The Anti-Docking Alliance Web Site Worldwide = Home Page | Archive The Anti-Docking Alliance Web Site in the UK Denmark | Archive


happy tails



February 1996


Report: the Animal Welfare Council regarding an assessment of reports on tail injuries. 


On 30 November 1994, The Animal Welfare Council submitted a letter with reference CAJ-0610 as well as enclosed documents to the undersigned, requesting that ”...comments on the data and their suitability are included in the assessment of departmental order no. 627 of 29/8/1991”. The enclosed documents consisted of 34 copies of completed report forms, a note containing the total number of dogs treated at the registered clinics during the second year of the study, and a copy of a letter from the Animal Welfare Council to the minister of justice dated 27/9/1991 in which the completed study is described:


”The Animal Welfare Council recommends that the issuing of the aforementioned departmental order is supplemented by a study comprising a sufficient number of animal hospitals and animal clinics with the aim of registering existing canine tail injuries over a period of at least two years – including the distribution of injuries across different breeds.


The Animal Welfare Council expects that such study, possibly supplemented by Swedish and Norwegian experiences, will provide a firmer foundation for the continuation or lifting of the departmental order. It is also expected that the level of risk facing each breed will be documented with greater certainty than that which is presently available”.


After some time, it was discovered that, erroneously, only copies of report forms from the second year of the study had been forwarded (1/10/93 – 30/9/94). Thus, in October 1995, I received a further 23 copies of completed report forms from the first year of the study (1/10/92 – 30/9/93).


From my understanding of the Animal Welfare Council’s enquiry, what is required is the following:


1)     comments on the data and their suitability for inclusion in the assessment of departmental order no. 627 of 29/8/1991;

2)     an assessment of injury incidence, including the distribution across different breeds;

3)     an illustration of the level of risk facing each breed.


Re 1): Comments on the data and their suitability for inclusion in the assessment of departmental order no. 627 of 29/8/1991


I assume that the wording ”assessment of departmental order” refers in particular to the assessment of the skewed incidence of tail injuries across breeds, which can be documented as well as expected in pursuance of the departmental order’s regulatory effect on the docking status of certain breeds, cf. items 2 and 3.


Comments on the data and their suitability for these assessments may be divided into the following subitems:


a)     Study design

b)    Data quality 

c)     Data scope (test sample size)


Re a) Study design


Besides the total number of dogs treated at participating clinics/hospitals during the second year of the study, only tail injury incidents are reported (a ”case-only” design), and so the data are primarily useful to describe the incidents, whereas any analysis of comparable risk between for example docked and non-docked breeds is limited. Nor is it possible to estimate tail injury frequency on the basis of the received data.


It is important to note that the ratios (for example in the form of percentages), which in the present data can be calculated as the number of injuries divided by the number of treated dogs, are numerical quantities (”proportional rates”) that cannot be compared with actual injury frequencies. Such are to be calculated on the basis of the number of animals within the actual canine population that may incur the injuries (the ”risk population”, to be included as the denominator). Interpretations of differences in proportional rates are difficult, as the number of treated dogs only constitutes one part of the risk population, and as many factors influence the composition and number of dogs treated at participating clinics/hospitals.


The applied design is therefore not suitable for actual numerical analyses. As a minimum, and in addition to a record of the injuries, an analytical design should have a record of breed and docking status of a comparable test sample of dogs without tail injuries, which is chosen from, for example, other treated dogs in the clinical data or from an existing canine register (”case-control” design). Alternatively, the study could be carried out by following a number of dogs of known and various breeds and docking status for a certain period of time and registering new incidents of tail injuries (”cohort” design). A third possibility is to investigate a large test sample of dogs of different breeds in terms of the presence of tail injuries at the time of the investigation (”cross section” design). In principle, these three types of analytical design would all provide information about incidence and risk with respect to breed and docking status, but they would entail different degrees of suitability and significant differences in terms of practical possibilities.


Notably, when assessing study design, any comparison between breeds and docking status must take into account the fact that the change to the docking regulations was introduced on 1/9/1991, and so non-docked individuals of breeds included in the docking ban belong to certain, limited age groups, and the relevant age groups change with each of the years during which the records take place. Comparisons between breeds and docking status must therefore take into consideration the age of the dogs, which means that a great number of the tail injuries to the older dogs will not be included in this part of the calculations. The analytical design must allow for this, and it will also influence the chances of reaching the required number of examined dogs (cf. c)), as these will be limited to the relevant age groups.


