Epidemiological studies of pig diseases: 1. Use of French protocols for risk factor assessment to predict the health status of Australian herds

Objective To determine in Australian pig herds the accuracy of French protocols for risk factor assessment.
Procedure Data on health indicators and risk factors were collected for three syndromes, 'pre-weaning diarrhoea', 'post-weaning diarrhoea' and 'respiratory problems', using the French protocols. The protocols were used on 118 occasions in 32 Western Australian pig herds during 3 years (1988 to 1991).
Results There was a wide variation in pre-weaning performance, for example growth rate was 107 to 273 g/day (< 200 g/day in 33% of herds). Respiratory lesions at weaning were associated with poor pre-weaning performance. Post-weaning (21 days after weaning) growth rate was 114 to 408 g/day (< 250 g/day in 54% of herds). In the grower herds, 91% of herds had pneumonia, and growth rate was 439 to 625 g/day (< 550 g/day in 54% of herds). Pleurisy as well as pneumonia was associated with reduced growth rate. The risk factor most closely associated with respiratory health status was air volume per pig.
Conclusion Risk factors were most accurate at predicting the health status in post-weaning problems. A weaning weight of at least 7.9 kg and weaning age of 30 days optimised weaner performance. Stocking densities and shed designs providing at least 3 m³ air volume and 0.6 m² floor space per pig throughout the growing phase should be considered for an improved respiratory health status. Australian pig sheds often do not provide a satisfactory environment for optimum health. The technique of risk factor assessment as an aid to the maintenance of health in pig herds is applicable in Australia, but further research is necessary to determine the most important Australian risk factors.     

Postweaning multisystemic wasting syndrome in Danish pig herds: productivity, clinical signs and pathology

A case-control study of 74 herds with postweaning multisystemic wasting syndrome (pmws) and 74 matched control herds was carried out. In the case herds the mortality rates of weaner and finisher pigs were 11.2 and 5.2 per cent respectively, compared with 3.1 and 3.2 per cent in the control herds. In most case herds, pmws developed within the first four weeks after weaning. Wasting, diarrhoea and respiratory signs were observed in 10 per cent of the weaner pigs (7 to 30 kg) in the case herds compared with 7 per cent in the control herds. The average daily gains of the weaner pigs and finisher pigs were 36 g and 52 g less in the case herds than in the control herds. By examining three weaner pigs from each herd the pmws diagnosis was confirmed by histopathology and immunohistochemistry in 78 per cent of the case herds, but at least one pmws-positive weaner pig was found in 19 of the control herds. The prevalence of pmws-positive pigs among illthriven weaner pigs was 45 per cent (101/222) in the case herds, and 12 per cent (27/222) in the control herds. Specific gross pathological findings were associated with a positive pmws diagnosis; pigs with heavy, rubber-like lungs, atonic intestines, and enlarged bronchial and inguinal lymph nodes, had a 0.7 probability of a positive pmws diagnosis by laboratory examinations. However, for illthriven pigs, this probability of having pmws was equal in the case herds and the control herds.     

Epidemiological studies of piglet diarrhoea in intensively managed Danish sow herds. II. Post-weaning diarrhoea

