Normal Findings – Anatomy and Terminology
For professionals involved in the evaluation of children for possible sexual abuse, it is crucial to develop a thorough understanding of the appearance of the anogenital anatomy both in abused and non-abused children. It is difficult to describe a finding as abnormal without fully understanding the range of normal at each developmental level. The increasing knowledge on medical diagnosis in sexual abuse has gone from early imprecise and misleading work [Cantwell, 1983, 1987] to detailed macrophotographic analysis of subtle anogenital findings [Kerns et al., 1992].

As a result of this evolution, the range of findings now interpreted as normal or unspecific (abnormal but not related to abuse) has continuously increased while the number of findings which clearly indicate abusive trauma has decreased [Kerns, 1998; McCann, 1998; Ricci, 1998; Berenson, 1998; Adams, 2001]. In the first decade of medical CSA research, many previously undescribed findings were interpreted as related to trauma in the light of a child’s history suggesting abuse. Many of these findings later were demonstrated in studies of non-abused children, mainly by McCann and Berenson [McCann et al., 1989, 1990b; Berenson et al., 1991, 1992, 1993; Berenson, 1993, 1995]. Although there are methodological considerations and limitations to non-abuse studies, this work has greatly enhanced our knowledge.

A main concern is the question if the screening tools are specific enough to exclude undisclosed sexual abuse (for discussion of limitations and methods of research, see Berenson [1998] and Ricci [1998]).

The evaluation and description of anogenital findings should be done on the basis of a descriptive and standardized terminology of normal structures.

Clinicians involved in describing findings of abused children should speak the same language and describe normal variations and pathologic changes in the same terminology. Although specialists in pediatric and adolescent gynecology are expected to be familiar with the specific terminology, it is worthwhile to have a look at the concepts developed in the recent child abuse literature [APSAC, 1995; Pokorny, 2000].

The documentation in the medical record should be as specific as possible and avoid general terms as ‘normal genitalia’, not reflecting the great variety of normal. ‘Vulva’ and ‘pudenda’ also lack specificity. ‘Virgo intacta’, ‘virginal introitus’, ‘marital hymen’, ‘gaping vulva’ or ‘enlarged vaginal opening’ are insufficient, imprecise and not descriptive terms and should be avoided.

The term ‘posterior fourchette’ may not be familiar in Europe. It describes the posterior joining of the labia minora (frenulum labiorum pudendi minorum), an area which is especially vulnerable to trauma by attempted penile penetration. Sometimes the area from the frenulum to the commisura posterior (posterior joining of the labia majora) is also subsumed under the expression ‘posterior fourchette’. The description of genital findings should begin with the external, superficial structures and move on to the inner structures according to the conduct of the examination.

Much of medical research on genital changes in CSA has focused on the hymen. This was especially true in the first decade of research in the 1980s, when considerable efforts were made to attribute the probability of sexual abuse according to measurements of the hymenal orifice’s transverse diameter [Cantwell, 1983, 1987]. Paradise [1989] summarized the debate at the end of the decade in a report titled ‘Predictive accuracy and the diagnosis of sexual abuse: A big issue about a little tissue.’ She pointed out that ‘the diagnosis of sexual abuse inevitably rests not on a genital measurement but on descriptive statements made by a child’.

The size of the hymenal opening increases with age and depends on a wide variety of factors like examination position, examination technique, amount of traction used and the degree of relaxation of the child, thus potentially varying within the same examination. Genital measurements have considerably lost significance in the evaluation of children being allegedly sexually abused [Berenson et al., 1992, 2002; Berenson, 1993; McCann et al., 1990a; McCann, 1998; Ingram et al., 2001a]. There even have been reports on post-traumatic scarring leading to entirely closing off the hymnal opening, resembling an imperforate hymen [Berkowitz et al., 1987a; Botash and Jean-Louis, 2001]. With the exception of complex genital abnormalities, all girls are born with hymens – several studies have failed to document a single case of an absent hymen at birth in more than 26,000 newborn girls. Thus, an entity known as ‘congenital absence of hymen’ does not exist as a sole congenital abnormality [Bays and Chadwick, 1993].

The appearance of the hymenal membrane is quite variable and is strongly influenced by the factors mentioned above but also by age and hormonal factors.

The whitish-pink hymen of newborn girls is mostly annular, rather thick and redundant due to maternal estrogen influence.

The appearance changes markedly due to withdrawal of estrogen thus creating the typical and most prevalent crescentic, semilunar configuration mostly found in children over 3 years of age. The hymen becomes a thin, more translucent and reddish membrane due to vascularization until pubertal raise of estrogen again creates a paler, redundant and fimbriated appearance. Hymenal configurations observed in children are: (1) Annular (circumferential, concentric) – mostly in newborns and infants. (2) Crescentic (semilunar, posterior rim type) – most frequent in prepubertal girls. (3) Fimbriated (denticular) – mainly in newborns, small infants and puberty. (4) Normal variants include: sleeve-like (hymen altus), septate, microperforate (= cribriform) or imperforate hymen. An asymmetric appearance is frequent, not abnormal and often results from asymmetric traction [Berenson et al., 1991, 1992; Berenson, 1993, 1995, 1998; McCann et al., 1990b; Emans, 2000].

