Forensic Science International 275 (2017) 195–202

Contents lists available at ScienceDirect

Forensic Science International

journal homepage: www.elsevier.com/locate/forsciint

Genital and anal injuries: A cross-sectional Australian study of

1266 women alleging recent sexual assault
Renate R. Zilkensa,** , Debbie A. Smithb , Maureen A. Phillipsa,b , S. Aqif Mukhtara ,
James B. Semmensa , Maire C. Kellyb,*
a
Centre for Population Health Research, Curtin University, Perth, Western Australia,Australia
b
Sexual Assault Resource Centre, Women and Newborn Health Service, Subiaco, Western Australia, Australia

A R T I C L E I N F O A B S T R A C T

Article history: Objectives: To describe the frequency of genital and anal injury and associated demographic and assault
Received 18 November 2016 characteristics in women alleging sexual assault.
Received in revised form 24 February 2017 Design: Cross-sectional study.
Accepted 20 March 2017
Setting: Sexual Assault Resource Centre (SARC), Western Australia.
Available online 29 March 2017
Participants: Total of 1266 women attending SARC from Jan-2009 to Mar-2015.
Methods: Women underwent a standardised data collection procedure by forensically trained doctors.
Keywords:
Multivariate logistic regression analyses were performed.
Sexual assault
Genital injury
Main outcome measures: (1) Frequency of genital and anal injuries by type of sexual assault. (2)
Anal injury Identification of independent factors associated with genital and anal injuries following, respectively,
Female completed vaginal and anal penetration.
Intimate partner violence Results: Genital injury was observed in 24.5% of all women with reported completed vaginal penetration;
Domestic violence in a subset with no prior sexual intercourse 52.1% had genital injury. Genital injury was more likely with
no prior sexual intercourse (adjusted odds ratio [adj. OR] 4.4, 95% confidence interval [95%CI] 2.4–8.0),
multiple types of penetrants (adj. OR 1.5, 95%CI 1.0–2.1), if general body injury present and less likely with
sedative use and delayed examination. Anal injury, observed in 27.0% of reported completed anal
penetrations, was more likely with multiple types of penetrants (adjusted OR 5.0, 95%CI 1.2–21.0), if
general body injury present and less likely with delayed examination.
Conclusion: This study separately quantifies the frequency of both genital and anal injuries in sexually
assaulted women. Genital injuries were absent in a large proportion of women regardless of prior vaginal
intercourse status. It is anticipated that findings will better inform the community, police and medico-
legal evidence to the criminal justice system.
© 2017 The Authors. Published by Elsevier Ireland Ltd. This is an open access article under the CC BY-NC-
ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction
The presence of general body injuries has been associated with
higher rates of laying charges [1] and prosecution of sexual assault
[2–5]. Conviction rates are also higher in women who sustain
genital injuries following sexual assault. In a large 2009 South
* Corresponding author at: Sexual Assault Resource Centre, PO Box 842, Subiaco,
WA, 6904, Australia.
** Corresponding author at: Curtin University, Room238, Building 400, GPO
Box U1987, Perth, Western Australia 6845, Australia.
E-mail addresses: r.zilkens@curtin.edu.au (R.R. Zilkens),
Debbie.Smith@health.wa.gov.au (D.A. Smith), Maureen.Phillips@health.wa.gov.au
(M.A. Phillips), Aqif.Mukhtar@curtin.edu.au (S. A. Mukhtar),
james.semmens@curtin.edu.au (J.B. Semmens), Maire.Kelly@health.wa.gov.au
(M.C. Kelly).
African study [6] both general and genital injury were strongly
associated with conviction. Although historically some courts have
relied upon the presence of genital injury to “prove” sexual assaults
[7], it is well recognised that genital injury is not seen in the
majority of women following sexual assault.
In providing expert testimony to the courts it is important to
have access to separate prevalence estimates for genital and anal
injuries following, respectively, non-consensual vaginal or anal
penetration. There are a number of reports in the literature
concerning the prevalence of genital injury following alleged
sexual assault. Unfortunately, the variety of examination/visuali-
sation techniques, participant inclusion criteria and injury
definitions used by many of these studies make them difficult to
apply in the Australian setting. In Australia, macroscopic (naked
eye) examinations are routinely used to detect genitoanal injuries
following sexual assault and genital redness and/or swelling are

