Statistical analysis

Data of two calves had to be removed from analyses as they died of perforated abomasal ulcers (confirmed by necropsy, one calf of EPI, another of INJ) on d 5 and 6 respectively.

Data of these calves are discussed elsewhere (Chapter 2). For final analyses, data of 27 calves (INH (n = 10), INJ (n = 9), EPI (n = 8)) remained.

Data were analysed using a commercial statistical analysis program (Statistical Analysing System, release 9.1 for Windows, SAS Institute Inc., Cary, N.C., USA). Data are presented as means with standard deviations. Results were analyzed by two-way analysis of variance for repeated measurements (Proc GLM, Repeated-Statement). Subsequently group means at different points of time were tested for statistical differences using the LSMEANS statement (pdiff/tdiff option). Differences between means and baseline within groups were tested using t-tests for paired observations. Results of the scoring systems were analysed with signed rank test for differences between each measurement and baseline. Groups were compared using the Wilcoxon test. For dichotomous traits Fisher’s exact test was used. The significance level was set at P = 0.05.

3.4. Results

The 27 evaluated calves went through surgical procedure (mean duration 65 min) without anaesthetic complications (Offinger 2010).

3.4.1. Milk intake, total energy intake and daily weight gain

When milk was offered for the first time after surgery (three h post op) all calves got up immediately, were able to stand firmly and drank the total amount of provided milk. Between d 1 and 6 milk intake was refused twice in INH (n = 10) group, twelve times in INJ group (n = 9) and once in EPI group (n = 8) with a significant difference between INJ and EPI group on d 3. Means of milk intake (kg kg BW^-1 d^-1) and total energy intake (MJ ME calf^-1 d^-1) decreased in all groups after surgery, in INJ group the difference was significant to baseline.

While mean daily weight gain before surgery (d -5 to -1) was highest in INJ compared to INH and EPI group (not significant) weight gain of INJ group was lowest (significant difference between INJ and INH) after surgery (d 1 to 6). In INJ group mean weight gain was significantly lower after surgery compared to before surgery (Table 2).

3.4.2. Clinical and behavioural monitoring

Mean rectal temperature in INJ group remained higher compared to other groups until d 5 with significant differences on d 1, 2 and 3 (Figure 3).

Additional antibiotic treatment, based on defined criteria, was required by none of ten calves of INH and four of nine calves of INJ group between d 4 and 6 (significant differences between groups on all days). In the EPI group two, three and four of eight calves received antibiotic treatment on d 4, 5 and 6 respectively (significant difference to INH group on d 6).

Wound inflammation score and VAS were significantly higher compared to baseline in all groups after surgery. Comparing the groups, the inflammation score was significantly higher in INJ compared to INH group on d 3, and higher in EPI compared to INJ group on d 5. The VAS score was significantly higher in INJ and EPI group compared to INH group on d 3 and in EPI compared to INH group on d 4.

The MPDS score was significantly increased compared to baseline in all groups after surgery.

INH obtained significantly lower scores than INJ and EPI group whereas highest scores were seen in INJ group (Figure 2).

3.4.3. Cortisol

Cortisol concentrations (Table 4) increased slightly in all groups with high dispersion about the mean. Significant differences to baseline were seen in group INJ on d 1 and in group INH on d 3, but the two-way analysis of variance showed no significant effects for time and group.

3.4.4. Examination of the respiratory tract

Clinical examination: The INJ group’s respiratory score (Table 3) was higher compared to INH and EPI group on d 3 to 6 with a significant difference on d 3. The auscultation score increased significantly compared to baseline in all groups after surgery and did not decrease

to baseline level during complete study period (Table 3). No significant group differences existed but highest scores were found in the INJ group.

Ultrasonography and arterial blood gas analysis: Ultrasonography and arterial blood gas analysis revealed no anaesthesia related differences between groups and blood gases remained in physiological range in all groups throughout the study period (Table 4).

