A survey of adult victims of dog bites in New Zealand

Abstract

AIM: To describe the circumstances of dog bites to adults in New Zealand, in order to better understand factors associated with these bites.

METHODS: A questionnaire was sent to 1,800 adults aged ≥16 years who had made claims to the Accident Compensation Corporation (ACC) in 2002 as a result of dog bites.

RESULTS: Five hundred and thirty-five questionnaires were returned; 50% of respondents were male, and 30% of bites were reported to local authorities. Most injuries were to the legs or hand. More people were bitten by male dogs than female dogs. A disproportionate number of bites took place in rural areas, but the most common locations for attacks were streets/walkways, and the victim's home. Protection of territory, accidental bites, fear, and pain were considered to be the most common reasons for dogs to bite.

CONCLUSION: Many victims were bitten in situations that could have been avoided.

KEY WORDS:

• Dog bite
• survey
• Accident Compensation Corporation

ACC	=	Accident Compensation Corporation

Introduction

Dog bites are frequently reported by news media in New Zealand, yet there is little information on the causes of these bites. An increase in the incidence of hospitalisation for dog bites was reported between 1979 to 1988 (Langley 1992), while the number of bites reported to regional councils decreased from 1996/1997 to 2001/2002 (Anonymous 2003). The incidence of dog bites was summarised based on statistics on death and hospitalisation for New Zealand from 1989 to 2001 (Marsh et al. 2004), and fatalities due to dog bites have been analysed (Healey 2007). While these studies provide information on the incidence of dog bites, the nature of injuries sustained and details of the bite victims, none shed much light on circumstances surrounding, or factors leading to, dog bites. Previous research in New Zealand has not explored the psychological impacts on recipients of dog bites.

The Department of Internal Affairs reported that 58 councils dealt with 3,020 attacks by dogs on humans in one year in 2001/2002 (Anonymous 2003). If these data are extrapolated to cover all 72 councils in the country, then there were about 3,400 reported bites for a 12-month period in 2001/2002. This number, however, is below the 8,677 claims made to ACC following dog bites in the same time period. In addition, the number of ACC claims due to dog bites has increased markedly from 6,300 claims in 1999/2000 (Anonymous 2003). The public would benefit from a reduction in the incidence of dog bites, and any legislative approach to dog control requires an understanding of the factors associated with aggression in dogs. It remains difficult to understand the circumstances surrounding dog bites in New Zealand. The experience of dog bites was previously analysed for a narrow sub-population, veterinary students at Massey University, Palmerston North, New Zealand (Wake et al. 2006).

The aim of this study was to describe the circumstances of dog bites to adults, in order to better understand factors associated with these bites. We surveyed adult ACC claimants who had suffered dog bites. The data gathered about the dogs that had bitten, the background to the attack, and the victims are presented here.

Materials and methods

Subjects, and sampling methods

Two thousand people were selected by the ACC from those who had made claims to the ACC in 2002 as a result of dog bites. The sample was limited to people who were 16 years of age and older at the time of the claim, and had not received serious injuries. Participants, however, had to have injuries that required medical attention. To preserve the claimant's anonymity, the ACC sent out questionnaires on behalf of the researchers, to 2,000 claimants. Two hundred surveys were returned as undeliverable. Thus, the survey was based on 1,800 delivered surveys. This study was approved by the Massey University Human Ethics Committee, Palmerston North, New Zealand.

The questionnaire consisted of 32 questions about the physical and psychological impacts of the bite, the dog, the circumstances around the incident, the participant's history and circumstances of dog ownership, and whether the bite was officially reported. A copy of the questionnaire is shown in Supplementary Table 1 ¹ .

Statistical analysis

Data were coded and analysed using SPSS v13 for Windows and v17 for Macintosh OS X (SPSS Inc, Chicago IL, USA). Variables were cross-tabulated, and further analysed using the χ² contingency test for nominal data, and the Kolmogorov-Smirnov twosample test for ordinal data. It is important to note that not every respondent answered every question, so the sample size was not the same for every analysis.