http://dx.doi.org/10.1016/j.forsciint.2017.03.013
0379-0738/© 2017 The Authors. Published by Elsevier Ireland Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-
nd/4.0/).
196 R.R. Zilkens et al. / Forensic Science International 275 (2017) 195–202
considered non-specific findings and excluded from genital injury
definitions. Of the 85 studies of genital injury prevalence data
reviewed by Lincoln et al. in 2013 [8] only fifteen used ‘naked eye’
macroscopic examination [4,7,9–19]. Only six of these separated
genital injuries from anal/peri-anal injuries [4,11,14,15,19,20] and
only three of the six [4,19,20] excluded genital redness and/or
swelling as an injury.
The aim of this study was to determine the frequency of both
genital and anal injuries in women attending a sexual assault
service according to the nature of the sexual assault (i.e. completed
penetration vs attempted vs unknown). We also sought to
determine which demographic and assault characteristics were
associated with the detection of genital injuries in women with
completed vaginal penetration and with anal injuries in those
women reporting completed anal penetration.
2. Methods
2.1. Definitions
Alcohol use refers to alcohol consumed in the 6 h period prior to
the assault.
Anal injury included injury to the perianal region, anus and
rectum.
Assailant types were categorized as stranger, intimate partner,
acquaintance/friend, accidental acquaintance (known <24 h),
unknown (no memory), relatives and other (including work
colleagues and carers). Intimate partner included current and
ex-partners (including husbands, de factos and boyfriends).
Current mental illness was based on the patient’s self-reported
history and included psychotic (e.g. schizophrenia, bipolar
disorder) and non-psychotic (e.g. anxiety, depression) disorders.
General body (non-genitoanal) injury included injuries found on
the head (scalp/hair, eyes, ears, facial), mouth (lips, teeth and oral
cavity), neck, torso (chest, breasts, upper back, abdomen, lower
Fig. 1. Diagram of the female external genito-anal region used to document injuries.
back and buttocks), arms (inner upper arms, remainder of arms,
hands, and fingernails), and legs (inner thighs, remainder of thighs,
lower legs, feet, knees).
Genital injury included injuries on the mons pubis, internal/
external genitalia and perineum (Fig. 1).
Indecent assault was a sexual act without consent in the absence
of completed or attempted penetration.
Injury types included bruises, abrasions, lacerations, incised
wounds, penetrating (stab) wounds and burns. Redness and/or
tenderness were not included due to their non-specific nature.
Injuries considered by the forensic clinician to be self-inflicted
were excluded.
Non-prescribed sedating agents include cannabinoids (marijuana
& synthetic), opiates (heroin) and benzodiazepines.
Sexual assaults included in this study were completed or
attempted penetration of the patient’s vagina or anus by a penis,
finger, hand or object without their consent. The nature of the
penetration was classified as unknown if the patient suspected
sexual assault but had no or incomplete memory of the incident.
Stimulants include amphetamine, ecstasy, cocaine (there were
no hallucinogens in this cohort).
Type of penetrant refers to the body part or object that is
penetrating the vagina or anus such as the penis, finger, hand, and/
or object.
2.2. Selection of study participants
The Sexual Assault Resource Centre (SARC) is the sole sexual
assault referral centre for police and other emergency providers in
Perth, the capital of Western Australia. Study participants included
post-pubertal adolescents and adult females aged 13 years and
older referred to SARC for an emergency consultation between
1 January 2009 and 31 March 2015 following alleged recent sexual
assault. Excluded from the study were patients who (i) did not give
consent for research, (ii) were solely indecently assaulted, (iii) did
 R.R. Zilkens et al. / Forensic Science International 275 (2017) 195–202
not know either the date of the sexual assault or could not estimate
time since assault, (iv) were referred to SARC for emergency
consultation more than 10 days after the sexual assault, (v) did not
consent to genital/anal examination, (vi) admitted that the report
was fallacious and/or the alleged assault was considered to be a
false report by the police or forensic clinician, (vii) only reported
oral assault with a penis, (viii) only reported vaginal and/or anal
assault with a tongue.
2.3. Forensic examination and data collection
A total of 24 clinicians examined patients during the study
period with 12–15 doctors on staff each year. On average, each
doctor examined 53 study patients (interquartile range 14–85).
The SARC doctors are forensically trained: six doctors have
completed the Master of Forensic Medicine, two are Fellows of
the Australasian College of Legal Medicine (FACLM), seven have
Fellowships of the Faculty of Clinical Forensic Medicine from the
Royal College of Pathologists Australasia. As part of their work at
SARC, all doctors take part in regular continuing education. The
physical examinations are conducted according to a standard
sexual assault examination protocol, as outlined in the SARC
Medical and Forensic Manual, Western Australia. This included
examination of the entire body (head to toe, front and back) and
genito-anal regions, with measurement and documentation of any
injuries and findings on standardized body diagrams in the SARC
Forensic Record. Fig. 1 shows the diagram used to document
external genito-anal injuries. Macroscopic visualization, rather
than colposcopy or staining, was used for genito-anal examination.
Patient and/or guardian informed consent was obtained for use of
de-identified data for research. The attending clinician entered
history and examination data into the SARC Medical Forensic
Services Clinical Information System [21].
Missing or inconsistent database data was queried and
amended where possible following clinician chart review.
2.4. Statistical analysis
Descriptive statistics were used to describe characteristics of the
women and the sexual assault and summarized as means  standard
deviations for continuous data and as percentages for categorical
data. Chi Square and Fisher Exact tests were used to compare
proportions by subgroup. Purposeful selection of covariates were
used to assess prognostic factors for injury in patients and to
construct multivariate logistic regression models. All initial models
included covariates that had a p-value <0.25 in bivariate analyses
along with factors of known clinical importance; model covariates
with p-values of 0.15 or less were retained in the model. A value of
25% was used as an indicator of an important change in a
coefficient. The purposeful selection model for the covariates and the
modelling processes were described by Hosmer et al. [22]. No two-
way interactions significant at p < 0.01 were detected. Odds ratios
and 95% confidence intervals were estimated. All statistical analyses
were performed using Stata version 14.1 (College Station, TX, USA).
2.5. Details of ethics approval
Ethics approval was obtained from the Women and Newborn
Health Services Human Research Ethics Committee (Approval
number 2014089EW) and Curtin University Human Research Ethics
Committee, Western Australia (Approval number HR98/2015).
3. Results
A total of 1755 women (and 103 men) presented to SARC for
emergency consultation during the 75 month study period, of
197
whom 1266 were included in the study. The exclusion rate was
27.9% (n = 489): 5.4% (n = 95) not consenting to research, 2.0%
(n = 35) presented following indecent assault, 1.1% (n = 19) sexual
assaults considered to be a false report, 4.0% (n = 70) presenting
>10 days after assault, 0.9% (n = 15) reported solely oral assault,
0.7% (n = 12) reported solely vaginal assault with a tongue, 13.8%
(n = 243) not consenting to genito/anal examination. Police
involvement at the time of emergency consultation was 69.4%
(n = 879). Eighteen women presented with two or more separate
sexual assaults during the study period with each consultation
included in the analysis. The Fig. 2 flow chart provides an overview
of the study design and details on the number participants
examined for site-specific injuries and the prevalence of injuries by
type of assault.
Table 1 provides details on the demographic and assault
characteristics of the 1266 study participants (mean age
26.5  10.9 years, range 13–88 years). Women with a current
mental illness accounted for 39.7% (n = 503) of study participants
and a history of consuming alcohol in the 6 h period prior to the
assault was elicited in 60.7% (n = 768). Fifty-one percent of women
presented to SARC within 24 h of the alleged sexual assault and
81.9% within 72 h. Completed vaginal penetration was the most
common form of sexual assault (n = 948, 74.9%), with 10.4%
(n = 132) reporting both completed vaginal and completed anal
assaults.
3.1. Genital injury
Forty-three (3.4%) of the 1266 study participants were not at
risk of genital injuries because they stated that neither attempted
nor completed vaginal penetration had occurred. Of the remaining
1223 women evaluated for genital injury, 77.5% (n = 948) reported
completed vaginal penetration, 3.3% (n = 40) reported attempted
vaginal penetration and 19.2% (n = 235) suspected sexual assault
but had no clear recollection of the incident.
Overall, genital injury was detected in 22.0% (269/1223)
examined for genital injury. Genital injury was detected in
24.5% (232/948) of those examined for alleged completed vaginal
penetration, 15.0% (6/40) of those following attempted vaginal
penetration and in 13.2% (31/235) of women with suspected sexual
assault but no clear recollection of the type of penetration (Fig. 2).
Of the 71 women with no prior sexual intercourse who reported
completed vaginal penetration, 52.1% (n = 37) had genital injury
and 47.9% (n = 34) did not. The 232 women with a genital injury
following completed vaginal penetration had, on average, 2.4 geni-
tal injuries each, with 50% (n = 117) having only one genital injury.
The most common type of genital injury in women reporting
completed vaginal penetration were lacerations followed by
abrasions; the most common sites with at least one injury were
the posterior fourchette, fossa navicularis, labia minora and
hymen/hymenal remnant (Table 2).
Table 3 details the frequency and odds of genital injury in
948 women reporting completed vaginal penetration with respect
to eight demographic and assault characteristics associated with
genital injury in bivariate analyses (p < 0.25). Logistic regression
modelling determined that four of these eight factors (including
use of sedatives, a history of prior sexual intercourse, time to
examination, number of vaginal penetrants) were independently
associated with genital injury. Sedative use within 6 h of the sexual
assault was protective of genital injury (adjusted OR = 0.3, 95% CI
0.1, 0.7). The factor with the greatest impact on risk of genital injury
was a history of no prior vaginal intercourse (adjusted OR = 4.7, 95%
CI 2.8, 8.1). The odds of observing a genital injury decreased with
increasing time to examination. Factors not related to genital
injury examined in univariate analysis are listed in Table 3
footnotes.
198 R.R. Zilkens et al. / Forensic Science International 275 (2017) 195–202

