North Carolina State University
Animal Science Departmental Report
2004-2005

Return to main menu
Return to Swine articles


The Humane Euthanasia—On-Farm Applications

 

W. E. M. Morrow, S. Miller, R. E. Meyer1, J. Roberts2 and D. Lascelles3

 

1Department of Clinical Sciences, College of Veterinary Medicine, Mississippi State University

2Department of Population Health and Pathobiology, College of Veterinary Medicine, North Carolina State University

3Department of Clinical Sciences, College of Veterinary Medicine, North Carolina State University

 

Introduction                                                                   

In general, an animal should be culled when it is no longer profitable or euthanatized when it is inhumane to allow it to continue living.  The difficulty all farm managers encounter is defining when animals become uneconomic and whether to treat or euthanatize the compromised animal.  Individual managers usually resort to a very subjective assessment often heavily weighted by the perceived ability of the animal to return a profit. Focus groups of North Carolina farm managers have told us that having clear criteria for when to euthanatize an animal would help reduce some of the job stress they feel. Unlike the companion animal arena, where there is much discussion and many suggested guidelines on the appropriateness and timing for euthanasia, there are relatively few guidelines for when a farm animal should be euthanatized. Many of the companion animal guidelines are very subjective (e.g., ability to enjoy food, ability to breathe freely and without difficulty, ability to eat and drink without pain, ability to respond to owner and family), but when taken together are helpful in creating a euthanasia profile. Other guidelines are more objective; Duncan (1988) recommends that companion animals should be euthanatized if they have:

 

·        Weight loss: 20-25% of total body weight, characterized by muscle wasting

·        Extreme weakness/Inability: no desire to eat or drink, persisting for 24 hours or more.

·        Moribund state: depression and body temperature below 99oF

·        Infection: involving one or more organ systems, which fails to respond to treatment within an appropriate amount of time

·        Respiratory/cardiovascular: failure of these systems, including blood loss or anemia resulting in a hematocrit below 20%

·        Nervous/musculoskeletal: injuries that cannot be healed, resulting in uncontrolled seizures or the loss of a limb

 

Similar guidelines could be adapted for farm animals. For example, the following general guidelines could apply to pigs of any weight or age:

 

·                    Weight loss of 20-25% of total body weight, characterized by muscle wasting.

·                    Extreme weakness or inability with a lack of desire to eat or drink persisting for 24 hours or more.

·                    Suffering from any infection/disease which fails to respond to treatment.

 

In addition, some farming systems have adopted specific protocols to help managers decide what to euthanatize and what to keep. For example, the “two-strike” system has two criteria that must be fulfilled before a weaner pig is euthanatized: underweight (e.g., less that 8 lb on a farm with 18 day weaning) and have a disability such as a rupture, or navel ill, or lameness, or poor body condition. This introduces a special category of concern for pork producers, the lightweight pig. It has been long accepted that lightweight piglets at birth are lightweight at weaning (England 1974). Others have established that lightweight pigs at weaning remain small and are a significant contributor to the variation in slaughter weight and, as such, a major problem in assembling slaughter loads. In 3-site production, where the system rewards nursery managers for dispatching more pigs, there tends to be more pigs shipped than there should be. Conversely, finishing managers struggle with the issue of how to handle the underweight/disadvantaged pigs they are shipped.

