Orphaned eastern cottontail rabbits (ECRs; Sylvilagus floridanus) often present to wildlife clinics within their geographic range and require considerable dedication of time and resources. The objective of this analytical cross-sectional study was to assess initial examination findings to be used as prognostic indicators for orphaned neonatal and juvenile ECRs. The medical records of the University of Illinois Wildlife Medical Clinic were searched for ECRs presenting between 2012 and 2018. This criterion identified 1,256 ECRs that were then classified as survivors (survived and released) or as nonsurvivors (euthanized or natural death) within 72 h of admission. Presenting weight, body system abnormalities, hydration status, intervention prior to presentation, and singleton versus group presentation were categorically recorded for each individual ECR. The data were modeled using a series of logistic regression models fitted using the general linear model. Individuals were significantly more likely to be nonsurvivors if they presented as singletons (P<0.0001), presented with moderate/severe (P<0.001) or mild integumentary signs (P=0.0261), presented with multi-organ disease (P<0.001), presented with neurologic signs (P<0.0003), or had treatment provided prior to presentation (P=0.031). Factors that did not predict survival status in ECRs included body weight (P=0.210), presence of respiratory signs (P=0.674), and presence of dehydration (P=0.356). These findings may be used at wildlife medical clinics to make triage criteria for euthanasia as well as dedicate limited funds and labor to cases with the best prognosis for survival.

The eastern cottontail rabbit (ECR; Sylvilagus floridanus) is the most ubiquitous species of rabbit within the genus and is widely distributed throughout North America (Chapman and Litvaitis 2003). This species preferentially inhabits peripheral shrubbery of open areas and pastures making the Midwest region of the US a preferred habitat (Mankin and Warner 1999; Bertolino et al. 2011). Eastern cottontail rabbits have the ability to thrive in urban and suburban environments, including parks, gardens, and city landscapes, frequently leading to human-wildlife conflicts (Chapman and Flux 1990; Messmer 2009). Interactions with companion animals and lawnmowers often lead to injuries and presence of nests in flowerbeds commonly leads to human inconvenience and displacement, making the ECR a common species presented to wildlife rehabilitation centers in their range (Hunt et al. 2014).

Another common cause for presentations of ECRs to medical clinics relates to a misunderstanding of their nesting habits. The nest, built into the ground for optimal camouflage, is the home of ECR kits for the first 3–4 wk of their life until they are fully weaned (Cherney and Nieves 1991; Bewig and Mitchell 2009). During the preweaning period, the doe exhibits crepuscular nursing behaviors and typically visits the nest only at dusk and dawn (Varga 2014). This lack of visible maternal care often leads to an anthropomorphic misinterpretation of nest abandonment, leading to presentation of neonatal ECRs as presumed orphans to wildlife medical and rehabilitation centers.

Neonatal and juvenile ECRs are commonly presented to wildlife medical centers in the Midwest and exhibit high rates of morbidity and mortality despite intensive care. Wildlife clinics strive to make appropriate triage decisions to maximize the capabilities of animal welfare, veterinary services, and funding, but there are no current data regarding accurate prediction of survival status of this species in captive care. The objectives of this study were to identify predictive factors for survival in orphaned neonatal and juvenile ECRs presented to a wildlife medical center. Ultimately this knowledge may be used to guide decision-making regarding care and triage of this species. Based upon clinical experiences, we hypothesized that survival would be negatively impacted by lower weight status, singleton presentation, and multiple body system abnormalities at the time of presentation.

Case selection

The medical record database of the University of Illinois Wildlife Medical Clinic (UIWMC, Urbana, Illinois, USA) was searched to retrieve records of orphaned ECRs. Each record was reviewed by a single reviewer (S.L.P.). Inclusion criteria included date of presentation (1 January 2012–31 December 2017), species group (ECR), and weight (40–120 g, inclusive). Weight parameters for this study were set by standards of intake at UIWMC, where ECRs of less than 40 g are euthanized on arrival due to historically poor survival and ECRs greater than 120 g are released or sent to rehabilitation on arrival if they are otherwise healthy due to their ability to self-feed. Cases were excluded from this study if the medical record was incomplete or if ECRs were dead on arrival or euthanized on arrival prior to medical intervention.