Total
Total Population
Population of
of Women
Women Alleging
Alleging Sexual
Sexual Assault
Assault
n=1755
n=1755

Excluded
Excluded (n=
(n= 489)
489)
Declined
Declined consent
consent to
to research
research (n=95)
(n=95)
Solely
Solely indecent
indecent assault
assault (n=35)
(n=35)
Report deemed
Report deemed fallacious
fallacious (n=19)
(n=19) Included
Included
Time
Time to
to examination
examination >10
>10 days
days (n=70)
(n=70) n=1266
n=1266
Only reported oral assault (n=15)
Only reported oral assault (n=15)
Only
Only reported vaginal assault with tongue
reported vaginal assault with tongue (n=12)
(n=12)
Did
Did not
not consent
consent to
to genito-anal
genito-anal exam
exam (n=243)
(n=243)

Part
Part 11 Part
Part 22
Genital
Genital Injury
Injury Study
Study Anal
Anal Injury
Injury Study
Study

Excluded:
Excluded: n=43
n=43 Excluded:
Excluded: n=803
n=803
Included:
Included: n=1223
n=1223 Included:
ncluded: n=463
n=463
Not
Not At
At Risk
Risk of
of Not
Not At
At Risk
Risk of
of
At
At Risk
Risk of
of At
At Risk
Risk of
of
Genital
Genital Injury
Injury Anal
Anal Injury
Injury
Genital Injury
Genital Injury Anal Injury
Anal Injury
(no
(no vaginal
vaginal penetration)
penetration) (no
(no anal
anal penetration)
penetration)

Completed
Completed Attempted
Attempted Suspected/
Suspected/ Completed
Completed Attempted
Attempted Suspected/
Suspected/
Vaginal
Vaginal Vaginal
Vaginal No
No Anal
Anal Anal
Anal No
No
Penetration
Penetration Penetration
Penetration Recollection
R eecollection Penetration
Penetration Penetration
Penetration Recollection
Recollection
n=948
n=948 n=40
n=40 n=235*
n=235* n=174
n=174 n=54
n=54 n=235*
n=235*

Genital
Genital Injury
Injury Genital
Genital Injury
Injury Genital
Genital Injury
Injury Anal
Anal Injury
Injury Anal
Anal Injury
Injury Anal
Anal Injury
Injury
Detected
Detected Detected
Detected Detected
Detected Detected
Detected Detected
Detected Detected
Detected
n=232
n=232 (24.5%)
(24.5%) n=6
n=6 (15.0%)
(15.0%) n=31
n=31 (13.2%)
(13.2%) n=47
n=47 (27.0%)
(27.0%) n=5
n=5 (9.3%)
(9.3%) n=14
n=14 (6.0%)
(6.0%)

Fig. 2. Flow diagram of study design showing prevalence of site-specific genito-anal injuries in women by vaginal and anal penetration status.
*235 patients suspected that a sexual assault took place but have no clear recollection of the incident.