This study comprised 48,931 litters in 89 sow herds. During the study (1976-82) weaning age decreased from approx. 42 days to approx. 30 days. The mean incidence of post-weaning diarrhoea was 6.0% of litters weaned, with little variation by year but with considerable variation among herds. Within the individual herd increased incidence occurred over limited periods, probably associated with specific infections. Litters with diarrhoea during the suckling period had increased risk of post-weaning diarrhoea. The incidence of post-weaning diarrhoea increased with litter size at weaning. Thus, a litter of 11-12 piglets at weaning had 1.2 times higher risk than litters with 8-10 piglets. In contrast to pre-weaning diarrhoea, there was no association between parity of the sow and diarrhoea in the litter after weaning. Litters weaned below 2 weeks of age had a 2-fold risk of developing diarrhoea after weaning and a 2.4-fold higher mortality rate than did litters weaned at 6-7 weeks. Similarly, litters weaned at an individual piglet weight below 3 kg bodyweight had a 3-fold higher risk of developing diarrhoea after weaning and a 5-fold higher mortality rate than did pigs from litters weaned at a bodyweight of 7-8 kg. The incidence of post-weaning diarrhoea decreased with increasing herd size. Piglets from litters with post-weaning diarrhoea had reduced weight gains after weaning and were 2.3 days older at 25 kg bodyweight than piglets from non-diarrhoeic litters. Likewise, diarrhoea after weaning was associated with an increased incidence of diseases of the skin and respiratory tract. Thus the risk of contracting respiratory disease was 4 times greater in diarrhoeic litters.     

Risk factors for post-weaning diarrhoea on piglet producing farms in Finland

Background

Post-weaning diarrhoea (PWD) is a significant gastrointestinal disease in pigs. It is considered a multifactorial disease associated with proliferation of enterotoxigenic Escherichia coli in the intestinal tract of affected pigs. The aim of this study was to analyse risk factors related to the occurrence of PWD on Finnish piglet producing farms.

Methods

The data of a follow-up study of 73 conventional piglet producing farms was used in the case-control study. The selection of the 41 PWD case and 28 control farms was based on the use of antimicrobials for treating diarrhoea in weaned pigs and the answers related to the occurrence of diarrhoea after weaning in the questionnaire. Four intermediate farms were excluded from the statistical analysis.Altogether 39 factors related to herd characteristics, weaner pig management and pig health were studied. The median number of sows was 59.0 (IQR = 44.0; 74.5) and 52.5 (IQR = 36.8; 61.5) on the case and the control farms, respectively.The significances of the univariable associations between the explanatory variables and the outcome variable were tested, and in the multivariate analysis quasibinomial generalized linear models were applied.

Results

An increased risk of PWD was associated with the regimen of twice a day feeding and feed restriction after weaning (P = 0.02; compared to feeding three or more meals a day or the use of ad libitum feeding) and with a higher number of sows on the farm (P = 0.02; risk increasing with increasing number of sows). Automatic temperature control was associated with a decreased risk of PWD (P = 0.03; compared to manual temperature control).

Conclusion

Twice a day feeding of newly-weaned pigs should be avoided if the amount of feed given is restricted. Variation in ambient temperature should be minimized in housing of newly-weaned pigs and this can be achieved by using automatic temperature control. With increasing number of sows in the herds the risk of PWD increases and more attention should be paid to prevention of post-weaning diarrhoea.     

An evaluation of the efficacy of cordial for preventing post-weaning diarrhoea in pigs

An evaluation of the efficacy of a commercial crystalline cordial concentrate for preventing post-weaning diarrhoea (PWD) and improving weight gains in pigs following weaning, was carried out in 3 herds using 861 pigs. The incidence of PWD in pigs drinking the cordial was reduced in 2 herds and its severity was reduced in one of these. There was no effect on either incidence or severity in the other herd. The pigs given cordial consumed considerably more water per day and less feed per day than pigs drinking plain water during the 10-day supplementation period. There was no improvement in average daily gain or feed conversion ratio attributable to adding cordial to drinking water in any of the herds. In one herd the growth rate of pigs receiving cordial was actually depressed during the supplementation period. There was no reduction in mortality or in number of treatments given for diarrhoea in pigs receiving cordial.     