Industry-specific guidelines for euthanasia, such as the National Pork Producers Council guide “On Farm Euthanasia of Swine” (NPPC 04259-4/97) and university-produced extension training materials such as “On-Farm Euthanasia: Better Ways” (Morrow and Meyer, 2001), generally agree closely with AVMA-accepted methods and processes. Unfortunately, these guidelines do not help producers decide if and when any individual animal should be euthanatized to end its suffering. Suffering can be conceptualized as the product of pain and its duration. By daily monitoring, farm managers can usually estimate duration but the difficulty of estimating pain remains. Overt pain behaviors in pigs can be difficult to evaluate. Pain may reduce normal pig social behaviors and vocalization, while vocalization in response to handling may be more pronounced. Changes in gait and reluctance to move may also be observed (Dombromylskyj et al in Flecknell and Waterman-Pearson 2000). Managers can usually identify those animals suffering the most because they exhibit aberrant behavior or the presence of visible lesions (e.g., burns, lacerations, compound fractures) make it obvious. However, the issue is clouded because a condition may be visually striking but less painful (e.g., prolapses) or inconspicuous but more painful (e.g., arthritis). Various methods of generating a pain score and assessing animal pain have been reviewed by Dombromylskyj and coworkers (in Flecknell and Waterman-Pearson, 2000). Objective measures, such as heart rate, respiratory rate, and temperature are unreliable guides to the presence of pain, as are humoral factors such as epinephrine, norepinephrine, and cortisol. These measures may be useful when integrated into a pain scoring system, but are of limited use when used alone as they are influenced by so many factors other than pain. 

Our study developed a protocol for handling compromised pigs.  We collected economic and welfare information that producers can use to make informed decisions on when pigs should be euthanatized.

 

Materials and Methods

The Study

The study was conducted in 2002-2003 in 5 commercial nursery sites (Farms 1-5) owned by 4 separate entities (A, B, C, and D.  Two farms were owned by the same entity) in North Carolina.  Each farm was part of a 3-site production system and regularly received nursery-aged pigs from their supplying owner and after growing for about 6 weeks the pigs were shipped to their respective finishing sites.  

Batches of pigs were assigned to treatment (heavy, medium or light euthanasia) before the pigs arrived on site but at the convenience of the owners (Table 1). Each owner was requested to complete at least 12 batches (4 heavy, 4 medium and 4 light) but not all were able to achieve this.  Pigs within batches were screened upon entering the nursery and either euthanised, tagged or penned normally depending on their condition and the assigned treatment.  The heavy protocol triggered euthanasia for conditions that were less severe (compromising) than the medium protocol, and the light protocol triggered euthanasia only for the most compromised pigs (Table 2).  Pigs with conditions not sever enough to trigger euthanasia were ear-tagged and monitored daily by farm staff for changes in their condition.

 

Farms

Farms varied in where they housed tagged pigs and the status of porcine reproductive and respiratory disease (PRRS) in the herds supplying the nursery pigs:

Farm A/1: Tagged pigs were left in the pens they were found in. The pig flow was a known PRRS positive flow with clinical problems. Two batches were processed and mean batch size was 3585 pigs. 200 pigs were euthanized and 80 tagged.

Farm B/2: Tagged pigs were grouped into set pens if small or lightweight, but were left in the pens they were found if large or normal weight. The pig flow was known to be PRRS positive but not a clinical problem at this time. 12 batches were processed and mean batch size was 587 pigs.

Farm B/3: Tagged pigs were grouped into set pens if small or lightweight, but were left in their original pens if of normal weight or heavier. This practice was used to prevent heavier lame pigs from beating up on small weak tagged pigs. The pig flow was known to be PRRS positive but not a clinical problem at this time.   Twelve batches were processed and mean batch size was 801 pigs.

Farm C/4: Tagged pigs were grouped into set pens. The pig flow was a healthy PRRS-free flow. Six batches were processed and mean batch size was 2207 pigs.

Farm D/5: Tagged pigs were grouped into set pens. The site was recently depopulated and the pig flow was a healthy PRRS-free flow. 15 batches were processed and mean batch size was 930 pigs.

Any pig requiring treatment was treated according to standard operating procedures on the study farms. The time taken to administer treatment was recorded as was the amount, type, and cost of drugs administered.  If more than one pig was treated at the same time (e.g., injecting 5 pigs with an antibiotic) the time taken to treat the group was averaged and the mean cost in time assigned to individual pigs. Tagged pigs were euthanized if their condition progressed to a level that triggered euthanasia for their treatment group.

The conditions monitored and the levels triggering euthanasia were decided in advance by consensus with the investigators and veterinarians responsible for the health of the pigs studied (Table 2).  If pigs died they were weighed and date noted.

