- Email: [email protected]
Short-term and long-term outcomes of ewes and their offspring after elective cesarean section
Theriogenology 79 (2013) 486–494
Contents lists available at SciVerse ScienceDirect
Theriogenology journal homepage: www.theriojournal.com
Short-term and long-term outcomes of ewes and their offspring after elective cesarean section Steinar Waage*, Guri Wangensteen 1 Department of Production Animal Clinical Sciences, Norwegian School of Veterinary Science, Oslo, Norway
a r t i c l e i n f o
a b s t r a c t
Article history: Received 18 April 2012 Received in revised form 29 October 2012 Accepted 1 November 2012
Postoperative outcomes of animals that have undergone cesarean delivery have been reported previously; however, in most studies results were influenced by a combination of surgery per se and the preoperative condition of the animal, which was frequently impaired because of the presence of dystocia. To evaluate the effects of the cesarean section itself we conducted a matched cohort study comparing postpartum complications and future reproductive performance of 162 ewes subjected to elective cesarean section and 162 ewes that had an unassisted vaginal delivery. Survival and subsequent growth of their lambs were also compared. Effect of mode of delivery on weight gain was estimated using linear mixed models. Case ewes, which underwent surgery during the period from 1996 through 2004, and control ewes were from the flock at the Animal Production Experimental Centre, Norway. Two ewes (1.2%) that underwent cesarean section died; one developed peritonitis and the other experienced uterine prolapse and did not recover. Postoperatively, four ewes suffered from metritis, three suffered a wound infection, and four a delayed wound healing; all recovered after treatment. One of the ewes that delivered vaginally died 3 days after lambing. The incidences of fetal and postnatal deaths did not differ significantly between the cesarean and the vaginal delivery groups (fetal deaths, 3.5% and 3.1%, and postnatal deaths, 9.9% and 7.1%, respectively). Survival rates and weight gains of the lambs the subsequent months were similar for the two groups. Seventy percent of the ewes that had a cesarean section and 72% of those that had a vaginal delivery were bred the next season; conception rates were 89% and 90%, respectively. However, the ewes subjected to surgery the previous year gave birth to significantly fewer live-born lambs (mean, 1.64) than those that had had a vaginal delivery (1.93). The difference was the result of a reduced litter size and an increased number of fetal deaths in the former group. Birth weights of the live-born lambs the second year did not differ between the groups. In conclusion, severe short-term complications were rare among the ewes that underwent elective cesarean section. Survival and growth of their lambs and their conception rate the subsequent season did not differ from the corresponding outcomes of the ewes that delivered vaginally, but their fertility was reduced in the sense that they gave birth to fewer live-born lambs the following lambing season. Ó 2013 Elsevier Inc. All rights reserved.
Keywords: Sheep Cesarean section Reproductive performance Lamb survival Lamb growth
1. Introduction
* Corresponding author. Tel.: þ47 22 964 875; fax: þ47 22 597 083. E-mail address: [email protected] (S. Waage). 1 Present address: Veterinary Surgeon, Fiskå, Norway. 0093-691X/$ – see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.theriogenology.2012.11.001
Dystocia is common in sheep. However, large breed differences in the incidence have been reported, varying from approximately 1% in a study of Scottish Blackface [1] to 34% in Dorset Horn ewes [2]. In a study of Texel Sheep, Milk Sheep, and various crossbreeds, in which the birth of
S. Waage, G. Wangensteen / Theriogenology 79 (2013) 486–494
each lamb was recorded, 32% of the lambs experienced a difficult parturition [3]. According to records for 2010 in the Norwegian Sheep Recording System, a complicated birth that required assistance occurred in 13% of 270,850 lambings, and some modest assistance (gentle traction) was provided in an additional 12% of the lambings (S. Waage, unpublished data). Apart from the occasional adverse consequences for the ewe, dystocia is a major cause of perinatal lamb mortality [4]. In most cases in which assistance is needed, manual correction and vaginal delivery is achieved. Examples are the common malposture dystocias involving the head or forelimbs and the more rare breech and transverse presentations. However, there are types of dystocia for which assisted vaginal delivery might fail and cesarean section is the alternative obstetric intervention. Fetopelvic disproportion, which occurs most frequently in singleton births of male lambs in primiparous ewes of certain breeds, is reported to be a major cause of cesarean section [5–7]. Also, failure of or incomplete dilatation of the cervix (“ring womb”) is a relatively frequent cause of cesarean section [7,8]. More rare indications are fetal retention, usually in cases of multiple births, irreducible uterine torsion, and fetal malformation. When vaginal birth fails, some might consider euthanasia an alternative to cesarean section. Of interest to the sheep farmer, apart from animal welfare aspects, is the economic outcome after cesarean section, which depends on the costs of surgery and the outcomes of the ewe and her offspring. Short-term effects with respect to complications and survival and subsequent performance must be taken into account. Of main interest in meat sheep are future fertility of the ewe and her mothering capabilities, which to some extent might be reflected by the growth rate of her lambs. Thus, to assess the effect of the cesarean section, cases should be monitored for a considerable time period after surgery. However, studies of cases subjected to cesarean section because of dystocia would obviously include some ewes with an impaired prognosis because of prolonged labor before surgery and results would reflect the combined effects of the surgery and potentially previous straining. Short-term outcomes reported after cesarean delivery in dystocia cases in small ruminants are generally not very encouraging. Three studies reported mortality rates for surgical cases of 5%, 17%, and 19%, respectively [5,6,8]. Two of these studies also recorded fetal survival, which was poor; 26% [5] and 58% [8] of the fetuses were dead. Future fertility of ewes subjected to cesarean section has been studied to a very limited extent. Two previous studies reported high pregnancy rates, but the proportion selected for subsequent breeding of those subjected to surgery was relatively small [5,8]. Postnatal survival and growth rate of the lambs of ewes that had undergone cesarean section do not seem to have been investigated. Of particular interest is to determine whether or not there are any adverse effects of the surgery itself. To evaluate the effects of cesarean delivery per se on postpartum health and survival and on future performance, we therefore conducted a retrospective study of ewes subjected to elective cesarean section. Detailed postoperative
487
and long-term outcomes for these ewes were compared with outcomes for ewes that had an unassisted vaginal delivery. 2. Materials and methods 2.1. Study design and animals included We conducted a retrospective matched cohort study including as cases all ewes that were subjected to elective cesarean section at the Animal Production Experimental Centre (APEC), Ås, Norway, during the period from 1996 through 2004. These ewes served as teaching cases for veterinary students in their surgical training in obstetrics. Ewes selected for this purpose were those with expected date of lambing close to the predetermined days of the teaching program; other criteria were not considered. No medical or obstetric indication for cesarean delivery was present. For each ewe subjected to surgery we used as control an ewe that had an unassisted vaginal birth. Among the candidates for being the control, we selected the ewe that lambed closest in time to the ewe that underwent cesarean section. All ewes included in the study were clinically healthy at the time of parturition. Permission to perform elective cesarean section as part of the student teaching program in obstetrics has been given by the Norwegian Animal Research Authority. 2.2. Experimental station, feeding, and management practices The APEC, owned by the University of Life Sciences and the Norwegian School of Veterinary Science, is the main experimental station for production animals in Norway and is located at the University Campus in Ås. The annual number of lambing ewes in the flock in the study period varied between 150 and 250. Nutrition and management routines at the APEC during pregnancy and around lambing did not change much from year to year during the study period. The ewes were removed from pasture in late autumn, approximately 1 month before breeding, and kept indoors in group pens during pregnancy with voluntary access to a limited outdoor area. Roughages fed to the pregnant ewes were grass silage, partly stored as round bales, and hay. The breeding season started in the end of November and continued until late December. Each day, a ram was led through the pens to detect ewes in estrus. Ewes in estrus were mated with a predetermined ram selected among the five to six rams available in the flock. Ewes mated were recorded by the staff; hence, the approximate day of lambing could be determined. Ultrasound scanning was performed, and the ewes were grouped according to age and the number of fetuses. Barley and a commercial concentrate product were given to nulliparous ewes during the entire pregnancy, and older ewes received concentrates only during the final weeks of pregnancy. Amounts of concentrate differed according to age of the ewe and the number of fetuses she carried. Body condition scoring was performed to check whether nutrient intake was sufficient. Mineral supplementation was provided using a commercial product.
488
S. Waage, G. Wangensteen / Theriogenology 79 (2013) 486–494
Shortly before expected lambing, ewes were placed in individual pens with floor covered with fresh, dry straw and were monitored day and night. Neonatal lambs were observed and sucking assistance was provided if a lamb was reluctant or failed to suck its dam. If considered necessary, lambs were given colostrum by stomach tube. When possible, cross-fostering of surplus lambs was practiced when litters were large. The sheep were put out to pasture in May. In mid-June, the flock was routinely collected for antiparasitic treatment, and the lambs were weighed. Lambs were also weighed approximately 2.5 months later. Antiparasitic treatment was repeated when the outdoor period ended in mid-October. 2.3. Cesarean section Cesarean section was performed on approximately day 147 of gestation; it was required that there was milk in the udder. The operations were performed under “field conditions”, using a storage room at the APEC, and were carried out by veterinary students as part of their final training in obstetrics. Two students performed the cesarean section in each ewe and different students were involved in each case. They were closely supervised by the Professor of Obstetrics, Department of Production Animal Clinical Sciences, and occasionally by an Associate Professor in the same department. The perioperative procedures, including the method used for anesthesia and the surgical technique, did not vary during the study period. The ewe was standing in a box with metal mesh floor, a metal front with an adjustable opening, giving the ewe access to a feeding trough, but preventing retraction of the head. The right side of the box was against the wall. To allow a proper working posture, the box was somewhat elevated above the surrounding floor. As is recommended [9] benzylpenicillin sodium iv (approximately 50,000 IU per kg) was administered before surgery. The left sublumbar region was sheared and shaved and the skin washed and disinfected. An inverted-L block was used for local anesthesia [9], administering a total of 30 to 40 mL 2% lidocaine hydrochloride with adrenaline. A sterile plastic drape was used to cover the flank, using an adhesive spray to achieve adherence to the skin. The incision, approximately 20 cm in length and running approximately vertically, was in the midparalumbar fossa. The uterine wall was incised along the greater curvature of a pregnant horn, avoiding placentomes. In cases of multiple pregnancies, all lambs were usually removed through a single incision; in a few cases both horns were incised. No attempt was made to remove fetal membranes. The uterine incision was closed with a continuous, inverting pattern, either Cushing suture or modified Lembert, using absorbable material. Peritoneum and the two inner muscle layers were closed with a single continuous suture and the external muscle layer with another continuous suture, using absorbable material. The skin incision was closed with a Ford interlocking suture pattern, using nonabsorbable material. The newborn lambs were put in the trough in front of the ewe and taken care of by the ewe.