Completed vaginal penetration with one, two and three types of
penetrants (e.g. penis, finger, object or hand) was reported in,
respectively, 74.0% (n = 701), 24.9% (n = 236) and 1.2% (n = 11) of
women reporting completed vaginal penetration. The proportion
with a genital injury increased with the use of multiple penetrants
(Table 3). After adjustment for other factors (Table 3), vaginal
penetration by multiple penetrants was associated with a 1.5 fold
increased risk (95% CI 1.1, 2.1) of genital injury compared to women
assaulted with one penetrant.
Frequency of genital injury by type of penetrant was examined
in the 948 women reporting completed vaginal penetration. The
two most common penetrants in the 701 “single” penetrant cases
were penis (n = 550, 78.5%) and finger (n = 133, 19.0%). The
difference in the proportion of genital injury due to penile (126/
550, 22.9%) and finger (22/133, 16.5%) penetration was not
statistically significant (p = 0.110). Vaginal penetration by hand
and object were reported in, respectively, 7 (1.0%) and 11 (1.6%) of
the 701 “single” penetrant cases. The prevalence of genital injuries
due to hand and object penetration was 71.4% (5/7) and 36.4% (4/
11), respectively.
In a sub-group of 807 women with completed vaginal
penetration consenting to both general body and genito-anal
examinations, 69.8% (n = 563) had general body injuries. Women
with a general body injury were more likely to present with a
genital injury (27.9% (157/563) vs 20.9% (51/244), p = 0.037).
Multivariate logistic regression modelling determined that, in
this sub-group, five factors, including (i) a history of prior sexual
intercourse, (ii) general body injury, (iii) number of vaginal
penetrants, (iv) sedative use and (v) time to examination, were
independently associated with genital injury. The factor with the
greatest impact on risk of genital injury was a history of no prior
vaginal intercourse (adjusted OR = 4.4, 95% CI 2.4, 8.0), followed by
the presence of a general body injury (adjusted OR = 1.6, 95%CI 1.1,
2.3) and use of multiple types of penetrants (adjusted OR = 1.5, 95%
CI 1.0, 2.1). Sedative use within 6 h of the sexual assault was
protective of genital injury (adjusted OR = 0.3, 95% CI 0.1, 0.7).
Increasing time to examination was associated with decreasing
genital injury.
In a sub-group analysis of 189 females aged 13–17 years old
examined after completed vaginal penetration, 24.3% (n = 46) had
no history of previous vaginal sex, and 70.4% (n = 133) had been
sexually active prior to the assault (missing information n = 10).
The prevalence of genital injury was significantly higher in the “no
prior sex group” when compared to their counterparts (52.2% (24/
46) vs 19.5% (26/133); p < 0.001).
Of the 46 young females with no history of previous vaginal sex
who reported completed vaginal penetration the site specific
injuries were as follows: hymen 30% (n = 14), labia minora 20%
 R.R. Zilkens et al. / Forensic Science International 275 (2017) 195–202 199

Table 1
Demographic and assault characteristics of 1266 women presenting to the Sexual Assault Resource Centre (Western Australia) and consenting to genito-anal examination
following alleged vaginal and/or anal assault.

Demographic characteristics n % Assault characteristics n %

Age Assailant type
13–14 55 4.3 Stranger 197 15.6
15–19 334 26.4 Intimate partner 201 15.9
20–29 489 38.6 Acquaintance/friend (>24 h) 442 34.9
30–39 230 18.2 Accidental acquaintance (<24 h) 243 19.2
40–49 112 8.9 Unknown (no memory) 106 8.4
>49 46 3.6 Relative 20 1.6
Current mental illness Other 57 4.5
No 763 60.3 Number of assailants
Yes 503 39.7 Single 1,096 86.6
Physical disability Multiple 103 8.1
No 1237 97.7 Uncertain 67 5.3
Yes 29 2.3 Location of assault
Intellectual disability Indoors 922 72.8
No 1218 96.2 Outdoor 300 23.7
Yes 48 3.8 No memory 23 1.8
Any disability No information 21 1.7
No 1196 94.5 Vaginal penetration
Yes 70 5.5 No 43 3.4
Alcohol usea Completed 948 74.9
No 486 38.4 Attempted 40 3.2
Yes (not intoxicated) 202 16.0 Suspected/no recollection 235 18.6
Yes (intoxicated) 566 44.7 Anal penetration
Patient uncertain/missing data 12 0.9 No 803 63.4
Use of sedating agentsb Completed 174 13.7
No 1166 92.1 Attempted 54 4.3
Yes 100 7.9 Suspected/no recollection 235 18.6
Use of stimulantsc Time to examination
No 1209 95.5 <24 h 646 51.0
Yes 57 4.5 24 to <72 h 391 30.9
3 to <5 days 137 10.8
5 to <10 days 92 7.3
a
Alcohol use refers to alcohol consumed in the 6 h period prior to the assault.
b
Non-prescribed sedating agents include cannabinoids, opiates and benzodiazepines.
c
Stimulants include amphetamine, ecstasy and cocaine.

