of 9 cases, respectively. An aerosol containing ethanol (Feliway; Ceva Animal Health) was effective in 4 of 9 cats when compared with an aerosol containing isopropanol (Adaptil; Ceva Animal Health, formerly known as DAP) which was effective in only 1 of 9. The aerosols were not chosen on the basis of their active ingredient but on the basis of the alcohol carrier (ethanol versus isopropanol) using products that would be likely to be found in a veterinary surgery. Four of the nine cats failed to stop purring using either aerosol.

The results from this pilot study led to the selection of the interventions tested in the clinical trial described below.

Purring prevalence study

This study was performed by CJLL over a period of approximately 1 year, over three continuous blocks of time (lasting for 1 to 3 months each); all cats presented as first opinion cases were included in the study. The sex, breed, age, reason for presentation and purring status of each cat were recorded. The illness status was determined from the reason for presentation. Cats presented for vaccination and check-ups following the resolution of an illness were determined to be “healthy” and all other cats presenting with a variety of clinical problems were determined to be “ill”.

Randomised controlled trial of methods to stop purring

A randomised controlled trial was designed to test the efficacy of the three most promising methods to halt purring, identified from the pilot studies. Blowing at the ear was performed by blowing with a force sufficient to produce visible movement of the guard hairs at the cavum concha from a distance of approximately 15 cm. Holding the cat near to a running tap was achieved by turning on a tap in a sink approximately 1 · 4 m from the centre of the consulting table on which the cat was located with the tap running at a rate that could fill a 250 mL cup in less than 10 seconds but more than 2 seconds. The cat was then carried over to the sink to within 0 · 5 m of the tap. Purring sometimes stopped as the tap was turned on and sometimes only as the cat was close to the tap. The range of proximities of the cats to the tap was approximately 0 · 5 to 2 · 0 m. The aerosol test was performed by spraying Feliway directly at the consultation table 15 cm away from the cat (Table 1). The trial was performed in a conventional small animal consulting room with the owners present. The interventions were tested during the conduct of a conventional veterinary clinical examination.

Table 2. Results of an analysis of factors on the likelihood of purring
 P valueOROR 95% CI
  1. The results from a univariate binary logistic regression analysis on the effect of pedigree status, reason for presentation (ill versus not ill), age and sex on the likelihood (odds ratio) of a cat purring. The P value, odds ratio (OR) and the 95% confidence limits (CI) on the odds ratios are listed alongside each factor

Pedigree cat = Yes0 · 840 · 930 · 46 to 1 · 88
Pedigree cat = No (reference category) 1 
Ill cat = Yes0 · 770 · 910 · 48 to 1 · 72
Ill cat = No (reference category) 1 
Effect of age (per year)0 · 141 · 050 · 98 to 1 · 12
Male neutered0 · 630 · 670 · 13 to 3 · 54
Female neutered0 · 770 · 770 · 14 to 4 · 22
Male entire0 · 651 · 60 · 21 to 11 · 90
Female entire (reference category) 1 

A stratified, modified three-dimensional semi-crossover design was used (Carlsen et al. 1993, Hektoen et al. 2004). Block randomisation was used to create a list of the interventions to be used on each cat. To ensure cats received the three interventions without bias in regard to the order of treatment, and to ensure that each of the three treatments were tested equally, sequences of interventions were created consisting of a random order of the six possible permutations (i.e. if the treatments were a, b and c: permutations are abc, acb, bca, bac, cba and cab). This randomisation procedure was repeated five times to create a final list of intervention sequences to be attempted on the 30 cats (included in Table 3).