After the cesarean or vaginal delivery, the ewes and their lambs were kept in individual pens for 2 days. Afterward they were kept in group pens with four to five ewes for approximately 5 days and then moved to larger pens containing approximately 20 ewes. After a short periodd the duration depending on the weather conditionsdthey were put out to pasture. The pasture used initially was located near the APEC. During this postparturient period the ewes were closely supervised by the APEC staff. 2.4. Data collection The APEC flock is enrolled in the Norwegian Sheep Recording System and individual animal characteristics and breeding, lambing, and disease events are registered in the database of this system. However, we decided to collect data mainly from protocols and electronic files kept at APEC, which contained all data reported to the Sheep Recording System and additional details on the health and performance of the individual animals. Disease episodes were recorded in standardized health cards used in flocks enrolled in the Sheep Recording System, containing date, diagnosis code, and additional information on clinical signs and treatment. Each animal wears an ear-tag and has a unique identification consisting of numbers for the flock and the year of birth and an individual number within year of birth. Identities of the dam and, if cross-fostered, the foster dam are recorded for each individual animal. Data available for the ewes included their breed, disease episodes, the date and course of lambing, litter size and, when fetal death occurred, the number of stillborn lambs. Records for the lambs included the date of birth and liveweights recorded at birth, in late spring and in the autumn. For lambs that died, the date and the cause of death were recorded. Veterinary service at APEC is provided by the Ambulatory Clinic at the Norwegian School of Veterinary Science. If there are signs of some disorder, the staff at APEC calls the Ambulatory Clinic. In the Ambulatory Clinic, computerized detailed case records concerning examinations and treatments of ewes and lambs were available. A systematic and thorough search of case records was performed to obtain possible additional information on the ewes and lambs in our study. 2.5. Outcome variables Several outcome variables were studied. Ewe mortality and complications after delivery were detailed and ewes that died while on pasture recorded. For the lambs in the two delivery groups, we compared the proportions of stillborn lambs (comprising antepartum and intrapartum fetal deaths) and lambs that died between the times of birth and weight recording in late spring and between late spring and weight recording in the autumn. We also compared growth over two time periods as measured by differences in single lamb’s weights at birth and in late spring and in late spring and the autumn, respectively. Which ewes that were culled the subsequent autumn were decided by the staff at the APEC and were based on general productivity and health criteria and not the previous mode
S. Waage, G. Wangensteen / Theriogenology 79 (2013) 486–494
of delivery. For the two delivery groups, we compared the proportion of ewes that were bred the next season, the proportion of those bred that became pregnant, and the number of lambs born to the ewes, distinguishing between stillborn and live-born lambs. We also compared birth weights of the lambs. 2.6. Statistical methods SD was used as measure of variability. Crude relationships between dichotomous responses and delivery group were tested using the chi-square test. Fisher’s exact test was used when sample size requirements for the chisquare test were not met. The Mantel–Haenszel chisquare statistic was used to compare the distributions of ordinally scaled categorical variables. General linear mixed models were used to estimate the relationship between mode of delivery of a ewe and weight gain of the lamb(s) she reared. All lambs reared by a ewe in the particular period, including those that were fostered, were included. With a few exceptions, weights at birth, in late spring, and in autumn were available for lambs that were alive at those times. Growth of lambs reared by the same ewe would be expected to be correlated, and ewe was included as a random factor in the models. Several important confounders had to be considered. In addition to mode of delivery, we included breed and age of the ewe, litter size, lamb sex, lamb weight at the start of the period, and number of days between the times of weight recording as fixed effects in the multivariable models. Breed was dichotomized as Texel and the combination of pure Dala and Dala Texel crosses, age of ewes dichotomized as 1 and >1 year and litter size categorized as one, two and three lambs (the latter as >2 lambs when the model included quadruplet lambs). Fractional polynomials of the continuous variables, i.e., number of days between the times of weight recording and lamb weights at the start of the periods in question, were tested to determine which ones were the best fitting functional forms in the models [10]. Biologically plausible interactions between variables were tested by adding interaction terms to main effects models. Two-sided statistical tests were used, and P < 0.05 was considered significant. Software used was the FREQ and MIXED Procedures in SAS [11].
489
Table 1 Percentage distribution by baseline characteristics of 162 ewes subjected to elective cesarean section and 162 control ewes that had a normal vaginal delivery. Characteristic Breed Dala Dala Texel Texel Age (y) 1 2 3 >3 Number of lambs 1 2 3 4
Cesarean section
Vaginal delivery
50.0 40.7 9.3
53.1 38.9 8.0
19.8 23.5 34.0 22.7
25.3 37.0 14.8 22.9
28.4 51.2 19.8 0.6
29.6 60.5 9.3 0.6
(P ¼ 0.23). The total number of lambs born to ewes in the cesarean and the vaginal delivery groups was 312 and 293, respectively. Thus, the mean number of lambs born to the ewes that underwent cesarean section (1.93), stillborn lambs included, was numerically greater than the corresponding mean for the vaginal delivery group (1.81); however, the distribution of the ewes by litter size did not differ significantly between the groups (P ¼ 0.23). The distributions by breed were rather similar. For one of the ewes that underwent cesarean the staff at the APEC had estimated an incorrect lambing date (and, apparently, the udder had not been properly checked for the presence of milk), and the operation was performed 2 weeks before the expected day of lambing. In the remaining cesarean ewes, gestation length was 146.7 1.7 days, which was close to the mean of the ewes that delivered vaginally (146.4 1.7 days); this difference was not significant (P ¼ 0.2). Mean for all ewes of the Dala breed was 146.3 1.8 days, for the Texel breed 146.8 1.4 days, and for the Dala Texel crosses 146.8 1.7 days. There were no significant differences between the delivery groups within each of the three breed groups (P > 0.1; data not shown). Means according to litter size were 146.7 1.7 days for all singleton pregnancies, 146.6 1.6 days for twin pregnancies, and 146.3 1.9 days for litters >2. Differences between the delivery groups within litter sizes were not significant (P > 0.1; data not shown).
3. Results 3.2. Postdelivery ewe mortality and complications 3.1. Baseline characteristics A total of 162 ewes were subjected to cesarean section in the study period and 162 ewes that had an unassisted vaginal delivery were selected as control ewes. The number of pairs of cesarean and control ewes per year varied from 12 (in 1997) to 29 (in 1999). Characteristics of the ewes that underwent cesarean section and those that delivered vaginally are shown in Table 1. Ewe lambs selected for breeding were bred at approximately 7 months of age; thus, parity equals age in years. The overall distribution of ewes by age differed between the groups (P ¼ 0.04); however, the proportion of 1-year-old ewes did not differ
Two ewes (1.2%) died after the cesarean section. One of them, subjected to surgery in 1998, had two stillborn lambs and became ill the day after the surgery. She was treated for metritis with procaine penicillin (Penovet, approximately 50,000 IU per kg im) daily for 4 days, but died 5 days after delivery. She also suffered from chronic mastitis. Necropsy revealed peritonitis, metritis, salpingitis, and pleuritis. Another ewe experienced uterine prolapse 1 day after surgery and died before treatment could be given. One of the ewes that delivered vaginally died for unknown reasons 3 days after lambing; necropsy was not performed.
490
S. Waage, G. Wangensteen / Theriogenology 79 (2013) 486–494
Four ewes that underwent cesarean section recovered after antibacterial treatment for metritis; three of these ewes were treated in 1998. Of the remaining ewes in the cesarean delivery group, three developed a wound infection and four experienced a delayed or somewhat complicated wound healing; all of these recovered after proper treatment. By accident, a somewhat curious cause of delayed healing of the surgical wound was observed; birds were seen pecking the wound of two ewes that were let out shortly after the surgery. Retained placenta was observed in one of the ewes that underwent surgery and another was treated for mastitis 4 days postpartum. One of the ewes in the vaginal birth group was treated for metritis, three were treated for infected traumatic wounds not related to lambing, one developed vaginal prolapse 12 days postpartum, and two experienced mastitis. 3.3. Ewe and lamb survival until the subsequent autumn The proportions of stillborn lambs in the cesarean delivery group (11 out of 312, 3.5%) and the vaginal delivery group (nine out of 293, 3.1%) did not differ (P ¼ 0.8) (Table 2). The ewe that, by mistake, underwent cesarean operation 2 weeks before the expected day of lambing had two premature stillborn lambs. The remaining stillborn lambs were born close to the expected day of lambing. Cross-fostering carried out in the postpartum period affected 17 of the 162 ewes in the cesarean delivery group and eight of the 162 ewes that delivered vaginally. Two lambs were removed from each of three ewes in the cesarean delivery group, one lamb was removed from each of eight ewes in the cesarean delivery group and each of five ewes that had a vaginal birth, one lamb was fostered onto each of five ewes in the cesarean delivery group and three ewes that delivered vaginally, and two lambs were fostered onto one ewe in the cesarean delivery group. When put out to pasture, 49, 88, and 33 of the remaining 160 ewes that underwent cesarean section reared one, two, and three lambs, respectively, and 49, 103, and nine of the
Table 2 Distribution of ewes that underwent cesarean section and ewes that had an unassisted vaginal delivery (VD) by number of stillborn lambs and number of lambs they lost during the periods from lambing to weight recording in late spring and from late spring to weight recording in the autumn. Time period
Delivery group
Number of lambs per ewe 0
1
2
Prepartum
Caesarean ewes VD ewes Cesarean ewes VD ewes Cesarean ewesc VD ewesd
153 153 140 144 153 154
7 9 15 16 6 7
2 0 7 2 1 0
Birth to springa Spring to autumnb
a Mean age ( SD) of the lambs when weighed in late spring, 50.4 7.5 days. b Mean age ( SD) of the lambs when weighed in the autumn, 125.8 10.6 days. c Two cesarean ewes that died postpartum not included. d One VD ewe that died postpartum not included.
