Between December 1999 and February 2001, two visits, eight weeks apart, were made to 90 herds of Danish finisher pigs. The prevalence of clinical signs was recorded by three veterinary technicians from the Danish Bacon and Meat Council according to a standardised procedure; they had been trained and their observations were monitored and validated before and during the study. A total of 154,347 finisher pigs were examined and 22,136 clinical signs were recorded. Vices accounted for 43 per cent of the signs. The highest mean prevalence was observed for ear necrosis (4·44 per cent), followed by respiratory signs (2·17 per cent), lameness (1·92 per cent), other skin diseases (1·73 per cent), tail bites (1·26 per cent), umbilical hernia (0·78 per cent), flank bites (0·52 per cent), diarrhoea (0·27 per cent), respiratory distress (0·12 per cent), atrophic rhinitis (0·10 per cent), recumbency (0·09 per cent) and central nervous disease (0·05 per cent). The prevalence of atrophic rhinitis was higher in conventional herds than in specific pathogen-free herds. The prevalence of clinical signs of atrophic rhinitis was higher among finishers weighing 51 to 75 kg than among finishers weighing up to 50 kg, and the prevalence of respiratory signs was higher among finishers weighing 51 to 75 kg then among finishers weighing 76 to 100 kg.
Statistics from Altmetric.com
INFORMATION about the prevalence of disease among pigs, and their productivity, is important when assessing their welfare and for the introduction of a health management approach to veterinary services for pig producers. National databases such as the Danish National Swine Herd Health Programme (Willeberg and others 1984) or the National Animal Health Monitoring System in the usa (King 1990) have been used for studying pig health (Christensen and others 1994). However, if the data are not derived from farms selected at random their value for disease surveillance at a national level may be limited (Dohoo 1988). Similarly, data from diagnostic laboratories may be accurate and reliable, but their value will be limited if the samples have not been collected at random from the whole population.
Slaughterhouse records have also been used to estimate the prevalence of disease in pig herds (Cleveland-Nielsen and others 2002), but the accuracy and completeness of the records may vary (Enøe and others 2003). Among pathological findings at slaughter, only pneumonia and pleuritis have been reported to be related directly to records of the treatments of individual pigs within 30 days of slaughter, and the treatment of groups of pigs and clinical observations did not reliably predict the pathological findings (Blocks and others 1994). Furthermore, slaughterhouse records may underestimate the prevalence of disease owing to healed lesions or pigs that have died or been euthanased.
The clinical signs that most frequently require antimicrobial treatment have also been used as an estimate of herd health (Blocks and others 1994, Christensen and others 1994, Heinonen and others 2001). However, the treatments recorded by farmers might underestimate the quantities of drugs used (Dunlop and others 1998). Since January 2000, the quantities of antibiotics used in Danish pig production have been monitored nationwide by gathering information directly from pharmacies and veterinarians (Stege and others 2003). However, the overall usage of antibiotics might not be directly related to the occurrence of clinical signs, because some infectious respiratory and intestinal diseases often result in the treatment of an entire batch of pigs although only some of them show clinical signs of disease. Furthermore, some clinical signs, such as ear necrosis and umbilical hernia, do not result in treatments, some producers may be more reluctant to initiate treatment than others, and some may treat the animals only once whereas others may treat them for several days, variations that will prevent the overall consumption of antibiotics from giving an accurate indication of the prevalence of clinical signs.
Accurate information about the prevalence of animal diseases is needed (Dohoo 1988, King 1990), but valid studies are rare and often limited by economic and practical restraints. The prevalences of clinical signs recorded in pig herds are available but are often limited to a region (Blocks and others 1994, Heinonen and others 2001), to certain farms cooperating with their veterinarians (Christensen and others 1994) or to certain clinical signs (Pearce 1999). Estimates based on a small number of herds may be seriously biased by changes in only one herd (Christensen and others 1994). To increase the number of herds, producers' records have been used (Blocks and others 1994, Christensen and others 1994, Heinonen and others 2001), but such records depend on the skill, experience, education, judgement and willingness of many observers to make a decision (Vaillancourt and others 1990, 1992). Clinical records made by different observers may be subject to wide variations (Gore 1981, Theodossi and others 1981, Bailey and others 1989, Pattinson and Theron 1989), but education and standardisation might increase the level of agreement (Klinkhoff and others 1988). Precise definitions of the clinical signs and a limitation on the number of observers may optimise the quality of the data.
