The incidence and economic impact of the Escherichia coli peritonitis syndrome in Dutch poultry farming

Abstract

The incidence and economic impact of the Escherichia coli peritonitis syndrome (EPS), characterized by acute mortality, were estimated in chicken egg-producing farms in the Netherlands in 2013. The incidence was significantly higher (P < 0.05) in the meat-sector (35% affected farms) compared to the layer-sector (7% affected farms). In consumption egg-producing farms EPS occurred on 12% of the free range and organic farms, while it was found on 1% and 4% of the cage and barn farms, respectively. Data from four layer and two broiler breeder flocks with EPS were used to estimate the overall economic impact of the disease. Mean numbers of eggs lost were 10 and 11 per hen housed (phh), while mean slaughter weight loss was 0.2 and 0.5 kg phh in the four layer and two broiler breeder flocks, respectively. Total losses including costs of destruction of dead hens, compensated for reduced feed intake due to a smaller flock size, ranged from €0.28 phh (cage farms) to €9.75 phh (grandparent farms) in the layer-sector and from €1.87 phh (parent farms) to €10.73 phh (grandparent farms) in the meat-sector. Antibiotics against EPS were given often and repeatedly especially in the meat-sector. Including the costs of antibiotics, total losses were estimated at €0.4 million, €3.3 million and €3.7 million for the layer-sector, the meat-sector and poultry farming as a whole, respectively. Research focusing on the prevention and treatment of EPS is justified by its severe clinical and economic impact.

Introduction

Colibacillosis in laying chickens induced by avian pathogenic Escherichia coli predominantly refers to salpingitis/peritonitis/salpingoperitonitis (SPS) and the E. coli peritonitis syndrome (EPS).

Coliform SPS (“egg peritonitis”), which has been observed since the beginning of poultry farming, is a chronic condition and forms a substantial proportion of the “normal” mortality (norm mortality as presented in the breeder's manual) in flocks of laying hens. The incidence of SPS is generally low and as a consequence its economic impact is limited, because mortality rates are low and egg production and egg quality are not or only slightly affected (Bisgaard & Dam, 1981; Jordan et al., 2005; Landman & Cornelissen, 2006; Stokholm et al., 2010). SPS is probably induced by opportunistic E. coli bacteria, which may ascend from the cloaca into the oviduct of hens predisposed to salpingitis due to the relaxation of the inner circular vaginal musculature as a consequence of intensive egg production and associated oestrogenic activity (Barnes et al., 2008).

In contrast to SPS, EPS is a (per)acute disease characterized by mortality soaring up to 10–15% above the norm in combination with severe septicaemia and fibrinous polyserositis lesions, while salpingitis rarely occurs (Dhillon & Jack, 1996; Zanella et al., 2000; Vandekerchove et al., 2004a; Landman & Cornelissen, 2006; Raviv et al., 2007; Landman et al., 2013, 2014). E. coli strains involved in EPS are likely primary virulent (Vandekerchove et al., 2004b; Boelm et al., 2011; Landman et al., 2013). The disease occurs after these E. coli strains enter the bloodstream inducing bacteraemia, probably via the respiratory tract, although entrance via the vaginal route cannot be discarded (Vandekerchove et al., 2004b; Landman et al., 2013). Following experimentally induced infections, three distinct forms of EPS were found: the septicaemic form (only signs of sepsis are seen and E. coli is isolated from the bone marrow), the acute peritonitis form (besides signs of sepsis, acute fibrinous peritonitis is also observed and E. coli is isolated from bone marrow) and the chronic peritonitis form (i.e. chronic peritonitis is found without signs of sepsis and E. coli is not isolated from the bone marrow) (Landman et al., 2013). EPS has been observed from the mid-1990s in the USA and in many European countries including the Netherlands. It seems to have increased substantially thereafter (Dhillon & Jack, 1996; Zanella et al., 2000; Vandekerchove et al., 2004a; Landman & Cornelissen, 2006; Raviv et al., 2007).

In view of the high mortality due to EPS, the disease results in great economic losses at the farm level and likely also for the poultry farming as a whole. However, published data on the quantification of mentioned losses are not available. Therefore, in the present work, the economic impact of EPS for both individual Dutch farms and poultry farming in the Netherlands was investigated.

The study consists of three parts. In Part I, the scale of the Dutch laying chicken farming and the EPS incidence referring to 2013 were assessed. In Part II, losses in four layer and two broiler breeder flocks (reference flocks), which had been used previously in a study performed in 2011–2012 on the molecular typing of avian pathogenic E. coli colonies obtained from the bone marrow of hens with EPS (Landman et al., 2014), were quantified. In Part III, losses in Dutch poultry farming as a whole were estimated based on data obtained in Parts I and II. Losses estimated in Parts II and III were calculated using prices of 2014 as prices of 2013 were not available anymore at the time (year 2014) calculations were performed, but likely prices in 2013 and 2014 will not differ substantially.