remaining 161 ewes in the vaginal delivery group reared one, two, and three lambs, respectively. Three ewes in the cesarean delivery group died during the grazing period. One of these by accident got access to barley and died from severe acidosis and other complications 15 days after the surgery. Before she died she had not shown any signs of postoperative complications. The second ewe died for unknown reasons while on pasture in the summer, 2 months after the surgery. She had not suffered any postoperative complications or shown any signs of disease before she was found dead. Thus, there were no indications that she died from complications related to surgery; however, because she was not subjected to postmortem examination, there might be some uncertainty. The third ewe showed signs of tetany 5 months after the surgery and was treated with calcium and magnesium for two consecutive days, but did not recover. Analysis of blood samples and postmortem examination did not establish a definitive diagnosis; however, there was no sign of any complications related to the cesarean section. One ewe that delivered vaginally died during the grazing period; the cause of death was not known. Between the times of birth and weight recording in late spring, 22 ewes that had undergone cesarean section lost a total of 29 lambs and 18 ewes in the vaginal birth group lost 20 lambs (Table 2). The proportions of lambs that died that were reared by ewes in the cesarean delivery group (29 out of 294, 9.9%) and by ewes in the vaginal delivery group (20 out of 282, 7.1%) did not differ (P ¼ 0.23). Only two lambs in each delivery group died later than 14 days after birth, and six lambs in each group died later than 7 days after birth. Causes of death varied, but no particular diagnosis was overrepresented in one of the delivery groups. Cases included lambs born weak that died shortly after birth (mainly in triplet litters), trauma experienced after birth, some cases of Escherichia coli intoxication (“watery mouth”), a few lambs that were euthanized because of leg problems, and some combinations of causes. Because of the close observation of newborn lambs and intervention when needed, starvation was not a primary cause of death. Between late spring and weight recording in the autumn, seven ewes in the cesarean delivery group lost eight lambs and seven ewes that delivered vaginally lost seven lambs (Table 2). 3.4. Lamb growth Means of weights and of age when weighed for the two groups of lambs are shown in Table 3. In the multivariable models with weight gains from birth to late spring and from spring to autumn as outcome variables, we included as explanatory variables mode of delivery and possible determinants available. Ewe was included as a random factor. Weight gains over the two periods were not significantly affected by mode of delivery (Tables 4 and 5). Several covariates were highly significant. Testing of fractional polynomials showed that the square root of the number of days between the weight recording events (from birth to spring and from spring to autumn) was the most appropriate form of this variable in the model. Birth
S. Waage, G. Wangensteen / Theriogenology 79 (2013) 486–494 Table 3 Descriptive data for lambs reared by ewes that underwent cesarean section and by ewes that had an unassisted vaginal delivery (VD), respectively. Variable
Weight at birth (kg) Age (d) when weighed in spring Weight in spring (kg) Age (d) when weighed in autumn Weight in autumn (kg)
Lambs of cesarean ewes
Lambs of VD ewes
N
N
Mean
277 263
5.15 50.9
SD 0.94 7.2
Mean
283 261
5.02 49.8
0.95 7.8
263 256
22.5 126.0
5.1 10.6
261 256
22.2 125.5
4.9 10.7
256
40.6
7.2
256
40.2
6.8
3.5. Subsequent reproductive efficiency of the ewes Forty-four ewes in the cesarean delivery group and 44 ewes in the vaginal birth group were slaughtered in the autumn. There was no record of any complications related to the cesarean operation among ewes that were culled in the cesarean delivery group. The total numbers of ewes that were rebred in the autumn were 113 in the cesarean delivery group and 116 in the vaginal delivery group. The proportions of ewes in the cesarean delivery group (49/162, 30.2%) and in the vaginal birth group (46/162, 28.4%) that died or were culled between delivery and the subsequent breeding season did not differ significantly (P ¼ 0.7). Of the ewes that were bred, 12 ewes that had undergone cesarean section (10.6%) and 12 ewes that had a vaginal birth (10.3%) did not conceive. The number of lambs born to those that became pregnant, including stillborn lambs, is Table 4 Multivariable adjusted effects of delivery mode and ewe and lamb characteristics on lamb weight gain between birth and weight recording in late spring estimated by fitting a linear mixed model, accounting for clustering of lambs reared by the same ewe.
Intercept Delivery Cesarean section Vaginal delivery Age of ewe 1y >1 y Breed of ewe Texel Dala and crosses Lambs per ewe 1 2 3 Sex of lamb Male Female Age of lamba Lamb birth weight (kg) a
Coefficient
28.7 0.27 Reference
95% Confidence interval Lower
Upper
33.2
24.2
0.88
Reference 2.17
0.33
Variable
Coefficient 95% Confidence interval
Intercept Delivery Cesarean section Vaginal delivery Age of ewe 1y >1 y Breed of ewe Texel Dala and crosses Lambs per ewe 1 2 3 Sex of lamb Male Female Days between weighingsa Lamb weight in spring (kg)
20.4
a
Lower
Upper
28.0
12.8
P
<0.0001
0.54
1.04
0.6
0.07
2.32
0.07
1.96 Reference
0.43
3.49
0.01
3.20 2.01 Reference
1.40 0.51
5.00 3.51
0.0006 0.009
1.79 Reference 3.60 0.07
1.14
2.44 <0.0001
2.78 0.02
4.41 <0.0001 0.16 0.12
0.25 Reference Reference 1.12
Scaled by the square root of number of days.
shown in Table 6. The ewes that had undergone cesarean section gave birth to significantly fewer lambs (1.76 0.63) than those in the vaginal delivery group (1.97 0.72); this was the result of a smaller proportion of triplets among the former. Measured potential confounders of this relationship, i.e., the distribution of the ewes by age and by litter size at the previous lambing, did not differ between the delivery groups. Twelve of the 101 ewes that had undergone cesarean section (11.9%) and four of the 104 ewes in the vaginal delivery group (3.8%) had a stillborn lamb; the proportions differed significantly (P ¼ 0.03, Mantel–Haenszel chisquare). Comparisons within subgroups according to litter size showed that the difference was mainly the result of
P
<0.0001 0.38
1.29
3.05
<0.0001
0.89 Reference
0.26
2.03
0.13
4.87 2.30 Reference
3.48 1.10
6.27 3.50
<0.0001 0.0002
1.21 Reference 4.80 1.22
0.73
1.69
<0.0001
4.26 0.90
5.35 1.55
<0.0001 <0.0001
Scaled by the square root of age in days.
Table 5 Multivariable adjusted effects of delivery mode and ewe and lamb characteristics on lamb growth between the times of weight recording in late spring and autumn estimated by fitting a linear mixed model, accounting for clustering of lambs reared by the same ewe.
SD
weight and weight in spring were entered without transformation.
Variable
491
Table 6 Distribution of 101 ewes subjected to cesarean section and 104 ewes that had a vaginal delivery (VD) by total number of lambs and number of liveborn lambs to which they gave birth in the subsequent year. Lambs per ewe
Litter sizea 1 2 3 4 Live-born lambsb 0 1 2 3 4
Cesarean ewes
VD ewes
N
%
N
%
35 55 11 0
34.7 54.5 10.9 0
27 54 22 1
26.0 51.9 21.2 1.0
2 38 55 6 0
2.0 37.6 54.5 5.9 0
0 29 54 20 1
0 27.9 51.9 19.2 1.0
a Caesarean versus vaginal delivery: P ¼ 0.03 (Mantel–Haenszel chisquare). b Caesarean versus vaginal delivery: P ¼ 0.003 (Mantel–Haenszel chisquare).
492
S. Waage, G. Wangensteen / Theriogenology 79 (2013) 486–494
a greater proportion of stillborn lambs in triplet or quadruplet litters (five out of 11 vs. two out of 23, P ¼ 0.02, Fisher’s exact test). Considering the reduced litter size and the greater number of fetal deaths, the mean number of live-born lambs per ewe was significantly lower in the cesarean delivery group (1.64 0.63) as compared with the vaginal delivery group (1.93 0.71). The distribution of live-born lambs by litter size is shown in Table 6. In ewes that had had a cesarean delivery, lambing was unaided in 73%, gentle traction was provided in 13%, and dystocia was present in 14%. The corresponding figures for ewes that had had a vaginal delivery the previous year were 74%, 6%, and 20%. The distributions within the groups were not significantly different (P ¼ 0.7, Mantel–Haenszel chisquare). Mean birth weight of the live-born lambs for which weights were available was 5.2 0.9 kg for the 129 lambs born to the ewes that had undergone cesarean section and 5.1 0.9 kg for the 153 lambs born to the ewes that delivered vaginally at the previous lambing. Estimates from a mixed model, which included ewe as a random factor and delivery group, litter size, ewe age, and lamb sex as fixed effects, are shown in Table 7. Mode of previous delivery was not significant. Breed and plausible interaction terms were not significant when added to the model. Neonatal losses within the first week of life did not differ between the ewes that had undergone cesarean section (seven out of 101 lost one lamb) and those that had had a vaginal delivery (seven out of 104 lost one lamb, two lost two lambs) (P ¼ 0.4, Mantel–Haenszel chi-square). Clinical disorders affecting the ewes that might have been related to the previous cesarean section were not observed the second year. 4. Discussion Very few studies seem to have been published on postoperative mortality and morbidity and on subsequent reproductive performance of animals subjected to elective Table 7 Multivariable adjusted effects of mode of delivery and ewe and lamb characteristics on birth weight of lambs born to ewes that had a cesarean (N ¼ 129 lambs) or an unassisted vaginal delivery (N ¼ 153 lambs) the previous lambing season, estimated by fitting a linear mixed model and accounting for clustering of lambs born to the same ewe. Variable
Coefficient
Intercept Delivery Cesarean section Vaginal delivery Age of ewe 1y >1 y Lambs per ewe 1 2 >2 Sex of lamb Male Female
4.14
95% Confidence interval Lower
P
Upper
3.81
4.47
<0.0001
0.21
0.22
0.95
0.07
0.04
0.02
1.50 0.59 Reference
1.18 0.31
1.82 0.87
<0.0001 <0.0001
0.43 Reference
0.27
0.59
<0.0001
0.007 Reference Reference 0.29
cesarean section. Some case reports are available and a few observational studies have been conducted regarding cesarean cases in which surgical intervention was considered necessary; however, in such cases, postpartum outcomes of the dam and her offspring are likely to be affected to a varying extent by prepartum disorders or intrapartum complications experienced before surgery. In this study, we therefore included as cesarean cases only ewes for which no medical or obstetrical indications for surgery were present. Two of the 162 ewes that underwent cesarean section died within a few days postsurgery. In one of them, necropsy revealed peritonitis, metritis, and salpingitis, which were likely related to surgery. Both of her fetuses were dead prepartum and she also suffered from chronic mastitis and pleuritis; the combined effects of these preoperative disorders might have contributed to the fatal outcome. The single case of uterine prolapse after cesarean section, which resulted in the death of the ewe, could be considered a matter of coincidence and, therefore, not related to the surgery. An incidence rate of this disorder in sheep, which would likely vary somewhat among different breeds, does not seem to have been reported. Nevertheless, it is noteworthy that also in a previous study, comprising 57 ewes that underwent cesarean section, a case of uterine prolapse occurred the day after surgery [5]. Thus, some connection with the surgery cannot be ruled out. Hypothetically, if an inversion of a part of a uterine horn were introduced and not corrected during the cesarean section this might have initiated the development of the full prolapse. Relatively high mortality rates have been reported for ewes that underwent cesarean section because of dystocia. Brounts et al. [8], who reported joint follow-up results for 85 ewes and 25 goats admitted to a veterinary teaching hospital because of dystocia, found that 19% died after surgery; the main causes were pregnancy toxemia and peritonitis. A field study found that 17% of ewes that underwent cesarean section died; however, the mortality rate was low (2%) for ewes that delivered live or freshly dead lambs [6]. In another field study, 5% of the ewes that underwent cesarean section died postoperatively [5]. The greater losses observed in these studies as compared with our study were likely the result of the impaired preoperative condition of many ewes. In the field, the risk of an adverse outcome increases when vaginal birth fails and there is a delay before surgery is carried out. Four ewes experienced metritis after cesarean delivery but recovered after antibiotic treatment. Three of these and the ewe that died from peritonitis and metritis underwent cesarean section one particular year, suggesting the possibility of some common risk factor for bacterial contamination of the operative field that year. The proportion of ewes that experienced metritis postoperatively (3.1%) is not particularly high. A study in Belgium comprising 1159 suckler beef cows, of which approximately 90% had had a cesarean delivery, found that 13% of the cows developed metritis [12]. In human medicine, metritis has been reported to complicate 3% to 20% of cesarean deliveries [13,14].