Table 3. Full dataset of results from the randomised controlled trial
CatDescriptorsSequenceFirst testSecond testThird test
  1. ND Not done, M Male, MN Male neutered, F Female, FN Female neutered, DSH Domestic short haired

  2. The complete dataset of results from the randomised controlled trial listing the sequence number of the cats, descriptors (their age, breed and sex), the predetermined sequence of tactics applied (1 = aerosol-tap-blow, 2 = aerosol-blow-tap, 3 = tap-blow-aerosol, 4 = tap-aerosol-blow, 5 = blow-aerosol-tap, 6 = blow-tap-aerosol), the sequence of tactics used, and the results

111 years, DLH, MN6“Blow”“Tap”“Aerosol”
FailedFailedFailed
26 months, DSH, FN2“Aerosol”“Blow”“Tap”
FailedFailedWorked
36 months, DSH, FN3“Tap”“Blow”“Aerosol”
WorkedNDND
4Aged, DSH, FN1“Aerosol”“Tap”“Blow”
WorkedNDND
59 years, Siamese, FN5“Blow”“Aerosol”“Tap”
FailedWorkedND
618 years, DSH, MN4“Tap”“Aerosol”“Blow”
FailedFailedFailed
710 years, DSH, MN3“Tap”“Blow”“Aerosol”
WorkedNDND
8Aged, DSH, MN2“Aerosol”“Blow”“Tap”
FailedFailedWorked
912 years, Norwegian forest Cat, MN6“Blow”“Tap”“Aerosol”
FailedWorkedND
1012 years, Persian, FN.1“Aerosol”“Tap”“Blow”
WorkedNDND
111 year, DSH, MN5“Blow”“Aerosol”“Tap”
FailedWorkedND
129 years, DSH, FN4“Tap”“Aerosol”“Blow”
WorkedNDND
1315 years, DSH, MN4“Tap”“Aerosol”“Blow”
FailedWorkedND
1415 years, British short hair, MN2“Aerosol”“Blow”“Tap”
WorkedNDND
1515 years, DSH, MN6“Blow”“Tap”“Aerosol”
WorkedNDND
1610 years, DSH, FN3“Tap”“Blow”“Aerosol”
FailedWorkedND
1714 years, DSH, FN5“Blow”“Aerosol”“Tap”
FailedFailedWorked
1815 years, DSH, MN1“Aerosol”“Tap”“Blow”
FailedWorkedND
191 year, DSH, F entire.4“Tap”“Aerosol”“Blow”
WorkedNDND
207 years, Burmese, MN6“Blow”“Tap”“Aerosol”
FailedWorkedND
219 months, DSH, F entire.3“Tap”“Blow”“Aerosol”
WorkedNDND
223 years, DSH, MN1“Aerosol”“Tap”“Blow”
WorkedNDND
23Aged, DSH, MN5“Blow”“Aerosol”“Tap”
FailedFailedWorked
244 months DSH, M entire.2“Aerosol”“Blow”“Tap”
WorkedNDND
253 years, DSH, FN6“Blow”“Tap”“Aerosol”
FailedWorkedND
2616 years, DSH, MN3“Tap”“Blow”“Aerosol”
WorkedNDND
2712 years, DSH, MN5“Blow”“Aerosol”“Tap”
FailedWorkedND
2810 years, DSH, FN2“Aerosol”“Blow”“Tap”
FailedFailedWorked
29Unknown (aged), DSH, FN1“Aerosol”“Tap”“Blow”
FailedWorkedND
3010 years, DSH, MN4“Tap”“Aerosol”“Blow”
WorkedNDND

Statistical analysis

A power calculation was performed to estimate the number of subjects required for the randomised controlled trial in order to have an 80% power (probability of obtaining a statistically significant result) using an Internet-based calculator (Lenth 2006–2009). For a chi-squared test on a 3 × 2 contingency table (3 tests, positive/negative result, 2 degrees of freedom) a chi-square value of 5 · 99 would equate to P = 0 · 05 and require n = 49. As each cat would provide up to three test results an estimate of the number of cats required was made on the assumption that on average the interventions would work 50% of the time, and so 30 cats would be subjected to the first test, 15 of 30 to a second test, and 7 of 15 cats to a third test leading to a total of 52 test results.