Materials and Methods

Definitions

Poultry sector (restricted to egg laying chickens) indicates either all layer chickens or all meat chickens. Further, within each sector different categories are considered. Categories within the layer-sector are grandparent, parent, layers housed in cages (layer cage), layers housed in barns (layer barn), layers kept in a free range system (layer free range), organic layers (layer organic) and layers kept for the production of vaccines (layer vaccine). Categories within the meat-sector are grandparent and parent. Flocks of laying hens were considered to have EPS if their mortality was increased and E. coli was isolated from bone marrow of fresh dead hens. In these flocks, peritonitis was often found.

All prices and costs are presented in euros excluding value-added tax (VAT) and refer to the Netherlands in 2014.

Part I

Estimation of the scale of Dutch poultry farming

Data on the number of farms populated in 2013 and the number of hens allocated on these farms provided by the Dutch Commodity Board for Poultry and Eggs were used to estimate the scale of Dutch poultry farming. The cycle length is essential to be able to calculate the total number of poultry farms and the average number of hens present per category. The cycle length was defined as the age of hens at slaughter (in days) minus the rearing period (in days) plus the downtime period (in days). The length of the rearing period was considered to be 119 days for the layer-sector and 140 days for the meat-sector. The downtime period was set at 20 days for both sectors. In the layer-sector, flocks are frequently moulted, which was taken into account for the calculation of the weighted average age at slaughter. Data on the number of farms populated in 2013, the number of hens allocated in 2013, the age at slaughter and the average weighted cycle length have been outlined in Tables 1 and 2. The total number of farms was calculated by multiplying the number of farms populated in 2013 by the cycle length (days) divided by 365. The average number of hens present per category was calculated by multiplying the number of hens placed in 2013 by .

Table 1. Estimation of the scale of the Dutch layer-sector farming (restricted to egg laying chickens) in 2013.

Categories	Number of farms populated in 2013	Number of hens allocated in 2013 × 1000	Age at slaughter (days)^a				Average weighted cycle length (days)^b	Estimation of the total number of farms^c	Estimation of the average number of hens present per category × 1000^d
			White layers		Brown layers
			Moulted (% of flocks)	Not moulted (% of flocks)	Moulted (% of flocks)	Not moulted (% of flocks)
Grandparent	5	97	(0)	511 (100)	(0)	511 (100)	412	5	104
Parent	29	663	(0)	504 (100)	(0)	490 (100)	398	32	687
Layer cage	46	3078	770 (60)	616 (40)	735 (60)	574 (40)	591	74	4815
Layer barn	473	16,440	735 (50)	560 (50)	714 (50)	546 (50)	540	700	23,421
Layer free range	178	4343	700 (30)	546 (70)	658 (30)	532 (70)	482	235	5497
Layer organic	117	1377	(0)	546 (100)	(0)	525 (100)	437	140	1573
Layer vaccine	13	533	(0)	450 (100)	(0)	450 (100)	351	13	483

^aPersonal communication, P. Arts, Hendrix Genetics B.V., Boxmeer, the Netherlands and B. van Nijhuis, Verbeek Hatchery, Lunteren, the Netherlands.

^bCycle length is age at slaughter (days) minus the length of the rearing period (on average 119 days) plus the downtime period (on average 20 days). The age at slaughter was estimated taking into account the percentage of flocks moulted and considering that an equal number of white and brown layers are kept.

^cThe total number of farms is the number of farms populated in 2013 multiplied by .

^dThe average number of hens present per category is the number of hens allocated in 2013 multiplied by .

Table 2. Estimation of the scale of Dutch meat-sector farming (restricted to egg laying chickens) in 2013.

Categories	Number of farms populated in 2013	Number of hens allocated in 2013 × 1000	Age at slaughter (days)^a	Average cycle length (days)^b	Estimation of the total number of farms^c	Estimation of the average number of hens present per category × 1000^d
Grandparent	24	594	444	324	21	495
Parent	200	5165	428	308	169	4075

^aPersonal communication, O. van Tuijl, Aviagen B.V., Roermond, the Netherlands.

^bCycle length is age at slaughter (days) minus the length of the rearing period (on average 140 days) plus the downtime period (on average 20 days).

^cThe total number of farms is the number of farms populated in 2013 multiplied by .

^dThe average number of hens present per category is the number of hens allocated in 2013 multiplied by .

Estimation of the EPS incidence in Dutch poultry farming

The incidence of EPS in 2013 in Dutch poultry on the farm and house level was estimated using data from submissions of birds from affected farms to GD - Animal Health, data from Integral Production Chain Monitoring – Central Registration Antibiotic Use (Integrale Keten Bewaking – Centrale Registratie Antibiotica (IKB – CRA)) and data from the Veterinary Monitoring of Poultry (VMP). Virtually all Dutch poultry veterinarians participate in IKB – CRA and must report disease cases with associated antibiotic use to the database. Furthermore, about 40% of Dutch poultry veterinarians participate in VMP in which system diseases are reported independent of antibiotic use. Double counts of disease outbreaks, that is, cases that had been registered through the submission of diseased birds to GD, but had also been reported through IKB – CRA and VMP, were avoided by using the Unique Identification Number of farms and the birthdate of the flock.

