The combination of diazepam and ketamine is a commonly described protocol for induction of general anaesthesia in healthy dogs of various ages. It may also be indicated in certain cases with cardiovascular compromise (Beteg et al. 2010; Boutureira, Trim & Cornell 2007; Fayyaz et al. 2009; Green et al. 1981; Haskins, Farver & Patz 1986; Hazra, De & Roy 2008; Hellyer, Freeman & Hubbell 1991; Kolata 1986; White, Shelton & Taylor 2001). The combination of diazepam and ketamine, at a dose range of 0.2 mg/kg – 0.5 mg/kg and 5 mg/kg – 10 mg/kg respectively, has generally been associated with excitement-free induction of anaesthesia in dogs. However, maintenance of pharyngeal and laryngeal reflexes as well as hypersalivation have resulted in difficult intubations being reported (Green et al. 1981; Hellyer et al. 1991; White et al. 2001). Recovery from diazepam-ketamine in dogs has been reported to be free of emergence excitation although it is commonly associated with ataxia (Beteg et al. 2010; White et al. 2001).

Propofol (2, 6-di-isopropylphenol) has a wide initial dose range (2 mg/kg – 8 mg/kg) in dogs and induces rapid central nervous system depression facilitating anaesthetic induction within 20–30 seconds after commencement of intravenous administration (Robinson & Borer-Weir 2013; Watkins, Hall & Clarke 1987). Propofol bypasses the early stages of anaesthetic depth that are often associated with involuntary clonic movements, thereby facilitating routine excitement-free anaesthetic inductions in up to 92.5% of cases (Branson 2007; Davies 1991; Glowaski & Wetmore 1999). However, adverse induction behaviour characterised by paddling of limbs, muscle twitches and pain on injection has previously been associated with propofol administration in dogs (Davies 1991; Smith et al. 1993).

Recovery from propofol is rapid and excitement free, with return to consciousness occurring approximately 20 minutes after bolus administration (Branson 2007). Poor recoveries from propofol are uncommon and are characterised by tremors, opisthotonus, excessive salivation and vomiting (Robertson, Johnston & Beemsterboer 1992; Smith et al. 1993). Diazepam-ketamine induction and recovery characteristics compared with propofol have not been reported previously. The aim of this prospective clinical study was to compare a diazepam-ketamine combination with propofol alone for anaesthetic induction and recovery.

Thirty-six healthy male small-breed dogs weighing a mean (± s.d.) of 5.5 kg ± 2.3 kg and with a mean age of 26 ± 13 months were randomly assigned to an induction regimen of either diazepam-ketamine (Group D/K) or propofol (Group P). Dogs were declared healthy based on physiologically normal haematological (haematocrit, total serum protein, blood smear) and serum chemistry (blood urea nitrogen, serum creatinine) profiles and clinical examination performed upon admission to the hospital.

No statistically significant differences (p > 0.05) between Group D/K and Group P were observed with regard to age and weight. Additionally, there were no statistically significant differences with regard to cage habitus score, sedation score and duration of anaesthetic maintenance (Table 3).

The difference in induction score between the two groups, however, was statistically significant (p = 0.014), as Group D/K had better induction scores depicted by shorter induction times (p = 0.018) and fewer follow-up boli required to achieve endotracheal intubation (Table 3, Figure 1). Clinically, however, both groups produced acceptable anaesthesia, with only seven dogs in Group P and one dog in Group D/K failing to score perfectly (exhibited varying amounts of paddling). The latter was an outlier (inferior score) (Figure 1). No statistically significant differences were observed with regard to intubation score.

Group P had a greater incidence of myoclonus than Group D/K (p = 0.003), with nine dogs in Group P (n = 18) observed to have myoclonus. Group D/K had a very low incidence of myoclonus, with only one dog having muscle tremors.

There was a statistically significant difference in scores for quality of recovery from anaesthesia between the groups (p = 0.00002) and Group P had significantly superior recoveries when compared with Group D/K (Figure 1). Group-D/K dogs had inferior recoveries from anaesthesia, with 15 of the 18 dogs showing ataxia of varying degrees (7 of the 15 evinced paddling on recovery). One dog in Group P was an outlier on recovery. There was no statistically significant difference in time from termination of isoflurane to extubation of the patients in Group P and Group D/K, but time to standing was significantly shorter in Group P (p = 0.035; Table 3). Clinically, however, both Group P and Group D/K had generally acceptable recovery from anaesthesia. No dogs in Group P and only seven dogs in Group D/K scored higher than 4 and all dogs were able to stand 23 minutes after termination of isoflurane administration.