The hymen is an elastic tissue, unlike the common lay misconception of a piece of paper irrevocably ‘broken’ by penetration or the idea of an impermeable membrane. The degree of elasticity is difficult to quantify, although penetration of a digit may occur without sequelae. The use of tampons may cause enlargement of the hymenal opening but no disruption of its integrity.

Masturbation in prepubertal girls is clitoral and does not cause injury of the hymen, especially because the prepubertal hymen is well innervated and quite pain-susceptible. Also, physical activities like gymnastics, running, jumping or splits do not lead to hymenal damage [Bays and Chadwick, 1993; Bays, 2001; Emans et al., 1994; Finkelstein et al., 1996].

Among the hymenal variations a ‘notch’ is an angular or V-shaped indentation on the edge of the hymenal membrane. It is referred to as ‘concavity’ when its appearance is curved or U-shaped. Superior (also referred to as ‘anterior’) notches in the supine position are common in non-abused children. The withdrawal of estrogen in infants and resulting involution of tissue in the periurethral area may lead to formation of new superior notches.

Notches that extend to the vestibular or vaginal wall have only been reported in abused children and have then been referred to as ‘transections’. No notches between 4 and 8 o’clock have been found in non-abused children, making it a finding consistent with penetrating trauma [Bays and Chadwick, 1993; Berenson, 1998; Finkel and Giardino, 2002].

Other hymenal variations include ‘external ridges’ on the external surface of the hymen. They are frequently found in newborns and tend to resolve in most children until 3 years of age. They have previously been misinterpreted as new scar formation. ‘Longitudinal intravaginal ridges’ or columns were reported in 25–89%, a variance maybe reflecting age or racial differences in the studies.

‘Bumps’ or ‘mounds’ (also referred to as ‘projections’) mostly occur when an intravaginal ridge attaches to the hymenal rim, but may also exist as isolated finding. They have been found in 7–34% of the non-abused population and also are no indicators of abuse as misconceived in earlier studies. ‘Tags’ or elongated projections of the hymenal rim are often incorrectly referred to as polyps.

The latter are benign mucosal tumors growing into a lumen, while tags may result from previous vaginal septa or intravaginal ridges protruding from the hymenal rim. Their prevalence ranges from 2 to 25%.

Vestibular and periurethral supporting bands also have been misinterpreted as scar formation while being found in 50–90% of non-abused girls [Emans et al., 1987; Berenson et al., 1991, 1992; Berenson, 1995, 1998; McCann et al., 1990b]. Other variations of normal described in McCann’s study were erythema of the vestibule (56%), labial adhesions (39%), lymphoid follicles in the fossa navicularis (34%), and urethral dilatation with labial traction (15%). A posterior fourchette midline avascular area also called ‘linea vestibularis’ is found in up to 25% and has the potential to be confused with scar tissue [McCann et al., 1990b; Kellog and Parra, 1991, 1993].

Normal perianal findings in both sexes include erythema (41%), increased pigmentation (30%), venous congestion (73%), midline skin tags (11%), and smooth midline wedge-shaped areas referred to as ‘diastasis ani’ (26%).

The latter results from a congenital anomaly of external sphincter muscle fibers. This finding also has a considerable potential for misinterpretation as scar tissue [McCann et al., 1989]. A recent study found significant differences in perianal anatomy between boys and girls as well as between different examination positions. The authors interpret the differences as a result of anatomical and muscular differences between boys and girls [Myhre et al., 2001].

Findings in Abused Children
The reasons for the paucity of specific findings in sexual abuse victims have been discussed earlier in this text, the abusive physical contact producing no injury being the most prominent reason. The timing of the examination crucially determines the frequency of abnormal findings in those cases where injury has been produced, mainly because of the rapid and often complete healing of anogenital injuries and typically delayed disclosure. The range of findings varies considerably, involving superficial mucosal abrasions and scratches to clear transecting lacerations of anogenital tissues.

As the offender frequently uses little physical force, many resulting injuries tend to be superficial and heal rapidly. Most of these findings resolve within 2–3 days after trauma, which is consistent with the characteristics of wound healing. Tissue damage heals by a predictable pathologic process. Superficial damage will heal completely within 5–7 days by regeneration without producing any scar tissue. Serious lacerating wounds heal by repair, leaving a scar of much lesser extent than the primary injury [Finkel and De Jong, 2001]. Thus the kind of findings vary considerably with the nature of the abuse, hereby involved objects, the degree of force used, the age of the child and the frequency of the abuse. The time since the last incidence and a history of pain and/or bleeding were the only two factors which significantly increased the likelihood of detecting abnormal physical findings in two studies [Kerns et al., 1993; Adams et al., 1994].