So, several circumstances relating to the applied design means that one cannot draw any real conclusions about the incidence and risks of tail injuries on the basis of the present data.


Re b) Data quality


The total of 57 completed record forms (23 submitted from the first year and 34 submitted from the second year) comprises:


Two forms with a message from the relevant clinic that there were no diagnosed tail injuries within the specified time period. Both originated from the same veterinary hospital.


A printout from one veterinary hospital with a very inadequate description of one incident.


54 more or less completed forms of which two however turned out to be repeats of one veterinary hospital’s first year submissions. (One of the copies even had the year 1994 added after the date, though the copy of that same form was submitted in 1993 without stating the year!).


Of the remaining 52 forms, a good deal were incomplete as regards individual pieces of information, for example date (four forms), the age of the dog (totally absent on one form, and without any indication of days/weeks/months/years on several others, entailing the risk of errors), and the sex of the dog (91 forms). Of the 48 forms that were dated, four were dated prior to the start of the first year (1/10/92) and two were dated after the end of the second year (30/9/94).


This sub-optimal data quality presumably reflects varying degrees of knowledge about and co-responsibility for the study, its aim, character and premises, as well as the fact that the recording and reporting at the hospitals/clinics often involves several people.


Hence, the present data are generally of sub-optimal or varying quality, which makes actual analyses and conclusions open to doubt.


Re c) Data scope (test sample size)


From the forwarded note, it emerges that the ten participating clinics/hospitals treated approximately 70,000 dogs within a year. That same year records 26 incidents of tail injuries, which constitutes a minimal proportion of the treated dogs (on average approximately four incidents per 10,000 treated dogs) (see table 1). The incidence of injuries is distributed across 20 stated breeds (se table 2). Only a very comprehensive study can provide a statistically sound assessment of differences between breeds for such relatively rare injury. A comparison of two breeds, of which one is expected to be twice as much at risk of tail injuries as the other, and where the incidence of the latter is expected to be like the average (four per 10,000 treated dogs), should comprise 64,000 treated dogs from each of the two breeds. In order for the difference to be statistically sound, this is equivalent to an expectation of approximately 50 incidents of tail injuries in the first breed and approximately 25 in the other. (Please note that also breed, docking status and age of all dogs included in the study should be registered and reported, and not just for the tail-injured dogs, cf. a) above about design). In comparison, the breed with the greatest number of reported tail injuries during the two years of available data is the German Shepard with seven incidents (table 2). The number of German Shepherds among all the treated dogs is unknown.


Hence, the limited data sample size of the present data also means that one cannot expect statistically sound conclusions.


Re 2) An assessment of injury incidence, including the distribution across different breeds


Table 1 shows the reported numbers distributed across the ten veterinary hospitals/clinics and time period.




                             PERIOD                                  PERIOD 2


HOSPITAL 0        1        2        3        Total  Total treat.   Prop. morb.*


ad          1          2          2          1          6          3000            7

          bd          0          1          2          0          3          1800            11

          dh          0          0          0          0          0          4500            0

          dk          1          3          1          0          5          6500            2

          gd          2          2          2          0          6          20000           1

          hd          0          2          5          0          7          8423            6

          ho          0          5          6          0          11          9000            7

          kd          0          3          3          0          6          6500            5

          rd          0          1          4          1          6          8000            5

          sd          0          2          1          0          3          3200            3


Total            4          21          26          2          53          70923           4


* Proportional morbidity rate: Number of tail injuries per 10,000 treated dogs during PERIOD 2.


Table 2 shows the distribution of injuries across breeds, comparing those breeds that have not been docked either prior to or after the introduction of the departmental order (DOCK TYPE = 0) with those breeds that are illegal to dock after 1/9/91 (DOCK TYPE = 1).