This paper describes the results of a large-scale study of the prevalence of clinical signs in herds of Danish finisher pigs.
MATERIALS AND METHODS
Selection of herds
Between December 1999 and February 2001, a cross-sectional study was made of 98 herds of finisher pigs selected from a stratified random sample of 176 herds (Fig 1) and each herd was followed up. A herd consisted of up to five sites owned by one farmer or cooperative. They were selected on the basis of employment (full-time or part-time), economic size, types of farming (pigs only or pigs and other types of farming), agricultural area, the farmer's age and location. The 98 herds in the final sample were on average larger than the average for the population, delivering more pigs for slaughter, and there were more specific pathogen-free (spf) herds than the average for the population (Table 1). All the herds were visited twice, with an average of eight weeks between the two visits.
Selection of pigs
Three veterinary technicians from the National Committee for Pig Production, Danish Bacon and Meat Council, visited all the herds. Veterinary technicians are lay staff trained to assist in health surveillance and data collection in pig herds; they were instructed and trained before and monitored during the study (Petersen and others 2004). Each observer was assigned to one region, and one observer took all the records on the two visits to each herd. In herds with up to 1000 slaughter pigs, the technicians examined all the pigs in all the pens. In herds with more than 1000 pigs, the pens were selected by systematic random sampling: in herds with 1001 to 1500 pigs, the pigs in every third pen were not examined, and in herds with more than 1500 pigs, the pen selection frequency was obtained by dividing the estimated total number of pigs (number of pens multiplied by the number of pigs in one pen) by 1000. Pens with more than 50 pigs were not included in the study owing to the risk of recording one pig more than once because of the difficulty in keeping track of the individual pigs.
Recording of clinical signs
Standard clinical records of each pen examined (Petersen 2002) were made during each visit to each herd (Table 2), the clinical signs recorded being those that would be evident to a layman. Tear staining was not included in the definition of atropic rhinitis. Sneezing is rare and was therefore not included. Clinical signs were recorded only if the diseased pig could be found: for example, evidence of diarrhoea on the floor of the pen was not recorded unless a pig with diarrhoea was observed. The pigs in each pen were counted and their average weight was estimated. Before entering the pen, the technician recorded any clinical signs observable from the aisle, and then entered the pen to make all the pigs stand and move around. The clinical signs were recorded without reference to their aetiology, as defined in Table 2.
Agreement between observers
All the technicians had been trained to make a standard clinical record, but during the first five herd visits each technician was accompanied by one of two veterinarians. One veterinarian participated in 30 of the herd visits during the data collection period. The agreement between the observers was assessed twice by letting the three technicians and one veterinarian make a standard clinical examination independently in the same herd on the same morning (Petersen and others 2004). Despite their training, there were some variations in the records of different clinical signs made by the observers, and these variations were used for further training of the observers during the data collection period.
Two of the participating herds were excluded because all of the pens contained more than 50 pigs, five were excluded because of missing records, and one was excluded because it would not allow the second visit. During the two visits to the 90 remaining herds, 154,347 finisher pigs were examined, and 22,136 clinical signs were recorded; the within-herd prevalences are shown in Table 3. Ear necrosis and other skin diseases, lameness and signs of respiratory disease were the most frequently observed clinical signs (Fig 2). The other skin diseases were primarily superficial abscesses and skin lesions caused by fighting. The signs with the lowest prevalence were atrophic rhinitis, central nervous signs, growth retardation, recumbency and respiratory distress (Fig 2, aggregated as other clinical signs). Vices, defined as ear necrosis, tail bites and flank bites, accounted for 43 per cent of the recorded clinical signs, whereas the apparently infectious conditions such as respiratory signs, lameness and diarrhoea accounted for 32 per cent. The prevalence of atrophic rhinitis was higher in the conventional herds than in the spf herds (Table 4). Finishers weighing 51 to 75 kg had a significantly higher prevalence of clinical signs of atrophic rhinitis than finishers weighing up to 50 kg, and a higher prevalence of respiratory signs than finishers weighing 76 to 110 kg (Table 5).