The level of agreement between observers was moderate to substantial (0.40–0.75) in all categories of induction and recovery scoring, barring the intubation score, where Kendall's Tau B test indicated fair agreement (0.32).

All dogs were client owned and consent was required in writing prior to being enrolled in the study. The dogs enrolled in the study were exposed to a moderate degree of discomfort as a result of the orchidectomy performed. The surgical procedure was performed by a specialist surgeon and all venepunctures as well as intravenous cannula placement for blood collection and drug administration, respectively, were performed by experienced veterinarians to limit the level of discomfort experienced. Appropriate analgesia including morphine and carprofen were provided during the peri-anaesthetic period. The present study was pre-approved by both the Animal Ethics Committee and the Research Committee of the Faculty of Veterinary Science, University of Pretoria (V017-33).

This study demonstrated that Group D/K was associated with better quality of induction and myoclonus scores when compared with Group P. Recovery from anaesthesia was observed to be inferior and of longer duration for Group D/K than for Group P.

The high quality of induction score associated with Group D/K in the present study supports current literature, which describes excitement-free dissociative anaesthesia with sufficient muscle relaxation to permit endotracheal intubation in dogs (Beteg et al. 2010; Hellyer et al. 1991; White et al. 2001). Maintenance of pharyngeal and laryngeal reflexes is expected with dissociative anaesthesia and although difficult endotracheal intubations have been reported previously, adequate doses (diazepam: 0.5 mg/kg – 1.0 mg/kg; ketamine: 5 mg/kg – 10 mg/kg) of both agents as described in the present study facilitated simple endotracheal intubation, favourable intubation scores and short induction times (Haskins et al. 1986; Jackson et al. 2004; White et al. 2001).

The observation of inferior quality of induction scores associated with Group P in the present study when compared with Group D/K was surprising and contrary to published literature (Amengual et al. 2013; Bufalari et al. 1998; Hall & Chambers 1987). The present study's design involved the comparison of two commonly used clinical protocols under clinical conditions. Statistically the quality of induction with propofol was inferior to diazepam-ketamine; however, both were clinically acceptable, demonstrating desirable anaesthetic induction characteristics in the majority of dogs anaesthetised.

Time to induction of anaesthesia in Group P was statistically longer and required more follow-up boli to achieve endotracheal intubation when compared with Group D/K. These findings are contrary to those of Watkins et al. (1987) and Amengual et al. (2013), who described rapid induction of anaesthesia facilitating easy endotracheal intubation in dogs anaesthetised with propofol administered at similar doses. Possible explanations for the induction characteristics associated with Group P include the following:

The propofol dose range described for anaesthetic induction in dogs is wide (Jiménez et al. 2012; Robinson & Borer-Weir 2013). The initial induction dose in the present study (2 mg/kg) reflected the minimum induction dose described by Robinson & Borer-Weir and the mean propofol induction dose observed in the study of 6 mg/kg was comparable to doses reported by Doebeli et al. (2013). The rate of propofol administration may influence the cardiorespiratory system, potentially resulting in hypotension and apnoea if administered rapidly (Keates & Whittem 2012). This was not the case in the present study. Conversely, in the present study propofol was administered at a rate of 2 mg/minute, half the dose rate reported by Bufalari et al. (1998) and Robinson and Borer-Weir et al. (2013). The slow rate of propofol administration may have resulted in partial anaesthetic recovery characterised by paddling, requiring numerous follow-up boli and delayed endotracheal intubation (Branson 2007; Musk et al. 2005; Zoran, Riedsel & Dyer 1993). As a result, fair comparison of induction, intubation and myoclonus scores of Group P to Group D/K is limited.

Adequate premedication provides anxiolysis, muscle relaxation and analgesia as well as decreasing induction agent dose requirements (Grint, Alderson & Dugdale 2010). Pre-anaesthetic administration of ACP and morphine is commonly utilised in dogs and produces mild to moderate sedation throughout a wide dose range (Brodbelt, Taylor and Stanway 1997; Grint et al. 2010; Henao-Guerrero & Riccó 2014; Heard, Webb & Daniels 1986; Monteiro et al. 2009; Robertson et al. 2001; Smith et al. 2001). The dose of ACP (0.02 mg/kg) and morphine (0.3 mg/kg) used in the present study were conservative and resulted in a light degree of sedation as demonstrated by the low sedation scores. Had there been more pronounced sedation, it is likely that the induction dose as well as the rate of administration used in the propofol group would have been adequate to facilitate easy endotracheal intubation as described by Amengual et al. (2013). In addition, this would have resulted in shorter and excitement-free anaesthetic inductions characterised by less myoclonus (Musk et al. 2005).