(n = 9), posterior fourchette 15% (n = 7), fossa navicularis 11% (n = 5), Of the 133 young females with a history of previous vaginal sex
clitoris 4% (n = 2), high vagina 4%(n = 2), low vagina 2% (n = 1) who reported completed vaginal penetration the site specific
urethral 2% (n = 1), labia majora 2% (n = 1), cervix 2% (n = 1). There injuries were: hymen 5% (n = 4), labia minora 2% (n = 3), posterior
were no injuries classified as vestibular or perineal in these fourchette 6% (n = 8), fossa navicularis 6% (n = 8), clitoris 2% (n = 3),
46 females. low vagina 2% (n = 2), labia majora 2% (n = 2), vestibule 1% (n = 1).
There were no injuries classified as perineal, high vagina, urethral
or cervical in these 133 females.
Table 2
Frequency of type and site of injury in women reporting completed penetration.
3.2. Anal injury
Genital injury Vaginal Anal injury Anal
penetration penetration Anal injury analysis was confined to 463 women who reported
(N = 948) (N = 174)
either attempted or completed anal penetration or suspected
n (%) n (%) sexual assault but had no clear recollection of the incident. Overall,
Type Type anal injury was detected in 14.3% (66/463) of these women. Anal
Laceration 124 (13.1) Laceration 37 (21.3) injury was detected in 27.0% (47/174) of cases with alleged
Abrasion 104 (11.0 Abrasion 5 (2.9)
completed anal penetration, in 9.3% (5/54) of cases with attempted
Bruise 54 (5.7) Bruise 15 (8.6)
Stab wound 1 (0.1) Stab wound 0 (0.0) anal penetration and in 6.0% (14/235) of women who suspected
Burn 1 (0.1) Burn 0 (0.0) sexual assault but had no clear recollection of the incident (Fig. 2).
The 47 women with an anal injury following completed anal
Site Site penetration had, on average, 2.1 anal injuries each, with 45%
Posterior fourchette 70 (7.4) Perianal 33 (19.0)
Fossa navicularis 64 (6.8) Anus 17 (9.8)
(n = 21) having only one anal injury. The most common type of anal
Labia minora 58 (6.1) Rectum 5 (2.9) injury in women reporting anal penetration were lacerations
Hymen/hymenal remnant 57 (6.0) followed by bruising; the most common site with at least one
Vagina 32 (3.4) injury was the perianal region (Table 2).
Clitoris 10 (1.1)
Completed anal penetration with one, two and three penetrants
Cervix 10 (1.1)
Vestibule 5 (0.5) was reported in, respectively 92.5% (161/174), 6.9% (12/174) and
Urethral 4 (0.4) 0.6% (1/174) of cases reporting completed anal penetration. The
Perineum 4 (0.4) two most common penetrants in “single” penetrant cases were
Mons pubis 2 (0.2) penis (123/161, 76.4%) and finger (32/161, 19.6%) with only
N = number of women; n = number of women with at least one injury. 6 women reporting single penetration with an object. Type of
200 R.R. Zilkens et al. / Forensic Science International 275 (2017) 195–202

Table 3
Frequencies and odds ratios for genital injury according to demographic and assault characteristics in women reporting completed vaginal penetration (n = 948).

All Genital injury x2 Relative odds of genital injury

(N = 948) No (n = 716) Yes (n = 232) Unadjusted Adjusted

n n % N % OR 95%CI OR 95%CI
Age
13–19 311 226 72.7 85 27.3 0.196 1.3 0.9–1.8
20–49 605 468 77.4 137 22.6 1 Ref
>49 32 22 68.8 10 31.2 1.6 0.7–3.4

Mental illnessa
No 580 428 73.8 152 26.2 0.119 1 Ref
Yes 368 288 78.3 80 21.7 0.8 0.6–1.1

Sedating agentsb
No 879 653 74.3 226 25.7 0.002 1 Ref 1 Ref
Yes 69 63 91.3 6 8.7 0.3 0.1–0.6 0.3 0.1–0.7

Prior intercoursec
No 71 34 47.9 37 52.1 <0.001 3.9 2.4–6.4 4.7 2.8–8.1
Yes 840 657 78.2 183 21.8 1 Ref 1 Ref
Not asked 37 25 67.6 12 32.4 1.7 0.8–3.5 1.9 0.9–3.9

Time to examination
<24 h 499 348 69.7 151 30.3 <0.001 1 Ref 1 Ref
24 to <72 h 281 222 79.0 59 21.0 0.6 0.4–0.9 0.6 0.4–0.8
3 to <5 days 100 83 83.0 17 17.0 0.5 0.3–0.8 0.4 0.2–0.7
5 to <10 days 68 63 92.7 5 7.4 0.2 0.1–0.5 0.1 0.1–0.4

Assault site
Indoor 703 537 76.4 166 23.6 0.147 1 Ref
Outdoor 232 167 72.0 65 28.0 1.3 0.9–1.8
Unknown 13 12 92.3 1 7.7 0.3 0.0–2.1

Vaginal penetrantsd
Single 701 544 77.6 157 22.4 0.012 1 Ref 1 Rref
Multiple 247 172 69.6 75 30.4 1.5 1.1–2.1 1.5 1.1–2.1

OR = odds ratio; 95%CI = 95% confidence interval x2 = chi-square p-value.