                                            DOCK TYPE

BREED                                    0          1          Total


Crossbreed                               15          0          15

Boxer                                       0          1          1

Cocker Spaniel                            0          3          3

Dalmatian                                  4          0          4

Doberman                                 0          1          1

Field Trial Spaniel                            0          1          1

Flat-Coated Retriever                2          0          2

Golden Retriever                          2          0          2

Grand Danois                             4          0          4

Dachshund                               2          0          2

Irish Setter                                1          0          1

Short-Haired Dachshund            1          0          1

Labrador Retriever                          3          0          3

Long-Haired Pointer                   1          0          1

Wire-Haired Dachshund            1          0          1

Wire-Haired Pointer                   0          1          1

Samoyed Spitz                      1          0          1

German Shepard                           7          0          7

Springer Spaniel                            0          1          1

West Highland White Terrier          1          0          1


Total                    45          8          53


However, the two incidents reported for Boxer and Doberman are injuries to docked tails, and so these two dogs are not included in the following assessments.


Moreover, a Wire-Haired Pointer has been placed in DOCK TYPE = 1, as this dog is reported as being non-docked, despite the fact that, according to the departmental order, this breed is exempt from the docking ban.


Table 3 shows the 51 injuries to the non-docked dogs distributed across the four time periods. It can be seen that, as expected, the proportion of dogs that are no longer docked (DOCK TYPE 1) increases after the departmental order comes into force.




                                                     DOCK TYPE

PERIOD                                   0                  1          Total


          0                                     4                  0          4       

                                       100.0%          0.0%


1                                     18                2          20

                                         90.0%            10.0%


2                                     22                3          25

                                        88.0%            12.0%


3                                     1                  1          2

                                        50.0%            50.0%       


          Total                               45                6          51

                                         88.2%            11.8%


Table 4 shows the age distribution of these injuries between the two types of canine breeds. As expected, DOCK TYPE 1 tail injuries are only seen in young dogs (0-2 years) born after the departmental order came into force.




                                                DOCK TYPE

AGE (years)                              0                  1          Total


          < 1                                  5                  4          9

                                                 55.6%  44.4%


1-2                                  3                  1          4

                                                75.0%  25.0%


2-3                                  6                  1          7

                                                85.7%  14.3%


3-4                                  3                  0          3

                                                100.0%  0.0%


4-5                                  11                0          11

                                                100.0%  0.0%


5-6                                  4                  0          4

                                                100.0%  0.0%


6-7                                  2                  0          2

                                                100.0%  0.0%


7-8                                  1                  0          1

                                                100.0%  0.0%


8-9                                  3                  0          3

                                                100.0%  0.0%


9-10                                3                  0          3

                                                100.0%  0.0%


11-12                              2                  0          2

                                                100.0%  0.0%


13-14                              1                  0          1

                                                100.0%  0.0%


UNKNOWN                  1                  0          1

                                                100.0%  0.0%


                    Total                      45                6          51

88.2%  11.8%


Table 5.1-5.2 show the same as table 4, but for each of the actual two record periods (the two periods with unsystematic records (PERIOD 0 and 3) are omitted here) and only for those age groups that are relevant to the comparison, that is, < 1 year for PERIOD 1 and 1-2 years for PERIOD 2.


Table 5.3 shows the sum of the two individual time periods. It can be seen that, when taking into consideration the effects of age and time, half of the tail injuries to non-docked dogs occurs in those breeds that have been included in the docking ban. In other words, a rough estimate is that the number of tail injuries treated at the participating clinics/hospitals will double from the time prior to until the time after the departmental order coming into force, where the entire age spectrum for tail injuries in all non-docked dogs is represented (i.e. approximately year 2004). Due to the previously mentioned circumstances, these estimates are however encumbered with considerable uncertainty, and the present basis does not allow for further specification of the size of this uncertainty. Moreover, one cannot be certain that the underlying assumption will actually hold up, that is, that the age distribution of injuries in the two DOCK TYPES will be the same.




PERIOD = 1: All dogs < 1 year at the time of treatment between 1/10/92 – 30/9/93 are born after 1/10/91.


                                       DOCK TYPE

AGE (years)                     0                  1          Total


< 1                                  2                  2          4

                               50.0%            50.0%




PERIOD = 2: All dogs younger than two years at the time of treatment between 1/10/93 – 30/9/94 are born after 1/10/91.


                                       DOCK TYPE

AGE (years)                     0                  1          Total


< 1                                  2                  2          4

                               50.0%            50.0%


1-2                                  1                  1          2

                               50.0%            50.0%


PERIOD = 2: All dogs younger than two years at the time of treatment between 1/10/93 – 30/9/94 are born after 1/10/91.




