Original communication
Macroscopically detected female genital injury after consensual and non-consensual vaginal penetration: A prospective comparison study

https://doi.org/10.1016/j.jflm.2013.06.025Get rights and content

Abstract

Objective

The aim of this study was to compare the prevalence, type and pattern of macroscopically detected female genital injury after consensual and non-consensual vaginal penetration to further an understanding of the forensic significance of genital injury in women reporting sexual assault. A secondary aim was to identify any effect of a range of possible variables upon the likelihood of genital injury resulting from vaginal penetrative sexual intercourse.

Study design

Two groups of reproductive age women (aged 18–45 years) were prospectively recruited within 72 h of a single episode of vaginal penetrative sex, and macroscopically examined for the presence of bruises, abrasions and lacerations at twelve external and internal genital sites. Forty one women who presented for forensic examination after reporting a sexual assault to police were recruited to the non-consensual group and 81 women who presented for routine cervical screening or with sexual health concerns to a primary health care service to the consensual group. Each group was examined by a different group of doctors, all of whom were experienced in both forensic genital examination and gynaecological examination of healthy and diseased sexually active women. Data collection and examination protocols were the same for both groups.

Results

The key finding was a statistically significant difference in genital injury prevalence between women who were vaginally penetrated non-consensually and consensually; 53.7% of the non-consensual group (22/41) and 9.9% of the consensual group (8/81) were found to have at least one genital injury [OR 10.57, CI (4.07, 27.42), p < 0.00001]. Penetration with finger/s and possible pre-existing genital ‘infection’ were found to be significantly associated with the presence of injury in the univariate analysis after adjusting for consent. Logistic regression demonstrated that women penetrated without consent were 19.5 times more likely to sustain at least one genital injury, than those penetrated consensually [OR 19.53, CI (6.03, 63.24)] and that a penetration scenario that included finger/s was 4.2 times more likely to result in at least one genital injury than penetration without finger involvement [OR 4.25, CI (1.42, 12.78)], when controlling for other variables in the model. Whilst a comparatively low injury prevalence in the consensual group limited interpretation, results revealed possible differences in genital injury typology and pattern resulting from non-consensual and consensual vaginal penetration. Lacerations were seen after both consensual and non-consensual vaginal penetration, while abrasions and bruises were seen exclusively in the non-consensual group.

Conclusion

This study demonstrated a significant consent group difference in genital injury prevalence and the highest macroscopically detected genital injury prevalence rate resulting from non-consensual vaginal penetration identified to date. Results also indicate that vaginal penetration with finger/s increases the likelihood of sex-related injury. The difference in type of injury sustained as a result of non-consensual and consensual vaginal penetration was an unexpected finding, and warrants further investigation. These results highlight the importance of a standardised means of detecting genital injury based on consistent injury definitions, examination protocols, and examiner experience and suggest that macroscopic genital examination may be uniquely placed to detect consent group differences in injury typology and pattern if they exist.

Introduction

Despite the frequent focus of sexual assault trials upon the issue of consent, the vast majority of female genital injury research to date has concentrated exclusively on injury arising non-consensually, and is therefore of little value in the interpretation of its significance in relation to consent. To date, considerable effort has been exerted investigating genital injury in sexual assault complainants using a range of different genital examination methods. Most of this research has been done by retrospective record review with outcomes which are difficult to interpret because of methodological inconsistencies. The relative paucity of robust genital injury studies in consensually sexually active women is a serious omission from the literature.

While most genital injury research is confined to the forensic field, useful information can be derived from other ‘non-forensic’ sources. Studies of genital injury after consensual sex can be found in sexual health and reproductive medicine, investigating the relationship of injury to risk of sexually transmitted infection1, 2 or the use of various contraceptive measures.3, 4, 5, 6 Although the interpretive value of these studies may be limited to some degree by the lack of a forensic approach to the genital examination, they cannot be ignored given the need for comparative data about genital injury resulting from consensual sex.