S. Waage, G. Wangensteen / Theriogenology 79 (2013) 486–494
Only one case of retained placenta was observed, and the ewe recovered after antibiotic treatment. Retained placenta has been reported to be a frequent complication in small ruminants that underwent cesarean section because of dystocia [8]. Three (2%) of the 160 surviving ewes suffered a wound infection after the cesarean section but were cured with antibiotic treatment. The occurrence of such complications was modest compared with what has been experienced with human cesarean cases, where rates of surgical site infections from 1.4% to 10% have been reported [14–17]. Factors related to surgical skills will likely affect prognosis of cesarean cases. In this study, the operations were performed by veterinary students with very limited surgical experience. Close professional supervision and guidance apparently have prevented major deviations from the appropriate operative procedure. However, the operation lasted longer, sometimes much longer, than what would have been the case if carried out by an experienced surgeon (Professor Emeritus of Obstetrics, Stig A. Ødegaard, personal communication, March 2012). The risk of postoperative complications, including infections, is likely to increase with increasing surgery time; however, in human medicine only a weak relation has been observed between the occurrence of complications and the duration of the cesarean delivery [18]. Fetal deaths are frequently observed when ewes undergo emergency cesarean delivery. A field study and a study based on small ruminants admitted to a teaching hospital found, respectively, that 26% and 58% of the lambs were dead at birth [5,8]. In our study, only 3.5% of the 312 lambs born to the ewes that underwent cesarean section were dead at birth, and a similar proportion occurred in the control ewes that delivered vaginally. According to the annual report from the Norwegian Sheep Recording System, the overall loss of lambs because of fetal death in the flocks enrolled in 2004 was 3.7% [19]. It could not be ruled out that the cesarean surgery and potential postoperative discomfort would interfere with the ewe’s mothering capabilities, ewe–lamb bonding and the ewe’s milk secretion and production, which might in turn affect neonatal survival and growth of the lambs. Studies addressing these questions have not been published. During the first 6 to 9 weeks after cesarean delivery the ewes in this group lost 10% of their live-born lambs. However, postnatal losses among the ewes that had a vaginal delivery were not significantly lower. According to the literature, similar or even higher neonatal mortality rates are not uncommon in sheep flocks [20,21]. Moreover, though growth rate of the lambs during this period was strongly influenced by litter size, birth weight and sex of the lamb and age of the ewe, and the mode of delivery did not have any significant effect. Similarly, lamb losses and weight gain from late spring until weight recording in the autumn did not differ between the delivery groups. Subsequent reproductive efficiency of ewes that underwent cesarean section has only to a very limited extent been addressed previously. Two studies reported high pregnancy rates for ewes that were bred the subsequent season. The proportions of the original cases that were bred were 36 out of 57 ewes [5] and only 16 out of 110
493
ewes [8], respectively. In our study, approximately 70% of the original cases in each of the two delivery groups were rebred the subsequent autumn. Of ewes that were bred, almost equal proportions of the ewes in the cesarean delivery group (10.6%) and the vaginal delivery group (10.3%) did not conceive. This is in contrast with results obtained in a study in beef cattle, where the subsequent pregnancy rate for 105 cows that had had a cesarean delivery was 27% lower than the herd average [22]. In women, there are indications that cesarean delivery is associated with increased risk of fetal death in subsequent pregnancy [23,24]. Only singleton births were investigated in these studies. Explanations proposed were placentation abnormalities caused by uterine scar and possible reduction in uterine blood supply because of ligation of vessels during the previous cesarean operation. Abdominal adhesions are not uncommon in women [25] and animals [26] that have undergone cesarean section, and subfertility is a possible consequence [27]. In our study, the mean number of live-born lambs per ewe was markedly lower for ewes that had a previous cesarean delivery than for those that had a vaginal birth (1.64 vs. 1.93). This difference was the result of a significantly lower number of lambs born per ewe in the former group and, additionally, an increased number of stillborn lambs in this group. The age distribution of the ewes and their reproductive capacity, as indicated by their litter size the previous year, might have confounded the observed difference; however, these variables did not differ between the two groups. Assuming that ovulation and fertilization were similar in the two delivery groups, unobserved early embryonic death or unsuccessful or inadequate placentation are possible explanations. Because the proportion of ewes that conceived in each of the two groups was similar, and the decreased number of live-born lambs in the cesarean delivery group was mainly the result of a reduced number of triplets, it is tempting to suggest that in some cases the previous cesarean operation caused structural changes that restricted the uterine capacity for triplet pregnancy. It might be a matter of discussion whether the reduced number of triplet births should be considered an important side effect of cesarean section. Most sheep holders seem to prefer twin births because rearing triplets might sometimes be too demanding for a ewe. In women who had a cesarean delivery, mean birth weight of babies in subsequent pregnancy has been reported to be somewhat lower than that of babies born to women who had previously delivered vaginally [23]. When comparing birth weights of the live-born lambs delivered the subsequent year by the ewes in the two delivery groups, and accounting for measured confounders, we found no difference between the groups. Several factors might affect the risk of peripartum complications and postpartum performance of ewes and their offspring. Therefore, when comparing short-term and long-term outcomes of ewes and lambs subjected to different modes of delivery, possible confounders must be considered. Nutrition, environmental conditions, and management practices are known to be important determinants of health and productivity of sheep. Because our study was limited to a single flock, kept at an experimental
494
S. Waage, G. Wangensteen / Theriogenology 79 (2013) 486–494
station with well-established common management practices, significant confounding because of differential exposure to such factors between the two study groups, which could have affected their outcomes differently, is not very likely. To further reduce potential time-related confounding we selected as the control for each cesarean case the ewe that lambed closest in time. Main residual confounding was assumed to be the result of individual animal differences, and individual characteristics expected to be important determinants for the outcome variables of interest were accounted for in the analyses where appropriate. When analyzing lamb weights and growth we specified ewes as clusters. In doing so we not only accounted for the potential intralitter correlation caused by common maternal effects but also the litter-specific environmental factors and attention paid to the lambs by the staff, in particular while kept in the individual pen. The quality of available data is essential in retrospective studies. Data used in this study were collected from records at the APEC. These were recorded by skilled personnel and were generally intended for scientific purposes. Disease data were cross-checked with records at the Ambulatory Clinic, the Norwegian School of Veterinary Science, who provided the daily veterinary care at APEC. Thus, there are good reasons to consider the data we used as being reliable and precise. 5. Conclusions There was a moderate frequency of postoperative complications experienced by ewes that underwent elective cesarean section. The proportion of stillborn lambs and postnatal survival and growth of lambs born to these ewes did not differ from corresponding outcomes for lambs born vaginally. Pregnancy rate in the subsequent breeding season was not reduced in ewes that had had an elective cesarean delivery at term; however, as a result of smaller litters and an increased number of fetal deaths, these ewes gave birth to fewer live-born lambs than ewes that had delivered vaginally. Acknowledgments The authors thank Stig Anders Ødegaard, Professor Emeritus of Obstetrics, Norwegian School of Veterinary Science, who was in charge of student education in obstetrics during the study period, and the staff at the Animal Production Experimental Centre, Ås, in particular the present leader, Agnes Klouman, for excellent management of the sheep flock in the study period. The study was funded mainly by the Norwegian School of Veterinary Science, with some additional contributions from the Agricultural Agreement Research Fund, the Foundation for Research Levy on Agricultural Products, Animalia – Norwegian Meat and Poultry Research Centre and the Norwegian Sheep and Goat Association.
References [1] Whitelaw A, Watchorn P. An investigation into dystocia in a south country Cheviot flock. Vet Rec 1975;97:489–92. [2] George JM. The incidence of dystocia in Dorset Horn ewes. Aust Vet J 1976;52:519–23. [3] Grommers FJ, Elving L, Van Eldik P. Parturition difficulties in sheep. Anim Reprod Sci 1985;9:365–74. [4] Wiener G, Woolliams C, Macleod NSM. The effects of breed, breeding system and other factors on lamb mortality. 1. Causes of death and effects on the incidence of losses. J Agric Sci (Camb) 1983; 100:539–51. [5] Mosdøl G. Indications for and fertility after caesarean section in the ewe in common practice. Norsk Veterinærtidsskrift 1986;98: 441–4. [6] Scott PR. Ovine caesarean operations: a study of 137 field cases. Br Vet J 1989;145:558–64. [7] Majeed AF, Taha MB, Azawi OI. Cesarean section in Iraqi Awassi ewes: a case study. Theriogenology 1993;40:435–9. [8] Brounts SH, Hawkins JF, Baird AN, Glickman LT. Outcome and subsequent fertility of sheep and goats undergoing cesarean section because of dystocia: 110 cases (1981-2001). J Am Vet Med Assoc 2004;224:275–9. [9] Noakes DE, Parkinson TJ, England GCW. Veterinary reproduction and obstetrics. Ninth Edition. Edinburgh: Saunders; 2009. p. 354, 367. [10] Royston P, Ambler G, Sauerbrei W. The use of fractional polynomials to model continuous risk variables in epidemiology. Int J Epidemiol 1999;28:964–74. [11] SAS Institute Inc. SAS/STAT 9.2 user’s guide. Second Edition. Cary, NC: SAS Institute Inc; 2009. p. 1675–831, 3891–4092. [12] Hanzen C, Laurent Y, Ward WR. Comparison of reproductive performance in Belgian dairy and beef cattle. Theriogenology 1994; 41:1099–114. [13] Chaim W, Bashiri A, Bar-David J, Shoham-Vardi I, Mazor M. Prevalence and clinical significance of postpartum endometritis and wound infection. Infect Dis Obstet Gynecol 2000;8:77–82. [14] Spandorfer SD, Graham E, Forouzan I. Postcesarean endometritis. Clinical risk factors predictive of positive blood cultures. J Reprod Med 1996;41:797–800. [15] Hannah ME, Hannah WJ, Hewson SA, Hodnett ED, Saigal S, Willan AR. Planned caesarean section versus planned vaginal birth for breech presentation at term: a randomised multicentre trial. Lancet 2000;356:1375–83. [16] Tran TS, Jamulitrat S, Chongsuvivatwong V, Geater A. Risk factors for postcesarean surgical site infection. Obstet Gynecol 2000;95:367–71. [17] Yokoe DS, Noskin GA, Cunnigham SM, Zuccotti G, Plaskett T, Fraser VJ, et al. Enhanced identification of postoperative infections among inpatients. Emerg Infect Dis 2004;10:1924–30. [18] Killian CA, Graffunder EM, Vinciguerra TJ, Venezia RA. Risk factors for surgical-site infections following cesarean section. Infect Control Hosp Epidemiol 2001;22:613–7. [19] Offerdal E, editor. The Norwegian Sheep Recording System, Annual Report 2004. Norwegian Meat and Poultry Research Centre, Oslo, 2005. [20] Binns SH, Cox IJ, Rizvi S, Green LE. Risk factors for lamb mortality on UK sheep farms. Prev Vet Med 2002;52:287–303. [21] Huffman EM, Kirk JH, Pappaioanou M. Factors associated with neonatal lamb mortality. Theriogenology 1985;24:163–71. [22] Patterson DJ, Bellows RA, Burfening PJ. Effects of caesarean section, retained placenta and vaginal or uterine prolapse on subsequent fertility in beef cattle. J Anim Sci 1981;53:916–21. [23] Smith GCS, Pell JP, Dobbie R. Caesarean section and risk of unexplained stillbirth in subsequent pregnancy. Lancet 2003;362:1779–84. [24] Gray R, Quigley MA, Hockley C, Kurinczuk JJ, Goldacre M, Brocklehurst P. Caesarean delivery and risk of stillbirth in subsequent pregnancy: a retrospective cohort study in an English population. BJOG 2007;114:264–70. [25] Lyell DJ. Adhesions and perioperative complications of repeat cesarean delivery. Am J Obstet Gynecol 2011;205(Suppl.):S11–8. [26] Sloss V, Dufty JH. Elective caesarean operation in Hereford cattle. Aust Vet J 1977;53:420–4. [27] Silver RM. Delivery after previous cesarean: long-term maternal outcomes. Semin Perinatol 2010;34:258–66.
