Animal Welfare Audits for Cattle, Pigs, and Chickens that use the HACCP Principles of Critical Control Points with Animal Based Outcome Measures

(Updated July 2013)

Temple Grandin
Department of Animal Science
Colorado State University, Fort Collins, Colorado U.S.A.


Restaurant audits of animal welfare at slaughter plants have been very effective, because they are based on relatively few numerically scored critical control points. They are much easier to measure than collecting data on a multitude of items that vary greatly in their importance. The five main critical control points at a beef or pork slaughter plant are:
  1. Percentage of animals stunned correctly on the first attempt.

  2. Percentage that remain insensible.

  3. Percentage that do not vocalize (moo, bellow, or squeal) during movement up the race and during handling and stunning. All vocalizations in the stun box or restrainer are counted.

  4. Percentage that do not fall or slip during handling. Score slips and falls as separate variables.

  5. Percentage moved with no electric prod (goad).

Each critical control point is scored on a yes/no basis for each individual animal. All five of these measures are animal based outcome measures. There are also five acts of abuse that would result in an automatic audit failure:

  1. Dragging sensible non-ambulatory animals.

  2. Poking the animal in sensitive areas such as the eyes, ears, nose, or rectum with an electric prod or other object.

  3. Deliberately driving animals over the top of other animals.

  4. Slamming gates on animals.

  5. Beating animals or breaking tails.

For the complete American Meat Institute guidelines and the critical limits for each control point, click here. To pass a welfare audit the plant must have a passing score on all 5 of the critical control points and have no acts of abuse.


When an auditing system is being designed, choosing the most important critical control points will make it more effective.

A good critical control point measures many problems

For example, there are many different causes of a poor stunning score. It could be caused by poor maintenance of the stunner, untrained employees, wet cartridges, or cattle slipping in the stun box and becoming agitated. A high percentage of animals vocalizing may be caused by a head restrainer that applies excessive pressure, electric prods, missed stuns, or slipping on the stun box floor.

The critical control point principles can also be used on farms. For example, on a dairy cow, condition and lameness are two of the most important critical control points for dairy cow welfare. Many different factors can cause high percentages of either skinny or lame dairy cows. Measuring the correct critical control points makes it possible to monitor welfare with an easy to use audit. The purpose of an audit is to identify a problem such as lameness. It is the job of managers, veterinarians, and animal scientists to correct the problem. Progress on correcting the problem can be monitored by measuring a decrease in the percentage of lame cows.

The scoring systems described in this paper are designed for auditing which serves as a welfare screening test. These measures are simpler than the measures that may be required for scientfic research. Simple measurements make it easier to train inspectors, auditors, and animal welfare officers. The author has trained many auditors and when standards get too complicated, training auditors and getting inter-observer reliability is difficult. A study conducted for the European Union (EU) Welfare Quality Program (2009) showed that inter-observer agreements were good to very good for slipping/falling and vocalization scores in cattle (Welfare Quality Report, No. 11, p. 74). They did not score electric goad (prod) use. This shows that people can be easily trained to score these two variables.

A good critical control point is specific

The vague words: properly, sufficient, and adequate, should be banned because they mean different things to different people. An example of a well worded guideline would be "All of the pigs must be able to lie down at the same time without being on top of each other." Critical control points are worded correctly when there is good agreement between two different auditors.

You manage what you measure

Numerical quantification enables managers to determine if practices are improving or deteriorating. Without measurement, practices have a tendency to slowly deteriorate and nobody realizes it. I call this "Bad becoming normal." Conducting audits with numerical scoring makes it possible to determine if practices or the condition of the animals is improving or worsening. The most effective critical control points can be measured by direct observation by an auditor. Whenever possible, direct observation is superior to auditing paper work or records. Farms and slaughter plants with the best welfare standards conduct their own welfare audits weekly or monthly. A well managed operation can attain a very high standard of performance. It will never be perfect, but it can keep getting better and better.


Questions about Numerical Scoring

Some people who are involved with animal welfare standards are concerned that a numerical scoring system does not require perfection. For example, the American Meat Institute scoring system allows 1% of the cattle or pigs to fall down during handling and stunning. Some critics said this would be poor welfare if this many animals wree allowed to fall. When management improves their operation the percentage of animals falling drops well below 1%. Data collected from plants that were audited by both McDonald's and Wendy's had 0% cattle or pigs falling during their 2003, 2004, and 2005 audits. More than 6000 cattle and 3000 pigs were observed.

Farm Critical Control Points

Conditions on farms are more variable than the conditions in slaughter plants. For slaughter plants, the same critical control points can be used around the world. On farms, either regulations or a customer's requirements may determine the type of housing that can be used. For example, gestation stalls for sows or cages for laying hens may not be permitted. When a farm is audited, it will either have the specified housing or not have it. The type of housing would be the first critical control point. Euthanasia is the second critical control point for all species. All farms must have euthanasia equipment that is approved for their species. Debilitated, emaciated, or non-ambulatory animals that will not recover should be euthanized on the farm. A third critical control point is access to clean water and all feeding devices must be functional and well maintained. Below is a list of the most important critical control points for the welfare of different animals on the farm. The items on these lists should be the most heavily weighted. The critical limits for each control point can be found on other parts of www.grandin.com and on other websites. The purpose of this paper is to introduce the concept of critical control points.