 

Value of pigs

The value of each welfare compromised (tagged) pig in each batch was established at shipping by partial budget.   Additional income was calculated as the product of the animal’s weight at shipping by the standard value of $0.80 per lb. The cost of drugs administered was set at a standard value per ml of $0.563 ceftiofur sodium, $0.04 tylosin, $0.03 penicillin G, $0.05 long acting oxytetracycline.  Cost of time was set at $10/hour. The value of the batch was the sum of the value of all compromised pigs in the batch.

 

Assigning a welfare score

At the time they were received at the nursery and each day thereafter, farm staff observed the pigs under their care and determined if they needed treatment, if euthanasia was warranted and their condition; specifically if they changed levels in the physical condition they were experiencing.  Conditions and levels were recorded for each pig.  To determine the status of welfare associated with each condition (e.g., lame, tail bitten, ruptures) and its level of severity (i.e, A, B, C, or D) an expert panel was asked to create a welfare score by ranking each level of each specific condition from 0 to 10 (0 no effect on welfare, 1 little effect on welfare, to 10 worst possible effect on welfare) (Table 3).

When a pig had 2 or more physical conditions then the welfare scores for each condition were summed to make a final welfare score. For example, if a pig had damaged digits at level B (score 3) and a rupture at level A (score 1) then that pigs total welfare score was 4 and the level of severity was reassigned to level A to reflect its compromised welfare state.  The welfare score assigned to pigs was the product of the assigned initial score for their condition and level and the number of days they experienced that condition, that is, until they recovered, died, were euthanized or left the nursery to go to the finishing stage (e.g., a weak pig, level A, for 30 days would score 90 (3 x 30). Thus a low welfare score indicated fewer pigs were compromised for less time and a high score indicated more pigs were compromised for longer.

 

Data analysis

The model was of hierarchical design. For each dependent variable (value and welfare) we performed an analysis of co-variance (ANCOVA) in the GLM procedure of SAS (Version 8.01 for Windows, 98; Statistical Analysis Systems Institute Inc, Cary North Carolina). The batch (one turn of a nursery room) was the experimental unit.  The model included farm, protocol, barn, reason and the level of reason as nominal (class) data.  Barn was nested in farm and level nested in reason. The dependent variables were the value ($) and welfare.

For sensitivity analysis on the economic value, welfare, cost of treatment etc., a decision tree was constructed using the results from the mixed model ANOVA in Data 4.0, published by TreeAge Software http://www.treeage.com

The decision tree diagram represented in chronological order the alternative states for the pigs for their duration of stay in the nursery.  When pigs first entered the nursery they would be allocated to the treatment protocol (light, medium or heavy).  Then pigs would be examined and determined whether they were in a well or compromised state.  If compromised, the reason (state) was determined by clinical examination and its level (state) of severity.  Depending on the protocol, reason, and level then the manager would determine if the pig was either euthanized immediately or tagged and observed daily.

To construct the tree, the value (utility) of each of the potential outcomes (e.g., no value if euthanized) and the probability of that outcome occurring was calculated from the farm data.  From the constructed tree the expected utility of each decision was calculated by weighting the value of each outcome with the probability that the outcome will occur.  Then the weighted values for all outcomes for any branch of the tree was summed (folding or rolling back) giving an estimate for the value of that decision. The branch with the highest expected value ($) or lowest expected value (welfare) is expected to be the most favorable outcome for the population studied. The sensitivity of the results to changing inputs was calculated by varying the cost of treatment, probability of euthanasia, probability of tagged pigs dying/surviving and final value of the pig. 

The distribution for the initial reason was set as the same for all protocol (disregarding the chance that some reasons were disproportionally allocated among the protocol).  The data from 1904 pigs were used in the decision tree analysis; 1931 were available but the 1 beaten (multiple fight wounds) pig and the 24 with reason “unknown” were deleted from the data set. 

 

Results and Discussion:

A total of 51,041 nursery age pigs in 47 batches were screened upon entering the 5 farms and of those, 819 (1.6%) were immediately euthanized and 1118 (2.19%) were tagged and followed. Of the tagged pigs, 186 (16.64%) died, 924 (82.65%) were shipped to finishing and 8 were lost to follow up (Table 1).