Results

Respondents

Five hundred and thirty-five (30%) questionnaires were returned, and 268 (50%) respondents were male. A copy of the response raw data is shown in Supplementary Table 2 ¹ . Of the respondents who specified their age, 68 (13%) were between the 16 and 29 years old at the time of the bite, 83 (16%) 30–39 years, 121 (23%) were in their 40s, 129 (24%) were 50–59 years, and 130 (24%) were older than 59 years. Two hundred and sixty-four (50%) respondents reported they were from a city. The remainder were from large towns (n=68; 13%), small towns or villages (n=118; 22%), or rural properties (n=80; 15%). One hundred and one (19%) respondents had never lived with or owned a dog. Respondents were more likely to have lived with or owned a dog if they were from a rural property or small town than if they were from a large town or city (χ²=4.59; df=1; p<0.05). Seventy-one (90%) people from rural properties had lived with a dog, and 41 (52%) had spent more than 30 years living with dogs. By contrast, 204 (78%) people from cities had lived with a dog, and 56 (22%) had lived with a dog for more than 30 years. Most respondents (n=346; 66%) had been bitten on one occasion, and 111 (21%) had been bitten on two separate occasions; the remaining 69 (13%) had been bitten on three or more occasions.

Nature of injuries

Moderate injury (requiring medical attention) was reported by 452 (84%) participants. Sixty-four (12%) bites were minor (little/no blood drawn). Many who rated their injuries as minor stated that they went to the doctor for a tetanus injection rather than to have their injuries treated. Thirteen (2.4%) bites were rated severe (requiring hospitalisation), and three were very severe (requiring surgery and lengthy rehabilitation). The legs (n=237; 45%), hands (n=186; 35%) and arms (n=110; 22%) were the most common parts of the body to be injured. Twenty-eight (5.3%) respondents were bitten on the buttocks, 23 (4.3%) on the face, and 21 (3.9%) on the torso. Sixty (11%) respondents were bitten on more than one part of the body. There were no significant associations between the part of the body injured and the severity of the injuries.

Three hundred and seventy-nine (72%) respondents reported psychological effects from their bites. Two hundred and forty-five (46%) reported minor effects (slightly shaken after the attack), 74 (14%) reported moderate effects (lasting up to one month), while 57 (11%) said their attack resulted in severe long-term psychological effects. Three people (0.6%) reported very severe psychological effects that required counselling. Women were more likely than men to report severe psychological effects. Eighty-six (33%) women rated the psychological effects of their bites as moderate to very severe, compared with only 47 (18%) men (Kolmogorov-Smirnov two-sample test; K=1.72; p<0.02). Bites to the legs or feet were rated as more severe psychologically than bites to other areas (Kolmogorov-Smirnov two-sample test; K=2.02; p<0.002).

The dogs

The relationships between the respondents (victims) and owners of the biting dogs are shown in Table 1. Dog owners classed as ‘other’ were reported by people who had gone to someone's house during the course of their work and were bitten by their client's dog. Five people were bitten by police dogs.

One hundred and fifty-six (29%) respondents were unsure of the sex of the dog that bit them; 248 (47%) said they were bitten by male dogs and 126 (24%) by female dogs. One hundred and seventeen (22%) of the dogs were de-sexed, 146 (28%) were not, and 266 (50%) respondents said they were unsure. People bitten on the hand were less likely to be unsure of the dog's sex (χ²=27.16; df=2; p<0.001) than people bitten on other parts of the body, and the opposite was true for those bitten on the legs/feet (Table 2; χ²=23.32; df=2; p<0.001).

Table 1. Relationship of the dog's owner to the bite victim (n=529), in a survey of adult victims of dog bites in New Zealand.

Dog's owner	Percentage
Stranger	36.2
Respondent (victim)	18.9
Friend	14.0
Other^a	11.0
Neighbour	10.0
Family member not living with the victim	5.1
Family member living with the victim	4.2
Flatmate	0.6
^a Other is a client of the victim

Table 2. Percentage of people bitten on the arm, hand, leg or multiple parts of the body, and the sex of the biting dog, in a survey of adult victims of dog bites in New Zealand^a.

Sex of dog	Site of dog bite
Arm	Hand	Leg	Other^b	Sub-total	Multiple bites^c	Total %
Male	50	52	42	55	48	41	47
Female	18	34	18	25	24	24	24
Unsure	32	14	40	21	28	36	29
Total n	74	154	196	44	468	59	527
^a Rounding of percentages results in some column totals ≠ 100%
^b Other includes bites to the buttocks
^c Multiple bites are listed separately, to maintain statistical independence

Fifty-four different breeds of dog were reported as having bitten respondents (Table 3). Nearly one-quarter of bites were identified as coming from four breeds. Two hundred and fifty-nine (49%) respondents were very sure that they had correctly identified the breed of dog, 63 (12%) were sure and 64 (12%) were reasonably sure, while 141 (27%) were unsure or very unsure that they had identified the breed correctly.