Table 2: Mean (± SD) daily milk intake, total energy intake and weight gain in INH (n = 10), INJ (n = 9) and EPI (n = 8) group, averaged over five days before surgery (pre op) and over six days after surgery (post op). * indicates significant (P < 0.05) difference to baseline (pre op). Different letters denote significant difference between groups (P < 0.05) at the specific

Figure 2: Mean (± SD) multiple pain and discomfort score (MPDS) in INH (n = 10), INJ (n = 9) and EPI (n = 8) group one day prior to surgery (d - 1) to six days after surgery (d 6). MPDS was assessed twice a day and averaged for further calculation. Symbols with a cross differ significantly (P < 0.05) from baseline (d -1). Different letters denote a significant difference between groups (P < 0.05) at the specific point of time.

Figure 3: Mean (± SD) rectal temperature in INH (n = 10), INJ (n = 9) and EPI (n = 8) group one day prior to surgery (d -1) to six days after surgery (d 6). Temperature was assessed twice a day and averaged for further calculation. Symbols with a cross differ significantly (P <

0.05) from baseline (d -1). Different letters denote a significant difference between groups (P

< 0.05) at the specific point of time.

Table 3: Mean (± SD) wound inflammation score, visual analogue scale (VAS), respiratory score and auscultation score in INH (n = 10), INJ (n = 9) and EPI (n = 8) group one day prior to surgery (d -1) to 6 days after surgery (d 6). Parameters were assessed twice a day and averaged for further calculation. * indicates significant (P < 0.05) difference to baseline (d -1). Different letters denote a significant difference between groups (P < 0.05) at the specific point of time.

Table 4 Mean (± SD) cortisol concentration, arterial partial pressure of oxygen (pa O2), arterial partial pressure of carbon dioxide (p_a CO₂), oxygen saturation (SaO₂) and ultrasonographic score in INH (n = 10), INJ (n = 9) and EPI (n = 8) group one day prior to surgery (d -1) and on d 1, 3 and 5 after surgery. * indicates significant (P < 0.05) difference to baseline (d -1). Different letters denote a significant difference between groups (P < 0.05) at the specific point of time.

3.5. Discussion

The results of the present study indicate significant differences in postoperative recovery after umbilical surgery in calves depending on the anaesthetic regime used. After passing through the identical surgical procedure calves that received anaesthesia by injection of xylazine and ketamine had a more complicated recovery due to pulmonary complications compared to those anesthetized by epidural and particularly by inhalation anaesthesia. Postoperative pulmonary complications led to significant impairment of well-being and decreased daily weight gain in calves.

The animals’ postoperative condition was measured by means of a multiple pain and discomfort scale (MPDS, Table 1). Score systems have proved to be a valuable instrument concerning the assessment of animal pain and discomfort (Molony & Kent 1997, French et al.

2000, Anil et al. 2002). The score system used here (modified according to Schulze 2010) is based on behavioural patterns considered as pain indicating in calves in earlier studies (Molony & Kent 1997, Hudson et al. 2008). In the current study the MPDS demonstrated a higher degree of postoperative pain and discomfort in calves anesthetized by injection of xylazine and ketamine (INJ) compared to calves after inhalation (INH) or epidural (EPI) anaesthesia. This was reflected in significantly decreased milk intake, total energy intake and daily weight gain in INJ group compared to before surgery. The decline in daily weight gain compared to before surgery averaged 7 % in INH group, 28 % in EPI group and 65 % in INJ group. This illustrates clear disparities in postoperative productivity.

Anderson & Muir (2005) state that producing depression (here decreased daily weight gain) may arise from severe stress, resulting from severe pain. In the present study, painfulness of the surgical wound was assessed twice a day via wound inflammation score and VAS (Table 3). Signs of inflammation and sensitivity of the wound area were significantly increased compared to baseline in all calves throughout the study period. A more pronounced affection in INJ group compared to other groups was not present. Wound inflammation and wound pain in the study population was generally low according to assessed scores and no unscheduled aftercare was necessary in any case. Thus, the wound healing process was considered as

normal in all calves. Mean plasma cortisol concentration, which is commonly used as indicator for pain induced stress (Molony & Kent 1997, Mellor et al. 2002, Stafford et al.

2002) was slightly but not significantly increased compared to baseline, in all groups at all times (Table 4). EPI calves showed lower cortisol concentrations compared to other groups, but no significant differences were found. The great dispersion about mean cortisol concentrations may be explained by the sensitivity of this parameter towards external influences. It is presumed, that inflammatory disturbance in the wound healing process and thereby excess of wound pain is not preferentially responsible for the INJ group’s low daily weight gain in the postoperative period. In this regard one has to consider that all calves in this study received pre-emptive and postoperative NSAID treatment. Thus, considerable pain reactions, by sensitization and/or inflammation, were not to be expected in any group.