Analysis of the potential association of factors with the purring phenomenon in cats from the prevalence study was undertaken using univariate binary logistic regression. The results of the randomised controlled trial were analysed using a Fisher's exact test for analysis of first attempts (for which independence could be assumed). Univariate binary logistic regression was used to analyse the full set of trial results using attempt number and intervention as the predictor variables. The level of statistical significance chosen was P < 0 · 05. Statistical analyses were performed using SPSS v.20, IBM Inc, New York, USA.

Ethical considerations

This study was not submitted for formal ethical review although advice was sought to determine if formal review was required. No cat was subjected to any intervention that went beyond what might reasonably be encountered during a routine veterinary examination, and no diagnostic or treatment option was influenced by any cat's involvement in this study. Client consent was obtained verbally during the consultations during which data was collected. No client information was included in any study records.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Prevalence study

Of the 341 cats examined, 62 were purring at the time of auscultation (18 · 2%). There were 180 male (15 entire, 165 neutered) and 161 female (25 entire, 136 neutered). One hundred and ninety-seven were brought to the practice because of illness and 144 for routine health visits or vaccinations. The majority of cats were domestic with shorthair or longhair (295). The remaining breeds were Persian (n = 11), British shorthair (n = 10), Siamese (n = 6), Burmese (n = 5), ragdoll (n = 5), Bengal (n = 3), Maine coon (n = 2) and one each of Tonkinese, Balanese, Norwegian forest and Turkish van.

The mean recorded age of the cats for which a reliable age was available, n = 296 (87%), was 8 · 2 years (range 0 · 5 to 19 years).

The results of a binary logistic regression analysis of age, sex, breed, pedigree status and illness status versus purring status failed to identify any of these factors as having a statistically significant association with purring (Table 2).

Randomised controlled trial

The results of the randomised controlled trial on interventions to stop purring are shown in Fig 1. On the basis of the first attempt only the success rates for the three interventions were: blowing 1 of 10 (10%), aerosol 5 of 10 (50%) and tap 7 of 10 (70%). A Fisher's exact test of these results gave a P value of 0 · 035 indicating that the differences between the three interventions were statistically significant. Based on the results of all attempts success rates were: blowing 2 of 15 (13%), aerosol 9 of 18 (50%) and tap 17 of 21 (81%). Because these results included multiple test results from the same cat statistical analysis was performed using binary logistic regression looking at the effect of cat, attempt number and the intervention on cessation of purring. The cat and attempt number effects were not statistically significant. The effect of the interventions yielded a P value of 0 · 003. Using “tap” as a reference intervention the odds ratio for the likelihood of stopping purring with “blow” was 0 · 039 with a P value of 0 · 001 and the odds ratio for “aerosol” was 0 · 26 with a P value of 0 · 069. Two out of a total of 30 (7%) purring cats failed to stop purring with any of the interventions. The remaining 28 cats ceased purring in response to one of the interventions attempted. Thirteen (46%) cats stopped purring with the first intervention attempted, 10 (36%) with the second intervention attempted and 5 (18%) with the third.

image

Figure 1. Success and failure rates for the three tactics tested in the randomised controlled trial. A total of 54 attempts to stop purring were made for the 30 cats included in the trial

Download figure to PowerPoint

The distribution of the attempt numbers for each of the interventions tested is shown in Fig 2. The success and failure rates for each tactic broken down by sequence number are shown in Fig 3.

image

Figure 2. Each tactic was used as the initial test for the same number of cats. For cats which failed to cease purring with the first intervention then a second predetermined tactic was used to try to stop the cat from purring. If that tactic also failed then the remaining intervention was used. This design means that effective tactics are tested more frequently than ineffective tactics. Table 1 shows all the results from the randomised controlled study

Download figure to PowerPoint

image

Figure 3. The results of the randomised controlled trial showing the numbers of successes and failures for each tactic tested broken down by attempt number. Success rates do not appear to vary significantly with attempt number

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

With all the interventions attempted the cat was held in light restraint (i.e. held gently); there was no systematic difference in the way the cats were held or carried associated with the different interventions although there was inevitably some undocumented variation.