Contents lists available at SciVerse ScienceDirect
Theriogenology journal homepage: www.theriojournal.com
Short-term and long-term outcomes of ewes and their offspring after elective cesarean section Steinar Waage*, Guri Wangensteen 1 Department of Production Animal Clinical Sciences, Norwegian School of Veterinary Science, Oslo, Norway
a r t i c l e i n f o
a b s t r a c t
Article history: Received 18 April 2012 Received in revised form 29 October 2012 Accepted 1 November 2012
Postoperative outcomes of animals that have undergone cesarean delivery have been reported previously; however, in most studies results were influenced by a combination of surgery per se and the preoperative condition of the animal, which was frequently impaired because of the presence of dystocia. To evaluate the effects of the cesarean section itself we conducted a matched cohort study comparing postpartum complications and future reproductive performance of 162 ewes subjected to elective cesarean section and 162 ewes that had an unassisted vaginal delivery. Survival and subsequent growth of their lambs were also compared. Effect of mode of delivery on weight gain was estimated using linear mixed models. Case ewes, which underwent surgery during the period from 1996 through 2004, and control ewes were from the flock at the Animal Production Experimental Centre, Norway. Two ewes (1.2%) that underwent cesarean section died; one developed peritonitis and the other experienced uterine prolapse and did not recover. Postoperatively, four ewes suffered from metritis, three suffered a wound infection, and four a delayed wound healing; all recovered after treatment. One of the ewes that delivered vaginally died 3 days after lambing. The incidences of fetal and postnatal deaths did not differ significantly between the cesarean and the vaginal delivery groups (fetal deaths, 3.5% and 3.1%, and postnatal deaths, 9.9% and 7.1%, respectively). Survival rates and weight gains of the lambs the subsequent months were similar for the two groups. Seventy percent of the ewes that had a cesarean section and 72% of those that had a vaginal delivery were bred the next season; conception rates were 89% and 90%, respectively. However, the ewes subjected to surgery the previous year gave birth to significantly fewer live-born lambs (mean, 1.64) than those that had had a vaginal delivery (1.93). The difference was the result of a reduced litter size and an increased number of fetal deaths in the former group. Birth weights of the live-born lambs the second year did not differ between the groups. In conclusion, severe short-term complications were rare among the ewes that underwent elective cesarean section. Survival and growth of their lambs and their conception rate the subsequent season did not differ from the corresponding outcomes of the ewes that delivered vaginally, but their fertility was reduced in the sense that they gave birth to fewer live-born lambs the following lambing season. Ó 2013 Elsevier Inc. All rights reserved.
Keywords: Sheep Cesarean section Reproductive performance Lamb survival Lamb growth
1. Introduction
* Corresponding author. Tel.: þ47 22 964 875; fax: þ47 22 597 083. E-mail address: [email protected] (S. Waage). 1 Present address: Veterinary Surgeon, Fiskå, Norway. 0093-691X/$ – see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.theriogenology.2012.11.001
Dystocia is common in sheep. However, large breed differences in the incidence have been reported, varying from approximately 1% in a study of Scottish Blackface [1] to 34% in Dorset Horn ewes [2]. In a study of Texel Sheep, Milk Sheep, and various crossbreeds, in which the birth of
S. Waage, G. Wangensteen / Theriogenology 79 (2013) 486–494
each lamb was recorded, 32% of the lambs experienced a difficult parturition [3]. According to records for 2010 in the Norwegian Sheep Recording System, a complicated birth that required assistance occurred in 13% of 270,850 lambings, and some modest assistance (gentle traction) was provided in an additional 12% of the lambings (S. Waage, unpublished data). Apart from the occasional adverse consequences for the ewe, dystocia is a major cause of perinatal lamb mortality [4]. In most cases in which assistance is needed, manual correction and vaginal delivery is achieved. Examples are the common malposture dystocias involving the head or forelimbs and the more rare breech and transverse presentations. However, there are types of dystocia for which assisted vaginal delivery might fail and cesarean section is the alternative obstetric intervention. Fetopelvic disproportion, which occurs most frequently in singleton births of male lambs in primiparous ewes of certain breeds, is reported to be a major cause of cesarean section [5–7]. Also, failure of or incomplete dilatation of the cervix (“ring womb”) is a relatively frequent cause of cesarean section [7,8]. More rare indications are fetal retention, usually in cases of multiple births, irreducible uterine torsion, and fetal malformation. When vaginal birth fails, some might consider euthanasia an alternative to cesarean section. Of interest to the sheep farmer, apart from animal welfare aspects, is the economic outcome after cesarean section, which depends on the costs of surgery and the outcomes of the ewe and her offspring. Short-term effects with respect to complications and survival and subsequent performance must be taken into account. Of main interest in meat sheep are future fertility of the ewe and her mothering capabilities, which to some extent might be reflected by the growth rate of her lambs. Thus, to assess the effect of the cesarean section, cases should be monitored for a considerable time period after surgery. However, studies of cases subjected to cesarean section because of dystocia would obviously include some ewes with an impaired prognosis because of prolonged labor before surgery and results would reflect the combined effects of the surgery and potentially previous straining. Short-term outcomes reported after cesarean delivery in dystocia cases in small ruminants are generally not very encouraging. Three studies reported mortality rates for surgical cases of 5%, 17%, and 19%, respectively [5,6,8]. Two of these studies also recorded fetal survival, which was poor; 26% [5] and 58% [8] of the fetuses were dead. Future fertility of ewes subjected to cesarean section has been studied to a very limited extent. Two previous studies reported high pregnancy rates, but the proportion selected for subsequent breeding of those subjected to surgery was relatively small [5,8]. Postnatal survival and growth rate of the lambs of ewes that had undergone cesarean section do not seem to have been investigated. Of particular interest is to determine whether or not there are any adverse effects of the surgery itself. To evaluate the effects of cesarean delivery per se on postpartum health and survival and on future performance, we therefore conducted a retrospective study of ewes subjected to elective cesarean section. Detailed postoperative
487
and long-term outcomes for these ewes were compared with outcomes for ewes that had an unassisted vaginal delivery. 2. Materials and methods 2.1. Study design and animals included We conducted a retrospective matched cohort study including as cases all ewes that were subjected to elective cesarean section at the Animal Production Experimental Centre (APEC), Ås, Norway, during the period from 1996 through 2004. These ewes served as teaching cases for veterinary students in their surgical training in obstetrics. Ewes selected for this purpose were those with expected date of lambing close to the predetermined days of the teaching program; other criteria were not considered. No medical or obstetric indication for cesarean delivery was present. For each ewe subjected to surgery we used as control an ewe that had an unassisted vaginal birth. Among the candidates for being the control, we selected the ewe that lambed closest in time to the ewe that underwent cesarean section. All ewes included in the study were clinically healthy at the time of parturition. Permission to perform elective cesarean section as part of the student teaching program in obstetrics has been given by the Norwegian Animal Research Authority. 2.2. Experimental station, feeding, and management practices The APEC, owned by the University of Life Sciences and the Norwegian School of Veterinary Science, is the main experimental station for production animals in Norway and is located at the University Campus in Ås. The annual number of lambing ewes in the flock in the study period varied between 150 and 250. Nutrition and management routines at the APEC during pregnancy and around lambing did not change much from year to year during the study period. The ewes were removed from pasture in late autumn, approximately 1 month before breeding, and kept indoors in group pens during pregnancy with voluntary access to a limited outdoor area. Roughages fed to the pregnant ewes were grass silage, partly stored as round bales, and hay. The breeding season started in the end of November and continued until late December. Each day, a ram was led through the pens to detect ewes in estrus. Ewes in estrus were mated with a predetermined ram selected among the five to six rams available in the flock. Ewes mated were recorded by the staff; hence, the approximate day of lambing could be determined. Ultrasound scanning was performed, and the ewes were grouped according to age and the number of fetuses. Barley and a commercial concentrate product were given to nulliparous ewes during the entire pregnancy, and older ewes received concentrates only during the final weeks of pregnancy. Amounts of concentrate differed according to age of the ewe and the number of fetuses she carried. Body condition scoring was performed to check whether nutrient intake was sufficient. Mineral supplementation was provided using a commercial product.
488
S. Waage, G. Wangensteen / Theriogenology 79 (2013) 486–494
Shortly before expected lambing, ewes were placed in individual pens with floor covered with fresh, dry straw and were monitored day and night. Neonatal lambs were observed and sucking assistance was provided if a lamb was reluctant or failed to suck its dam. If considered necessary, lambs were given colostrum by stomach tube. When possible, cross-fostering of surplus lambs was practiced when litters were large. The sheep were put out to pasture in May. In mid-June, the flock was routinely collected for antiparasitic treatment, and the lambs were weighed. Lambs were also weighed approximately 2.5 months later. Antiparasitic treatment was repeated when the outdoor period ended in mid-October. 2.3. Cesarean section Cesarean section was performed on approximately day 147 of gestation; it was required that there was milk in the udder. The operations were performed under “field conditions”, using a storage room at the APEC, and were carried out by veterinary students as part of their final training in obstetrics. Two students performed the cesarean section in each ewe and different students were involved in each case. They were closely supervised by the Professor of Obstetrics, Department of Production Animal Clinical Sciences, and occasionally by an Associate Professor in the same department. The perioperative procedures, including the method used for anesthesia and the surgical technique, did not vary during the study period. The ewe was standing in a box with metal mesh floor, a metal front with an adjustable opening, giving the ewe access to a feeding trough, but preventing retraction of the head. The right side of the box was against the wall. To allow a proper working posture, the box was somewhat elevated above the surrounding floor. As is recommended [9] benzylpenicillin sodium iv (approximately 50,000 IU per kg) was administered before surgery. The left sublumbar region was sheared and shaved and the skin washed and disinfected. An inverted-L block was used for local anesthesia [9], administering a total of 30 to 40 mL 2% lidocaine hydrochloride with adrenaline. A sterile plastic drape was used to cover the flank, using an adhesive spray to achieve adherence to the skin. The incision, approximately 20 cm in length and running approximately vertically, was in the midparalumbar fossa. The uterine wall was incised along the greater curvature of a pregnant horn, avoiding placentomes. In cases of multiple pregnancies, all lambs were usually removed through a single incision; in a few cases both horns were incised. No attempt was made to remove fetal membranes. The uterine incision was closed with a continuous, inverting pattern, either Cushing suture or modified Lembert, using absorbable material. Peritoneum and the two inner muscle layers were closed with a single continuous suture and the external muscle layer with another continuous suture, using absorbable material. The skin incision was closed with a Ford interlocking suture pattern, using nonabsorbable material. The newborn lambs were put in the trough in front of the ewe and taken care of by the ewe.