In Australia, adult sexual assault complainants are forensically examined without magnification or genital staining; clinicians routinely perform ‘naked eye’ or macroscopic genital examinations to detect injury. For this reason, the results of non-consensual genital injury studies using colposcopy and/or staining methods have little relevance for medical witnesses tasked with interpreting genital examination findings for the courts in this country. However, because macroscopic genital examination is the routine method for examining healthy, consensually sexually active women in the primary health care setting, it is the only genital examination method with a readily available means of collecting comparative data in relation to genital injury resulting from consensual sex. As such, macroscopic genital examination studies have the greatest potential to assist the courts with respect to the issue of genital injury and consent.

A total of 85 papers citing female genital injury prevalence data were identified by systematic PubMed and National Library of Medicine (US) searches; study methodology and outcomes were reviewed and key features summarised to assess interpretive value. Seventy papers considered injury due to non-consensual sex, 8 injury due to consensual sex and 7 compared injury in non-consensual and consensual groups. The primary aim of nine of the 70 non-consensual sex studies was to explore the relationship between genital injury and legal outcome. Two important papers which did not provide genital injury prevalence data were considered separately; one explored changes in patterns of genital injury resulting from consensual sex over time7 and the other, a consent group comparison study, investigated the possibility of a predictive model for genital injury and consent.8

To assist interpretation, papers were first categorised by genital examination protocol used. Many studies lacked clarity about the methods by which injuries were detected and the order in which they were performed if more than one examination method was used. Because detection of genital abrasions and lacerations is enhanced by toluidine blue staining while bruise detection is reduced,9 the timing of genital stain application is important and has the potential to affect injury detection rates. Colposcopic studies did not always specify the magnification used and two studies used colposcopic equipment during genital examination solely for ‘photo-documentation’ of macroscopically detected injury.10, 11 Five different genital examination protocols were identified as follows;

  • i.

    Macroscopic.

  • ii.

    Staining and macroscopic.

  • iii.

    Colposcopic.

  • iv.

    Staining and colposcopic.

  • v.

    Combination of macroscopic, colposcopic and staining.

The following methodological criteria were then used to select 46 key studies considered to be of greater interpretive value;

  • i.

    Prospective design.

  • ii.

    Penetration to examination time (PET) of less than 72 h.

  • iii.

    Exclusion of prepubertal and postmenopausal subjects.

  • iv.

    Exclusion of non-specific examination findings ‘redness’ and ‘swelling’.

  • v.

    Exclusion of anal penetration scenario and anal/perianal examination findings.

No study met all five criteria; however all met at least one and most at least two. Table 1 summarises this information for 39 key studies providing non-consensual genital injury prevalence data by method of examination11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48 and Table 2 does the same for 14 key studies citing consensual genital injury prevalence data.1, 3, 4, 9, 23, 25, 26, 27, 34, 39, 44, 49, 50, 51 Seven consent group comparison studies cited both consensual and non-consensual genital injury rates23, 25, 26, 27, 34, 39, 44 and therefore appear in both tables.

Table 3 summarises the ranges of genital injury prevalence data found as a result of non-consensual and consensual sex, for each of the five identified examination protocols in the 46 ‘key’ studies. Significant variation in injury prevalence can be seen between studies, even when using the same genital examination protocols. Injury prevalence ranges appear similar in both consent groups using most examination protocols, with the possible exception of the staining/macroscopic examination studies. This may relate to the use of differing study methodology for consent groups in the majority of these studies. Only 7 studies investigated both consent groups simultaneously; however, not all applied the same methodology to both consent groups.23, 44

Table 4 lists features of the seven consent group comparison studies citing injury prevalence data published to date. Of the body of work addressing this subject, it is these studies which provide the most valuable contribution to the debate about the significance of sex-related genital injury in relation to consent.

Although the need for comparative genital injury data after non-consensual and consensual sex was recognised by researchers investigating the use of genital staining techniques in detecting injury as early as the 1980s,25, 26, 27 these relatively small studies utilising toluidine blue, were never progressed with larger sample sizes. They did, however, provide the first evidence of a difference in genital injury prevalence after consensual and non-consensual sex.