Thousands of cats must be examined every day by veterinarians all over the world. If a cat is purring during an examination it becomes a challenge for the clinician to obtain even basic information such as heart rate, or to identify heart murmurs, dysrhythmias, or the presence of more subtle signs such as adventitious respiratory sounds or cardiac gallop sounds. It is surprising that no studies have been published describing the frequency with which cats purr whilst they are being examined. Nor, to the authors’ knowledge, have studies been performed to establish tactics that can be used to interrupt or stop a cat from purring. This study was performed to bridge that gap.

The findings demonstrated that in a single veterinary clinic, 18% of 341 cats presented for clinical examination were purring. Purring in the consulting room occurred in a substantial proportion of cats apparently irrespective of their health, age or sex. A larger study, or analysis of other variables, might have been able to identify some association but a tendency to purr may well just reflect the personality or demeanour of any particular cat.

The study design employed was relatively complex but reflects what might reasonably be expected to occur in a practice setting (i.e. if the first tactic fails to halt purring then try another tactic). This design complicates the statistical analysis as a simple chi-squared or Fisher's exact test cannot be used when tests are repeated on the same cat although these tests could be used to analyse the first attempts. Randomisation should ensure that any cats that spontaneously stopped purring, or cats that had an increased likelihood of stopping, were equally distributed among the three intervention groups. In an ideal world this trial would have used a negative control in order to establish the rate of spontaneous cessation of purring, but this was not one of the hypotheses of this study. It was not feasible to blind this trial but it is reasonable to assume that the investigator had no particular reason to favour a particular intervention. The fact that this was performed by a single veterinary surgeon in a single practice may reduce the external validity of the results, although it also ensures that the influence of operator or environmental variation on the outcomes was less than in a multi-centred study. The study was designed to minimise the bias that might occur due to a particular intervention being used more as a second or third attempt to stop purring. Figure 2 shows that each of the tactics was used first on an equal number of occasions. However, due to the different efficacies of the tactics there is some variation in the number of times each tactic was used as a second or third intervention, and inevitably a more successful intervention will end up being tested more frequently than a less effective one. It is also possible that there were other factors associated with the use of the interventions, that were not recorded, that could have accounted for the differences in their efficacy. For example, if the amount of handling or the type of handling was consistently different for each intervention this may be the explanation for the difference between interventions. Similarly the choice of the aerosol used may be important. The hypothesis was based on the possible aversion to the use of any ethanol-based aerosol but it cannot be discounted that the active ingredient (a pheromone in this case) may have been responsible for the effect observed.

The overall success rate in stopping purring in cats using one or more of the tactics under test was 93% (28 of 30). The most successful tactic was to hold the cat fairly close to a running tap which was successful in 17 of 21 of cats (81%) compared to use of the aerosol in 9 of 18 (50%) or blowing in 2 of 14 (14%). The statistical analysis of the first interventions attempted showed that the differences between the success rates were significant but did not show that the use of the tap was statistically significantly different from the use of the aerosol. There is no clear evidence that the use of other tactics previous to a successful intervention had any effect on the outcome (Fig 3) as the success rates broken down by attempt number appear similar for each tactic for each of the attempt sequence numbers and in a binary logistic model there was no statistically significant effect associated with attempt number. This analysis also enabled an investigation for the statistical significance of the individual interventions and showed that proximity to a running tap was significantly different to blowing at the ear (P = 0 · 001) but not significantly different to use of an aerosol (P = 0 · 057, a possibility of this result occurring by chance of 1:18). The inclusion of more cats in the trial would have resolved this uncertainty. It is unclear why a running tap works, and it has also not been established whether this has any negative consequences. It is possible that cats find the sound and sense of a running tap distracting, perhaps because of their acute sense of hearing or possibly because of a dislike of running water observed in most breeds and individuals. There does not appear to be a good explanation as to why the sound of a running tap should be more effective than blowing at the ear, or proximity to an ethanol-based aerosol spray. If it is just the sound of running water then a sound recording played to the cat might be equally effective as the real thing. This supposition would need to be tested in a further study, but if it were effective this could perhaps be incorporated into a “smart phone” application for veterinarians.