Part II

Reference farms and flocks

Four chicken farms with outbreaks of EPS were included in this part of the study (farms A, B, C and D). Farms A, B and C were layer farms with different housing systems (cage, floor and free range, respectively), while on farm D broiler breeders were kept in a floor system. Farm C was the only organic and only multiple age farm (Table 3). In farm B, EPS was observed in the second laying period after forced moulting at 70 weeks of age. Two successive flocks of farms C and D showed EPS.

Table 3. The economic impact of EPS in six reference flocks of laying chickens expressed phh and per flock.

Farm-flock	Breed	Housing	Flock size^a	Period of increased mortality^b	Mortality (norm)^c	Number of eggs lost (value = I)^d	Egg mass loss in kg^d	Slaughter weight loss of flock in kg (value = II)^d	Reduction of feed consumption in kg (value = III)^d	Costs of destruction dead hens (IV)^d	Total loss (I + II – III + IV = V)^d	Loss per flock (flock size × V)
A-1^e	brown layers	cage	51,932	41–57	4.3 (1.6)	2.68 (0.14)	0.171	0.053 (0.02)	0.326 (0.09)	0.00	0.07	3635
B-1^f	white layers	floor	26,646	80–106^g	14.1 (3.8)^h	8.89 (0.54)	0.569^h	0.178^h (0.07)	1.062^h (0.29)	0.01	0.33	8793
C^i-1	brown layers	free range	15,826	40–73^g	18.0 (5.8)	10.38 (1.40)	0.681	0.242 (0.11)	1.609 (0.70)	0.01	0.82	12,997
C^i-2^j	brown layers	free range	15,547	36–77^g	19.8 (3.4)	16.71 (2.26)	1.062	0.341 (0.15)	2.370 (1.03)	0.02	1.40	21,766
D-1	broiler breeders	floor	13,890	24–57^g	22.5 (6.8)	15.19 (2.96)^k	irrelevant	0.627 (0.45)	3.374 (0.95)	0.03	2.49	34,586
D-2^j	broiler breeders	floor	13,680	25–59^g	13.9 (7.0)	6.20 (1.21)	irrelevant	0.279 (0.20)	1.317 (0.37)	0.01	1.05	14,364

Notes: values, costs of destruction, total loss and loss per flock are expressed in euros. Health care costs (diagnostics, medication (flock D1) and vaccination (flock D2)) are not included. Prices of eggs, feed, spent hens and destruction of dead hens are given in Table 4 and refer to the Netherlands in 2014. Calculation procedures can be found in the text.

^aAt 20 weeks of age except for farm B for which flock size at 75 weeks of age is given.

^bExpressed as age of birds in weeks.

^cCumulative mortality during the period of increased mortality in percentage of the number of birds at 20 weeks; for farm B the number of birds at 75 weeks is used.

^dExpressed phh at 20 weeks of age; for farm B at 75 weeks of age.

^e10,560 hens of this flock were sold at 54 weeks of age.

^fFlock was moulted at 70 weeks of age.

^gThe latter age represents the end of the production period.

^hNorm dates from 2002 and concerned white layers housed in cages. Recent data are not available.

ⁱOrganic farm.

^jFlock was allocated in the same house as the previous flock on this farm included in the present study.

^kLoss for this flock was calculated from 25 weeks of age onwards as the weight of the first eggs is below the required weight (≥50g) for a hatching egg.

Samples of birds originated from one and the same house per farm and were taken once from farms A and B, while two successive flocks with EPS of farms C and D were each sampled once. The age of the hens at sampling and the number of selected dead birds per flock ranged from 28 to 82 weeks and from 12 to 25, respectively. In total, 102 hens were examined; 98 (96%) had died due to EPS and from 92% (90/98) of these hens with EPS E. coli was isolated from the bone marrow (Landman et al., 2014).

Flock sizes, the age periods with increased mortality and mortality data are presented in Table 3. The start of the period of increased mortality in hens ranged from 24 weeks of age (farm D, flock 1) to 80 weeks of age (farm B, moulted flock). Increased mortality continued until the end of the production period, except for one flock in which elevated mortality was observed from 41 to 57 weeks of age, while the production period ended at 68 weeks of age (farm A). Mortality exceeding the norm ranged from 2.7% to 16.4% in hens.

Economic impact of EPS on various parameters due to increased mortality

The impact of mortality due to EPS on egg production (number and kg of eggs) and slaughter weight of the flock was calculated. Losses were corrected for the decreased feed consumption resulting from a smaller flock size.

All EPS effects are expressed per hen housed (phh) at 20 weeks of age and for flock B1 at 75 weeks of age. Except for loss of number of eggs, all other effects are presented in kg.

Reference data for egg production, egg weight, body weight, body weight of spent hens and feed consumption

Egg production based on hens present is not or only slightly affected by EPS as the mortality from this disease is (per)acute (Zanella et al., 2000; Vandekerchove et al., 2004a, b; Landman et al., 2014). Therefore, the actual egg production based on hens present of the flocks concerned was used as reference. “Normal” mortality rate, egg weight, body weight, body weight of spent hens and feed consumption were obtained from the breeder's management guides. The feed norm of flock B1 dates from 2002 and concerned housing in cages; recent data were not available.