After the cesarean or vaginal delivery, the ewes and their lambs were kept in individual pens for 2 days. Afterward they were kept in group pens with four to five ewes for approximately 5 days and then moved to larger pens containing approximately 20 ewes. After a short periodd the duration depending on the weather conditionsdthey were put out to pasture. The pasture used initially was located near the APEC. During this postparturient period the ewes were closely supervised by the APEC staff. 2.4. Data collection The APEC flock is enrolled in the Norwegian Sheep Recording System and individual animal characteristics and breeding, lambing, and disease events are registered in the database of this system. However, we decided to collect data mainly from protocols and electronic files kept at APEC, which contained all data reported to the Sheep Recording System and additional details on the health and performance of the individual animals. Disease episodes were recorded in standardized health cards used in flocks enrolled in the Sheep Recording System, containing date, diagnosis code, and additional information on clinical signs and treatment. Each animal wears an ear-tag and has a unique identification consisting of numbers for the flock and the year of birth and an individual number within year of birth. Identities of the dam and, if cross-fostered, the foster dam are recorded for each individual animal. Data available for the ewes included their breed, disease episodes, the date and course of lambing, litter size and, when fetal death occurred, the number of stillborn lambs. Records for the lambs included the date of birth and liveweights recorded at birth, in late spring and in the autumn. For lambs that died, the date and the cause of death were recorded. Veterinary service at APEC is provided by the Ambulatory Clinic at the Norwegian School of Veterinary Science. If there are signs of some disorder, the staff at APEC calls the Ambulatory Clinic. In the Ambulatory Clinic, computerized detailed case records concerning examinations and treatments of ewes and lambs were available. A systematic and thorough search of case records was performed to obtain possible additional information on the ewes and lambs in our study. 2.5. Outcome variables Several outcome variables were studied. Ewe mortality and complications after delivery were detailed and ewes that died while on pasture recorded. For the lambs in the two delivery groups, we compared the proportions of stillborn lambs (comprising antepartum and intrapartum fetal deaths) and lambs that died between the times of birth and weight recording in late spring and between late spring and weight recording in the autumn. We also compared growth over two time periods as measured by differences in single lamb’s weights at birth and in late spring and in late spring and the autumn, respectively. Which ewes that were culled the subsequent autumn were decided by the staff at the APEC and were based on general productivity and health criteria and not the previous mode
S. Waage, G. Wangensteen / Theriogenology 79 (2013) 486–494
of delivery. For the two delivery groups, we compared the proportion of ewes that were bred the next season, the proportion of those bred that became pregnant, and the number of lambs born to the ewes, distinguishing between stillborn and live-born lambs. We also compared birth weights of the lambs. 2.6. Statistical methods SD was used as measure of variability. Crude relationships between dichotomous responses and delivery group were tested using the chi-square test. Fisher’s exact test was used when sample size requirements for the chisquare test were not met. The Mantel–Haenszel chisquare statistic was used to compare the distributions of ordinally scaled categorical variables. General linear mixed models were used to estimate the relationship between mode of delivery of a ewe and weight gain of the lamb(s) she reared. All lambs reared by a ewe in the particular period, including those that were fostered, were included. With a few exceptions, weights at birth, in late spring, and in autumn were available for lambs that were alive at those times. Growth of lambs reared by the same ewe would be expected to be correlated, and ewe was included as a random factor in the models. Several important confounders had to be considered. In addition to mode of delivery, we included breed and age of the ewe, litter size, lamb sex, lamb weight at the start of the period, and number of days between the times of weight recording as fixed effects in the multivariable models. Breed was dichotomized as Texel and the combination of pure Dala and Dala Texel crosses, age of ewes dichotomized as 1 and >1 year and litter size categorized as one, two and three lambs (the latter as >2 lambs when the model included quadruplet lambs). Fractional polynomials of the continuous variables, i.e., number of days between the times of weight recording and lamb weights at the start of the periods in question, were tested to determine which ones were the best fitting functional forms in the models [10]. Biologically plausible interactions between variables were tested by adding interaction terms to main effects models. Two-sided statistical tests were used, and P < 0.05 was considered significant. Software used was the FREQ and MIXED Procedures in SAS [11].
489
Table 1 Percentage distribution by baseline characteristics of 162 ewes subjected to elective cesarean section and 162 control ewes that had a normal vaginal delivery. Characteristic Breed Dala Dala Texel Texel Age (y) 1 2 3 >3 Number of lambs 1 2 3 4
Cesarean section
Vaginal delivery
50.0 40.7 9.3
53.1 38.9 8.0
19.8 23.5 34.0 22.7
25.3 37.0 14.8 22.9
28.4 51.2 19.8 0.6
29.6 60.5 9.3 0.6
(P ¼ 0.23). The total number of lambs born to ewes in the cesarean and the vaginal delivery groups was 312 and 293, respectively. Thus, the mean number of lambs born to the ewes that underwent cesarean section (1.93), stillborn lambs included, was numerically greater than the corresponding mean for the vaginal delivery group (1.81); however, the distribution of the ewes by litter size did not differ significantly between the groups (P ¼ 0.23). The distributions by breed were rather similar. For one of the ewes that underwent cesarean the staff at the APEC had estimated an incorrect lambing date (and, apparently, the udder had not been properly checked for the presence of milk), and the operation was performed 2 weeks before the expected day of lambing. In the remaining cesarean ewes, gestation length was 146.7 1.7 days, which was close to the mean of the ewes that delivered vaginally (146.4 1.7 days); this difference was not significant (P ¼ 0.2). Mean for all ewes of the Dala breed was 146.3 1.8 days, for the Texel breed 146.8 1.4 days, and for the Dala Texel crosses 146.8 1.7 days. There were no significant differences between the delivery groups within each of the three breed groups (P > 0.1; data not shown). Means according to litter size were 146.7 1.7 days for all singleton pregnancies, 146.6 1.6 days for twin pregnancies, and 146.3 1.9 days for litters >2. Differences between the delivery groups within litter sizes were not significant (P > 0.1; data not shown).
3. Results 3.2. Postdelivery ewe mortality and complications 3.1. Baseline characteristics A total of 162 ewes were subjected to cesarean section in the study period and 162 ewes that had an unassisted vaginal delivery were selected as control ewes. The number of pairs of cesarean and control ewes per year varied from 12 (in 1997) to 29 (in 1999). Characteristics of the ewes that underwent cesarean section and those that delivered vaginally are shown in Table 1. Ewe lambs selected for breeding were bred at approximately 7 months of age; thus, parity equals age in years. The overall distribution of ewes by age differed between the groups (P ¼ 0.04); however, the proportion of 1-year-old ewes did not differ
Two ewes (1.2%) died after the cesarean section. One of them, subjected to surgery in 1998, had two stillborn lambs and became ill the day after the surgery. She was treated for metritis with procaine penicillin (Penovet, approximately 50,000 IU per kg im) daily for 4 days, but died 5 days after delivery. She also suffered from chronic mastitis. Necropsy revealed peritonitis, metritis, salpingitis, and pleuritis. Another ewe experienced uterine prolapse 1 day after surgery and died before treatment could be given. One of the ewes that delivered vaginally died for unknown reasons 3 days after lambing; necropsy was not performed.
490
S. Waage, G. Wangensteen / Theriogenology 79 (2013) 486–494
Four ewes that underwent cesarean section recovered after antibacterial treatment for metritis; three of these ewes were treated in 1998. Of the remaining ewes in the cesarean delivery group, three developed a wound infection and four experienced a delayed or somewhat complicated wound healing; all of these recovered after proper treatment. By accident, a somewhat curious cause of delayed healing of the surgical wound was observed; birds were seen pecking the wound of two ewes that were let out shortly after the surgery. Retained placenta was observed in one of the ewes that underwent surgery and another was treated for mastitis 4 days postpartum. One of the ewes in the vaginal birth group was treated for metritis, three were treated for infected traumatic wounds not related to lambing, one developed vaginal prolapse 12 days postpartum, and two experienced mastitis. 3.3. Ewe and lamb survival until the subsequent autumn The proportions of stillborn lambs in the cesarean delivery group (11 out of 312, 3.5%) and the vaginal delivery group (nine out of 293, 3.1%) did not differ (P ¼ 0.8) (Table 2). The ewe that, by mistake, underwent cesarean operation 2 weeks before the expected day of lambing had two premature stillborn lambs. The remaining stillborn lambs were born close to the expected day of lambing. Cross-fostering carried out in the postpartum period affected 17 of the 162 ewes in the cesarean delivery group and eight of the 162 ewes that delivered vaginally. Two lambs were removed from each of three ewes in the cesarean delivery group, one lamb was removed from each of eight ewes in the cesarean delivery group and each of five ewes that had a vaginal birth, one lamb was fostered onto each of five ewes in the cesarean delivery group and three ewes that delivered vaginally, and two lambs were fostered onto one ewe in the cesarean delivery group. When put out to pasture, 49, 88, and 33 of the remaining 160 ewes that underwent cesarean section reared one, two, and three lambs, respectively, and 49, 103, and nine of the
Table 2 Distribution of ewes that underwent cesarean section and ewes that had an unassisted vaginal delivery (VD) by number of stillborn lambs and number of lambs they lost during the periods from lambing to weight recording in late spring and from late spring to weight recording in the autumn. Time period
Delivery group
Number of lambs per ewe 0
1
2
Prepartum
Caesarean ewes VD ewes Cesarean ewes VD ewes Cesarean ewesc VD ewesd
153 153 140 144 153 154
7 9 15 16 6 7
2 0 7 2 1 0
Birth to springa Spring to autumnb
a Mean age ( SD) of the lambs when weighed in late spring, 50.4 7.5 days. b Mean age ( SD) of the lambs when weighed in the autumn, 125.8 10.6 days. c Two cesarean ewes that died postpartum not included. d One VD ewe that died postpartum not included.
remaining 161 ewes in the vaginal delivery group reared one, two, and three lambs, respectively. Three ewes in the cesarean delivery group died during the grazing period. One of these by accident got access to barley and died from severe acidosis and other complications 15 days after the surgery. Before she died she had not shown any signs of postoperative complications. The second ewe died for unknown reasons while on pasture in the summer, 2 months after the surgery. She had not suffered any postoperative complications or shown any signs of disease before she was found dead. Thus, there were no indications that she died from complications related to surgery; however, because she was not subjected to postmortem examination, there might be some uncertainty. The third ewe showed signs of tetany 5 months after the surgery and was treated with calcium and magnesium for two consecutive days, but did not recover. Analysis of blood samples and postmortem examination did not establish a definitive diagnosis; however, there was no sign of any complications related to the cesarean section. One ewe that delivered vaginally died during the grazing period; the cause of death was not known. Between the times of birth and weight recording in late spring, 22 ewes that had undergone cesarean section lost a total of 29 lambs and 18 ewes in the vaginal birth group lost 20 lambs (Table 2). The proportions of lambs that died that were reared by ewes in the cesarean delivery group (29 out of 294, 9.9%) and by ewes in the vaginal delivery group (20 out of 282, 7.1%) did not differ (P ¼ 0.23). Only two lambs in each delivery group died later than 14 days after birth, and six lambs in each group died later than 7 days after birth. Causes of death varied, but no particular diagnosis was overrepresented in one of the delivery groups. Cases included lambs born weak that died shortly after birth (mainly in triplet litters), trauma experienced after birth, some cases of Escherichia coli intoxication (“watery mouth”), a few lambs that were euthanized because of leg problems, and some combinations of causes. Because of the close observation of newborn lambs and intervention when needed, starvation was not a primary cause of death. Between late spring and weight recording in the autumn, seven ewes in the cesarean delivery group lost eight lambs and seven ewes that delivered vaginally lost seven lambs (Table 2). 3.4. Lamb growth Means of weights and of age when weighed for the two groups of lambs are shown in Table 3. In the multivariable models with weight gains from birth to late spring and from spring to autumn as outcome variables, we included as explanatory variables mode of delivery and possible determinants available. Ewe was included as a random factor. Weight gains over the two periods were not significantly affected by mode of delivery (Tables 4 and 5). Several covariates were highly significant. Testing of fractional polynomials showed that the square root of the number of days between the weight recording events (from birth to spring and from spring to autumn) was the most appropriate form of this variable in the model. Birth
S. Waage, G. Wangensteen / Theriogenology 79 (2013) 486–494 Table 3 Descriptive data for lambs reared by ewes that underwent cesarean section and by ewes that had an unassisted vaginal delivery (VD), respectively. Variable
Weight at birth (kg) Age (d) when weighed in spring Weight in spring (kg) Age (d) when weighed in autumn Weight in autumn (kg)
Lambs of cesarean ewes
Lambs of VD ewes
N
N
Mean
277 263
5.15 50.9
SD 0.94 7.2
Mean
283 261
5.02 49.8
0.95 7.8
263 256
22.5 126.0
5.1 10.6
261 256
22.2 125.5
4.9 10.7
256
40.6
7.2
256
40.2
6.8
3.5. Subsequent reproductive efficiency of the ewes Forty-four ewes in the cesarean delivery group and 44 ewes in the vaginal birth group were slaughtered in the autumn. There was no record of any complications related to the cesarean operation among ewes that were culled in the cesarean delivery group. The total numbers of ewes that were rebred in the autumn were 113 in the cesarean delivery group and 116 in the vaginal delivery group. The proportions of ewes in the cesarean delivery group (49/162, 30.2%) and in the vaginal birth group (46/162, 28.4%) that died or were culled between delivery and the subsequent breeding season did not differ significantly (P ¼ 0.7). Of the ewes that were bred, 12 ewes that had undergone cesarean section (10.6%) and 12 ewes that had a vaginal birth (10.3%) did not conceive. The number of lambs born to those that became pregnant, including stillborn lambs, is Table 4 Multivariable adjusted effects of delivery mode and ewe and lamb characteristics on lamb weight gain between birth and weight recording in late spring estimated by fitting a linear mixed model, accounting for clustering of lambs reared by the same ewe.
