Effect of non-steroidal anti-inflammatory drugs on postoperative respiratory and heart rate in cats subjected to ovariohysterectomy

Abstract

Objectives

The use of physiological parameters such as respiratory rate and heart rate to assess pain has long been discussed. The aim of the study was to compare postoperative respiratory rate and heart rate in cats subjected to flank ovariohysterectomy treated with a preoperative non-steroidal anti-inflammatory drug (NSAID) or no NSAID, and determine whether these parameters are suitable for postoperative pain assessment in cats. We hypothesised that cats without an NSAID would experience more postoperative pain, which may increase heart rate and respiratory rate.

Methods

A total of 168 female privately owned cats were studied. All cats were premedicated with medetomidine (0.08 mg/kg) and butorphanol (0.4 mg/kg) subcutaneously and anaesthesia was induced with intramuscular ketamine (5 mg/kg). Cats were divided into subgroups; controls (no NSAID) or cats given an NSAID, carprofen (4 mg/kg) or meloxicam (0.3 mg/kg), at premedication or induction of anaesthesia. Cats were subjected to flank ovariohysterectomy by the same surgeon. Atipamezole was administered 2.5 h after induction of anaesthesia. Respiratory rate and heart rate were measured 3.5 h after the induction of anaesthesia. Data were analysed using one-way ANOVA with mixed procedure and Tukey’s adjustment method for multiplicity.

Results

The postoperative respiratory rate and heart rate per minute for all cats were 34.0 ± 8.6 and 167.5 ± 27.4, respectively. Neither respiratory rate nor heart rate differed significantly between the control group and the NSAID groups or between different time points of administration of NSAIDs.

Conclusion and relevance

Assuming there was less postoperative pain in the group administered NSAIDs, the results of the study presented no support for use of respiratory rate and heart rate as parameters for postoperative pain assessment in individual cats. Study limitations included a lack of pain scoring and baseline data for respiratory rate and heart rate.

Introduction

Assessment and relief of pain in cats is important for animal welfare reasons and integrity of normal physiological processes; however, recognising pain in cats remains a challenge.^1–6 The understanding and recognition of pain is less understood in cats than in dogs.^7,8 In one study, 93% of veterinary paraprofessionals who responded to a questionnaire considered that their knowledge of pain, as well as assessment of pain, could be enhanced. ⁹ In another study, non-prescribers of perioperative non-steroidal anti-inflammatory drugs (NSAIDs) for neutering procedures in cats and dogs assigned lower pain scores to neutering compared with prescribers, and a minority used pain assessment tools, ¹⁰ which indicated clinicians assessed pain differently. This calls for an increased understanding of pain assessment.

Currently there is no optimal pain scoring technique for cats, although different pain assessment scales have been developed for this use. New methodologies have been investigated and one study suggested facial features as a tool for recognising pain in cats, but the study concluded further testing was needed to develop a clinical tool. ⁵ A multidimensional composite pain scale to assess postoperative pain in cats has also been developed; ¹¹ however, a quick, simple method to assess pain would be valuable.

The usefulness of physiological parameters such as heart rate (HR) and respiratory rate (RR) for pain assessment in small animal practice has long been discussed.^12–16 A previous study found that HR and RR were not reliable indicators of postoperative pain in cats. ¹⁵ However, that study was based on only 18 cats, including six controls. It has been suggested that further studies are required to provide stronger evidence for methods used to assess pain in cats. ¹ Because HR and RR are easily observed parameters that require little clinical skill, the potential use of those parameters for pain assessment would be valuable.

The aim of this study was to compare postoperative HR and RR in cats subjected to flank ovariohysterectomy, between controls without an NSAID and patients given preoperative NSAID treatment, and determine whether these parameters are suitable for postoperative pain assessment in cats. We hypothesized that cats without an NSAID would experience more postoperative pain, ¹⁷ which would potentially increase HR and RR.

Materials and methods

The prospective, randomised controlled study was performed at a small animal veterinary clinic (Arninge Djurklinik, Stockholm, Sweden). An ethical approval was obtained from the Animal Ethics Committee of Stockholm North, Sweden (reference 335/03). Privately owned cats admitted for ovariohysterectomy, minimum age 6 months, defined as adults,^18,19 were recruited. A preclinical examination was performed on admittance, and cats deemed unfit for ovariohysterectomy were excluded.

Anaesthesia

All cats were premedicated with medetomidine (Domitor Vet; Orion Pharma Animal Health) at 0.08 mg/kg and butorphanol (Torbugesic; Fort Dodge Animal Health) at 0.4 mg/kg administered subcutaneously (SC). About 20 mins later, anaesthesia was induced with ketamine (Ketaminol Vet; Intervet) at 5 mg/kg IM. After surgery, cats were kept in individual cages with blankets. At 2.5 h after induction of anaesthesia, atipamezole (Antisedan Vet; Orion Pharma Animal Health) at 0.2 mg/kg was administered IM.