Heart disease is common in cats and it can be a diagnosis that is difficult to make with certainty (Stepien 1998). In contrast to dogs, many healthy cats have heart murmurs on auscultation (Cote et al. 2004, Wagner et al. 2010). Furthermore, cats with heart disease or even heart failure may have no discernible heart murmurs (Ferasin et al. 2009, Paige 2009, Smith & Dukes-McEwan 2012). In the face of this some clinicians might be tempted to ignore auscultation when a cat is presented for examination if it is purring. This would, however, be unwise as auscultation should enable the examiner to assess and establish clinical parameters such as heart rate and to identify amongst others the presence or absence of dysrhythmias, gallop sounds and the respiratory sounds. The study presented here shows that running a tap nearby, or failing that, spraying an ethanol-based aerosol onto the examination table near to the animal will induce most cats to stop purring for a sufficient time period to enable the clinician to perform a satisfactory auscultation on that animal and that this should be successful in 9 of 10 cats.

Conflict of interest

None of the authors of this article has a financial or personal relationship with other people or organisations that could inappropriately influence or bias the content of the paper.

References

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. References
  • Carlsen, K. H., Kramer, J., Fagertun, H. E., et al. (1993) Loratadine and terfenadine in perennial allergic rhinitis. Treatment of nonresponders to the one drug with the other drug. Allergy 48, 431436
  • Côté, E., Manning, A. M., Emerson, D., et al. (2004) Assessment of the prevalence of heart murmurs in overtly healthy cats. Journal of the American Veterinary Medical Association 225, 384388
  • Ferasin, L. (2009) Feline myocardial disease 1: classification, pathophysiology and clinical presentation. Journal of Feline Medicine & Surgery 11, 313
  • Frazer-Sissom, D. E., Rice, D. A. & Peters, G. (1991) How cats purr. Journal of Zoology (London) 223, 6778
  • Hektoen, L., Larsen, S., Ødegaard, S. A., et al. (2004) Comparison of homeopathy, placebo and antibiotic treatment of clinical mastitis in dairy cows – methodological issues and results from a randomized-clinical trial. Journal of Veterinary Medicine Series A 51, 439446
  • Lenth, R. V. (2006–2009). Java Applets for Power and Sample Size [Computer software]. http://www.stat.uiowa.edu/~rlenth/Power Accessed March 13, 2013
  • Paige, C. F., Abbott, J. A., Elvinger, F., et al. (2009) Prevalence of cardiomyopathy in apparently healthy cats. Journal of the American Veterinary Medical Association 234, 1398403
  • Peters, G. (2002) Purring and similar vocalisations in mammals. Mammal Review 32, 245271
  • Remmers, J. E. & Gautier, H. (1972) Neural and mechanical mechanisms of feline purring. Respiratory Physiology 16, 351361
  • Smith, S. & Dukes-McEwan, J. (2012) Clinical signs and left atrial size in cats with cardiovascular disease in general practice. Journal of Small Animal Practice 53, 2733
  • Stepien, R. (1998) “Specific feline cardiopulmonary conditions. In: Manual of small animal cardiorespiratory medicine and surgery. Eds V. Luis Fuentes and S. Swift. BSAVA, Cheltenham, UK. pp 254257
  • Stogdale, L. & Delack, J. B. (1985) Feline purring. Compendium of Continuing Education for the Practicing Veterinarian 7, 551553
  • Turner, D. C. & Bateson P. (2000) The Domestic Cat: The Biology of Its Behaviour. 2nd edn. Cambridge University Press, Cambridge, UK
  • Wagner, T., Luis Fuentes, V., Payne, J. R., et al. (2010) Comparison of auscultatory and echocardiographic findings in healthy adult cats. Journal of Veterinary Cardiology 12, 171182