Across all breeds there was no significant relationship between the breed of dog and severity of the bite, but bites involving Rottweilers were associated with more severe psychological effects than other dogs in the survey (χ²=6.16; df=1; p=0.013). Moreover, 13/29 (45%) bites by Rottweilers were rated moderate, severe or very severe, compared with 120/500 (24%) bites by dogs of other breeds. German Shepherd dogs were more likely to be responsible for injuries to multiple parts of the body than other dogs in the survey (χ2=10.27; df=1; p=0.001). Eleven of 42 (26%) bites by German Shepherd dogs, and 48/481 (10%) bites by other breeds involved multiple body parts.

Ninety-two (17%) respondents were bitten by small dogs (<40 cm at the shoulder), while 285 (54%) were bitten by mediumsized dogs (40–60 cm at the shoulder) and 151 (29%) were bitten by large dogs (taller than 60 cm at the shoulder). Psychological injuries associated with bites by larger dogs were rated as more severe than bites by small or medium-sized dogs (χ²=10.90; df=2; p=0.004). Twelve of 91 (13%) bites by small dogs were rated moderate to very severe, compared with 73/284 (26%) bites by medium-sized dogs and 48/150 (32%) bites by large dogs. Large dogs were responsible for half of all severe or very severe physical injuries (?2=11.63; df=4; p=0.02). Large dogs, although responsible for only 28% of all bites, were more likely to cause injury to multiple parts of the body than small and medium-sized dogs (Kolmogorov-Smirnov two-sample test; K=1.59; p<0.02), and were responsible for 28/59 (48%) cases of multiple bites.

Seventy-one (13%) dogs were destroyed as a result of attacks on respondents. Rottweilers and Pitbull Terriers were more likely to be destroyed than other breeds. Eight of 29 (28%) Rottweilers involved in attacks were destroyed, compared with 63/495 (13%) of other breeds (χ²=5.16; df=1; p=0.023), and 11/34 (32%) Pitbull Terriers were destroyed, compared with 60/490 (12%) of other dogs (χ2=10.97; df=1; p=0.001). Twenty-one (4%) dogs were re-homed, while 34 (6%) were given obedience or behaviour–modification training. In 291 (55%) cases, nothing happened to the dog. Seventy-one (13%) respondents were unsure about the fate of the dog.

Table 3. Frequency and percentage of bites by breed of dog (n=529), in a survey of adult victims of dog bites in New Zealand.

Breed	n	Percentage
Mixed	140	26.5
Other breed (n=44)	135	25.5
Unknown	75	14.2
German Shepherd	42	7.9
Pitbull Terrier	34	6.4
Rottweiler	30	5.7
Jack Russell Terrier	22	4.1
Labrador	14	2.6
Border Collie	13	2.5
Blue Heeler	12	2.3
Bull Terrier	12	2.3

Circumstances of the bites

One hundred and sixty-seven (32%) bites took place in large cities, 73 (14%) in provincial cities, 61 (11%) in large towns, 103 (19%) in small towns, and 127 (24%) bites occurred on rural properties. Compared with the proportion of people who said they were from rural and urban areas, people were more likely to be bitten in rural areas (χ²=251.06; df=1; p<0.001). One hundred and ninety-eight (37%) people were from rural areas but 230 (43%) were bitten in rural areas, whereas 332 (62%) respondents were from urban areas and 300 (56%) respondents were bitten in urban areas.

The most common locations for dog bites were public streets/walkways (n=138; 26%), the victim's home (n=110; 21%), a stranger's property (n=77; 14%), or a friend's property (n=66; 12%). For the remainder, 25 (5%) took place at a family member's property, 22 (4%) in parks, 21 (4%) at a neighbour's property, 41 (8%) were recorded as other private property, and 30 (6%) as other public area.