Female Genitalia
The majority of studies describe findings of female genitalia. Minor trauma is frequently caused by genital fondling, rubbing or vulvar coitus (rubbing of the penis between the labia). It mostly leads to unspecific changes like erythema or superficial abrasions of the inner aspects of labia minora, periurethral area, vestibule or the clitoris which heal to integrity within a few days.

In cases of digital manipulation, these findings are frequently found in the superior-anterior region in supine position (between 9 and 3 o’clock). A history of dysuria is a valuable corroborating history of this type of sexual abuse.

Dysuria may also follow coitus however. Penetration of the hymenal orifice with a small object like a finger may possibly leave no physical sequelae due to the elastic nature of the hymenal tissue [Finkel and De Jong, 2001].

When an object forcefully penetrates into the vagina, residual signs are obvious when examined acutely. The extent of injury increases with lower age.

The majority of abuse-related findings are found in the posterior area involving the posterior part of the hymen, fossa navicularis, posterior fourchette, the posterior commissure and the lateral walls of the vagina. In rare cases, a perforation of the posterior fornix into the peritoneum may occur. A penile penetration causes disruption of the hymenal rim especially between 5 and 7 o’clock but can be found anywhere between the 3 and 9 o’clock position. This results in incomplete or complete transections and V-shaped notches or clefts of the hymenal membrane.

Within a few weeks the notches lose their clear-cut and sharp appearance and develop into a smoother, U-shaped aspect, referred to as concavities. In some cases a narrowing of the posterior rim develops. The edges of the hymenal rim may become thickened or rolled and intravaginal structures may be more readily exposed. Even if repeated penetration may lead to an enlargement of the hymenal transverse diameter, it is no longer considered to be a reliable diagnostic sign as an isolated finding [Paradise, 1989; Paradise et al., 1994; Kerns et al., 1992; McCann et al., 1992; McCann, 1998; McCann and Kerns, 1999; Bays and Chadwick, 1993; AAP, 1999; Adams, 2001; Finkel and Giardino, 2002].

Repeated irritation of genital mucosa may result in chronic inflammation (recurring vulvovaginitis) or in labial agglutination (labial synechia) as an acquired post-inflammatory condition. Both findings may gain significance with a corroborating history [Berkowitz et al., 1987b; McCann et al., 1988; Vandeven and Emans, 1993]. Due to their unspecific nature and prevalence in the prepubertal child, caution should be used when they present as supposed isolated and sole indicators of possible sexual abuse.

Anal Findings
The interpretation of anal signs of abuse (sodomy) in boys and girls is far more controversial than genital signs of trauma. The frequency and significance of findings are subject to substantial disagreement in the literature [Hobbs and Wynne, 1989b; Hobbs et al., 1999b; McCann et al., 1989; Finkel and De Jong, 2001; Finkel and Giardino, 2002]. The ability of the external sphincter to dilate considerably when passing large bolus of fecal matter without any injury to the anal tissues is a major contributing factor. Variables influencing the presence of physical signs include the size of the object introduced, the amount of force used, the age of the victim, the use of lubricants, frequency of episodes and time elapsed since the last episode.

Again, acute and extensive findings are not very problematic to interpret.

Deep lacerations and significant trauma to the anus are obvious results of anal penetration. In these cases, anoscopy may be helpful for identification of internal injuries like bruising, petechiae, or lacerations and for collection of seminal products [Ernst et al., 2000]. In the absence of acute findings, anoscopy is not indicated. The significance of chronic anal signs remains more problematic.

Hobbs and Wynne [1986, 1989a] reported a high incidence of 40–50% of abnormal anal findings in abused children. A significant number of findings seen in allegedly abused children in their study have been demonstrated also to occur in a non-abused population evaluated by McCann et al. [1989].

Anal fissures may result from constipation, but are not frequently reported in constipated children. Flattened anal folds should arouse some concern beyond the diaper age. Anal skin tags are frequently seen in the midline in non-abused children but may result from healed trauma if present outside the midline [Finkel and De Jong, 2001; Finkel and Giardino, 2002; Hobbs et al., 1999b]. Special controversy has evolved about the sign of ‘reflex anal dilatation’ (RAD). Although having confidence about the significance of this finding (‘dilatation >0.5 cm’) in earlier studies [Hobbs and Wynne, 1989a], in the recent revised edition of ‘Child Abuse and Neglect – A Clinician’s Handbook’, Hobbs and Wynne [1999b] state that ‘further research is needed to improve the understanding and… significance of the association of RAD with respect to age, constipation, general anesthesia, post-mortem change and neurologic disorders…’.

Especially the absence or presence of stool in the rectal ampulla when interpreting this finding is subject to controversy. The paucity of studies to date does not allow final conclusions on the significance of chronic anal findings. In the opinion of the authors, they may be overestimated in British studies [Hobbs and Wynne, 1986, 1989a, 1999a,b; RCP, 1997] and underestimated in American overviews [Finkel and De Jong, 2001]. There is only one article in the American literature specifically examining findings after anal abuse [Muram, 1989b]. As an isolated sign, we do not interpret RAD as diagnostic of CSA. A dilatation > 15–20mm without visible stool however is concerning and warrants further evaluation.