Calculation procedures

Loss of eggs and egg mass

Loss of eggs due to EPS is the number of eggs that would have been produced by hens with EPS if they had remained healthy. Calculation of loss of eggs was performed as follows. The number of hens that died in the first week in which mortality had increased was multiplied with the number of days between the middle of mentioned week and the end of the production period. Subsequently, this number was multiplied with the mean daily egg production per hen from the middle of mentioned week until the end of the production period. The latter was calculated by dividing added weekly production percentages by the number of weeks and by hundred. The same procedure was performed for the number of hens that died in the second week of increased mortality and so forth, until and including the last week of increased mortality. Egg losses due to mortality for each of the weeks of the period of increased mortality were added and thereafter multiplied by

(*mortality rate refers to the period of increased mortality) to compensate for “normal” mortality. Finally, the result was divided by the number of hens housed.

Loss of egg mass in layer flocks was determined as outlined next. First, the weighted average age at which the mortality occurred in the period of increased mortality was calculated. Egg weight at that age was averaged with egg weight at the end of the production period. Multiplying the number of eggs lost phh with mentioned average egg weight results in egg mass loss phh.

Loss of slaughter weight

The number of hens that died due to EPS was multiplied with the body weight of their flock mates at the end of the production period and divided by the number of hens housed.

Feed saved

The feed saved was calculated by multiplying the number of hens that died due to EPS by the number of days between the weighted average age at which mortality occurred in the period of increased mortality and the end of the production period by average daily feed intake per hen. The average daily feed intake per hen was the mean of the intake at mentioned weighted average age and that at the end of the production period. Finally, this calculation result was divided by the number of hens housed.

Medication and E. coli vaccination

Only flock D1 was medicated. Medication was as follows: doxycycline (Soludox 50% (Eurovet Animal Health B.V., Bladel, the Netherlands)), enrofloxacin (Enrox oraal 10% (Eurovet Animal Health B.V.)), colistin (Coliplus (AUV, Cuijk, the Netherlands)), doxycycline (Soludox 50%) plus tylosin (Tylan^® WO (Elanco Animal Health, Houten, the Netherlands)) and flumequin (Enterflume^® kalf kip (Dopharma B.V., Raamsdonkveer, the Netherlands)) at the age of 22 (preventive treatment), 25, 27, 28 and 34 weeks, respectively. Medications were given in the drinking water and the manufacturer's recommendations were followed for dosages and treatment duration (4–7 days).

Flock D2 was vaccinated twice with an E. coli autovaccine. The costs of medication (flock D1) and of E. coli vaccination (flock D2) were provided by the practitioner and are presented for the flock as a whole and phh.

Destruction of dead hens

The collection and destruction of dead farm birds in the Netherlands is carried out by Rendac (Son, the Netherlands) according to a routine schedule and itinerary. Dead chickens can be deposited at the farm in containers of 200, 750 or 950 l. The price to empty any of these containers by Rendac is €12.98, while the processing costs increase with the size of the container volume being €2.97, €13.33 and €16.90, respectively. Evidently, it is more costly to empty containers with dead birds if extra journeys are needed. Calculation of the costs involved with the collection and destruction of dead hens due to increased mortality was performed considering that containers with dead hens were emptied within a routine schedule and itinerary and that containers of 750 l, which can harbour 560 kg chicken, were used. Costs of destruction per flock were obtained by multiplying the number of hens that died due to EPS by their mean body weight in kg at the weighted average age this mortality occurred. Subsequently, the outcome was divided by 560, which yields the number of extra containers to be emptied. The latter number was multiplied by €26.31 (12.98 + 13.33) and subsequently divided by the number of hens housed at 20 weeks of age and rounded-off to euro cents, yielding the extra costs phh.

EPS effects in euros

Prices of eggs, spent hens, feed and destruction of dead hens used in the present study to calculate the economic effects of EPS are presented in Table 4. The effect phh was obtained by subtracting the saved feed costs from the sum of costs of lost eggs, lost slaughter weight and destruction of dead hens.

Table 4. Prices in euros excluding VAT of eggs (per 100), of spent hens (per kg body weight), of feed (per 100 kg) and of destruction of dead hens (per 750 l = 560 kg) used to calculate the economic effect of EPS occurring in six reference flocks of laying chickens.

Farm-flock	Breed	Housing	Eggs	Spent hens	Feed	Destruction of dead hens
A-1	brown layers	cage	5.28^a	0.45^b	27.25^c	26.31^d
B-1	white layers	floor	6.07^e	0.41^f	27.25^c	26.31
C^g-1	brown layers	free range	13.50^h	0.45^b	43.57^i	26.31
C^g-2	brown layers	free range	13.50^h	0.45^b	43.57^i	26.31
D-1	broiler breeders	floor	19.47^j	0.72^k	28.10^l	26.31
D-2	broiler breeders	floor	19.47^j	0.72^k	28.10^l	26.31

Note: prices refer to the Netherlands in 2014.