When asked how the dog was restrained prior to the bite, 33 (6%) victims said the dog was on a lead in a public place, 50 (10%) said it was free in a public place but supervised by the owner, and 63 (12%) said it was running free and unsupervised in a public place. One hundred and twenty-eight (24%) respondents said the dog was fenced on private property, although some said the dog jumped the fence or was allowed access to public areas through an open gate, 37 (7%) of the dogs were chained on private property, and 156 (30%) were on private property but not fenced or chained. Respondents were unsure how the dog was restrained in five (1%) cases, and 55 (10%) were recorded as ‘other’. Most respondents who recorded restraint as ‘other’ said the dog was inside a car or house at the time of the bite. In one case, the dog was inside a kennel.

On private property, dogs in rural areas were less likely to be restrained (n=49 (30%) fenced; n=20 (12%) tied) compared with dogs in urban settings (n=79 (48%) fenced; n=17 (10%) tied; χ²=7.51; df=2; p=0.023). One hundred and fifty-three (75%) people bitten in rural areas were bitten by dogs on private property, compared with 168 (64%) people bitten in urban areas (χ2=7.05; df=1; p=0.008).

Three hundred and seventy-five (71%) respondents said the bite was unprovoked. Eighty-eight (16%) said the dog was provoked, and 68 (13%) were unsure. However, 414 (77%) respondents identified a likely reason for the dog to bite them (Table 4). The reason given by the respondent was also compared with their description of the incident, and where necessary the researchers made a note of an alternative explanation for the bite. For the remainder of the analysis the researchers' reason for the bite was used.

Dogs that caused injury to the face or head were less likely to bite because of territorial defence or by accident (Table 5), and more likely to bite due to fear or pain than dogs that injured other parts of the body (χ²=33.89; df=4; p<0.001). People bitten on the hand were less likely to have been bitten due to protection of territory or unknown reasons, and more likely to have been bitten by accident or because of fear or pain (χ²=81.57; df=4; p<0.001). In contrast, dogs that bit the legs or feet were more likely to bite because of defence of territory, and less likely to bite because of fear, pain or by accident (χ²=52.82; df=4; p<0.001). Dogs that bit multiple parts of the body were less likely to have bitten by accident, and more likely to have been in pain (?2=12.53; df=4; p=0.014).

Respondents bitten by dogs due to defence of territory were more likely to be unsure of the dog's sex, and less likely to have been bitten by a female dog, whereas people bitten by accident and by dogs in pain were less likely to be unsure of the dog's sex (Table 6; χ²=58.56; df=8; p<0.001).

Table 4. The percentage of dog bites attributed to different reasons, as suggested by the respondent or the researcher, in a survey of adult victims of dog bites in New Zealand^a.

Reason	Percentage of bites
	Respondent^b	Researcher
Protect territory	21.7	31.4
Unsure	21.5	16.2
Accident	14.2	16.0
Dog afraid	8.2	9.0
Dog in pain	5.7	6.6
Protect person	3.4	5.3
Play	4.5	5.1
Protect item	3.2	2.6
Protect puppies	1.9	1.9
Other	15.7	6.0
n	530	532
^a Rounding of percentages results in some column totals ≠ 100%
^b Not all respondents gave a reason for their bites but in two cases the researchers were able to use the respondent's description of events to give a likely reason for the bite

Table 5. Percentage of people bitten by a dog on different parts of the body, with the reason for the bite, in a survey of adult victims of dog bites in New Zealand^a.

Reason^b	Part of body bitten						n
	Arm	Hand	Leg	Face	Other^c	Multiple
Territorial	14	12	51	1	8	14	167
Unsure	17	14	45	6	11	8	85
Accident	12	61	24	0	0	4	84
Fear	17	46	19	6	2	10	48
Pain	12	44	3	18	0	24	34
Other	12	31	41	3	3	12	111
n	72	154	198	18	27	60	529
^a Rounding of percentages results in some column totals ≠ 100%
^b Determined by researchers based on the respondent's description of events
^c Other includes bites to the buttocks

Table 6. The top five reasons, expressed as a percentage, for dogs to bite, with the sex of the dog responsible, in a survey of adult victims of dog bites in New Zealand.