Dairy Cows

  1. Percentage of lactating cows that are lame; use published lameness scoring methods. Scoring of lesions and swellings on the legs should also be tabulated because they are easy to observe indicators of leg problems. The goal is 5% or less of the cows are lame. Lameness is a serious welfare problem because sore feet cause long term pain.

  2. Percentage of lactating cows that are too thin and skinny; use published body condition scoring methods.

  3. Percentage of calves that have not received colostrum. Percentage of calves that are dirty.

  4. Prevention of downer non-ambulatory cows; calculate the percentage of downers.

  5. Dragging of sensible downers is prohibited and animals that will not recover should be euthanized on the farm.

  6. Welfare guidelines for surgical procedures such as tail docking and dehorning.

  7. Ammonia levels in indoor housing. 10 ppm is the goal; 25 ppm is the maximum. 20 ppm for European guidelines.

  8. Space requirements for cows and specific welfare housing specifications.

  9. Percentage of dirty cows use published scoring systems.

  10. Percentage of cows with obvious neglected health problems such as bad eyes, bald spots due to a failure to treat for external parasites, warble grubs, or severe injuries.

  11. Percentage of cows with swelling on their legs.

  12. Emergency power for running milking machines when there is a power failure. On very small farms cows could be milked by hand.

Beef Ranch and Feedlot

  1. Cattle handling - Score with an audit similar to the AMI guidelines:

    Percentage of cattle moved with no electric prod.
    Percentage that do not slip or fall.
    Percentage that do not run into fences.
    Percentage moved at a walk or trot.
    Percentage that do not vocalize (moo or bellow) when they are moved into restraint devices or in direct response to being caught by the head stanchion, body squeeze, or other holding device.

  2. Percentage of cows that are too thin and skinny; use published body condition scoring charts.

  3. Percentage of animals that are lame; use published lameness scoring methods. Lameness is a serious welfare problem because sore feet cause long term pain.

  4. Welfare guidelines for surgical procedures.

  5. Percentage of animals that appear to be in good health with no obvious problems such as bad eyes, injuries, warble grubs, swollen legs, runny noses, or bald spots due to failure to treat for external parasites.

  6. Heat stress in feedyards. This is one of the most important critical control points and conditions are highly variable around the world. It is also one of the more difficult things to measure. Heat stress can be measured by scoring respiration rate and panting. Animals with open mouth breathing (panting) are heat stressed.

  7. Mud in feedlots. Can be monitored by mud scoring; use published scoring systems for mud on cattle.

  8. Calf weaning methods. Weaning calves immediately before transport is not acceptable.

Pork Farm

  1. Percentage of sows that are too thin and skinny; use published body condition scoring methods. If the vertebrae on her back show she is too thin. This will look like bumps down her back. A sow in good body condition will have a smooth back.

  2. Percentage of sows that are lame and scoring of lesions and swellings on the legs. The goal is 5% or less of the sows are lame. Score as lame all sows that do not walk normally.

  3. Percentage of market pigs that are lame. In some herds, failure to select for good feet and legs has resulted in high percentages of lame pigs. Lameness is a serious welfare problem because sore feet cause long term pain.

  4. Percentage of sows that have injuries. Count all injuries that are more severe than scratches that do not break the skin. Examples of injuries that should be counted are: shoulder pressure sores with a scab, tail bites, lacerations, and abrasions that break the skin. It is recommended to score different types of injuries as separate percentage scores, because injuries such as pressure sores, tail bites, and deep scratches from fighting have different causes.

  5. Percentage of market pigs and piglets that have injuries.

  6. Ammonia levels in the buildings. 10ppm is the goal. 25 ppm is the maximum. 20 ppm for European guidelines.

  7. Welfare guidelines on surgical procedures.

  8. Life support backup procedures when the electricity fails in a mechanically ventilated building.

  9. Space requirements. Score sows and pigs separately. Percentage of sows and market pigs that have enough space so that they can lie down in full lateral recumbancy all at the same time, without being on top of each other. Some welfare codes require even more space.

  10. Specific housing specifications for welfare. Example: sow gestation stalls are banned. One program may require straw bedding and another program may not require it.

  11. Percentage of animals engaged in abnormal behaviors such as bar biting or belly nosing.

  12. Transport and handling. Use handling measures similar to cattle. In addition, measure the percentage of DOA (Dead On Arrival) pigs and the percentage of stressed or non-ambulatory pigs. See pig transport guidelines.

  13. Fear test. Measures the animal's willingness to approach people. Low fear pigs are more productive and less stressed. Use published tests.

Broiler Chicken and Turkeys

  1. Handling Measures - Done at the slaughter plant. A separate slaughter audit is on this website.

    Percentage of birds with broken wings (includes dislocated wings). Score with the feathers on to avoid confusion with damage due to feather removal.
    Percentage of broken legs.
    Percentage of bruised birds.
    Percentage of DOAs.
    Percentage of birds that are not able to lie down in the transport crates without being on top of other birds.
    Percentage of birds with green discolored thighs caused by leg tendon breakage. Score after feather removal. This is mainly a problem in heavy male turkeys.