The managers of the herds had to volunteer in writing before participating in the study, and this procedure might have interfered with the randomisation; 98 of the 176 farmers (56 per cent) originally selected at random participated, and their herds constituted 2 per cent of the total production of slaughter pigs in Denmark in 2000. However, the frequency of participation varied between the different regions (Fig 1), and the participating herds were larger in terms of the number of pigs delivered for slaughter and there were more herds with spf health status (Zhuang 2002) than in the general population.
At both visits, pens were selected at random. However, some of the younger pigs examined during the first visit could have been examined again at the second visit; for example, if the average daily weight gain was 1 kg and the pigs were slaughtered at 100 kg, then pigs weighing less than 40 kg at the first visit would still have been present at the second visit two months later. A total of 22,136 clinical signs were recorded in the 154,347 pigs examined, but because one pig could have had more than one clinical sign, the total number of pigs with clinical signs could have been lower.
Clinical signs due to vices were the most frequently recorded signs (Fig 2). Vices were observed in most herds, but the observed prevalence differed widely (Table 3). Vices seemed to be more frequent among young finishers, but no significant effect of health status or age was observed. The prevalence of ear lesions was lower (P=0·09) among the oldest finishers; the lack of significance, despite large differences in the mean values, was due to the large standard errors. Severe and acute ear lesions were more common in young finishers and more chronic, healing lesions with crusts were more common among older finishers. In a survey of Finnish high-health finishing units, tail biting accounted for 2 per cent of the antimicrobial treatments (Heinonen and others 2001). In a Dutch study, ear necrosis and tail bites were among the most frequently observed clinical signs (28·3 per cent and 16·4 per cent, respectively) (Blocks and others 1994). In the present study, vices, defined as ear necrosis, flank bites and tail bites, accounted for 43 per cent of the clinical signs recorded, with ear necrosis and tail bites being most frequent.
Christensen and others (1994) reported that lameness accounted for 11 per cent of the clinical signs recorded in Danish finishers. In the Netherlands, the prevalence of lameness has been reported to be 10·7 per cent (Blocks and others 1994). In the present study, lameness was the second most common clinical sign, and accounted for 15 per cent of those recorded; the health status and age of the pigs had no significant effect on the prevalence of lameness.
Christensen and others (1994) reported that respiratory signs were the cause of 54·6 per cent of the recorded treatments of Danish fattening pigs with antibiotics, and in the Netherlands, respiratory disorders were the cause of 39·7 per cent of the recorded treatments with antibiotics (Blocks and others 1994). In the present study, respiratory signs accounted for only 15 per cent of the recorded clinical signs, and severe respiratory distress was rarely recorded. There was a borderline significantly higher prevalence of clinical respiratory signs in the conventional herds than in the spf herds (P=0·05) (Table 4), and in the finishers weighing 51 to 75 kg than in the finishers weighing 76 to 110 kg (Table 5). However, 59 per cent of the conventional herds were free from Actinobacillus pleuropneumoniae serotype 2, and 9·7 per cent of them had been vaccinated against Mycoplasma hyopneumoniae, factors that could have reduced the prevalence of clinical signs of respiratory disease to a level close to that in the spf herds. The prevalence of respiratory disease is known to vary with season (Elbers and others 1992), but the effect of season could not be differentiated from the effect of herd because the two visits to each herd were made within eight weeks. The prevalence of respiratory signs appeared to be low in comparison with the frequency of treatments for respiratory signs reported by Christensen and others (1994); the difference may be real, or it may be due to the repeated treatments of individual pigs or the treatment of batches of pigs in the earlier study.
The other skin diseases accounted for 14 per cent of the observed clinical signs, and were mainly wounds other than vices due to fighting, and superficial abscesses. The health status and age of the pigs had no significant effect on the prevalence of these diseases.