Lauber's study with 22 participants in each consent group found 40% non-consensual vs 4.5% consensual with injury,25 and McCauley's 1987 study with 24 in the non-consensual group and 48 in the consensual group identified 58% non-consensual vs 10% consensual with injury.27 Both studies were prospective in design with consensual group women recruited from emergency department or outpatient clinics at the time of presentation, if they reported consensual sex within the previous 48 h. While participant age and penetration scenario was unclear in Lauber's study, McCauley's participants were aged 19–55 yrs and were all penetrated vaginally. Both studies examined the external genitalia only.

McCauley published a similar study in paediatric and adolescent subjects the previous year.26 Her adolescent participants were recruited prospectively in the same way with 25 in each consent group; age range was 11–18 years and all were examined within 48 h of vaginal penetration. Of note is that genital injury was macroscopically detected after staining in 28% of each consent group; that is, despite finding significantly different consent group injury prevalence rates in her adult study, there was no difference in injury prevalence between her adolescent consent groups.

Later, those pioneering the use of colposcopy in the forensic examination of sexual assault complainants attempted to compare genital injury rates in different consent groups,34, 39 but recruited their consensual subjects from those presenting to sexual assault services. Slaughter's34 consensual group comprised 75 subjects who had initially presented to a sexual assault service alleging recent non-consensual sex, but who later withdrew their allegations and said that sexual intercourse had been consensual rather than non-consensual. Jones's 2003 study39 recruited 51 adolescents to a consensual group from those brought to a sexual assault service by their parents for statutory reasons after sex which the adolescents themselves viewed as consensual. The uncertainty as to consent associated with this means of recruitment to a consensual group is problematic and compromises interpretation. Both studies examined subjects colposcopically after genital staining with toluidine blue; Slaughter's study population were aged 11–85 years and Jones's 13–17 years. Slaughter found genital injury in 89% of non-consensual and 11% of consensual subjects. As was the case with the adolescent staining/macroscopic consent group comparison study described earlier,26 Jones identified similar rates of genital injury in his adolescent non-consensual and consensual groups; 85% non-consensual vs 73% consensual subjects with injury.

More recently, Anderson44 compared genital injury detected colposcopically after toluidine blue staining in 46 consensual subjects recruited prospectively by advertisement, with data from retrospective record review of 56 non-consensual cases. Women aged 16–54 years were examined within 24 h of sexual intercourse, presumed to be penile vaginal penetration although not specifically stated; six external and internal sites were examined for bruises, abrasions and lacerations. She found little difference in genital injury prevalence between consent groups; 32.1% non-consensual vs 30.4% consensual subjects with injury.

Fortuitously for those jurisdictions where staining or colposcopic examination is not used routinely for sexual assault forensic examinations, the most recently published consent group comparison study23 has been done using only macroscopic genital examination. McLean's study, like Anderson's,44 looked at genital injury in a prospectively recruited consensual group and compared findings with data from retrospectively reviewed sexual assault complainant records. Prospective consensual group recruitment was achieved by flyer sent out with a Pap smear reminder letter. Participants in this study were aged 18 years and over, with no upper age limit specified; 8% of 500 non-consensual subjects and 28% of 68 consensual subjects were aged over 45 years. Examination for bruises, abrasions and lacerations at five external and internal genital sites within 48 h of penile vaginal penetration revealed 22.8% of non-consensual cases and 5.9% of consensual cases with injury.

For some researchers, the lack of a difference between consent group injury prevalence appears to have been the catalyst for a closer look at genital injury typology and patterns, in an effort to distinguish between injuries seen as a result of consensual and non-consensual sex. Three consent group comparison studies using examination enhancement techniques, identified little difference in genital injury prevalence rate between groups26, 39, 44; however, two of these focused exclusively on adolescent populations.26, 39

Whilst not the first to consider the concept of a ‘consent group specific’ genital injury pattern,33, 52, 53, 54 Anderson's study of reproductive age women which identified similar consent group injury prevalence rates, was the first to attempt a predictive model for ‘consent’ based on genital injury pattern using a logistic regression model. She claimed that 85% of non-consensual cases and 90% of consensual cases could be correctly identified by considering ‘total surface area of injury’ and ‘number of sites with tears, bruises, abrasions and redness’, while controlling for time from penetration to examination.8 Furthermore, ‘total injury surface area’ and ‘number of sites with redness and bruises’ were found to be ‘individually predictive’ for consent. Her efforts were, however, hampered by small sample sizes with only 40 subjects in each consent group, and by the inclusion of colposcopically-detected ‘redness’ as a positive injury finding. This non-specific genital finding was seen more frequently and at a greater number of sites in the consensual than the non-consensual group. Her inclusion of ‘total surface area of injury’ as an injury pattern variable also compromised results by failing to acknowledge the variation in morphology of bruises, abrasions and lacerations.