^aNederlandse Organisatie van Pluimveehouders (NOP = Dutch Organisation of Poultry Farmers) quotation for brown eggs (63–64 g) of week 10.

^bNOP slachtkippen (chickens for slaughter) quotation for brown layer hens (>1.9 kg) of 13 March.

^cBinternet: prijzen volgens prijs-informatie desk (Landbouw Economisch Instituut (LEI)) (Binternet: prices according to the price information desk (Agricultural Economics Research Institute)). Current prices all mash feeds February. Layer ration phase 2 for lots of 8 metric tons (mt).

^dCosts of collection during routine itinerary and processing of dead hens deposited in crane containers of 750 l (= 560 kg chicken) (Rendac, Son, the Netherlands).

^eNOP quotation for white barn eggs (64–65 g) of week 10.

^fNOP slachtkippen (chickens for slaughter) quotation for white layer hens (>1.7 kg) of 13 March.

^gMultiple age organic farm.

^hBiomeerwaarde Ei, Tholen, the Netherlands, personal communication, W. Thomassen (organic eggs).

ⁱPersonal communication, D. Keppels (Reudink B.V., Lochem, the Netherlands).

^jNOP/Nederlandse Vereniging van Pluimveehouders (Dutch Trade Union of Poultry Farmers) quotation for hatching eggs of week 9.

^kNOP slachtkippen (chickens for slaughter) quotation for broiler breeder hens (>3.5 kg) of 13 March.

^lBinternet: prices according to the price information desk (LEI). Current prices all mash feeds February. Laying broiler breeder mash for lots of 8 mt.

The financial impact on the farmer's labour income was obtained by multiplying this amount of money with the number of hens housed.

Part III

Extrapolation of the impact of EPS observed in Part II to all poultry categories described in Part I

The incidence of several diseases including E. coli infections seems to be dependent on the housing system (bacterial infections are more frequently found in laying hens in litter based housing systems including free range in comparison with caged birds (Kreienbrock et al., 2003; Petermann, 2003; Fossum et al., 2009); however, the severity of EPS seems to be housing independent (Dhillon & Jack, 1996; Zanella et al., 2000; Vandekerchove et al., 2004a)). The latter justifies averaging the data of the flocks described in Part II of the study; averages were calculated for layer and broiler breeder flocks separately. Average losses of eggs, of slaughter weight, average reduction in feed consumption and average costs of destruction of dead hens phh were calculated. Lost eggs were rounded-off to whole numbers, reduction of slaughter weight and feed consumption to hundreds of grams and costs of destruction of dead hens to euro cents. In order to calculate the financial impact of EPS on the whole of the Dutch laying chicken farming, the average data obtained from the four reference layer and two reference broiler breeder flocks were applied to all categories of layer- and meat-type chickens, respectively (Table 5).

Table 5. The economic impact of EPS on layer and meat chickens phh in the Netherlands in 2013.

Poultry sector and categories	Number of eggs lost (value = I)	Slaughter weight loss of flock in kg (value = II)	Reduction of feed consumption in kg (value = III)	Costs of destruction dead hens (value = IV)	Total loss (I + II – III + IV)
Layer-sector
Grandparent	10 (10.00)	0.2 (0.09)	1.3 (0.35)	0.01	9.75
Parent	10 (1.50)	0.2 (0.09)	1.3 (0.35)	0.01	1.25
Layer cage	10 (0.53)	0.2 (0.09)	1.3 (0.35)	0.01	0.28
Layer barn	10 (0.61)	0.2 (0.09)	1.3 (0.35)	0.01	0.36
Layer free range	10 (0.71)	0.2 (0.09)	1.3 (0.35)	0.01	0.46
Layer organic	10 (1.35)	0.2 (0.09)	1.3 (0.57)	0.01	0.88
Layer vaccine	10 (0.81)	0.2 (0.09)	1.3 (0.35)	0.01	0.56
Meat-sector
Grandparent	11 (11.00)	0.5 (0.36)	2.3 (0.65)	0.02	10.73
Parent	11 (2.14)	0.5 (0.36)	2.3 (0.65)	0.02	1.87

Notes: values and total loss are given in euros. Health care costs (diagnostics, medication and E. coli vaccination) are not included. Prices of eggs, feed, spent hens and destruction of dead hens are given in Table 4. Prices of categories of eggs not presented in Table 4 were kindly provided by P. van Horne (LEI = Agricultural Economics Research Institute), the Netherlands (personal communication). All prices are exclusive of VAT and refer to the Netherlands in 2014. Calculation procedures can be found in the text. The economic impact was calculated based on the assumption that layer- and meat-type flocks suffered the same production losses as the average production losses observed in respectively the four reference layer and the two reference broiler parent flocks with EPS described in Table 3. Total losses were used to estimate the financial impact of EPS for Dutch poultry farming in 2013 (Table 8).