Intercept Delivery Cesarean section Vaginal delivery Age of ewe 1y >1 y Breed of ewe Texel Dala and crosses Lambs per ewe 1 2 3 Sex of lamb Male Female Age of lamba Lamb birth weight (kg) a
Coefficient
28.7 0.27 Reference
95% Confidence interval Lower
Upper
33.2
24.2
0.88
Reference 2.17
0.33
Variable
Coefficient 95% Confidence interval
Intercept Delivery Cesarean section Vaginal delivery Age of ewe 1y >1 y Breed of ewe Texel Dala and crosses Lambs per ewe 1 2 3 Sex of lamb Male Female Days between weighingsa Lamb weight in spring (kg)
20.4
a
Lower
Upper
28.0
12.8
P
<0.0001
0.54
1.04
0.6
0.07
2.32
0.07
1.96 Reference
0.43
3.49
0.01
3.20 2.01 Reference
1.40 0.51
5.00 3.51
0.0006 0.009
1.79 Reference 3.60 0.07
1.14
2.44 <0.0001
2.78 0.02
4.41 <0.0001 0.16 0.12
0.25 Reference Reference 1.12
Scaled by the square root of number of days.
shown in Table 6. The ewes that had undergone cesarean section gave birth to significantly fewer lambs (1.76 0.63) than those in the vaginal delivery group (1.97 0.72); this was the result of a smaller proportion of triplets among the former. Measured potential confounders of this relationship, i.e., the distribution of the ewes by age and by litter size at the previous lambing, did not differ between the delivery groups. Twelve of the 101 ewes that had undergone cesarean section (11.9%) and four of the 104 ewes in the vaginal delivery group (3.8%) had a stillborn lamb; the proportions differed significantly (P ¼ 0.03, Mantel–Haenszel chisquare). Comparisons within subgroups according to litter size showed that the difference was mainly the result of
P
<0.0001 0.38
1.29
3.05
<0.0001
0.89 Reference
0.26
2.03
0.13
4.87 2.30 Reference
3.48 1.10
6.27 3.50
<0.0001 0.0002
1.21 Reference 4.80 1.22
0.73
1.69
<0.0001
4.26 0.90
5.35 1.55
<0.0001 <0.0001
Scaled by the square root of age in days.
Table 5 Multivariable adjusted effects of delivery mode and ewe and lamb characteristics on lamb growth between the times of weight recording in late spring and autumn estimated by fitting a linear mixed model, accounting for clustering of lambs reared by the same ewe.
SD
weight and weight in spring were entered without transformation.
Variable
491
Table 6 Distribution of 101 ewes subjected to cesarean section and 104 ewes that had a vaginal delivery (VD) by total number of lambs and number of liveborn lambs to which they gave birth in the subsequent year. Lambs per ewe
Litter sizea 1 2 3 4 Live-born lambsb 0 1 2 3 4
Cesarean ewes
VD ewes
N
%
N
%
35 55 11 0
34.7 54.5 10.9 0
27 54 22 1
26.0 51.9 21.2 1.0
2 38 55 6 0
2.0 37.6 54.5 5.9 0
0 29 54 20 1
0 27.9 51.9 19.2 1.0
a Caesarean versus vaginal delivery: P ¼ 0.03 (Mantel–Haenszel chisquare). b Caesarean versus vaginal delivery: P ¼ 0.003 (Mantel–Haenszel chisquare).
492
S. Waage, G. Wangensteen / Theriogenology 79 (2013) 486–494
a greater proportion of stillborn lambs in triplet or quadruplet litters (five out of 11 vs. two out of 23, P ¼ 0.02, Fisher’s exact test). Considering the reduced litter size and the greater number of fetal deaths, the mean number of live-born lambs per ewe was significantly lower in the cesarean delivery group (1.64 0.63) as compared with the vaginal delivery group (1.93 0.71). The distribution of live-born lambs by litter size is shown in Table 6. In ewes that had had a cesarean delivery, lambing was unaided in 73%, gentle traction was provided in 13%, and dystocia was present in 14%. The corresponding figures for ewes that had had a vaginal delivery the previous year were 74%, 6%, and 20%. The distributions within the groups were not significantly different (P ¼ 0.7, Mantel–Haenszel chisquare). Mean birth weight of the live-born lambs for which weights were available was 5.2 0.9 kg for the 129 lambs born to the ewes that had undergone cesarean section and 5.1 0.9 kg for the 153 lambs born to the ewes that delivered vaginally at the previous lambing. Estimates from a mixed model, which included ewe as a random factor and delivery group, litter size, ewe age, and lamb sex as fixed effects, are shown in Table 7. Mode of previous delivery was not significant. Breed and plausible interaction terms were not significant when added to the model. Neonatal losses within the first week of life did not differ between the ewes that had undergone cesarean section (seven out of 101 lost one lamb) and those that had had a vaginal delivery (seven out of 104 lost one lamb, two lost two lambs) (P ¼ 0.4, Mantel–Haenszel chi-square). Clinical disorders affecting the ewes that might have been related to the previous cesarean section were not observed the second year. 4. Discussion Very few studies seem to have been published on postoperative mortality and morbidity and on subsequent reproductive performance of animals subjected to elective Table 7 Multivariable adjusted effects of mode of delivery and ewe and lamb characteristics on birth weight of lambs born to ewes that had a cesarean (N ¼ 129 lambs) or an unassisted vaginal delivery (N ¼ 153 lambs) the previous lambing season, estimated by fitting a linear mixed model and accounting for clustering of lambs born to the same ewe. Variable
Coefficient
Intercept Delivery Cesarean section Vaginal delivery Age of ewe 1y >1 y Lambs per ewe 1 2 >2 Sex of lamb Male Female
4.14
95% Confidence interval Lower
P
Upper
3.81
4.47
<0.0001
0.21
0.22
0.95
0.07
0.04
0.02
1.50 0.59 Reference
1.18 0.31
1.82 0.87
<0.0001 <0.0001
0.43 Reference
0.27
0.59
<0.0001
0.007 Reference Reference 0.29
cesarean section. Some case reports are available and a few observational studies have been conducted regarding cesarean cases in which surgical intervention was considered necessary; however, in such cases, postpartum outcomes of the dam and her offspring are likely to be affected to a varying extent by prepartum disorders or intrapartum complications experienced before surgery. In this study, we therefore included as cesarean cases only ewes for which no medical or obstetrical indications for surgery were present. Two of the 162 ewes that underwent cesarean section died within a few days postsurgery. In one of them, necropsy revealed peritonitis, metritis, and salpingitis, which were likely related to surgery. Both of her fetuses were dead prepartum and she also suffered from chronic mastitis and pleuritis; the combined effects of these preoperative disorders might have contributed to the fatal outcome. The single case of uterine prolapse after cesarean section, which resulted in the death of the ewe, could be considered a matter of coincidence and, therefore, not related to the surgery. An incidence rate of this disorder in sheep, which would likely vary somewhat among different breeds, does not seem to have been reported. Nevertheless, it is noteworthy that also in a previous study, comprising 57 ewes that underwent cesarean section, a case of uterine prolapse occurred the day after surgery [5]. Thus, some connection with the surgery cannot be ruled out. Hypothetically, if an inversion of a part of a uterine horn were introduced and not corrected during the cesarean section this might have initiated the development of the full prolapse. Relatively high mortality rates have been reported for ewes that underwent cesarean section because of dystocia. Brounts et al. [8], who reported joint follow-up results for 85 ewes and 25 goats admitted to a veterinary teaching hospital because of dystocia, found that 19% died after surgery; the main causes were pregnancy toxemia and peritonitis. A field study found that 17% of ewes that underwent cesarean section died; however, the mortality rate was low (2%) for ewes that delivered live or freshly dead lambs [6]. In another field study, 5% of the ewes that underwent cesarean section died postoperatively [5]. The greater losses observed in these studies as compared with our study were likely the result of the impaired preoperative condition of many ewes. In the field, the risk of an adverse outcome increases when vaginal birth fails and there is a delay before surgery is carried out. Four ewes experienced metritis after cesarean delivery but recovered after antibiotic treatment. Three of these and the ewe that died from peritonitis and metritis underwent cesarean section one particular year, suggesting the possibility of some common risk factor for bacterial contamination of the operative field that year. The proportion of ewes that experienced metritis postoperatively (3.1%) is not particularly high. A study in Belgium comprising 1159 suckler beef cows, of which approximately 90% had had a cesarean delivery, found that 13% of the cows developed metritis [12]. In human medicine, metritis has been reported to complicate 3% to 20% of cesarean deliveries [13,14].