Sex of dogs	Reason for bite^a					Total %
	Protect territory	Unsure	Accident	Dog afraid	Dog in pain
Male	43	48	49	48	51	47
Female	13	19	39	21	49	23
Unsure	44	33	12	31	0	30
Total n	164	85	84	48	35	416
^a Determined by researchers based on the respondent's description of events

One hundred and fifty-seven (30%) bites were reported to authorities, 365 (68%) were not reported, and respondents were unsure in 11 (2%) cases. Sixty-nine (13%) said they reported the bite to a dog-control officer, 71 (13%) were reported to the local council, and 18 (3%) to police. Some bites were reported to both the local council and the police. Bites resulting in more severe psychological effects were more likely to be reported (Kolmogorov-Smirnov two-sample test; K=2.81; p<0.02). Eighty-seven (23%) respondents with very minor or minor psychological effects reported the bite, compared with 68 (52%) people suffering moderate to very severe effects.

Of the 165 dogs that bit due to fear, pain or by accident, 10 (6%) were destroyed. In contrast, 48/248 (19%) dogs that bit because of territorial defence or an unknown reason were killed (?2=14.5; df=1; p<0.001). Twenty-two (65%) bites by Pitbull Terriers were reported to authorities, whereas 132 (27%) bites by all other breeds were reported (χ2=21.12; df=1; p<0.001).

Discussion

The data for this study came from a survey of ACC claimants across New Zealand. This survey population is likely to provide fewer biased data than from some other sources. For example, reports to local councils of dog bites would unlikely be made by the dogs' owners, their relatives or close friends. Claims to ACC are confidential, and isolated from dog-controlling authorities. Moreover, a claim to ACC is necessary to receive compensation for medical treatment. Together, these features of the list of ACC claimants provide the best approximation of incidents requiring any medical treatment. Thus, the study is founded on a representative population of potential respondents. As with any survey, study was limited by the individuals who chose not to respond. Moreover, the findings of this study cannot be applied to children. Injuries to children tend to be more severe than injuries to adults (Brogan et al. 1995; Overall and Love 2001; Marsh et al. 2004; Wake et al. 2006). It is also likely that the nature and severity of psychological trauma and the circumstances of incidents of bites may be different for adults and children.

Most studies of dog attacks on people have found that males are bitten more often than females (Overall and Love 2001; Marsh et al. 2004; Wake et al. 2006) so the finding here that equal numbers of respondents in this study were male and female is unusual, and could be explained by the methods used. It is possible that women are more likely to complete and return postal surveys, which would artificially elevate their numbers. Alternatively, the lack of a difference in the proportion of men and women bitten in this study may be a result of the sample being made up of adults. The Department of Internal Affairs (Anonymous 2003) reported that between the ages of 15 and 40 men were more likely than women in the same age group to be treated for dog bites in hospitals in New Zealand, but over the age of 40 the proportions of men and women were similar.

The age distribution of people in this survey is also likely to have been affected by the methods used. Only 13% of respondents were between the ages of 16 and 29 when they were bitten. In contrast people in this age range made up 38% of adults admitted to hospitals in New Zealand from 1989 to 2001 because of dog bites (Marsh et al. 2004). Younger people in this study may be less likely to respond to the survey, and may be more likely to have moved house between the time their claims were made and the time the surveys were sent out, meaning they would be less likely to receive their surveys. Compared with the numbers of respondents from rural and urban areas, people were more likely to be bitten in rural areas (Wake et al. 2006). The higher dog-to-human ratio in rural areas of New Zealand (Anonymous 2003) may also have contributed to this trend.

Because the ACC excluded severe cases from the sample, data from this study on the severity of injuries caused by dog bites cannot be considered representative of all dog-bite injuries. Nevertheless, 16/532 (3%) respondents claimed to have sustained severe injuries, and this proportion was similar to the reported proportion of bites rated moderate to severe (2.7%) by the ACC in 2003 (Fox 2004).

The anatomical distribution of bites in this study was somewhat different from those reported elsewhere (Marsh et al. 2004; Wake et al. 2006), with fewer bites to the head area and more to the legs or feet. However, other studies included children in their samples, which will affect the distribution of bites to different parts of the body. Children are more likely to be bitten about the head region, whereas bites to the extremities are more common in adults (Marsh et al. 2004; Wake et al. 2006).