As prices of eggs and feed vary between breeds and housing systems, the financial impact of EPS was calculated for each of these. Calculation of the price of spent hens was made using the mean price of hens weighing >1.7 and >1.9 kg (€0.43 per kg body weight).

Use and costs of antibiotics

Based on IKB – CRA data, nine antibiotics were used (Table 6). The respective compounds used were amoxicillin trihydrate (Paracilline oplosbaar poeder (MSD Animal Health, Boxmeer, the Netherlands) and Octacilline (Eurovet Animal Health B.V.)), colistin sulphate (Coliplus (AUV) and Colisol^® (Dopharma B.V.)), doxycycline hyclate (Soludox 50% (Eurovet Animal Health B.V.) and Doxycycline 50% WSP (Dopharma B.V.)), enrofloxacin (Lanflox (Dopharma) and Enrox oraal 10% (Eurovet Animal Health B.V.)), flumequine (Enterflume^® Kalf Kip (Dopharma B.V.)), oxytetracycline hydrochloride (Dopharma B.V.), phenoxymethylpenicillin potassium (Phenoxypen^® WSP (Dopharma B.V.)), sulfachloropyridazine sodium and trimethoprim (Cosumix^® Plus (Dopharma B.V.)), and tylosin tartrate (Tylogran^® WSP (Dopharma B.V.) and Tylan^® WO (Elanco Animal Health)).

Table 6. Antibiotics used against EPS in Dutch laying chicken flocks in 2013: price, dose, duration of treatment and costs of treatment of layer- and meat-type hens.

Antibiotic	Price (€/kg)	Dose (mg/kg BW/day)	Duration treatment (days)	Costs of treatment in € per
				1000 layer-type hens (BW = 1.85 kg)	1000 meat-type hens (BW = 3.5 kg)
Amoxicillin trihydrate	199	23	4	34	64
Colistin sulphate	87/10^9 IU	75,000 IU	4	48	91
Doxycline hyclate	128	25	4	24	45
Enrofloxacin	780	10	4	58	110
Flumequine	107	10	4	8	15
Oxytetracycline hydrochloride	30	30	5	8	16
Phenoxymethylpenicillin potassium	191	20	5	35	67
Sulfachloropyridazine sodium and trimethoprim	228	30 and 6	4	51	96
Tylosin tartrate	122	100	4	90	171

Note: prices and costs exclude VAT and refer to the Netherlands in 2014.

Distributor prices of one packing unit (550 g to 5 kg; colistin sulphate: 1 l) were used in calculations and prices are expressed per kg pure antibiotic (colistin sulphate: per 10⁹ IU) (Table 6). Dose and duration of medication used in calculations were according to the manufacturer's recommendations; in cases where a dose and/or a duration interval was given, averaged values were used. All antibiotics were supplied by drinking water. If two products with the same antibiotic were administrated, prices, recommended dose per kg BW per day and recommended duration of treatment were averaged. Costs of antibiotic treatment of layer- and meat-type hens were calculated and expressed per 1000 birds, taking into account a BW of 1.85 and 3.5 kg, respectively (Table 6). Multiplying the number of treated hens (rounded-off to thousands) in each of the categories of layer- and meat-type farming with the price of the antibiotic treatment per 1000 hens resulted in the costs of antibiotic use in each of the categories. Summarizing these costs yields the total costs of antibiotic treatments against EPS for Dutch layer- and meat-type farming in 2013. The percentage of hens in houses in which EPS occurred that was treated with antibiotics was calculated as well (Table 7). Prices of antibiotics per kg (colistin sulphate per 10⁹ IU), treatment costs per 1000 hens and costs of treatment of each of the chicken categories were rounded-off to whole euros. Costs of medication for each of the chicken categories were rounded-off to thousands of euros (Table 8).

Table 7. Use and costs of antibiotics administered against EPS in Dutch laying chicken flocks in 2013.

Poultry sector and categories	Number of hens in houses in which EPS occurred (× 1000)	Number of hens treated (× 1000) (%^a)	Number of hens (× 1000) treated with (treatment costs in €)										Total costs (€)
			Amoxicillin trihydrate	Colistin sulphate	Doxycycline hyclate	Enrofloxacin	Flumequine	Oxytetracycline hydrochloride	Phenoxymethylpenicillin potassium	Sulfachloropyridazine sodium and trimethoprim	Tylosin tartrate
Layer-sector
Grandparent	11	3 (27)	3 (102)			6 (348)							450
Parent	106	42 (40)			28 (672)		14 (112)						784
Layer cage	51	0 (0)											0
Layer barn	621	335 (54)		372 (17,856)							96 (8640)		26,496
Layer free range	696	99 (14)		99 (4752)									4752
Layer organic	248	28 (11)		28 (1344)									1344
Layer vaccine	80	0 (0)											0
													33,826
Meat-sector
Grandparent	118	98 (83)			36 (1620)	30 (3300)			7 (469)	95 (9120)			14,509
Parent	1026	685 (67)	17 (1088)	87 (7917)	302 (13,590)	101 (11,110)	47 (705)	164 (2624)	6 (402)	273 (26,208)			63,644
													78,153

Note: costs are excluding VAT and have been estimated using the data presented in Table 6.