Factors not related to genital injury examined in univariate analysis (defined as Chi-square p-values >0.25) include: physical and intellectual disability, alcohol use,
stimulating agents, assailant type, number of assailants with sexual contact, anal penetration.
Adjusted odds ratios are adjusted for sedating agents, a history of prior vaginal intercourse, time to examination and number of types of vaginal penetrants. Score test for trend
p = 0.001 provides evidence of a trend in the decreasing odds of genital injury with increasing time to examination.
a
Mental illness, refers to current mental illness.
b
Sedating agents include marijuana, benzodiazepines and heroin.
c
Prior intercourse, refers to vaginal intercourse.
d
Vaginal penetrants refer to number of types of penetrants (e.g. penis, finger, hand, and object).

penetrant in single penetrant cases was not associated with anal
injury (Fisher’s exact p-value 0.584). Single penetrant cases had
less risk of anal injury compared with multiple penetrants (24.8%
(40/161) vs 53.9% (7/13), Fisher’s exact p-value 0.045).
Table 4 gives the frequency and relative odds of anal injury in
women examined following completed anal penetration for the
four factors associated with anal injury in univariate analysis (i.e.
time to examination, age and intellectual disability, number of
penetrants). Only time to examination and number of penetrants
were independently associated with anal injury. Factors not related
to anal injury examined in univariate analysis are listed in Table 4
footnotes.
In a sub-group of 151 women with completed anal penetration
consenting to both general body and genito-anal examinations,
74.2% (112/151) had general body injuries. Women with a general
body injury were more likely to present with an anal injury (31.3%
(35/112) vs 15.4% (6/39), p = 0.055). Multivariate logistic regression
modelling determined that, in this sub-group, three factors,
including (i) number of anal penetrants (ii) general body injury,
and (iii) time to examination, were independently associated with
genital injury. The factor with the greatest impact on risk of anal
injury was the use of multiple types of penetrants (adjusted
OR = 5.0, 95% CI 1.2, 21.0) followed by the presence of a general
body injury (adjusted OR = 2.7, 95%CI 1.0, 7.3). Increasing time to
examination was associated with decreasing anal injury.
4. Discussion
Due to the large number of women in our study we were able to
assess the individual frequencies of genital and anal injury
associated with specific types of sexual assaults. Genital and anal
injuries have been analysed by penetrant type (penis, finger, hand
or object), number of penetrants and completed or attempted
penetration. Separate analyses of genital and anal injuries are not
well documented in the literature.
We examined post-pubertal adolescent and adult female sexual
assault complainants without magnification or genital staining to
detect genital and anal injuries. Comparisons with studies using
colposcopy and/or staining methods are complex, as these studies
frequently report much higher injury rates than those seen by the
‘naked eye’. Of the fifteen [4,7,9–20,23] studies that used
macroscopic examination reviewed by Lincoln et al. [8] in 2013,
only six excluded/separated genital injuries from anal/peri-anal
injuries [4,11,14,15,19,20]. Of these six, only three [4,19,20] in
addition to the Australian study conducted by Lincoln et al. [8]
excluded redness and swelling as injuries and therefore are
 R.R. Zilkens et al. / Forensic Science International 275 (2017) 195–202 201

Table 4
Frequencies and odds ratios for anal injury according to demographic and assault characteristics in women reporting completed anal penetration (n = 174).

All Anal injury x2 Relative odds of anal injury

N = 174 No (n = 127) Yes (n = 47) Unadjusted Adjusted

n N % n % OR 95%CI OR 95%CI
Age
13–19 50 40 80.0 10 20.0 0.057 1 Ref
20–29 54 33 61.1 21 38.9 2.5 1.1–6.2
>29 70 54 77.1 16 22.9 1.2 0.5–2.9

Intellectual disability
No 166 123 74.1 43 25.9 0.134 1 Ref
Yes 8 4 50.0 4 50.0 2.9 0.7–11.9

Time to examinationa
<24 h 85 54 63.5 31 36.5 0.009 1 Ref 1 Ref
24 to <72 h 57 44 77.2 13 22.8 0.5 0.2–1.1 0.5 0.2–1.0
3 to <10 days 32 29 90.6 3 9.4 0.2 0.1–0.6 0.1 0.0–0.5

Anal penetrants
Single 161 121 75.2 40 24.8 0.023 1 Ref 1.0 Ref
Multiple 13 6 46.2 7 53.8 3.5 1.1–11.1 4.0 1.2–13.6

OR = odds ratio; 95%CI = 95% confidence interval x2 = chi-square p-value. Anal penetrants refer to number of types of penetrants (e.g. penis, finger, hand, and object).
Factors not related to anal injury in bivariate analysis (defined as Chi-square p-values >0.25) include: current mental illness, physical disability, alcohol use, sedative use,
stimulants, assailant type, number of assailants with sexual contact, location of assault (indoor versus outdoor), prior anal intercourse.
a
Score test for trend p = 0.002 provides evidence of a trend in the decreasing odds of anal injury with increasing time to examination.