Medical records review

For each case, categoric information was recorded, including survival status; weight class; reason for presentation; presence of abnormal examination findings of the respiratory, neurologic, or integumentary body systems; presentation number; hydration status; and intervention status prior to admission. Survival status was categorized as survivor if the case survived at least 72 h or was released or transferred within 72 h after intake. Survival status was categorized as nonsurvivor if the case was euthanized or died within 72 h after intake. Medical records for nonsurvivors were additionally evaluated for the general reason for euthanasia or cause of death. Weight class on admission was categorized as lower weight class (40–70 g) or higher weight class (71–120 g). In the UIWMC, ECRs less than 70 g are typically preweaning and need to be tube-fed once to twice daily, versus ECRs greater 71 g, which are displaying early weaning activity of eating solid foods. Presentation was categorized and scored as healthy (e.g., found in yard), known history of trauma (e.g., nest or individual being attacked by animals, hit by lawnmowers, or dropped from heights), or other (e.g., nest flooding or transfer from another medical clinic). Body system abnormalities on initial examination for the integumentary system were categorized as no abnormal findings, mild integument abnormalities (e.g., single puncture wound, superficial abrasion, or flaky skin), or moderate to severe integument abnormalities (e.g., more than one puncture wound or degloving injury). The presence or absence of respiratory abnormalities (e.g., labored breathing, nasal or ocular discharge), neurologic abnormalities (e.g., abnormal head posture or position, depressed or obtunded mentation, ataxia), or multi-organ system abnormalities with or without other findings (e.g., external parasites, corneal ulcer, or findings in multiple body systems) was recorded. Presentation number was categorized as singleton if the ECR was presented alone, or as group if the ECR presented with at least one other ECR. Hydration status was categorized as either euhydration or dehydration as determined by assessment of skin turgor. Intervention prior to presentation was categorized as either no known supportive care provided prior to presentation or documentation that the case was provided nutritional, thermal, or medical care prior to presentation.

Data analysis

The survival status (survivor or nonsurvivor) was modeled using a logistic regression model. Fixed effects considered were year of presentation, weight class, presentation number, body system abnormalities, hydration status, and intervention status. All models were fitted using the general linear model function in the effects package in commercial software (R Development Core Team 2016). An information theoretic approach (Burnham and Anderson 1998) was used to determine which model from the candidate set performed best using commercially available software (AICcmodavg package in R studio, Mazerolle 2015); this commonly used technique allows comparison between competing models. Models with the lowest value are ranked higher and more likely to account for the variation seen in the data (Dohoo et al. 2012). Odds ratios (ORs) were calculated from the coefficients of the highest-ranking model. Odds ratios were also calculated from bivariate comparisons of all parameters significant in the multivariable model. Statistical significance was set at P<0.05.

Of the 30 models tested, the model with all categoric variables except year ranked the highest (Table 1). A total of 1,256 individual ECRs met the criteria for inclusion (Table 2). Individuals with mild integument abnormalities (mortality: 34%, 22/64) were nearly two times more likely to be nonsurvivors (OR: 1.84; 95% confidence interval [CI]: 1.07–3.15) and those with moderate to severe integument abnormalities (mortality: 42%, 26/61) were approximately 4.5 times more likely to be nonsurvivors (OR: 4.52; 95% CI: 2.56–7.97) than individuals with no integumentary system abnormalities. Individuals with neurologic system abnormalities (mortality: 48%, 30/62; OR: 4.82; 95% CI: 2.69–8.64) and multiple system abnormalities (mortality: 42%, 37/88; OR: 3.47; 95% CI: 2.10–5.73) were more likely to be nonsurvivors compared to ECRs without these findings on initial assessment. Individuals with no prior treatment were more likely to survive (mortality: 7.8%, 79/1018; OR: 0.59; 95% CI: 0.37–0.95) than those with prior treatment (mortality: 13.4%, 32/238). Individuals that presented as singletons (mortality: 37%, 102/276) were twice as likely to be nonsurvivors as those that presented as a group (mortality: 13.8%, 136/980; OR: 0.44; 95% CI: 0.28–0.71). In further evaluation, those individuals that were classified in the lower weight class (40–70 g) and presented as singletons were twice as likely to be nonsurvivors (39%, 42/108) as those in the same weight class that presented as a group (18%, 111/628). Additionally, those individuals in the higher weight class (71–120 g) and presented as singletons were five times more likely to be classified as nonsurvivors (mortality 36%, 60/168) than those in the same weight class that presented as a group (7%, 25/352). Factors that were not found to predict survival status included hydration status and presence of respiratory system abnormalities, despite these factors being included in the final model (Table 3).