^aPercentage of hens in houses in which EPS occurred that were treated. Note that several flocks were treated with antibiotics more than once.

Table 8. Scale of the Dutch laying chicken farming and an estimation of the incidence and economic impact of EPS in 2013.

Poultry sector and categories	Number of farms	Average number of hens present per category × 1000	Number of farms with EPS (% of number of farms)	Number of hens in houses where EPS occurred × 1000 (% of average number of hens present)	Loss in € × 1000 due to loss of eggs, loss of slaughter weight and destruction costs of dead hens compensated for reduced feed intake^a	Costs of medication in € × 1000^b	Total losses in € × 1000
Layer-sector
Grandparent	5	104	2 (40)^A	11 (11)	107	0^c	107
Parent	32	687	8 (25)^B,C	106 (15)	133	1	134
Layer cage	74	4815	1 (1)^B	51 (1)	14	0	14
Layer barn	700	23,421	30 (4)^C,D,E	621 (3)	22	26	48
Layer free range	235	5497	28 (12)^D	696 (13)	32	5	37
Layer organic	140	1573	17 (12)^E	248 (16)	22	1	23
Layer vaccine	13	483	3 (23)^F	80 (17)	45	0	45
	1198	36,580	89 (7)^A	1813 (5)	375	33	408
Meat-sector
Grandparent	21	495	11 (52)^A	118 (24)	1266	15	1281
Parent	169	4075	55 (33)^B	1026 (25)	1919	64	1983
	190	4570	66 (35)^A	1144 (25)	3185	79	3264

Notes: within a poultry sector, data with a superscript uppercase letter in common differ significantly (P  <  0.05). Data with a subscript uppercase letter in common differ significantly (P < 0.05). All prices used to calculate losses and costs are excluding VAT and refer to the Netherlands in 2014.

^aLoss phh is given in Table 5.

^bCosts of medication can be found in Table 7.

^cValue is zero due to rounding-off to thousands of euros.

Statistical analysis

Differences in the proportion of farms with EPS between the layer- and meat-sector and between categories within each of the sectors were analysed using Fisher's exact test (Statistix^®, 2010). Bonferroni correction was applied for the multiple comparisons performed within the layer-sector. Differences were considered significant if P < 0.05.

Results

Part I

The scale of the Dutch laying chicken farming in 2013 is outlined in Table 8.

The incidence of EPS at farm level was significantly lower (P < 0.01) in the layer-sector in comparison with the meat-sector (7% and 35% affected farms, respectively). Farms with EPS harboured 5% and 25% of the hens that on average were present in the respective poultry sector (Table 8).

The EPS incidence in consumption egg-producing farms was low in cage and barn farms (1% and 4%, respectively) compared to free range and organic farms (12% in each of these categories). The difference in EPS incidence between barn farms and farms with free range (free range farms and organic farms) was significant (P < 0.05). Between the remaining categories of the layer-sector, the incidence of EPS ranged from 2–3% to 40% of the farms; differences being not significant (P > 0.05).

In the meat-sector, the EPS incidence ranged from 33% to 52% of broiler parent and broiler grandparent farms, respectively. Difference in EPS incidence between these categories was not significant (P > 0.05; Table 8).

Part II

Losses due to EPS in the six reference flocks are presented in Table 3. In layers, egg and slaughter weight losses phh ranged from 2.68 eggs to 16.71 eggs and from 0.053 to 0.341 kg, respectively. These losses resulted after correction for the reduced feed consumption, which ranged from 0.326 to 2.370 kg phh, in a loss of €0.07–1.40 phh. Losses per flock ranged from €3635 to €21,766.

In the two broiler breeder flocks losses were 15.19 and 6.20 eggs phh, while slaughter weight losses were 0.627 kg and 0.279 kg phh. Reduced feed consumptions were 3.374 kg and 1.317 kg phh. Costs of medication of flock D1 amounted to €2769, which was €0.20 phh. The costs of E. coli vaccinations with autovaccine of flock D2 were €2610 for the vaccine and €684 for the application, which results in €3294 for the flock and €0.24 phh. Including costs of medication and vaccination losses were €2.69 and €1.29 phh for flock D1 and D2, respectively, and €37,355 and €17,658 for flock D1 and D2 as a whole, respectively.

Part III

Mean values of production losses observed in the six reference flocks with EPS (Table 3) are outlined in Table 5. In layer chickens, mean losses phh due to EPS amounted to 10 eggs and 0.2 kg slaughter weight; in meat chickens, these values were 11 and 0.5, respectively. As a consequence of large differences in prices of eggs between categories of birds, total financial losses vary considerably between categories. Total loss due to diminished egg production, loss of slaughter weight and additional destruction costs compensated for reduced feed consumption ranged from €0.28 phh (layer cage) to €9.75 phh (grandparent layers) in the layer-sector and was €1.87 phh in broiler parents and €10.73 in broiler grandparents.