To date, researchers have employed a broad range of variables when referring to the concept of ‘genital injury pattern’; injury type, genital site,33, 34 injury severity,19, 55, 56, 57, 58, 59 number of injuries, number of sites injured, number of sites with a given type of injury, and injury size or total surface area of injury.7, 8 Anderson's attempt at a predictive model for consent using injury pattern variables, although not without problems, is a valuable contribution to research in the area, not only for proposing a statistical means of approaching the question of consent, but also because it clearly identifies the problem with defining ‘redness’ as injury in the genital area. It is, however, clear that larger, more methodologically robust studies are needed to determine whether such a model might be achievable in the future.

Whilst a considerable body of work exists in relation to sex-related female genital injury, methodological variation has rendered much of it of little interpretive value. The inclusion of redness and swelling, anal findings, adolescent or postmenopausal subjects, different penetration scenarios and long penetration to examination times distort research outcomes and contribute to inconsistencies in injury prevalence results. Failure to exclude more than one recent episode of sexual penetration, and any other recent penetrative acts that may have resulted in injury, such as the insertion of tampons,1, 60, 61, 62, 63, 64, 65, 66, 67 intra-vaginal contraception,1, 68 or even a speculum1, 27, 41 may affect study outcomes. Lack of information about factors with the potential to affect whether injury will occur or not, such as skin pigmentation,43, 69, 70, 71, 72 use of condoms and/or lubricant,3, 4 use of hormonal preparations, previous vaginal deliveries or genital surgery, and genital infection or inflammation at the time of penetration, is less than ideal.

More recently, genital injury studies have sought to address some of these issues by using prospective study design, standardised pathological injury definitions, clearly defined examination protocols, and acknowledging that genital tissue is subject to hormonal effects on genital tissue by considering adolescent and postmenopausal women separately, significantly improving the ability to interpret study findings.

The considerable challenges involved in recruiting subjects for examination after consensual sex, have led to a number of creative approaches in this respect. Uncertainty about consent status with early studies using sexual assault complainants who ‘withdrew’ allegations34 prompted researchers to seek volunteers to attend for genital examination after an episode of consensual sex, solely for the purpose of the study51 or when they attended for a routine Pap smear.23 It is possible that consensual sex that is, in effect ‘pre-meditated’ or carried out with the knowledge of subsequent participation in research, may differ from spontaneous sex and as such, may alter outcomes. Ideally, consent group comparison studies should prospectively recruit women to both groups at the time of presentation for genital examination, following the index episode of sexual intercourse rather than prior to it.

The legal implications of research outcomes in this area demand a robust methodological approach. To address the need for reliable comparative information about female genital injury resulting from consensual and non-consensual vaginal penetration, and ensure relevance to Australian jurisdictions, this study was therefore designed to incorporate the following elements;-

  • i.

    Prospectivity and recruitment of all participants at the time of presentation for genital examination.

  • ii.

    All examining clinicians experienced in examination of normal, healthy and diseased female genitalia as well as forensic sexual assault examinations.

  • iii.

    All participants of reproductive age without heavily pigmented skin.

  • iv.

    The same information about a range of possible confounding variables sought from participants at the time of examination.

  • v.

    All genital examinations performed within 72 h of a single episode of vaginal penetrative sex, consensual or non-consensual in nature.

  • vi.

    The same macroscopic examination protocol used for all participants

Section snippets

Aims

The specific aim of this study was to compare the prevalence, type, and pattern of genital injury found in two cohorts of reproductive age women; a ‘consensual group’ presenting for genital examination in a primary care setting who had had recent consensual vaginal penetrative sex and a ‘non-consensual group’ attending for a forensic examination after reporting recent sexual assault involving vaginal penetration to the police. A secondary aim was to identify any effect of a range of possible

Participant characteristics

A total of 122 participants were recruited to the study; 81 to the consensual group and 41 to the non-consensual group. Comparison of consent groups is summarised in Table 5.