Table 1

Akaike information criterion (AIC) rankings comparing multiple linear models relating categoric variables of 1,256 eastern cottontail rabbit (Sylvilagus floridanus) cases presented to the Wildlife Medical Clinic at the University of Illinois from 2012 to 2017.

Akaike information criterion (AIC) rankings comparing multiple linear models relating categoric variables of 1,256 eastern cottontail rabbit (Sylvilagus floridanus) cases presented to the Wildlife Medical Clinic at the University of Illinois from 2012 to 2017.
Akaike information criterion (AIC) rankings comparing multiple linear models relating categoric variables of 1,256 eastern cottontail rabbit (Sylvilagus floridanus) cases presented to the Wildlife Medical Clinic at the University of Illinois from 2012 to 2017.
Table 2

Frequency of categoric variables and survivor status of 1,256 eastern cottontail rabbit (Sylvilagus floridanus) cases presented to the Wildlife Medical Clinic at the University of Illinois from 2012 to 2017.

Frequency of categoric variables and survivor status of 1,256 eastern cottontail rabbit (Sylvilagus floridanus) cases presented to the Wildlife Medical Clinic at the University of Illinois from 2012 to 2017.
Frequency of categoric variables and survivor status of 1,256 eastern cottontail rabbit (Sylvilagus floridanus) cases presented to the Wildlife Medical Clinic at the University of Illinois from 2012 to 2017.
Table 3

Parameter estimates and odds ratios (OR) for nonsurvival status from the best general linear model as determined by Akaike information criterion for categoric variables related to 1,256 orphaned neonatal and juvenile eastern cottontail rabbits (Sylvilagus floridanus) presented to the Wildlife Medical Clinic at the University of Illinois (2012–17). The baseline parameters for each category were 40–70 g, no history of trauma, no skin lesion, normal respiratory and neurologic assessment, single organ involvement, dehydration, known prior care, and single presentation. Bivariate OR with 95% confidence interval (CI) is presented for comparison. — indicates not applicable.

Parameter estimates and odds ratios (OR) for nonsurvival status from the best general linear model as determined by Akaike information criterion for categoric variables related to 1,256 orphaned neonatal and juvenile eastern cottontail rabbits (Sylvilagus floridanus) presented to the Wildlife Medical Clinic at the University of Illinois (2012–17). The baseline parameters for each category were 40–70 g, no history of trauma, no skin lesion, normal respiratory and neurologic assessment, single organ involvement, dehydration, known prior care, and single presentation. Bivariate OR with 95% confidence interval (CI) is presented for comparison. — indicates not applicable.
Parameter estimates and odds ratios (OR) for nonsurvival status from the best general linear model as determined by Akaike information criterion for categoric variables related to 1,256 orphaned neonatal and juvenile eastern cottontail rabbits (Sylvilagus floridanus) presented to the Wildlife Medical Clinic at the University of Illinois (2012–17). The baseline parameters for each category were 40–70 g, no history of trauma, no skin lesion, normal respiratory and neurologic assessment, single organ involvement, dehydration, known prior care, and single presentation. Bivariate OR with 95% confidence interval (CI) is presented for comparison. — indicates not applicable.

The overall mortality rate within 72 h of admission for this population was 18.9% (238/1,256); 131 ECRs died naturally while 107 were euthanized. Reasons for death included severe trauma (0.8%, 11/1,256), obvious aspiration (1.0%, 13/1,256), obvious esophageal perforation during tube feeding (1.8%, 23/1,256), development of or persistent neurologic signs (2.3%, 29/1,256), and open for causation (12.9%, 162/1,256).