Furthermore, it could be shown that disturbance of gastrointestinal mucosa after surgical intervention did not emerge subjected to the anaesthetic regime used (Chapter 2).

Most noticeable postoperative findings in all groups and especially in the INJ group were clinical respiratory symptoms. The increase in rectal temperature (Figure 3) and in respiratory and auscultation scores (Table 3) gave evidence of acute pulmonary affection although all calves received perioperative antibiotic treatment with enrofloxacin. Enrofloxacin accumulates in respiratory tissue and little antibiotic resistance of infectious microbiological respiratory agents against enrofloxacin is reported. According to pharmacokinetic studies it may be that the used dose of 2.5 mg enrofloxacin kg BW^-1 was too low for reliable prevention of pulmonary infections (Davis et al. 2007). Unfortunately, no tracheobronchial lavage was performed (in the studied calves) so that no conclusive evidence can be provided regarding possibly responsible microbiological agents. Declined general condition, respiratory signs and increased body temperature were criteria for the extension of antibiotic treatment which appeared most frequently necessary in the INJ group. However, pulmonary affection appeared to be mild. This judgement was supported by in average undisturbed blood gas values and unremarkable results of lung ultrasonography (Table 4).

Extended affection of lung tissue (pneumonia) leads to impairment of pulmonary gas exchange, which is initially presented in decreased paO2 followed by an increase in paCO2

(Donawick & Baue 1968). In the current study respiratory affection was clinically present but not yet detectable via blood gas analysis. The same applies to ultrasonography, which is able to detect pathological processes that replace inflated lung tissue in superficial areas e.g.

atelectasis, consolidation or hepatisation (Braun et al. 1997, Rabeling et al. 1998, Flöck 2004). Acute obstructive pulmonary affection with bronchial constriction and oedematous swelling of mucus membranes as well as mucus accumulation in the trachea and bronchial tubes, as it was probably the case in calves of the present study, is not detectable by ultrasonography.

The development of only mild postoperative pulmonary complications (no changes in blood gases and ultrasonography) was probably supported by the effect of flunixin-meglumine, which has proved to be efficient in treatment of bovine respiratory disease, reducing pyrexia and the extent of lung consolidation in calves (Lockwood 2003). However, the observed acute respiratory affections after anaesthesia in calves of this study seemed still compromising enough to impair postoperative recovery.

The risk of aspiration of saliva or ruminal fluid, which may cause pulmonary affection, is always present during surgery in dorsal recumbency. Calves of the INH group were the only ones intubated which contributes to the prevention of aspiration pneumonia. However, all calves were placed on a cranially tilted table with the head positioned downwards and the shoulder being the highest point of the spinal column to avoid reflux or aspiration (Meyer et al. 2009). Thus, in the view of the authors, it appears improbable that respiratory affections were caused by aspiration.

Furthermore, all calves were housed in individual tiled pens without direct contact to other animals, wherefore new infections in the postoperative period appear to be unlikely. The authors assume that sub clinical infections, already existing at the start of the study, were stress activated by the effects of surgical intervention.

The occurrence of postoperative pulmonary complications is well known and discussed in human (Brooks-Brunn 1997, Warner 2000, Magnusson & Spahn 2003, Duggan & Kavanagh 2005) and veterinary medical literature (Alwood et al. 2006, Brainard et al. 2006, Staffieri 2007 and 2009). Adverse positioning is presumed to play a major role concerning the

development of postoperative pulmonary complications. Lateral and especially dorsal recumbency affects pulmonary health and results in hypoxemia during surgery (Klein &

Fisher 1988, Wagner et al. 1990, Nyman et al. 1990, Meyer et al. 2007). Via computed tomography in accordant positioning the prompt development of densities in pulmonary tissue could be detected in horses (Nyman et al. 1990, Staffieri et al. 2009) and dogs (Staffieri et al.