Mean age for the non-consensual group was slightly younger (27.51 yrs) than that of the consensual group (31.05 yrs). Three age categories within the specified participant age range of 18–45 years were considered; 18–21 years, 22–29 years and 30–45 years; there were similar proportions of women aged 22–29 years and 30–45

Genital injury prevalence and consent

The key finding in this study was a significant difference in genital injury prevalence between women who were vaginally penetrated non-consensually and consensually; 53.7% of the non-consensual group vs 9.9% of the consensual group had any genital injury. The only other study comparing genital injury in consent groups using macroscopic examination, found 22.8% of non-consensual cases and 5.9% of consensual cases to have any genital injury.23 All examinations in this UK study by McLean, with a

Conclusion

This study has identified useful information in relation to macroscopically detected genital injury, vaginal penetration and consent. Genital injury was significantly more likely to occur with non-consensual than consensual penetration, with penetration involving finger/s and when pre-existing genital ‘infection’ was likely. When controlling for finger/s penetration and pre-existing infection, injury was approximately twenty times more likely to occur if penetration was non-consensual than if

Ethical approval

This research was conducted as the basis for a Doctor of Philosophy thesis at the University of Western Australia (UWA) and as such, was approved by the UWA Human Research Ethics Committee (HREC), which is bound by the National Health and Medical Research Council (NHMRC) Ethics Guidelines. In addition, approval for the study was granted by each hospital and primary health care service at which examinations were conducted.

Funding

The first six months of this research project were supported by a National Health and Medical Research Council (NHMRC) Postgraduate Research Scholarship and an Amy and Athelstan Saw Medical Research Fellowship grant.

Conflict of interest

None declared.

Acknowledgements

The significant efforts of all recruiting doctors, and the assistance of Dr Pamela McBride with study design and Peter Black with WAQID development are gratefully acknowledged.

References (101)

  • C. White et al.

    Adolescent complainants of sexual assault: injury patterns in virgin and non-virgin groups

    J Clin Forensic Med

    (2006)
  • I. McLean et al.

    Female genital injuries resulting from consensual and non-consensual vaginal intercourse

    Forensic Sci Int

    (2011)
  • J. McCauley et al.

    Toluidine blue in the corroboration of rape in the adult victim

    Am J Emerg Med

    (1987)
  • R.B. Baker et al.

    Relationship of genital injuries and age in adolescent and young adult rape survivors

    J Obstet Gynecol Neonatal Nurs

    (2008)
  • W. Maguire et al.

    Injury in adult female sexual assault complainants and related factors

    Eur J Obstet Gynecol Reprod Biol

    (2009)
  • C. O'Brien

    Improved forensic documentation of genital injuries with colposcopy

    J Emerg Nurs

    (1997)
  • L.C. Lenahan et al.

    Colposcopy in evaluation of the adult sexual assault victim

    Am J Emerg Med

    (1998)
  • L. Slaughter et al.

    Colposcopy to establish physical findings in rape victims

    Am J Obstet Gynecol

    (1992)
  • L. Slaughter et al.

    Patterns of genital injury in female sexual assault victims

    Am J Obstet Gynecol

    (1997)
  • J.A. Adams et al.

    Adolescent sexual assault: documentation of acute injuries using photocolposcopy

    J Pediatr Adolesc Gynecol

    (2001)
  • J.S. Jones et al.

    Significance of toluidine blue positive findings after speculum examination for sexual assault

    Am J Emerg Med

    (2004)
  • J.S. Jones et al.

    Comparison of sexual assaults by strangers versus known assailants in a community-based population

    Am J Emerg Med

    (2004)
  • M.S. Sommers et al.

    The effects of age and ethnicity on physical injury from rape

    J Obstet Gynecol Neonatal Nurs

    (2006)
  • S. Stears et al.

    Anogenital injury in adolescent sexual assault: analysis of 1,024 cases using colposcopy and nuclear staining [abstract]

    Ann Emerg Med

    (2008)
  • J.G. Jones et al.