For the 7 categories (damaged digits, lame, light weight, repaired rupture, rupture, weak, or two or more) where more than 10 pigs were tagged the percentage mortality ranged from 5.16% (repaired rupture) to 64.3% (weak). For the conditions where more than 10 pigs died the percentage of deaths numerically increased as the level increased; lame/A 27.27% and lame/B 53.57%, light weight/B 9.54% and light weight/C 17.73% (Table 4). There was considerable variation in the reasons for which pigs were compromised, (e.g., 1100 lightweight compared with 41 damaged digits) (Table 4).

 

References

Duncan, J. C. (1988). Careers in Veterinary Medicine, Rosen Pub. Group (New York, NY).

England D.C., (1974) Husbandry components in prenatal and perinatal development in swine. J. An Sci 38:1045.

On Farm Euthanasia of Swine - Options for the Producer. Booklet #04259-4/97. American Association of Swine Practitioners and the National Pork Producers Council, Des Moines, IA, 1997.

On-Farm Euthanasia: Better Ways. Produced by W.E. Morgan Morrow and Robert E. Meyer, North Carolina State University College of Veterinary Medicine Biomedical Communications, 2001.

Dobromylskyj P, Flecknell PA, Lascelles BD, Livingston A, Taylor P, Waterman-Pearson A. Pain Assessment. In: Pain Management in Animals, Flecknell PA and Waterman-Pearson A, eds. W.B. Saunders, London, 2000.

Holton LL, Scott EM. Nolan, AM, Reid J, Welsh E, Flaherty D. Comparison of three methods used for assessment of pain in dogs. J Am Vet Med Assoc 1998; 212:61-66.

Christley RM, Reid S WJ, No significant difference: use of statistical methods for testing equivalence in clinical veterinary literature J Am Vet Med Assoc 2003: 222:433-437.



Table 1.  Summary information for the Euthanasia trial

Farm (Owner/Site)

First batch

Last Batch

Number of batches

Number of hogs screened

Number of pigs euthanized (%)

Number of pigs tagged (%)

Number of tagged pigs that died. (%)

Number of tagged pigs shipped. (%)

A/1

July 2002

September 2002

2

7171

200 (2.79)

80 (1.11)

25 (31.25)

49

B/2

July 2002

October 2002

12

7052

17 (0.24)

125 (1.77)

31 (24.8)

92

B/3

July 2002

September 2002

12

9618

12 (0.12)

343 (0.12)

18 (5.25)

325

C/4

August 2002

September 2002

6

13239

432 (3.26)

390 (2.94)

 

18 (4.61)

372

D/5

November 2002

February 2003

15

13961

158 (1.13)

180 (1.29)

94  (52.22)

86

Total

 

 

47

51041

819

1118

186

924

 


Table 2. Criteria for euthanasia by treatment protocol.

Level

 

 

CONDITION

 

 

Euthanasia Action (by protocol)

 

Lightweight:

 

 

 

A

Less that 40% under normal barn average weight (don’t tag)

No

No

No

B

40-49% under normal barn average weight

No

No

Yes

C

50-59% under normal barn average weight

No

Yes

Yes

D

More than 60% under normal barn average weight

Yes

Yes

Yes

 

 

 

 

 

 


Table 3. Welfare score by condition and level.

Level

CONDITION

Welfare score

 

 

 

 

Lightweight:

 

A

Less that 40% under normal barn average weight

0

B

40-49% under normal barn average weight

1

C

50-59% under normal barn average weight

2

D

More than 60% under normal barn average weight

3

 

 

 

 


Table 4. Final disposition of lightweight pigs by condition and level

 

Euthanized on entry

Tagged Pigs

Level

Number euthanized

Number

Tagged

Number died (as % of tagged)

Number shipped

A

5

14

 

14

B

131

325

31 (9.54)

294

C

317

141

25 (17.73)

116

D

167

0