suitable for comparison of genital injury frequency following
completed vaginal penetration with our study. A further four more
recent studies not reviewed by Lincoln, have significant methodo-
logical differences to our study precluding comparisons with our
findings [24–27].
Of the four comparable genital injury studies, the earliest study
(1992) reported a 9% prevalence of genital injury [4], much lower
than our overall prevalence of 22%. This large difference may be
due to the use of forensically trained clinicians in our study, there
being no mention of specific physician training in the US study. The
second study is confined to adolescents stratified by whether they
had ever had prior vaginal sexual intercourse [20]. Genital injury
prevalence in this study in adolescents with no prior vaginal sexual
intercourse was 53%, similar to the 52% in our study. The third
study, by the same Manchester research group [19], reported that
the prevalence of genital injury in adult complainants was 23%, the
same as for women older than 19 years with completed vaginal
penetration in our study. The fourth study, conducted in Queens-
land (Australia) reported a 54% prevalence of genital injuries in 41
women aged 18–44 years, double the prevalence in our 20–49 year
old women [8]. The higher prevalence in the Queensland study
may be because only cases examined within 72 h of the sexual
assault were included whereas our study included women
examined up to 10 days after alleged sexual assault. However,
this cannot be the entire explanation since only 30% of our study
participants presenting within 72 h had genital injuries. Another
factor that may have contributed to the higher genital injury rates
in the Queensland study is that all the women had reported their
sexual assault to the police compared to only 69% of the women in
our study. We postulate that women who report to police may be
motivated to do so because of general body injury and/or genito-
anal injury. We have shown that general body physical injury is
associated with increased risk of genital injury. However one US
study determined that general body injury is not a motivator for
women reporting to the police [28]. Whether this applies to the
Australian setting is not known.
An interesting finding was that although the number of
assailants with sexual contact was not associated with increased
genital and/or anal injury the number of different penetrant types
increased genito-anal injury frequency.
We were able to study anal injury in 463 women who we
considered to be at risk of this type of injury. We established that
27% of women reporting completed anal penetration and 9% of
women reporting attempted anal penetration sustained visible
anal injuries. Half the women examined for anal injury had no clear
recollection of the type of assault and 6% of these women had anal
injuries. Unfortunately, direct comparison of our findings with
other studies is difficult as published data describing the presence
of anal injury following anal penetration is scarce. This is because
most studies only present the findings broadly as genito-anal
injury following anal (and/or vaginal) penetration [9,10,16–18].
While there are two studies that report anal injury prevalence of 8–
11% this is in all examined sexual assault cases regardless of anal
penetration status [29,30].
A Swedish study found that odds of anal injuries were higher in
victims of intimate partners and acquaintances when compared
with victims of strangers [24]. We did not find assailant type to be
associated with anal injury. Indeed, only time to examination,
number of anal penetrants (single vs multiple) and general body
injuries were independently associated with anal injuries in
women reporting completed anal penetration.
4.1. Strengths and limitations
This study has a number of strengths. By including a large
number of study participants we were able to estimate site-specific
injury prevalence estimates by different types of assault. The large
study size also allowed us to use multivariate regression to identify
factors independently associated with injuries. Furthermore, the
exclusive use of a standard protocol by forensically trained
clinicians to examine all study participants provides confidence
in the data quality.
A number of study limitations exist. The study results are not
generalizable to all women following sexual assault for a number
of reasons. Firstly, 5% of women referred to SARC did not consent to
use of their data for research. This group may include some women
incapacitated with severe injuries where consent was unable to be
obtained. Also 11% of those referred within 10 days of the sexual
assault declined all forensic examinations. Secondly, sexual assault
is under-reported to police and health services [31,32]. Although
202 R.R. Zilkens et al. / Forensic Science International 275 (2017) 195–202
there is evidence that injury does not motivate US women to report
to police [28], it is possible that Australian women referred to SARC
may have more injuries than those not referred to SARC. If so, then
our injury frequencies may over estimate population rates.
Alternatively, some women with more severe non-life threatening
injuries sustained from intimate partner assaults may be reluctant
to attend SARC causing us to underestimate injury prevalence.
Women with life-threatening injuries treated in hospital may not
be identified as having been sexually assaulted.
As with all studies examining associations between sexual
assault and injury the history of the assault was subjective and
reliant on the patient. In an attempt to minimise error we excluded
women with confirmed false allegations.
We have deliberately presented our results in such a way that
places our findings into context with the self-reported history of
the women. This has been done to enable clinicians attending court
to relate their particular case findings to cases with similar
characteristics.
5. Conclusion
The prevalence of genital injury following non-consensual
vaginal penetration is similar to the prevalence of anal injury
following non-consensual anal penetration. Genital and anal injury
were both found in women who suspected a sexual assault but had
no clear recollection of the incident. We therefore recommend that
these women also require health and forensic service provision.