Identifying obstacles to reducing the use of antibiotics to control porcine proliferative enteropathy

Proliferative enteropathy (PE) is an enteric disease of pigs that results in diarrhoea, reduced growth rate, reduced feed conversion efficiency and sometimes death. A survey of 13 pig veterinary practitioners in Australia was conducted to determine: (1) PE control strategies (antibiotics and vaccination), (2) how the efficacies of these strategies are evaluated and (3) how reliance on antibiotics could be reduced by increasing vaccine adoption. Antibiotics were routinely prescribed in the diets of weaner, grower and finisher pigs by 9/13, 10/13 and 8/13 veterinarians, respectively, if vaccination was not implemented. Water-soluble antibiotics were prescribed less frequently than in-feed antibiotics. Efficacy of control strategies was assessed most often through reduced clinical signs (diarrhoea, 'tail-ender' pigs, death) and reduced lesion incidence at abattoir postmortem inspection. Twelve practitioners had recommended PE vaccination to their clients in the previous 6 months. Barriers to continued vaccine implementation included perceived lack of efficacy for pigs housed in bedded systems, high cost relative to medication and difficulties in vaccinating pigs post-weaning.     

Pre-weaning supplementary feed and porcine post-weaning diarrhoea

Attempts were made to induce an intestinal hypersensitivity response to weaner diet by feeding pigs with small quantities of this material before weaning. In two trials using different weaner diets piglets subjected to this regimen showed no significant differences in small intestinal structure, in ability to absorb xylose, in bodyweight gain, in incidence of diarrhoea or excretion of enteropathogens after weaning compared with pigs not given any of the diet before weaning, or fed with a different diet before weaning. When post-weaning diarrhoea occurred it was associated with an earlier, more prolonged and greater proliferation of enterotoxigenic Escherichia coli in the small intestines than occurred in healthy pigs after weaning. The greater proliferation in pigs which developed diarrhoea could not be attributed either to an excessive dietary intake after weaning, or to a specific proliferation of rotaviruses.     

Antimicrobial use in the Australian pig industry: results of a national survey

Objective   To describe how various antimicrobials are used in commercial pig herds in Australia and for what disease conditions.
Procedure   Managers of large pig herds (> 200 sows) across Australia and their veterinarians participated in an internet-based survey in 2006. Questions were asked about herd management, the occurrence of bacterial diseases and the type and frequency of antimicrobial use. An antimicrobial usage index for each herd was derived as a summary of the risk of selection for antimicrobial resistance. Relationships between responses were explored with univariate and multivariate analysis.
Results   Responses were received for 197 herds estimated to represent at least 51% of all large pig herds in Australia. Most piggeries relied on drugs of low importance in human medicine (e.g. tetracyclines, penicillins and sulfonamides). For the two drugs of high importance in human medicine that can be legally prescribed to pigs in Australia, ceftiofur use was reported in 25% of herds and virginiamycin in none. Infections attributed to Lawsonia , Mycoplasma and Escherichia coli motivated the most use of antimicrobials. No useful association was found between management factors and the antimicrobial use index.
Conclusion   Most antimicrobial use in the Australian pig industry is based on drugs of low importance to public health. Enhanced control of E. coli infections without reliance on antimicrobials would further reduce the risk of selecting for antimicrobial resistance relevant to public health. The amount of variation in the usage index between herds suggests that antimicrobial use should be constantly reviewed on a herd by herd basis.     

Challenges posed to the traceability of weaner pigs following live auction

Objectives   To gather demographic data on live pig sales through a peri-urban saleyard in Camden, New South Wales, and to demonstrate the difficulties in tracing the subsequent movements of pigs, particularly weaner pigs.
Procedure   Records of pig sales held weekly at the Camden saleyards for the 2003/2004 and 2004/2005 financial years were analysed. Saleyard data on a number of variables were entered into a purpose-designed database. Distributions of pig sales according to pig class (weaner/porker/baconer/backfatter), purchaser type (butcher/non-butcher), and transaction type (cash/account) were determined.
Results   More weaners (3192 in 2003/2004 and 3940 in 2004/2005) were sold than any other class of pig, accounting for 45% of the total pigs sold during this period. During 2003/2004, 3802 pigs were bought by 329 non-butcher purchasers including 1631 weaners (43%) purchased by 153 non-butchers (47%). The majority of these non-butchers during the study period (86%) paid cash for their pigs and did not provide the necessary information to allow the end destination of pigs to be determined. Location data was available for all vendors but only 25% of purchasers.
Conclusions   This study highlights the challenges posed in tracing movement of pigs following sale by auction. A high proportion of weaners sold at this peri-urban saleyard would not have been slaughtered immediately. This study highlights the potential difficulties in tracing pig movements after sale, for disease control purposes. We recommend that legislation be amended requiring the identification of weaner pigs sold live at auction in all states of Australia and the recording of the property identification code of all vendors and purchasers of pigs sold live at auction.     