  2. Condition of litter.

    At the slaughter plant, measure the percentage of birds that have foot pad lesions.
    Percentage of dirty birds that have been soiled by dirty litter.
    Percentage of birds with hock burn.

  3. Ammonia levels in the buildings. 10ppm is the goal. 25 ppm is the maximum. 20 ppm for European guidelines.

  4. Welfare guidelines on surgical procedures.

  5. Percentage of market ready birds that are lame. Use published gait scoring methods. Must be measured on the farm by walking through the birds. Birds with good legs will move away from the person. Score as lame the birds that can not walk or only walk 1 to 4 steps before laying down. Birds are too stressed to measure gait at the slaughter plant.

    In the Field Poultry Gait Scoring Method, Dawkins et al (2004) contain an easy to use gait scoring system that can be used in the field. They use a three point scoring system.

    0 = Normal - Walks at least 10 steps with ease and is well balanced

    1 = Walks abnormally for at least 10 steps with an uneven stride and is unbalanced

    2 = Reluctant to walk or not able to walk. Birds that walk only 1 to 4 steps would be scored as 2's.

    Dawkins data showed that 73% of the chickens had normal gait. Accurately counting chickens during gait scoring in a large house can be made easier by penning approximately 100 chickens in two locations with either wire or cardboard panels. Gait can be scored as the chickens walk out through a gap in the pen. Penning along the wall is often easier and chickens can be scored as they walk out between an open end of the panel and the wall. All chickens caught in the pen must be scored. In well managed flocks 95% to 99% will pass the gait score test.

  6. Life support backup procedures when the electricity fails in a mechanically ventilated building.

  7. Percentage of injured hens in the breeder flock.

  8. Other specific welfare specifications for housing.

Egg Laying Hens

  1. Handling measures - Same as broilers.

  2. Space requirements in cages or on litter must be calculated with the number of hens placed in the house and not on house averages due to death losses. Space requirements must also be based on the actual usable full height space in slant back cages. All birds must be able to lie down at the same time without being on top of each other.

  3. Ammonia levels in the building. 10ppm is the goal. 25 ppm is the maximum. 20 ppm for European guidelines.

  4. Life support backup procedures when the electricity fails in a mechanically ventilated building.

  5. Welfare guidelines on surgical procedures. If hens are beak trimmed, the percentage of birds that have been correctly trimmed should be measured.

  6. Specific housing specifications for welfare.

  7. In cage systems, measure the percentage of cages where all the birds have enough space to be able to eat all at the same time. This helps to reduce fighting.

  8. Percentage of injured birds.

  9. Percentage of birds with good feather condition at the end of the laying period. Use published feather scoring systems. Separate feather condition scoring should be done for feather damage due to wear and feather damage and injury due to being pecked by another bird.

  10. Life support backup procedures when the electricity fails in a mechanically vented building.

Types of Standards

There are three basic types of standards for assessing conditions on farms and slaughter plants. They are:
  1. Animal Based Outcome Measures such as lameness and body condition scoring. Numerically scored animal based standards should receive the most emphasis. Scores on these continuous variables can never be 100% perfect. It is impossible to have 0% lame animals. Boththe OIE (World Organization for Animal Health) and the European Union are emphasizing the use of animal based standards. The main numerically scored animal based variables are:

    These numerically based standards should receive MAJOR EMPHASIS.

  2. Yes/No Standards List Prohibited Methods or Housing: Examples are:

    These items should receive MAJOR EMPHASIS.

  3. Input Based or Engineering (design) Standards: They are sometimes called resource based parameters. Some examples are:

    Input based standards should specify a few basic minimum requirements. Detailed specifications on exact methods or designs for building housing or vehicles should be avoided.

    These items should receive LESS EMPHASIS, but there are a few really important ones, which are listed above.

  4. Record Keeping Requirements

    The emphasis for effective animal welfare auditing should be on things that are directly observable instead of paperwork.

    Record keeping should receive LESS EMPHASIS. Record keeping is more important for disease control than for welfare.

Animal Based outcome Measures that can be Numerically Scored at the Slaughter Plant

It is much easier for an auditor or inspector to observe animal conditions at the slaughter plant instead of visiting every farm. Below is a list that shows many of the conditions that can be monitored at slaugther:

References

Grandin, T. (Editor) 2010. Improving Animal Welfare: A Practical Approach. CABI Publishing, Wallingford, Oxfordshire, UK.

Grandin, T. 2010. Auditing animal welfare at slaughter plants. Meat Science. 86:56-65.

Marion Stamp Dawkins, Christi A. Donnelly, and Tracey A. Jones, 2004. Chicken welfare is influenced more by housing conditions than by stocking density. Nature. Vol. 427, pp.342-344.


Click here to return to the Homepage for more information on animal behavior, welfare, and care.

Click here to get more information on T. Grandin (2010) Improving Animal Welfare: A Practical Approach, CABI Publishing..