TOTAL PERIOD 1-2: The sum of numbers from table 5.1 and 5.2


                                       DOCK TYPE

AGE (years)                     0                  1          Total


< 1                                  4                  4          8

                               50.0%            50.0%


1-2                                  1                  1          2

                               50.0%            50.0%


TOTAL                           5                  5          10                                                                 50.0%            50.0%


The breed distribution of the six injuries among previously docked breeds (including the previously mentioned Wire-Haired Pointer) is shown in table 6. Though involving very few data, it should be noted that they are all breeds used for hunting, whereas breeds as Boxer, Doberman etc. do not (yet) figure in the statistic in terms of injuries among non-docked breeds. Moreover, one individual clinic is apparently over-represented with four out of six incidents.




                             VETERINARY HOSPITAL

          BREED                 ad           hd          ho                Total


Cocker Spaniel                  1          2          0                  3

Field Trial Spaniel           0          1          0                  1

Wire-Haired Pointer          0          1          0                  1

Springer Spaniel                  0          0          1                  1


Total              1          4          1                  6


Only in the case of three out of the five dogs - where ”use” was completed - does it state that it concerns a hunting dog. The remaining two are characterized as ”family dogs” (both Cocker Spaniels) (table 7).




TYPE                    Freq             Percent       Cum


Hunting          3                  60.0%          60.0%

Family          2                  40.0%          100.0%


Total            5                  100.0%


Only one of the six injuries had occurred during hunting, while four had occurred in the home and one while being exercised (table 8).




OCCURRED                 Freq             Percent       Cum


Hunting                            1                  16.7%          16.7%

Family                             1                  16.7%          33.3%         

Home                              4                  67.5%          100.0%


Total                               6                  100.0%


The position of the tail injury is shown in table 9.




POSITION                      Freq             Percent       Cum


Tip                                  2                  40.0%          40.0%

Caudal 1/3                       3                  60.0%          100.0%     


Total                               5                  100.0%


Table 10 shoes the nature of the injury. It is worth noticing that two incidents are caused by a parasitic infection among newborn Cocker Spaniel puppies.




NATURE                        Freq             Percent       Cum


Deep wound                             1                  66.7%          66.7%

Superficial                       1                  16.7%          83.3%         

Other                               4                  16.7%          100.0%


Total                               6                  100.0%


The age of the injuries is shown in table 11.




AGE OF INJURY                Freq             Percent       Cum


Unknown                         1                  20.0%          20.0%

7 days                             1                  20.0%          40.0%         

8 days                             1                  20.0%          60.0%

14 days                            1                  20.0%          80.0%

Over 100 days             1                  20.0%          100.0%


Total                               5                  100.0%


The four oldest dogs had all had their tails docked, while one of the puppies was put down (table 12).




TREATMENT               Freq             Percent       Cum


Amputation                      4                  66.7%          66.7%

Put down                         1                  16.7%          83.3%         

Unknown                         1                  16.7%          100.0%


Total                               6                  100.0%


Re 3) An illustration of the level of risk facing each breed


As seen in 1), the nature, quality and scope of the data are insufficient to secure statistically sound conclusions.


The following, potentially interesting observations must therefore be viewed with some reservation and merely as possible hypotheses that should undergo further investigation at a later stage.


The frequency of tail injuries is limited when they, as is the case here, are calculated on the basis of records through veterinary treatments.


On the basis of the very slight data that relate to incidents for animals born after certain breeds were included in the 1991 docking ban (ten incidents), there seems to be an overall incidence among these breeds, which is equivalent to the number of incidents among the traditionally non-docked breeds (five incidents within each group). Of the former breeds, only the hunting dog types are represented among the tail injuries. Whether these observations are significantly different than expected on the basis of the breeds’ relative prevalence cannot be determined in this study. Hence, on the basis of the existing data, one cannot throw any further light on the level of risk facing the individual breeds.



Preben Willeberg,


Veterinary Forensic Medicine and Epidemiology

The Royal Veterinary and Agricultural University

The Campaign Against the Docking of Dogs' Tails

Please keep supporting anti docking: this website has been rescued for posterity by HistoryX a dog lover and proud Cocker spaniel owner 2024.