Thirty-six percent of respondents were bitten by a stranger's dog, and 24% of bites in this study involved dogs living with the respondent. Similarly, 23% of bites reported in the city of Guelph, Canada, were bites to the owner of the dog, and a further 20% of reported bites took place during a delivery operation (Szpakowski et al. 1989). In the study reported here, the 11% (58) of respondents who named the owner of the dog as ‘other’ worked as meter readers, couriers, plumbers or lawn-mowing professionals, and were on the client's property at the time of the bite.

As in this study, it is often reported that male dogs are responsible for more bites than female dogs, and more entire male dogs than neutered male dogs exhibit aggression towards people. Studies have also found that neutered female dogs were more aggressive than intact bitches (Podberscek and Blackshaw 1993; Guy et al. 2001; Overall and Love 2001; Takeuchi et al. 2001). Findings regarding sex and, in particular, neuter status in this and other studies must be interpreted with caution as they may simply reflect the numbers of each category in the total population of dogs.

While only 27% of respondents were unsure if they had identified the breed of dog correctly, data on breeds that bit, based on the victim's identification, must be interpreted with caution. Misidentification of breed by bite victims is a real possibility, and we came across a number of instances where the reported size of the dog did not appear to match the breed. In addition, the frequency of each breed in the dog population as a whole must also be taken into account. Large dogs in this study were likely to produce more severe physical and psychological injuries, and were more likely to injure multiple parts of the victim's body, possibly because larger dogs may be more difficult to fend off in an attack or because they have greater access to the upper extremities in comparison with a smaller dog. Some small breeds considered to be very aggressive are the Corgi and Chihuahua (Stafford 1996). While inappropriate aggression in any dog is a serious issue, larger dogs have the potential to do much more damage when they react aggressively.

Public streets or walkways were the most common locations for dog bites in this study (26%). However, only 6% of people hospitalised in New Zealand because of dog bites said they were bitten whilst on a street (Marsh et al. 2004). In that study, nearly one-third of bites took place in the home, not necessarily the home of the bite victim, and 1% of people were bitten on a farm.

Twelve percent of the dogs in this study were loose and unsupervised in a public place when they attacked. In contrast, over half of all dogs in Adelaide, Australia, that had bitten people were free and uncontrolled in public places (Thompson 1997). While in the current study the proportion of dogs that were fenced on private property was relatively high (24%), almost 30% of dogs were on private property but not confined in any way.

Respondents said their attack was unprovoked in 70% of cases. While canine aggression is often said to be unprovoked, on closer inspection of the circumstances of the bite there was almost always an identifiable eliciting stimulus (Borchelt 1983). Indeed, many respondents in this study said their bite was unprovoked but most respondents (79%) were able to give a plausible reason for the dog to bite them, and the researchers were able to assign a reason for 84% of bites. Defence of territory was the most common reason for dogs to bite, followed by accidental bites, most of which were inflicted while respondents were breaking up dog fights, or the dog was fearful or in pain. Very few previous studies on dog bites have looked at what caused dogs to bite; protective behaviour could have explained 46% of reported bites in Guelph (Szpakowski et al. 1989), compared with 41% of bites in the current study.

While data on the causes of dog attacks are scarce, they provide valuable information that serves to highlight particularly high-risk situations and how they can be avoided. Most of the 84 accidental bites in the current study occurred when dog owners were attempting to break up fights between two dogs. There is an obvious risk of getting bitten in this situation. Another situation that could easily be avoided is people getting bitten by dogs that are in pain. Typically, respondents bitten by dogs that were in pain were attempting to help (and in most cases were in the act of picking up) a dog that had been hit by a car, caught in an electric or barbed-wire fence, or injured in a fight with another dog. These situations often resulted in bites to the face or hand.

Dog bites are an escalating problem in New Zealand; there are an average of 240 hospitalisations due to dog bites every year (Marsh et al. 2004), at a considerable cost to the ACC, i.e. the taxpayer (Healey 2007). While data on the incidence and financial cost of dog bites in New Zealand serve to highlight the magnitude of the problem, it is information on the circumstances of the bites that could have the greatest potential to reduce the numbers of bites that take place. The data presented in this study contribute to an understanding of why dog bites occur, and thus provide guidance to individuals seeking to reduce the risk of personal injury, and to legislators framing law to protect the general public.

Acknowledgements

We thank the ACC, especially Peter Larking, for assistance with this research, and all the people who completed and returned questionnaires.

Notes

http://www.sciquest.org.nz/node/60928