While just over half of the women with no prior sexual intercourse
had genital injuries after completed vaginal penetration during the
sexual assault, almost a half did not. As with all prior published
research in this area, a large number of females alleging recent
sexual assault have no genito-anal injuries and the absence of
genito-anal evidence does not exclude sexual assault.
Disclosure of interests
The authors declare that they have no competing interests.
Contribution to authorship
The study was designed by RZ, DS, MK, SM, JS and MP. RZ analysed
the data, and interpreted the results with help from DS, MK and MP.
RZ wrote the article and all authors commented on and contributed
to the revision of drafts and the final version of the article.
Funding
R. Zilkens and S. Mukhtar were funded by the Lotterywest Data
Linkage Infrastructure Grant (Grant Application number
#421003516).
Acknowledgement
The authors would like to acknowledge Ms Rae Cummins,
administrative Assistant, SARC Medical and Forensic Service, SARC
Western Australia for her assistance with the project.
References
[1] M.J. McGregor, J. Du Mont, T.L. Myhr, Sexual assault forensic medical
examination: is evidence related to successful prosecution, Ann. Emerg.
Med. 39 (2002) 639–647.
[2] D. Lievore, Prosecutorial Decisions in Adult Sexual Assault Cases: An Australian
Study, Australian Governement’s Office of the Status of Women, Australian
Institute of Criminology, Canberra, 2004.
[3] R. Kingsnorth, D. Cummings, J. Lopez, J. Wentworth, Criminal sentencing and
the court probation office: the myth of individualized justice revisited, Justice
Syst. J. 20 (1999) 255–273.
[4] B. Rambow, C. Adkinson, T.H. Frost, G.F. Peterson, Female sexual assault:
medical and legal implications, Ann. Emerg. Med. 21 (1992) 727–731.
[5] C. Spohn, D. Beichner, E. Davis-Frenzel, Prosecutorial justifications for sexual
assault case rejection: guarding the “gateway to justice”, Soc. Probl. 48 (2001)
206–235.
[6] R. Jewkes, N. Christofides, L. Vetten, R. Jina, R. Sigsworth, L. Loots, Medico-legal
findings, legal case progression, and outcomes in South African rape cases:
retrospective review, PLoS Med. 6 (2009).
[7] L. Bowyer, M.E. Dalton, Female victims of rape and their genital injuries, Br. J.
Obstet. Gynaecol. 104 (1997) 617–620.
[8] C. Lincoln, R. Perera, I. Jacobs, A. Ward, Macroscopically detected female genital
injury after consensual and non-consensual vaginal penetration: a prospective
comparison study, J. Forensic Leg. Med. 20 (2013) 884–901.
[9] C.M. Palmer, A.M. McNulty, C. D’Este, B. Donovan, Genital injuries in women
reporting sexual assault, Sex. Health 1 (2004) 55–59.
[10] N.F. Sugar, D.N. Fine, L.O. Eckert, Physical injury after sexual assault: findings of
a large case series, Am. J. Obstet. Gynecol. 190 (2004) 71–76.
[11] J.E. Tintinalli, M. Hoelzer, Clinical findings and legal resolution in sexual
assault, Ann. Emerg. Med. 14 (1985) 447–453.
[12] P.S. Cartwright, R.A. Moore, J.R. Anderson, D.H. Brown, Genital injury and
implied consent to alleged rape, J. Reprod. Med. 31 (1986) 1043–1044.
[13] P.S. Cartwright, Factors that correlate with injury sustained by survivors of
sexual assault, Obstet. Gynecol. 70 (1987) 44–46.
[14] T. Manser, Findings in medical examinations of victims and offenders in cases
of serious sexual offences: a survey, Police Surg. 38 (1991) 4–27.
[15] S.M. Ramin, A.J. Satin, I.C. Stone Jr., G.D. Wendel Jr., Sexual assault in
postmenopausal women, Obstet. Gynecol. 80 (1992) 860–864.
[16] J. Wiley, N. Sugar, D. Fine, L.O. Eckert, Legal outcomes of sexual assault, Am. J.
Obstet. Gynecol. 188 (2003) 1638–1641.
[17] M. Hilden, B. Schei, K. Sidenius, Genitoanal injury in adult female victims of
sexual assault, Forensic Sci. Int. 154 (2005) 200–205.
[18] P.A. Crane, Predictors of injury associated with rape, J. Forensic Nurs. 2 (2006)
75–83 9.
[19] I. McLean, S.A. Roberts, C. White, S. Paul, Female genital injuries resulting from
consensual and non-consensual vaginal intercourse, Forensic Sci. Int. 204
(2011) 27–33.
[20] C. White, I. McLean, Adolescent complainants of sexual assault; injury
patterns in virgin and non-virgin groups, J. Clin. Forensic Med. 13 (2006) 172–
180.
[21] S.A. Mukhtar, D.A. Smith, M.A. Phillips, M.C. Kelly, R.R. Zilkens, J.B. Semmens,
Capturing sexual assault data: an information system designed by forensic
clinicians and healthcare researchers, Health Inf. Manag. J. (2017) 1–10.
[22] D.W. Hosmer, S. Lemeshow, R.X. Sturdivant, Applied Logistic Regression, John
Wiley & Sons, Inc, New York, NY, USA, 2013.
[23] L. Morgan, A. Dill, J. Welch, Sexual assault of postmenopausal women: a
retrospective review, BJOG 118 (2011) 832–843.
[24] A.S. Moller, T. Backstrom, H.P. Sondergaard, L. Helstrom, Patterns of injury and
reported violence depending on relationship to assailant in female Swedish
sexual assault victims, J. Interpers. Violence 27 (2012) 3131–3148.
[25] B.S. Astrup, P. Ravn, J.L. Thomsen, J. Lauritsen, Patterned genital injury in
cases of rape—a case-control study, J. Forensic Leg. Med. 20 (2013) 525–
529.
[26] R. Karanfil, A. Keten, C. Zeren, M.M. Arslan, A. Eren, Evaluation of sexual
assaults in Turkey, J. Forensic Leg. Med. 20 (2013) 404–407.
[27] M.L. Larsen, M. Hilden, O. Lidegaard, Sexual assault: a descriptive study of
2500 female victims over a 10-year period, BJOG 122 (2015) 577–584.
[28] J.S. Jones, C. Alexander, B.N. Wynn, L. Rossman, C. Dunnuck, Why women don’t
report sexual assault to the police: the influence of psychosocial variables and
traumatic injury, J. Emerg. Med. 36 (2009) 417–424.
[29] R. Jina, R. Jewkes, L. Vetten, N. Christofides, R. Sigsworth, L. Loots, Genito-
anal injury patterns and associated factors in rape survivors in an urban
province of South Africa: a cross-sectional study, BMC Womens Health 15
(2015) 29.
[30] C. Grossin, I. Sibille, G. Lorin de la Grandmaison, A. Banasr, F. Brion, M.
Durigon, Analysis of 418 cases of sexual assault, Forensic Sci. Int. 131 (2003)
125–130.
[31] D. Lievore, Non-reporting and Hidden Recording of Sexual Assault: An
International Literature Review, Commonwealth office of the status of women,
Canberra, 2003.
[32] J. Mouzos, T. Makkai, Women’s Experience of Male Violence: Findings From the
Australian Component of the International Violence Against Women Survey
(IVAWS) Research and Public Policies Series No 56, Australian Institute of
Criminology, 2004.