Adverse effects of soluble non-starch polysaccharide (guar gum) on piglet growth and experimental colibacillosis immediately after weaning

This study evaluated the effects of adding soluble fibre to the diet of healthy weaner pigs and weaner pigs experimentally infected with enterotoxigenic Escherichia coli (ETEC) in a model of post-weaning colibacillosis. Bodyweight gain, intestinal changes and proliferation of ETEC were measured 7 days following weaning. The basal diet consisted of pregelatinised rice fortified with animal protein. Addition of guar gum to this diet elevated the soluble fibre content from 1 to 6 per cent, and was associated with reduced bodyweight gains, increased large intestinal weights and fermentation, and increased proliferation of ETEC in the small intestine. The optimal levels and type of dietary fibre used for weaner pig diets require further evaluation.     

Diarrhoea in the growing pig - a comparison of clinical,morphological and microbial findings between animals from good and poor performance herds

Diarrhoea among growing pigs (8-13 weeks old) is a significant problem in many herds. Nine herds with poor performance and diarrhoea among growing pigs were selected on the basis of their piglet mean age at a body weight of 25 kg, compared to the overall mean age in Swedish herds. In addition, four herds with good average performance and no problems with diarrhoea were selected. Pigs were necropsied and samples for histology and microbiology were collected. Based on the necropsy findings, the pigs from the good performing herds were all judged to be healthy. The presence of Brachyspira pilosicoli and Lawsonia intracellularis was significantly correlated to poor performing herds and the results indicate that these microbes are main pathogens involved in enteric diseases among Swedish grower pigs. In addition, concomitant infections with other presumptive pathogens were commonly found.     

Epidemiological and genetical studies in Norwegian pig herds. III. Herd effects

Herd differences in disease incidence in 70 Norwegian pig herds were assessed by health card registrations. Such differences proved to be considerable. The incidence of the MMA-syndrome, mastitis, metritis and neonatal diarrhoea was higher in herds producing only weaners ("weaner herds") than in herds with combined production. The mean disease incidence in herds providing breeding stock for the Norwegian Pig Breeders' Association ("breeding herds") was at the same level as in ordinary combined production herds. There was no consistent relationship between changes in disease incidence and herd size, though the incidence of the MMA-syndrome was lower in the largest herds.     

A national serological survey to verify Australia's freedom from porcine reproductive and respiratory syndrome

Objective To provide serological data to support Australia's claim of freedom from porcine reproductive and respiratory syndrome.
Design A national serological survey was designed to provide 99% confidence of detecting at least one infected pig herd in Australia, assuming that at least 5% of herds would have been exposed to porcine reproductive and respiratory syndrome virus and that at least 25% of the 'finisher' pigs in these herds would have antibodies to the virus.
Procedure A two-stage testing regime was used. All samples were tested with a commercially available enzyme-linked immunosorbent assay. If assay reactions were found, all samples from the herd were to be tested using the indirect immunofluorescence antibody assay.
Results Of the 875 samples from 163 herds from all States in Australia, there was some evidence of reactivity in only four samples from four herds on the enzyme-linked immunosorbent assay. Further testing using the indirect immunofluorescence antibody assay according to the study protocol demonstrated that the reactions were not due to the presence of specific porcine reproductive and respiratory syndrome virus antibodies in the sera.
Conclusion The results of this study support the view that Australian pigs are free of porcine reproductive and respiratory syndrome virus.     