In this study, ECRs that presented as singletons were twice as likely to be nonsurvivors as ECRs that presented in a group. Investigating the factors that contributed to a singleton's increased mortality was outside the scope of this study. However, we predict that the potential reasons for an ECR to be isolated from its littermates, such as predation, illness, or nest destruction, are also potential causes for decreased survival. The ECRs that presented as a group may be more likely to be victims of human-wildlife conflicts of inconvenient nest location or misconceptions of nest abandonment, and these animals would be more likely to be presenting without medical issues. Future investigations of survival status associated with singletons could include evaluating stress responses to better understand the effects of social nesting structure on immune health and stress response.

Although weight class was not a significant predictor of nonsurvival status, ECRs presenting as singletons in the lower weight class were more likely to be nonsurvivors compared to ECRs presenting as a group. The assignment of weight classes used here was influenced by the estimated weaning weight of ECRs (Orr 1998) and experience with ECRs at the UIWMC. Young ECRs are completely dependent on milk from the doe for the first 10–15 d of life and the weaning process begins at 2 wk of age (about 70–140 g; Cherney and Nieves 1991; Bewig and Mitchell 2009); thus, weight class is an indirect indicator of weaning status. A full understanding of the mechanisms behind poor survival of preweaned ECRs in the wildlife clinic setting is outside the scope of this study. However, it is likely that gastrointestinal microbiome considerations and lack of readily available species-specific milk replacement may play a role in decreased survival. Preweaned rabbits have very few microorganisms in their gastrointestinal tract and rely on an antimicrobial milk oil in the doe's milk to control their gastrointestinal microflora (Varga 2014). For orphaned ECRs in a medical clinic, it is commonly recommended to feed commercially available puppy and kitten milk replacers which do not contain the enzymes found naturally in doe's milk that are needed to make this protective milk oil (Mullineaux 2014; Varga 2014). Additionally, the stomach pH in preweaned rabbits is much higher (pH 5–6.5) in comparison to the much more acidic gastric environment of adult rabbits (pH 1–2; Varga 2014). The less acidic neonatal stomach in conjunction with the lack of protective milk enzymes places hand-reared ECRs at increased risk for passage of enteric pathogens such as Escherichia coli, which can cause diarrhea and mortality events (Cherney and Nieves 1991; Varga 2014). Given the poor survival of singletons presenting as preweaned ECRs, future studies should be aimed at understanding the microbiome changes in neonatal ECRs, in particular those presenting as a singleton, to aid in the development of a species-specific milk replacer or to serve as a guideline for use of appropriate probiotic agents.

Orphaned ECRs with mild abnormalities of the integument system were found to have a nearly doubled likelihood of nonsurvival status compared to their counterparts with no dermatologic abnormalities. The likelihood of nonsurvival quadrupled if the integument abnormalities were classified as moderate to severe. When the types of integument abnormalities recorded in this study are taken into consideration (abrasions, puncture wounds, degloving injuries), it is highly likely that those described are secondary to trauma. We suspect that many of the nonsurvivors suffering from the more severe integumentary abnormalities noted on initial examination ultimately died secondary to internal hemorrhage, herniation, or septicemia. Crushing and tensile forces of a predator's bite can cause internal injuries of various severities, including organ herniation, internal bleeding, and orthopedic trauma (Holt and Griffin 2000). The appearance of a skin wound externally may not be indicative of the extensive tissue damage below the skin (Holt and Griffin 2000). Additionally, infections caused by predator bite wounds can lead to septicemia, as commonly isolated bacteria in dog and cat bite wounds are Pasteurella spp., Staphylococcus intermedius, Enterococcus spp., Bacillus spp., and Streptococcus spp. (Griffin and Holt 2001; Abrahamian and Goldstein 2011). It is likely that many of these fatal integument abnormalities occurred secondary to human or domestic animal-wildlife conflicts, as well as wounds from predators, which have been previously documented as common reasons for presentation to a wildlife clinic setting (Hunt et al. 2014). Wildlife medical clinics presented with orphaned ECRs suffering from integumentary abnormalities on initial examination should consider aggressive antimicrobial protocols or potentially reexamine euthanasia guidelines when considering the poor survival status of this group.