Surgery

All cats were subjected to left flank ovariohysterectomy by the same experienced surgeon (OVH).^{20, 21}

Study groups

Cats were randomly allocated to seven subgroups: one control group and six NSAID treatment groups (Table 1).

Table 1.

Mean and SD of postoperative heart rate (HR), respiratory rate (RR) and body weight in cats subjected to flank ovariohysterectomy in the different non-steroidal anti-inflammatory drug (NSAID) groups and a control group. HR and RR did not differ significantly between groups

Group	n	HR (beats/min)	RR (/min)	Body weight (kg)
Carp-pre	22	167.6 ± 19.2	34.7 ± 9.0	3.3 ± 0.6
Carp-ind	21	170.2 ± 23.1	33.5 ± 10.6	3.1 ± 0.3
Carp-un	22	161.2 ± 39.7	36.6 ± 11.0	3.5 ± 0.6
Melox-pre	21	170.7 ± 26.0	33.9 ± 8.6	3.3 ± 0.5
Melox-ind	22	181.0 ± 24.7	31.0 ± 3.6	3.1 ± 0.4
Melox-un	26	157.9 ± 26.9	31.5 ± 6.4	3.4 ± 0.5
Controls	34	167.5 ± 22.6	36.2 ± 7.6	3.3 ± 0.5

Carp = carprofen; Melox = meloxicam; pre: NSAID given at premedication 20 mins prior to induction of anaesthesia; ind: drug given at induction of anaesthesia; un: timing was unspecified, drug given with premedication or at induction of anaesthesia

Syringes were prepared with carprofen and meloxicam at the appropriate dosage for the patient and were labelled accordingly. In addition, a placebo (saline) dosage was prepared at a volume equivalent to one of the NSAID drugs. The veterinary nurse randomised choice of the NSAID, time-point of treatment and administered one NSAID or placebo.

Carprofen (Rimadyl; Orion Pharma Animal Health) was administered SC at 4 mg/kg to 65 cats, of which 22 cats were given the NSAID at premedication 20 mins prior to induction of anaesthesia (Carp-pre) and 21 cats at induction of anaesthesia (Carp-ind). For 22 cats, the timing was unspecified, given with premedication or at induction (Carp-un).

Meloxicam (Metacam; Boehringer Ingelheim Vetmedica) was administered SC at 0.3 mg/kg to 69 cats. In 21 cats, meloxicam was administered at premedication 20 mins prior to induction of anaesthesia (Melox-pre) and in 22 cats, meloxicam was administered at induction of anaesthesia (Melox-ind). In 26 cats, the timing was unspecified, meloxicam was given with premedication or at induction (Melox-un).

In total, 34 cats were allocated to the control group and received a saline injection at premedication or induction as a placebo treatment. These controls were given an NSAID after measurement of RR and HR.

One hour after administration of atipamezole, corresponding to 3.5 h after administration of ketamine, RR and HR were assessed by the operating surgeon, who was blinded to treatment. RR was determined by observing the thoracoabdominal movements of the cat from a distance of 1 m. After this, the cat was approached and HR was measured by cardiac auscultation. The HR was counted at auscultation with a stethoscope after about 20–30 s of contact with the cat to allow the HR to stabilise. After assessment of RR and HR, the labelled syringes were checked and an NSAID was administered to the cat if it belonged to the placebo group.

Statistics

Data were analysed using one-way ANOVA with the mixed procedure of the SAS (2013) package. Post-hoc comparisons between groups were adjusted for multiplicity using Tukey’s method. P <0.05 was considered significant. No evident deviations from normality or homoscedasticity were detected.

Results

The mean and standard deviation of the body weight of all cats were 3.2 ± 0.4 kg (Table 1). The mean body weight of the cats did not differ between groups (P ⩾0.12). At examination of RR and HR all cats were awake in sternal recumbency, or were sitting or standing.

The postoperative RR and HR per minute for all cats were 34.0 ± 8.6 and 167.5 ± 27.4, respectively. There were no significant differences in RR (P ⩾0.28) or HR (P ⩾0.53) between controls and cats that received an NSAID or between different time-points of administration of an NSAID (P ⩾0.92 and P ⩾0.87, for RR and HR, respectively).

The RR or HR did not differ between cats that received carprofen or cats that received meloxicam (P ⩾0.31 and P ⩾0.2, for RR and HR, respectively).

The RR for early vs late administration of carprofen and meloxicam did not differ (P = 1 and P = 0.92, respectively), neither did the HR for early vs late administration of carprofen and meloxicam (P = 1 and P = 0.87, respectively).