Diagnoses and treatments in health-classified fattening herds rearing pigs all in-all out

This study describes diseases encountered, medications used and veterinary involvement in all in--all out finishing herds belonging to one pork production system. The finishing herds had a particular management and housing regime. The pigs originated from health classified farrowing units. Information on 207,442 pigs was collected from 595 log books. Altogether 91% of the pigs received no treatments. Four percent of the batches of pigs were given antimicrobial mass medications. The local veterinarian visited the herds on average 2.6 times during the finishing period and made the diagnoses in more than half of the cases. At least one pig was affected with arthritis or tail biting in more than half of the batches, whereas locomotory diseases were recorded in one third of the batches. All other diagnoses were encountered in 1%-13% of the batches. Only a few pigs were treated individually in the affected groups. Antimicrobial drugs were given to 8% and other medicines to 0.7% of the pigs. The diagnosis was missing at least for one pig in 29% of the batches and the information about the medicine use in 8% of the treatments was missing. The study shows that it is possible to rear finishing pigs with only a small proportion of the animals needing treatments. The need of mass medications was low, because infectious diseases affecting the whole herd were uncommon. The recommendations for antimicrobial use given by the authorities had been followed quite well. The farmers and the veterinarians should be educated in order to realise the importance of proper record keeping.     

Descriptive summary of an outbreak of porcine post-weaning multisystemic wasting syndrome (PMWS ) in New Zealand

CASE HISTORY: Investigations were conducted to determine the cause of an acute, multi-farm outbreak of porcine respiratory disease that included diarrhoea and subsequent loss of body condition in affected pigs. A definition for post-weaning multisystemic wasting syndrome (PMWS) including both clinical and pathological features, previously developed for the pig industry in New Zealand, was applied to the current outbreak. In addition to self-reporting by owners of affected farms, local veterinarians, disease and epidemiology consultants, and animal health officials from the Ministry of Agriculture and Forestry (MAF) were involved in conducting farm visits and submission of diagnostic specimens. CLINICAL FINDINGS AND DIAGNOSIS: Pathogens known to be endemic in the pig industry in New Zealand as well as likely exotic diseases were excluded as causative agents of the outbreak. Clinical signs including dyspnoea, diarrhoea, and rapid loss of body condition were consistent with the New Zealand case definition for PMWS. Interstitial pneumonia, pulmonary oedema, generalised lymph-node enlargement, and presence of porcine circovirus type 2 (PCV2) inclusion bodies were consistently identified in affected pigs. Classical swine fever virus (CSFv), Porcine reproductive and respiratory syndrome virus (PRRSv), and Influenza virus were ruled out, using molecular and traditional virological techniques. Spread of the disease between farms was hypothesised to be facilitated by locally migrating flocks of black-backed seagulls. The original source of the disease incursion was not identified. DIAGNOSIS: Based on the consistent presence of circovirus-associated lesions in lymphoid tissues in combination with generalised enlargement of lymph nodes, histiocytic interstitial pneumonia, clinical wasting, and poor response to antibiotic therapy, a diagnosis of PMWS was made. CLINICAL RELEVANCE: PMWS should be considered in the differential diagnoses of sudden onset of respiratory dyspnoea, diarrhoea, and rapid loss of body condition in young pigs in New Zealand pig herds.     

Haematological and clinicochemical blood profiles in slaughter pigs.