S. Waage, G. Wangensteen / Theriogenology 79 (2013) 486–494
Only one case of retained placenta was observed, and the ewe recovered after antibiotic treatment. Retained placenta has been reported to be a frequent complication in small ruminants that underwent cesarean section because of dystocia [8]. Three (2%) of the 160 surviving ewes suffered a wound infection after the cesarean section but were cured with antibiotic treatment. The occurrence of such complications was modest compared with what has been experienced with human cesarean cases, where rates of surgical site infections from 1.4% to 10% have been reported [14–17]. Factors related to surgical skills will likely affect prognosis of cesarean cases. In this study, the operations were performed by veterinary students with very limited surgical experience. Close professional supervision and guidance apparently have prevented major deviations from the appropriate operative procedure. However, the operation lasted longer, sometimes much longer, than what would have been the case if carried out by an experienced surgeon (Professor Emeritus of Obstetrics, Stig A. Ødegaard, personal communication, March 2012). The risk of postoperative complications, including infections, is likely to increase with increasing surgery time; however, in human medicine only a weak relation has been observed between the occurrence of complications and the duration of the cesarean delivery [18]. Fetal deaths are frequently observed when ewes undergo emergency cesarean delivery. A field study and a study based on small ruminants admitted to a teaching hospital found, respectively, that 26% and 58% of the lambs were dead at birth [5,8]. In our study, only 3.5% of the 312 lambs born to the ewes that underwent cesarean section were dead at birth, and a similar proportion occurred in the control ewes that delivered vaginally. According to the annual report from the Norwegian Sheep Recording System, the overall loss of lambs because of fetal death in the flocks enrolled in 2004 was 3.7% [19]. It could not be ruled out that the cesarean surgery and potential postoperative discomfort would interfere with the ewe’s mothering capabilities, ewe–lamb bonding and the ewe’s milk secretion and production, which might in turn affect neonatal survival and growth of the lambs. Studies addressing these questions have not been published. During the first 6 to 9 weeks after cesarean delivery the ewes in this group lost 10% of their live-born lambs. However, postnatal losses among the ewes that had a vaginal delivery were not significantly lower. According to the literature, similar or even higher neonatal mortality rates are not uncommon in sheep flocks [20,21]. Moreover, though growth rate of the lambs during this period was strongly influenced by litter size, birth weight and sex of the lamb and age of the ewe, and the mode of delivery did not have any significant effect. Similarly, lamb losses and weight gain from late spring until weight recording in the autumn did not differ between the delivery groups. Subsequent reproductive efficiency of ewes that underwent cesarean section has only to a very limited extent been addressed previously. Two studies reported high pregnancy rates for ewes that were bred the subsequent season. The proportions of the original cases that were bred were 36 out of 57 ewes [5] and only 16 out of 110
493
ewes [8], respectively. In our study, approximately 70% of the original cases in each of the two delivery groups were rebred the subsequent autumn. Of ewes that were bred, almost equal proportions of the ewes in the cesarean delivery group (10.6%) and the vaginal delivery group (10.3%) did not conceive. This is in contrast with results obtained in a study in beef cattle, where the subsequent pregnancy rate for 105 cows that had had a cesarean delivery was 27% lower than the herd average [22]. In women, there are indications that cesarean delivery is associated with increased risk of fetal death in subsequent pregnancy [23,24]. Only singleton births were investigated in these studies. Explanations proposed were placentation abnormalities caused by uterine scar and possible reduction in uterine blood supply because of ligation of vessels during the previous cesarean operation. Abdominal adhesions are not uncommon in women [25] and animals [26] that have undergone cesarean section, and subfertility is a possible consequence [27]. In our study, the mean number of live-born lambs per ewe was markedly lower for ewes that had a previous cesarean delivery than for those that had a vaginal birth (1.64 vs. 1.93). This difference was the result of a significantly lower number of lambs born per ewe in the former group and, additionally, an increased number of stillborn lambs in this group. The age distribution of the ewes and their reproductive capacity, as indicated by their litter size the previous year, might have confounded the observed difference; however, these variables did not differ between the two groups. Assuming that ovulation and fertilization were similar in the two delivery groups, unobserved early embryonic death or unsuccessful or inadequate placentation are possible explanations. Because the proportion of ewes that conceived in each of the two groups was similar, and the decreased number of live-born lambs in the cesarean delivery group was mainly the result of a reduced number of triplets, it is tempting to suggest that in some cases the previous cesarean operation caused structural changes that restricted the uterine capacity for triplet pregnancy. It might be a matter of discussion whether the reduced number of triplet births should be considered an important side effect of cesarean section. Most sheep holders seem to prefer twin births because rearing triplets might sometimes be too demanding for a ewe. In women who had a cesarean delivery, mean birth weight of babies in subsequent pregnancy has been reported to be somewhat lower than that of babies born to women who had previously delivered vaginally [23]. When comparing birth weights of the live-born lambs delivered the subsequent year by the ewes in the two delivery groups, and accounting for measured confounders, we found no difference between the groups. Several factors might affect the risk of peripartum complications and postpartum performance of ewes and their offspring. Therefore, when comparing short-term and long-term outcomes of ewes and lambs subjected to different modes of delivery, possible confounders must be considered. Nutrition, environmental conditions, and management practices are known to be important determinants of health and productivity of sheep. Because our study was limited to a single flock, kept at an experimental
494
S. Waage, G. Wangensteen / Theriogenology 79 (2013) 486–494
station with well-established common management practices, significant confounding because of differential exposure to such factors between the two study groups, which could have affected their outcomes differently, is not very likely. To further reduce potential time-related confounding we selected as the control for each cesarean case the ewe that lambed closest in time. Main residual confounding was assumed to be the result of individual animal differences, and individual characteristics expected to be important determinants for the outcome variables of interest were accounted for in the analyses where appropriate. When analyzing lamb weights and growth we specified ewes as clusters. In doing so we not only accounted for the potential intralitter correlation caused by common maternal effects but also the litter-specific environmental factors and attention paid to the lambs by the staff, in particular while kept in the individual pen. The quality of available data is essential in retrospective studies. Data used in this study were collected from records at the APEC. These were recorded by skilled personnel and were generally intended for scientific purposes. Disease data were cross-checked with records at the Ambulatory Clinic, the Norwegian School of Veterinary Science, who provided the daily veterinary care at APEC. Thus, there are good reasons to consider the data we used as being reliable and precise. 5. Conclusions There was a moderate frequency of postoperative complications experienced by ewes that underwent elective cesarean section. The proportion of stillborn lambs and postnatal survival and growth of lambs born to these ewes did not differ from corresponding outcomes for lambs born vaginally. Pregnancy rate in the subsequent breeding season was not reduced in ewes that had had an elective cesarean delivery at term; however, as a result of smaller litters and an increased number of fetal deaths, these ewes gave birth to fewer live-born lambs than ewes that had delivered vaginally. Acknowledgments The authors thank Stig Anders Ødegaard, Professor Emeritus of Obstetrics, Norwegian School of Veterinary Science, who was in charge of student education in obstetrics during the study period, and the staff at the Animal Production Experimental Centre, Ås, in particular the present leader, Agnes Klouman, for excellent management of the sheep flock in the study period. The study was funded mainly by the Norwegian School of Veterinary Science, with some additional contributions from the Agricultural Agreement Research Fund, the Foundation for Research Levy on Agricultural Products, Animalia – Norwegian Meat and Poultry Research Centre and the Norwegian Sheep and Goat Association.
References [1] Whitelaw A, Watchorn P. An investigation into dystocia in a south country Cheviot flock. Vet Rec 1975;97:489–92. [2] George JM. The incidence of dystocia in Dorset Horn ewes. Aust Vet J 1976;52:519–23. [3] Grommers FJ, Elving L, Van Eldik P. Parturition difficulties in sheep. Anim Reprod Sci 1985;9:365–74. [4] Wiener G, Woolliams C, Macleod NSM. The effects of breed, breeding system and other factors on lamb mortality. 1. Causes of death and effects on the incidence of losses. J Agric Sci (Camb) 1983; 100:539–51. [5] Mosdøl G. Indications for and fertility after caesarean section in the ewe in common practice. Norsk Veterinærtidsskrift 1986;98: 441–4. [6] Scott PR. Ovine caesarean operations: a study of 137 field cases. Br Vet J 1989;145:558–64. [7] Majeed AF, Taha MB, Azawi OI. Cesarean section in Iraqi Awassi ewes: a case study. Theriogenology 1993;40:435–9. [8] Brounts SH, Hawkins JF, Baird AN, Glickman LT. Outcome and subsequent fertility of sheep and goats undergoing cesarean section because of dystocia: 110 cases (1981-2001). J Am Vet Med Assoc 2004;224:275–9. [9] Noakes DE, Parkinson TJ, England GCW. Veterinary reproduction and obstetrics. Ninth Edition. Edinburgh: Saunders; 2009. p. 354, 367. [10] Royston P, Ambler G, Sauerbrei W. The use of fractional polynomials to model continuous risk variables in epidemiology. Int J Epidemiol 1999;28:964–74. [11] SAS Institute Inc. SAS/STAT 9.2 user’s guide. Second Edition. Cary, NC: SAS Institute Inc; 2009. p. 1675–831, 3891–4092. [12] Hanzen C, Laurent Y, Ward WR. Comparison of reproductive performance in Belgian dairy and beef cattle. Theriogenology 1994; 41:1099–114. [13] Chaim W, Bashiri A, Bar-David J, Shoham-Vardi I, Mazor M. Prevalence and clinical significance of postpartum endometritis and wound infection. Infect Dis Obstet Gynecol 2000;8:77–82. [14] Spandorfer SD, Graham E, Forouzan I. Postcesarean endometritis. Clinical risk factors predictive of positive blood cultures. J Reprod Med 1996;41:797–800. [15] Hannah ME, Hannah WJ, Hewson SA, Hodnett ED, Saigal S, Willan AR. Planned caesarean section versus planned vaginal birth for breech presentation at term: a randomised multicentre trial. Lancet 2000;356:1375–83. [16] Tran TS, Jamulitrat S, Chongsuvivatwong V, Geater A. Risk factors for postcesarean surgical site infection. Obstet Gynecol 2000;95:367–71. [17] Yokoe DS, Noskin GA, Cunnigham SM, Zuccotti G, Plaskett T, Fraser VJ, et al. Enhanced identification of postoperative infections among inpatients. Emerg Infect Dis 2004;10:1924–30. [18] Killian CA, Graffunder EM, Vinciguerra TJ, Venezia RA. Risk factors for surgical-site infections following cesarean section. Infect Control Hosp Epidemiol 2001;22:613–7. [19] Offerdal E, editor. The Norwegian Sheep Recording System, Annual Report 2004. Norwegian Meat and Poultry Research Centre, Oslo, 2005. [20] Binns SH, Cox IJ, Rizvi S, Green LE. Risk factors for lamb mortality on UK sheep farms. Prev Vet Med 2002;52:287–303. [21] Huffman EM, Kirk JH, Pappaioanou M. Factors associated with neonatal lamb mortality. Theriogenology 1985;24:163–71. [22] Patterson DJ, Bellows RA, Burfening PJ. Effects of caesarean section, retained placenta and vaginal or uterine prolapse on subsequent fertility in beef cattle. J Anim Sci 1981;53:916–21. [23] Smith GCS, Pell JP, Dobbie R. Caesarean section and risk of unexplained stillbirth in subsequent pregnancy. Lancet 2003;362:1779–84. [24] Gray R, Quigley MA, Hockley C, Kurinczuk JJ, Goldacre M, Brocklehurst P. Caesarean delivery and risk of stillbirth in subsequent pregnancy: a retrospective cohort study in an English population. BJOG 2007;114:264–70. [25] Lyell DJ. Adhesions and perioperative complications of repeat cesarean delivery. Am J Obstet Gynecol 2011;205(Suppl.):S11–8. [26] Sloss V, Dufty JH. Elective caesarean operation in Hereford cattle. Aust Vet J 1977;53:420–4. [27] Silver RM. Delivery after previous cesarean: long-term maternal outcomes. Semin Perinatol 2010;34:258–66.