Discussion

In this prospective, randomised controlled study there were no statistical differences between the control group and the NSAID groups in RR or HR measured postoperatively. There was also no difference in postoperative RR and HR between different time-points of administration of an NSAID. The results of the study presented no support for the use of RR and HR as parameters for postoperative pain assessment in individual cats, assuming cats without a NSAID at the time of observation did experience more pain.

The results were in agreement with a previous study where HR and RR did not differ between control cats and cats subjected to surgery. In that study, only visual analogue scale scores and response to palpation scores differed significantly between control and surgical groups. ¹⁵ The results were also in agreement with a study in dogs that investigated the potential use of physiological parameters for scoring pain. That study concluded that HR and RR were not useful indicators of pain in hospitalised dogs. ¹³

Objective clinical variables such as HR and RR that were evaluated in a study of cats were not found to be consistent indicators of pain, which is in agreement with our results. ²² In the guidelines ²³ for recognising, assessing and treating pain in companion animals it is stated that HR and RR have not been consistently correlated with signs of pain in cats.^11,17 The Melbourne pain scale uses HR and RR for the assessment of pain in dogs. ²⁴ Blood pressure is used in the UNESP-Botucatu multidimensional composite pain scale for assessing postoperative pain in cats, but not HR or RR.^11,17

Studies have suggested that cats neutered by the flank approach tend to suffer from more postoperative pain than cats subjected to midline incision.^25–27 In another study, no significant difference was found in postoperative pain between midline and flank groups. ²⁸ In the present study, the flank approach was used and we hypothesised our control cats experienced more pain than cats given an NSAID. However, HR and RR did not differ. An NSAID was given before surgery, which has been recommended for optimal effect. ²⁹

In previous studies comparing meloxicam and carprofen for postoperative analgesia in cats subjected to ovariohysterectomy, pain scores did not differ significantly between groups.^30,31 This would suggest analgesia was similar in the two NSAID groups of the present study. We hypothesised that cats given an NSAID would experience less postoperative pain. However, a confounder may be that all cats had received butorphanol, which may have caused some analgesic effect at time of assessment. Several clinical studies support the use of butorphanol in cats.^32–39 Butorphanol has a rapid and relatively short duration of action with a suggested analgesic effect between 45 mins and 4 h. ⁴⁰ One study reported a duration of 165 mins by use of the thermal threshold method in healthy cats. ⁴¹ Pathways for thermal stimuli may, however, differ from stimuli from a surgical wound, which is why conclusions about duration of analgesic effect on surgical pain should be made cautiously. However, in the present study measurements of RR and HR were performed almost 4 h post-butorphanol injection, which suggests the drug’s effects were minor at that time.

The resting and sleeping mean RR in the home environment of healthy cats is <30 breaths/min. ⁴² Visiting a veterinary clinic is a stressful event that may affect HRs and RRs, and should be considered at examination. ⁴³ In the present study stress was not scored or measured, but it is not surprising that the observed RR in our study was higher compared with previously reported resting or sleeping RR in a home environment. ⁴² The HRs and RRs of the present study were lower than reported results of cats in hospital environment after a 10 min acclimatization period, but the present study did not include baseline data. However, in that previous study blood pressure had been measured before obtaining RR and HR, which potentially caused a stress reaction. ⁴³ Therefore, comparisons between studies should be made cautiously.

Concerns about the effect on kidneys and cardiovascular system by an NSAID given preoperatively have been expressed; hence, clinicians may prefer to administer the drug postoperatively. In healthy dogs given an NSAID, no perioperative adverse effects on the cardiovascular system have been reported at recommended dosages.^29,44,45 Most likely, in healthy cats the risk for drug-induced organ damage is minimal but caution is encouraged in patients with hypovolaemia or kidney-liver pathology.

Study limitations

In the present study, pain was not scored using pain scales, and it is not known whether controls experienced more pain or whether there was still an analgesic effect from butorphanol. The surgeon was experienced, and postoperative pain perception may have been low due to the short duration of the procedure and low nociceptive stimulation. Subcutaneous administration of butorphanol and medetomidine may have slowed their clearance, causing analgesia to be longer than expected. However, at the time of measurement of HR and RR, the effect of butorphanol was expected to be minor. A difference in pain perception between controls and cats given an NSAID was therefore expected, but not possible to validate with the design of the study. Data for individual postoperative temperatures or exact duration of surgery were not available, which is another study limitation, because postoperative hypothermia may have affected HR and RR and masked the influence of pain on these parameters. Unspecified timing of the NSAID for some cats, given with premedication or at induction, was another study limitation. However, the results did not differ between groups. Furthermore, no power analysis was performed prior to the study and baseline data was not available for analysis.

Conclusions

Assuming there was less postoperative pain in the group administered NSAIDs, the results of the study presented no support for use of RR and HR as sole parameters for postoperative pain assessment in individual cats.