In a cohort study, 40 pig finishing herds were selected: twenty pig herds with a low and twenty pig herds with a high prevalence of several pathological lesions recorded at slaughter in a six-month period before the start of the study. Blood samples were taken from 20 pigs per herd at the end of the finishing period to investigate haematological and clinicochemical profiles. There was only a significant difference in serum albumin concentration between the low and high lesion prevalence groups. There were distinct differences in blood profiles between pig herds, but also between batches of pigs within a herd, housed in different compartments. Differences between castrated males and gilts were also demonstrated with respect to mean values of the blood variables haemoglobin, iron, copper, beta-globulin, eosinophils and segmented neutrophils. However, the differences were not of a biologically important magnitude. The mean values of the blood variables pepsinogen and lymphocytes differed significantly between pig herds from the two participating integration groups. Pigs with a higher albumin concentration, a lower gamma-globulin concentration, a higher copper concentration and a higher creatine kinase activity in serum showed a higher daily weight gain.     

Detection and genetic diversity of porcine rotavirus A,B and C in eastern Australian piggeries

Rotaviruses (RV) have a high prevalence in piggeries worldwide and are one of the major pathogens causing severe diarrhoea in young pigs. RV species A, B, and C have been linked to piglet diarrhoea in Australian pig herds, but their genetic diversity has not been studied in detail. Based on sequencing of the structural viral protein 7 (VP7) RVA G genotypes G3, G4 and G5, and RVC types G1, G3, G5, and G6 have been identified in Australian piggeries in previous studies. Although occurrence of RVB was reported in Australia in 1988, no further genetic analysis has been conducted. To improve health management decisions in Australian pig herds, more information on RV prevalence and genetic diversity is needed. Here, 243 enteric samples collected from 20 pig farms within Eastern Australia were analysed for the presence of RV in different age groups using a novel PCR-based multiplex assay (Pork MultiPath™ enteric panel). RVA, RVB, and RVC were detected in 10, 14, and 14 farms, respectively. Further sequencing of VP7 in selected RV-positive samples revealed G genotypes G2, G5, G9 (RVA), G6, G8, G14, G16, G20 (RVB), and G1, G3, G5, G6 (RVC) present. RVA was only detected in young (<10 weeks old) pigs whereas RVB and RVC were also detected in older animals (>11 weeks old). Interestingly, RVB and RVC G-type occurrence differed between age groups. In conclusion, this study provides new insights on the prevalence and diversity of different RV species in pig herds of Eastern Australia whilst demonstrating the ability of the Pork MultiPath™ technology to accurately differentiate between these RV species.     

Risk factors for post-weaning mortality of Merino sheep in south-eastern Australia

Objective   To measure associations between body weight, growth rate, sex, time of shearing and post-weaning mortality of Merino sheep.
Design   Uni- and multivariable survival analyses of sheep mortality during the first year after weaning, using records (n = 3657) from two field experiments conducted in Western Victoria from 1996 to 2003.
Results   Overall mortality was 14.3% (range 4.5–26.8%) and mean maximum mortality rate was 29 deaths/1000 weaners/month. Increased mortality risk was associated with decreases in fleece-free body weight and mean weaner growth rate, particularly at low weights and growth rates. Weaners in the lightest weaning weight quintile had a hazard ratio of 3.5, compared with the middle quintile. The hazard ratio for a 2-kg decrease in weaning weight was 1.2 to 1.7 for weaners lighter than 22 kg. The hazard ratio for a reduction in mean weaner growth rate in the first 5 months after weaning of 0.25 kg/month was 1.1 to 6.8 if mean growth rate was less than 1 kg/month, but did not differ significantly from 1 at greater growth rates. The hazard ratio for wether weaners was approximately 1.5 compared with ewe weaners. The hazard ratio for weaners shorn between December and May, compared with unshorn weaners, was 1.2 to 3.5, with the greatest risk difference associated with shearing in March (45 deaths/1000 weaners/month).
Conclusion   Improving the body weight and mean growth rate of weaner sheep is likely to reduce post-weaning mortality. Lightweight weaners in a flock should be managed separately from the main portion after weaning. In southern Australia, not shearing spring-born Merino weaners between December and May may assist in reducing overall post-weaning mortality.     