Several neurologic exam findings exhibited by ECRs were observed, but identifying whether these findings were secondary to infectious, traumatic, congenital, or other causes was outside the scope of this study; however, given the high number of orphaned ECRs in this study with a history of trauma (443/1,256; Table 2), there is a high likelihood that many of these cases suffered from traumatic brain injury. Traumatic brain injury is common in companion domestic species and is associated with high mortality when severe (Dewey 2000; Platt et al. 2001). Treatment and prognosis for patients suffering from traumatic brain injury may be influenced by their score using the modified Glasgow Coma Scale (Platt et al. 2001). Survival status associated with neurologic exam findings in ECRs may be better evaluated with the development of an ECR-specific Glasgow-style scoring system based on retrospective and clinical studies.

Rabbits that had not received treatment prior to presentation were twice as likely to survive as those that had received some form of treatment by a Good Samaritan. It is probable that ECRs that received treatment may have been in captivity for a longer period of time prior to presenting to the UIWMC, thus reducing their access to medical care in a timely manner. As many ECRs are removed from nests for presumably being orphans, there is also potential that treatments provided by a layperson, including over-the-counter medications or inappropriate food items, may have caused harm. These findings support that wildlife patients should be presented to wildlife rehabilitation and veterinary professionals as soon as possible after identification of their need to be removed from their natural surroundings.

The vast majority of the ECRs reported here did not have a definitive diagnosis for the cause of death or reason for decline. Postmortem evaluations are a mainstay of obtaining this information in a clinical situation; however, none of the ECRs received a postmortem evaluation due to funding limitations at our institution. Our findings support the prognostic value of several initial physical examination findings; however, there is a lack of understanding for why ECRs with specific poor prognostic indicators die or decline and require euthanasia. Future prospective research should focus on specific causes of death and diagnoses for ECR in rehabilitation facilities.

The inherent limitations of a retrospective analysis must be considered, including the variability in exam findings by various individuals examining ECRs at this clinic. The UIWMC is primarily operated by veterinary student volunteers under direct supervision of faculty, thus initial examinations of ECRs are performed by multiple individuals with varying levels of experience. To limit the biases that could be introduced by varying veterinary experience levels, simple categoric information was extracted from each of the individual records that would mirror what a low-experience-level student would be expected to be able to discern, including weight, presence or absence of respiratory, neurologic, or integumentary signs, etc. In addition, the data presented here may not adequately represent survival of ECRs presenting to other institutions if differing diagnostic and treatment protocols are utilized. Lastly, there may be other physical examination findings or diagnostic results that act as prognostic indicators in this population. Future studies should evaluate the role of temperature and blood glucose aberrations, as these have been shown to be prognostic indicators in domestic rabbits (Oryctolagus cuniculi; Di Girolamo et al. 2016; Harcourt-Brown and Harcourt-Brown 2012), and evaluation of other physical examination findings, including body condition scoring.

Results of the present study are intended to serve as a resource for wildlife medical clinics and rehabilitation centers so that they may better assess prognosis for survival of orphaned ECRs given initial examination assessments. The information in the present study aims to assist in identification of cases with the best chance of survival outcome to allow for appropriate allocation of resources, as well as to aid in identification of patients with the least likely chance of survival so that more intensive medical intervention or humane euthanasia can be initiated.