    Sexual assault in postmenopausal women: epidemiology and patterns of genital injury

    Am J Emerg Med

    (2009)
  • A.J. Rogers et al.

    The forensic rape examination: is colposcopy really necessary? [abstract]

    Ann Emerg Med

    (2008)
  • M.S. Sommers et al.

    Forensic sexual assault examination and genital injury: is skin color a source of health disparity?

    Am J Emerg Med

    (2008)
  • B.S. Astrup et al.

    Nature, frequency and duration of genital lesions after consensual sexual intercourse – implications for legal proceedings

    Forensic Sci Int

    (2012)
  • D. Alempijevic et al.

    Severity of injuries among sexual assault victims

    J Forensic Leg Med

    (2007)
  • M.J. McGregor et al.

    Sexual assault forensic medical examination: is evidence related to successful prosecution

    Ann Emerg Med

    (2002)
  • J.G. Jones et al.

    Genital and anal injuries requiring surgical repair in females less than 21 years of age

    J Pediatr Adolesc Gynecol

    (2008)
  • H. Wessells et al.

    Penile and genital injuries

    Urol Clin North Am

    (2006)
  • K.F. Barrett et al.

    Tampon-induced vaginal and cervical ulceration

    Am J Obstet Gynecol

    (1977)
  • A.J. Nordin et al.

    Tampon-induced vaginal bleeding presenting as intermenstrual bleeding

    Int J Gynaecol Obstet

    (1995)
  • D. Raudrant et al.

    Study of the vaginal mucous membrane following tampon utilisation; aspect on colposcopy, scanning electron microscopy and transmission electron microscopy

    Eur J Obstet Gynecol Reprod Biol

    (1989)
  • W. Bounds et al.

    Preliminary report of unexpected local reactions to a progestagen-releasing contraceptive vaginal ring

    Eur J Obstet Gynecol Reprod Biol

    (1993)
  • L.M. Hopkins et al.

    Racial/ethnic differences in perineal, vaginal and cervical lacerations

    Am J Obstet Gynecol

    (2005)
  • C. Rechtin et al.

    The effects of skin pigmentation on the detection of genital injury from sexual assault: a population-based study [abstract]

    Ann Emerg Med

    (2009)
  • D. Howard et al.

    Differences in perineal lacerations in black and white primiparas

    Obstet Gynecol

    (2000)
  • L. Rossman et al.

    Genital trauma associated with forced digital penetration

    Am J Emerg Med

    (2004)
  • E.A. Sturgiss et al.

    Characteristics of sexual assaults in which adult victims report penetration by a foreign object

    J Forensic Leg Med

    (2010)
  • L.O. Eckert et al.

    Factors impacting injury documentation after sexual assault: role of examiner experience and gender

    Am J Obstet Gynecol

    (2004)
  • J. Stacey et al.

    Anogenital injury following sexual assault in women after recent consensual intercourse [abstract]

    Ann Emerg Med

    (2007)
  • B.M. Peckham et al.

    Focal vulvitis: a characteristic syndrome and cause of dyspareunia

    Am J Obstetrics Gynecol

    (1986)
  • L. Rossman et al.

    Accuracy and reliability of photo-colposcopy in the documentation of anogenital injuries after sexual assault (abstract)

    Ann Emerg Med

    (2005)
  • J.A. Adams et al.

    Signs of genital trauma in adolescent rape victims examined acutely [abstract]

    J Pediatr Adolesc Gynecol

    (2000)
  • P.A.W. Brennan

    The medical and ethical aspects of photography in the sexual assault examination: why does it offend?

    J Clin Forensic Med

    (2006)
  • B. Woolley et al.

    Comparative analysis of incapacitated versus forcible sexual assault in a community-based population [abstract]

    Ann Emerg Med

    (2007)
  • A. Oostema et al.

    Alcohol-related sexual assault victimization among adolescents [abstract]

    Ann Emerg Med

    (2009)
  • J. Hammerstein et al.

    Alcohol consumption and adult sexual assault in a community-based population of women

    Ann Emerg Med

    (2004)
  • Cited by (0)

    View full text