Timeous endotracheal intubation after induction of anaesthesia with a short-acting induction agent may be challenging, as previously reported in dogs anaesthetised with propofol (Clarke & Hall 1990; Davies 1991). Robinson and Borer-Weir (2013) reported that prolonged intubation times and a greater number of follow-up boli of induction agent were required when intubations were performed by inexperienced personnel. In this study, assessment of anaesthetic depth and subsequent intubation were performed by a relatively inexperienced clinician, potentially contributing to the inferior induction and intubation scores obtained in Group P.

Signalment plays a role in the induction dose requirements in dogs. A study performed by Boveri, Brearley and Dugdale (2013) highlighted the importance of calculating propofol dose requirements based on lean body mass. Body condition score was not assessed during the pre-anaesthetic clinical examination, which is a limitation in the present study as it may have provided a more accurate dose requirement for the dogs induced with propofol.

One dog in Group D/K that scored poorly on induction and, similar to an outlier in the study by White et al. (2001), may have been given insufficient pre-anaesthetic sedation.

Recovery was statistically superior and shorter in Group P when compared with Group D/K. Historically propofol generally has been associated with acceptable recoveries and the results of the present study further support published literature (Bufalari et al. 1998; Smith et al. 1993; Suarez et al. 2012). One dog in Group P, however, did receive an inferior recovery score, evincing muscle tremors, inability to maintain sternal recumbency and vocalisation. Propofol previously has been associated with myoclonus, paddling and opisthotonus during the recovery period, although vocalisation has not been generally reported (Davies 1991; Robertson et al. 1992; Tsai, Wang & Yeh 2007).

Dogs induced with diazepam-ketamine were associated with statistically inferior and prolonged recoveries when compared with propofol. Clinically, however, the quality of anaesthesia was acceptable. Recovery from diazepam-ketamine demon­strated an unremarkable return to cons­ciousness, routine extubation, occasional paddling and vocalisation. Such a recovery, in a clinical setting, is considered acceptable as dogs were not at risk of self-inflicted injury as a result of trauma and most dogs would be able to stand after a relatively short period. Subjectively superior recoveries when compared with the present study's results were described by White et al. (2001) in dogs recovering from anaesthesia induced with diazepam-ketamine and maintained on isoflurane for 103 ± 43 minutes. Compared with that study, the present study had a relatively short duration of anaesthetic maintenance (31.3 ± 8 minutes). The shorter duration may not have provided sufficient time for metabolism and redistribution to significantly decrease the diazepam-ketamine plasma concentration prior to commencement of anaesthetic recovery, resulting in subjectively poorer recovery quality being observed (Jiménez et al. 2012).

The subjective scoring systems incorporated in the present study successfully differentiated the quality of induction and recovery between two groups of healthy dogs. A moderate to good level of agreement between observers was achieved but only after sufficient training on the correct use of the scales. The SDS scoring systems used to score induction and recovery from anaesthesia have been described previously in literature but have not been fully validated (Amengual et al. 2013; Jiménez et al. 2012). Additionally, the validity of the SDS scores used in the study by Jiménez et al. was recently questioned by Ferchichi et al. (2013) in terms of:

These arguments raised valid concerns and demonstrate limitations in the present study as well as the study performed by Jiménez et al. (2012). The results in the present study, however, are still relevant in the clinical setting, where subjectively assessed anaesthetic depth provides the end-point for titration of induction agents prior to intubation. Similarly, the inferior recovery scores consistently associated with diazepam-ketamine during the recovery period in the present study when compared with the near perfect recoveries observed in dogs induced with propofol remain clinically relevant. Separate score allocations for the period prior to extubation (early phase) and the period from extubation to standing (late phase) may have improved the accuracy of evaluating recovery quality and would be useful additions to follow-up research.

The use of propofol alone and a diazepam-ketamine combination both produced clinically acceptable induction of anaesthesia in the dogs in this study; however, propofol administered at a low dose rate produced measurably inferior induction characteristics. Recovery from anaesthesia induced with both of these protocols was satisfactory; however, recovery from propofol was more rapid and associated with less excitement and ataxia.