Growth-promoting antibiotics were withdrawn from Danish pig production from January 2000. As a result, there was a reduction in the total quantity of antibiotics used in spite of a simultaneous increase in the quantity prescribed for the treatment of diarrhoea (Anon 2002). There was a reduction in the average growth rate, and increases in mortality and in the prevalence of diarrhoea in weaners after growth promoters were withdrawn, but no such changes were observed in finishers (Anon 2002). In this study the average herd prevalence of diarrhoea among finishers was low, and there were no significant differences in prevalence between herds of different health status or pigs of different ages. Blocks and others (1994) observed that in 21 Dutch growing and finishing herds, the most common clinical sign reported by the farmers was gelatinous faeces, and the treatment of batches of pigs for digestive tract disorders accounted for 53·4 per cent of the treatments. Christensen and others (1994) reported that diarrhoea was the second most common reason for the treatment of Danish fattening pigs with antibiotics, and accounted for 23 per cent of the treatments. In the present study there was a relatively low prevalence of diarrhoea, which may have been due either to effective treatment regimens, or to a low prevalence of clinical signs due to subclinical infections, or to difficulties in assessing cases of diarrhoea with a high sensitivity. Diarrhoeic faeces on the floor of a pen of pigs was not recorded unless individual pigs with diarrhoea could be identified; during one examination of a pen not all the pigs with diarrhoea would have been showing clinical signs, and so the prevalence is likely to have been underestimated. As for the respiratory signs, the relatively high frequency of treatments compared with the low frequency of the clinical signs might have been due to the treatment of batches of pigs rather than individual pigs. However, despite the low in-herd prevalence, diarrhoea was observed in 57 per cent of the herds, with a median prevalence of 0·03 per cent, indicating that causative agents were present in many of the herds where few clinical signs were observed.
The low prevalence of atrophic rhinitis observed, particularly in the spf herds, was in accordance with the improved methods for controlling the disease by management strategies and vaccination and with the Danish spf classification. Finishers weighing 51 to 75 kg had a significantly higher prevalence of clinical signs of atrophic rhinitis than finishers weighing up to 50 kg, an increase consistent with the increase in the irreversible clinical signs with age; however, the prevalence observed among finishers weighing 76 to 110 kg was not significantly different from the prevalence in the other age groups. This paradox may be at least partially explained by a few herds having a high prevalence of atrophic rhinitis (Table 4). In these herds there happened to be outbreaks of clinical disease among the pigs weighing 51 to 75 kg at the time of the herd visits; if these herds had been visited later, the older/heavier pigs might have been the pigs with signs of atrophic rhinitis. However, as all the herds were examined twice within eight weeks, no certain explanation can be given.
The prevalence of umbilical hernia was low, and did not vary significantly with the health status or age of the pigs. Growth retardation was rare, but the prevalence varied between herds; in most herds the prevalence was close to zero, but a maximum prevalence of 6·6 per cent was recorded in one herd. Severe clinical signs, such as recumbency and respiratory distress, were also rare. However, the farmers were given notice of the herd visits, and they might have elected to euthanase pigs with severe clinical signs before the visits.
Before drawing conclusions on the basis of the observed prevalences, it must be considered that when examining the pigs on a farm, only a limited number of clinical problems may have been observed. The prevalences might also have been influenced by the nature of the clinical signs (Petersen and others 2004). Vices are revealed by clinical signs that are consistent over time, whereas signs such as watery faeces, indicating diarrhoea, or coughing, indicating respiratory disease, may be observable for only a short time. Similarly, the sensitivity and specificity of producer-recorded causes of preweaning mortality have been reported to vary with the cause of the mortality and the age of the pigs (Vaillancourt and others 1990), and the ability to correctly diagnose the cause was low. Before this study began, the observers were instructed and trained, and during the study their observations were validated and the interobserver agreement was assessed to try to achieve consistency (Petersen and others 2004).
This study was the first nationwide investigation of the prevalence of multiple clinical signs in herds of Danish finisher pigs. Previous papers have been based on a smaller number of herds and the recording of only one clinical sign per animal (Christensen and others 1994), or on herds in only one region (Blocks and others 1994). The observations were validated by training the observers and monitoring their records throughout the study.
The authors thank the veterinary technicians, Anders Mariager, Bjarne Nielsen and Søren Justesen, National Committee for Danish Pig Production, Danish Bacon and Meat Council (ncdpp-dbmc) for making the herd visits, Mrs Sanne Venneberg, Royal Veterinary and Agricultural University, and Mrs Susanne Søgaard Sørensen, ncdpp-dbmc, for typing the data, and the participating pig producers for access to their farms. The study was supported by the Research Centre for the Management of Animal Production and Health (cep 97-13). The authors also thank the Forsøgsleder R. Nørtoft Thomsens Foundation for financial support.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.