Abrahamian
FM
,
Goldstein
EJ
.
2011
.
Microbiology of animal bite wound infections.
Clin Microbiol Rev
24
:
231
246
.
Bertolino
S
,
di Montezemolo
NC
,
Perrone
A
.
2011
.
Daytime habitat selection by introduced eastern cottontail Sylvilagus floridanus and native European hare Lepus europaeus in northern Italy.
Zoolog Sci
28
:
414
419
.
Bewig
M
,
Mitchell
MA
.
2009
.
Wildlife.
In
:
Manual of exotic pet practice
, 1st Ed.,
Mitchell
MA
,
Tully
TN
Jr
, editors.
Saunders Elsevier
,
St. Louis, Missouri
, pp.
493
529
.
Burnham
KP
,
Anderson
DR
.
1998
.
Model selection and multimodal inference: A practical information-theoretic approach.
Springer-Verlag
,
New York, New York
,
488
pp.
Chapman
JA
,
Flux
JEC
.
1990
.
Rabbits, hares and pikas: Status survey and conservation action plan.
International Union for Conservation of Nature and Natural Resources
,
Gland, Switzerland
, pp.
95
110
.
Chapman
JA
,
Litvaitis
JA
.
2003
.
Eastern cottontail (Sylvilagus floridanus and allies).
In
:
Wild mammals of North America: Biology, management, and conservation
, 2nd Ed.,
Feldhamer
GA
,
Thompson
BC
,
Chapman
JA
, editors.
Johns Hopkins University Press
,
Baltimore, Maryland
, pp.
101
125
.
Cherney
L
,
Nieves
MA
.
1991
.
How to care for orphaned wild mammals.
Iowa State Univ Vet
53
:
94
99
.
Dewey
CW.
2000
.
Emergency management of the head trauma patient.
Vet Clin North Am Small Anim Pract
30
:
207
225
.
Di Girolamo
N
,
Toth
G
,
Selleri
P
.
2016
.
Prognostic value of rectal temperature at hospital admission in client-owned rabbits.
J Am Vet Med Assoc
248
:
288
297
.
Dohoo
I
,
Martin
W
,
Stryhn
H
.
2012
.
Model-building strategies.
In
:
Methods in epidemiologic research
,
Dohoo
IR
,
Martin
WS
,
Stryhn
H
, editors.
Ver. Inc.
,
Charlottetown, Prince Edward Island, Canada
, pp.
401
428
.
Griffin
GM
,
Holt
DE
.
2001
.
Dog-bite wounds: Bacteriology and treatment outcome in 37 cases.
J Am Anim Hosp Assoc
37
:
453
460
.
Harcourt-Brown
FM
,
Harcourt-Brown
SF
.
2012
.
Clinical value of blood glucose measurement in pet rabbits.
Vet Rec
170
:
674
.
Holt
DE
,
Griffin
G
.
2000
.
Bite wounds in dogs and cats.
Vet Clin North Am Small Anim Pract
30
:
669
679
.
Hunt
VM
,
Magle
SB
,
Vargas
C
,
Brown
AW
,
Lonsdorf
EV
,
Sacerdote
AB
,
Sorley
EJ
,
Santymire
RM
.
2014
.
Survival, abundance, and capture rate of eastern cottontail rabbits in an urban park.
Urban Ecosyst
17
:
547
560
.
Mankin
PC
,
Warner
RE
.
1999
.
A regional model of the eastern cottontail and land-use changes in Illinois.
J Wildl Manag
63
:
956
963
.
Mazerolle
MJ.
2015
.
Estimating detectability and biological parameters of interest with the use of the R environment.
J Herpetol
49
:
541
559
.
Messmer
TA.
2009
.
Human-wildlife conflicts: emerging challenges and opportunities.
Hum-Wildl Conflict
3
:
10
17
.
Mullineaux
E.
2014
.
Veterinary treatment and rehabilitation of indigenous wildlife.
J Small Anim Pract
55
:
293
300
.
Orr
D.
1998
.
Rehabilitation and release of the eastern cottontail.
.
Platt
SR
,
Radaelli
ST
,
McDonnell
JJ
.
2001
.
Prognostic value of the modified Glasgow Coma Scale in head trauma in dogs.
J Vet Intern Med
15
:
581
584
.
R Development Core Team
.
2016
.
R: A language and environment for statistical computing.
R Foundation for Statistical Computing
,
Vienna, Austria
.
http://www.R-project.org. Accessed February 2017
.
Varga
M.
2014
.
Basic rabbit science.
In
:
Textbook of rabbit medicine
, 2nd Ed.,
Varga
M
, editor.
Elsevier
,
Edinburgh, Scotland
, pp.
3
108
.