EPS affected layers housed in cages (one flock with 51,000 hens) and layers kept for vaccine production were not treated with antibiotics. In other categories of layers, the percentage of hens in EPS affected flocks that was treated with antibiotics ranged from 11 (layer organic) to 54 (layer barn). Predominantly, colistin sulphate was used (Table 7). In EPS affected meat chickens, the percentage of antibiotic treated birds was 83 in grandparents and 67 in parents. In total, eight antibiotics were used; most frequently sulfachloropyridazine sodium/trimethoprim and doxycycline hyclate were applied. Especially in the meat-sector, several flocks were treated with antibiotics more than once (Table 7). Costs of antibiotic cures per 1000 layer-type hens ranged from €8 (flumequine and oxytetracycline hydrochloride) to €90 (tylosin tartrate) and per 1000 meat-type hens from €15 (flumequine) to €110 (enrofloxacin) (Table 7) (tylosin tartrate was not used in meat-type hens; Table 6). In the layer- and meat-sector costs of antibiotics amounted to €33,000 and €79,000, respectively (Table 8). Including costs of medication with antibiotics, total financial loss of EPS in the layer-sector was estimated at €408,000 and in the meat-sector at €3,264,000, resulting in an estimated loss of €3.7 million in laying chickens in the Netherlands in 2013 (Table 8).

Discussion

EPS is a relatively new condition in laying chickens that seems to have emerged in the mid-1990s. It is characterized by acute mortality that lingers for weeks on end, soaring up to 10% or more in combination with severe septicaemia and fibrinous polyserositis lesions, while salpingitis rarely occurs (Dhillon & Jack, 1996; Zanella et al., 2000; Vandekerchove et al., 2004a; Landman & Cornelissen, 2006; Raviv et al., 2007; Landman et al., 2013, 2014). Due to the mortality and the extensive use of antibiotics to treat the condition, the economic impact of EPS is considered to be high. However, it has never been quantified before.

The incidence of EPS at farm level was significantly higher (P < 0.05) in the meat-sector compared to the layer-sector (35% and 7% affected farms, respectively). This is not unexpected as meat-type chickens are more susceptible to colibacillosis than layer-type chickens (Goren, 1991). In the layer-sector, the incidence of EPS seems to be related to the housing system, cage and barn farms showing clearly less outbreaks of the disease. Although the low incidence of EPS in cage housed layer flocks is in agreement with literature on the occurrence of bacterial diseases including colibacillosis in various housing systems (Kreienbrock et al., 2003; Petermann, 2003; Fossum et al., 2009), the also low incidence of EPS in barn-housed layers found in the present study deviates from mentioned literature. An explanation for the unexpected low incidence of EPS in barn-housed layers was not found.

The labour income on layer farms varies considerably between farms and years. The average income during the last five years for a farm with 50,000 layer hens kept in barns was approximately €30,000 per year (Van Horne & Wisman, 2014). In Part II of the present study, it appeared that losses in flocks with approximately 16,000–52,000 hens ranged from €3653 to €21,766, which represents a very substantial proportion of the labour income of farmers. These losses may even comprise their whole labour income. Besides these losses, farmers have to face a higher workload as extended culling and disposing of dead hens is required, which means that they have to work longer for less.

Despite the lack of a norm for the labour income in the meat-sector, it is reasonable to assume that the labour income here is in the same order of magnitude as in the layer-sector. Therefore, what has been stated earlier about the income losses on layer farms is also likely applicable to meat-type farms.

The calculation of the economic impact of EPS in Part II of the study could be made rather accurately. However, as the estimations of the economic significance of EPS in the various chicken categories were based on the four reference layer and two reference broiler breeder flocks of Part II of the study, the accuracy of these estimations largely depends on how representative are the six reference flocks with EPS previously mentioned. Since the extent to which the reference flocks are representative of all EPS cases in the field is unknown, the calculated losses of EPS given in Part III should be considered as gross estimations.

A number of possible relevant aspects that would likely increase the economic impact of EPS were not considered as they could not be reliably assessed. For example, (1) egg production drops of about 1% to 2% based on hens present as suggested by others (Zanella et al., 2000; Vandekerchove et al., 2004a, b) and explained by the occurrence of the chronic peritonitis form, which does not result in acute mortality, but in emaciation and arrested egg production (Landman et al., 2013); (2) the costs of diagnostic work and (autogenous) vaccines; (3) the costs of alternative treatments with acids, herbal compounds, chlorine, etc.

Antibiotic treatments were applied frequently and repeatedly, especially in meat-type hens, suggesting that they were not sufficiently effective. Indications for the shortcomings of antibiotic treatment to cure EPS were found earlier (Landman et al., 2014).

The significant economic impact of EPS and its negative effects on the welfare of both, the poultry farmer and his birds justifies intensive research with focus on the pathogenesis, prevention and treatment of this disease.

Acknowledgements

We thank colleagues Mr R. Nijland and Mrs I. Jorna for their skilful technical assistance.