2007). These densities were identified as atelectasis and pulmonary congestion by histological analysis (Nyman et al. 1990). The extend of atelectatic areas increases with thoracic dimensions and body mass and therewith correlates positively with hypoxemia in dorsal recumbency as it has been demonstrated in horses (Nyman et al. 1990, Mansel et al. 2008) and humans (Eichenberger et al. 2002). According to the size of atelectatic areas, they disappear short after repositioning or may persist for up to 24 h and more (Nyman et al. 1990, Eichenberger et al. 2002). Considering the short duration of dorsal recumbency in the current study (65 minutes on average) and the low body mass of the calves, it is presumed that intraoperatively occurred atelectasis resolved prompt after repositioning in sternal recumbency. This is supported by p_aO₂, p_aCO₂ and S_aO₂, which recovered within two to three h after surgery (Offinger 2010) and stayed then within reference range (Kiorpes et al. 1978), indicating adequate gas exchange ability, for the entire postoperative study period.

Furthermore, no evidence of atelectasis (homogeneously hyperechoic areas (Braun et al.

1997) could be detected via ultrasonographic examination of the lung on d 1, 3, and 5.

However, in the present study, ultrasonography was not performed during or directly after dorsal recumbency. Moreover, development of atelectasis was not assessed by other methods, e.g. computed tomography.

The temporary (intraoperative) presence of atelectasis might clear the way for the (postoperative) occurrence of pneumonia (Warner 2000, Duggan & Kavanagh 2005). Exact correlation remains unexplained, but bacterial entrapment in alveoli during atelectasis may be causal (Staffieri et al. 2007). Furthermore, an inhibition of the microbicidal oxidative activity of pulmonary alveolar macrophages might play a role in the development of pneumonia. This effect was demonstrated for inhalation and injection anaesthesia in humans (Welch 1985, Kotani et al. 1998). Moreover, a general stress related immunosuppression after surgical intervention might also contribute to pulmonary infection (Anderson & Muir 2005).

Regarding respiratory and auscultation scores (Table 3) as well as antibiotic requirement, postoperative pulmonary complications were most pronounced in the INJ group. As dorsal recumbency affected all calves in the same way, intraoperative effects of the anaesthetic regimes are presumed to be primarily responsible for the difference between groups.

Severe intraoperative cardiopulmonary depression after IM and IV bolus injections of xylazine were frequently confirmed in literature (Campbell 1979, Waterman 1981, Greene &

Thurmon 1988, Rioja et al. 2008) and may amplify pulmonary side effects of dorsal recumbency. While calves of the INH group received xylazine in a dose of 0.1 mg kg^-1 IM calves of INJ and EPI group were treated with 0.2 mg kg^-1 IM or epidurally, respectively.

Xylazine is absorbed from the epidural space in small amounts and leads to mild sedation and only mild depression of respiratory function. Accordingly, intraoperative stress responses in general and respiratory depression in particular were significantly lower in the EPI compared to INH and INJ group and cardiopulmonary parameters remained mostly within reference ranges (Offinger 2010). This observation is in accordance with results of previous reports (Skarda et al. 1990, Picavat et al. 2004, Meyer et al. 2009). During surgical intervention lowest mean arterial S_aO₂ and highest values for pulmonary shunt fraction were found in calves of the INJ group compared to INH and EPI. Furthermore, even respiratory arrest was seen repeatedly in calves of INJ group during anaesthesia (Offinger 2010). Thus, the high dose of IM administered xylazine combined with repeated ketamine injections, which were necessary for maintenance of anaesthesia, seem to be relevant for the occurrence of most pronounced postoperative pulmonary affection in the INJ group.

Results of the present study demonstrate that good postoperative recovery in EPI calves, probably due to the intraoperative compatibility of high-volume epidural anaesthesia, is comparable to INH calves and less complicated compared to those of the INJ group.

Conclusion and clinical relevance Pulmonary complications affecting postoperative recovery are a coherent risk of abdominal surgery in dorsal recumbency in calves, in particular after anaesthesia by injection of xylazine and ketamine. High-volume epidural anaesthesia with

xylazine and procaine is an efficient alternative anaesthetic regime for abdominal surgery in calves under field conditions; postoperative recovery is comparable to inhalation anaesthesia with isoflurane and it does not require considerable financial, technical or personnel expenditures.

4. Kapitel 2: Impact of umbilical surgery in dorsal recumbency under inhalation