Risk factors for clinical signs of PMWS and PDNS in pigs in The Netherlands: a case-control study

Potential risk factors for clinical signs of post-weaning multisystemic wasting syndrome (PMWS) and porcine dermatitis and nephropathy syndrome (PDNS) in pigs in the Netherlands were investigated in a matched case-control study using a questionnaire (personal interview). Eighty-two pig farmers were questioned about management, hygiene, husbandry systems, disease history, and preventive health care. In this study, 30 pig herds with (cases) and 30 pig herds without (controls) characteristic clinical signs of PMWS were compared. For PDNS, 11 pig herds with (cases) and II pig herds without (controls) characteristic clinical signs of PDNS were compared. Univariate analysis (P < 0.10) showed that the following occurred relatively more often in the PMWS case herds than in the control herds: 1) clinical signs of PDNS, porcine reproductive and respiratory syndrome (PRRS), porcine parvovirus (PPV) infections, meningitis, coccidiosis, and pre-weaning diarrhoea observed by the farmer; 2) vaccination against PRRS and mycoplasma; 3) non-optimal climatic conditions in the nursery rooms, a large variation in weaning age, a high occurrence of cross-fostering of piglets, a large number of sows with lactation problems, poor colostrum intake by piglets; and 4) (historical) use of breeding stock (including semen for artificial insemination) of Anglo-Saxon origin. In the final multivariate statistical model, one variable remained significantly associated with PMWS case herds, namely, the presence of clinical signs of PRRS (and/or the associated use of vaccination against PRRS). It should be noted that in almost all cases animals were vaccinated against PRRS because of clinical signs of PRRS that appeared a few months after the first occurrence of clinical signs of PMWS. This excludes PRRS vaccination as a primary factor in causing PMWS. Analysis of the PDNS case-control data showed comparable results with those of the PMWS study. In the final statistical model, the presence of clinical signs of PRRS (and/or the associated use of vaccination against PRRS) was significantly associated with PDNS case herds.     

The effect of antimicrobial growth promoter withdrawal on the health of weaned pigs in Finland

The use of the antimicrobial growth promoters (AGPs) carbadox and olaquindox has been banned in the European Union (EU) since September 1999. We studied the effects of the withdrawal on the health of weaned piglets on two types of piglet-producing farms (farrowing herds and farrow-to-finish herds) from the different regions of Finland. Farms with no major problems with post-weaning diarrhoea were selected for the study to better evaluate the effect of AGPs alone. Data on production, medication and incidence of diarrhoea were collected from 73 farms during 1 year after the withdrawal. On 29 of these farms, the data collection began 4 months before the withdrawal.

The health management of the pigs is considered good in Finland, and special attention has been paid to improve the husbandry practices and management of the farms. Eighty-two percent of the farms in the study were free of both Mycoplasma hyopneumoniae and Sarcoptes scabiei infection. Brachyspira hyodysenteriae infection was not detected in any of the farms. The median number of sows in the herds was 56.0 (IQR = 43.0; 72.5) in 2000.

The level of antimicrobial use in each herd was classified as low, moderate and high when the percentage of weaned pigs treated for diarrhoea during a 4-month period was 0–5%, 6–19% and ≥20%, respectively. Only on four herds (14%), there was an increase in the level of antibiotic use after the AGP withdrawal, when seasonally corresponding 4-month periods were compared. Fourty-one percent of these 29 farms were categorized as low users of antimicrobials, 38% as moderate users and 21% as high users.

The level of antimicrobial use for treatment of diarrhoea after weaning (and the incidence of diarrhoea in weaned piglets) did not increase significantly after the withdrawal of AGPs from weaner feeds according to farmers’ evaluations. In this study, the Escherichia coli infection was the most-common cause of diarrhoea in weaned pigs. The age at weaning did not change after the withdrawal of AGPs.