The search yielded 1269 articles, of which 69 full text articles were reviewed in full for eligibility. Of these, 17 met the inclusion criteria [4, 6, 14, 22,23,24,25,26,27,28,29,30,31,32,33,34,35]. All studies reporting primary data on the medical relevance and/or impact of virginity testing on examinee were included (n = 17). Studies with inappropriate study design were excluded (n = 44). This included editorials, opinions, and any study that did not report primary data on virginity testing and/or hymen examination. Studies with inappropriate study population were also excluded, including those that did not study females with a history of vaginal penetration (n = 4). Studies reporting on surgical interventions of the hymen not associated with virginity testing were excluded (n = 4). Ten studies reported on the medical utility of virginity testing and key findings are presented in Table 1 [22,23,24,25,26,27,28,29,30,31]. Eight studies reported on the impact of virginity testing on the examinee and key findings are presented in Table 2 and presented again in Table 3 by theme identified [4, 6, 14, 30, 32,33,34,35].
Ten studies reported on the medical relevance of hymen examination as a method to determine history of vaginal intercourse, the most common type of virginity testing . The study characteristics and key findings are summarized in Table 1 [22,23,24,25,26,27,28,29,30,31]. The available research on this topic comes chiefly from physician examination of prepubertal and adolescent girls after sexual abuse allegations to determine if evidence of vaginal penetration existed. Seven of the included papers studied the accuracy of abnormal genital examinations as an indicator for history of vaginal penetration [22,23,24,25,26,27,28]. Abnormal genital exams included findings such as a hymenal transection, laceration, enlarged opening, or scar. Two studies reviewed physician’s accuracy in determination of virginity by exam [29, 30], and one study reported on pediatric chief residents’ ability to correctly identify the hymen by examination .
In a case-control study by Berenson et al. in the United States, genital features were compared between 192 girls with a history of vaginal penetration from sexual abuse and 200 who denied past penetration . Presence or absence of 21 different hymenal or vulvar features was compared between the two groups, such as presence of hymenal tissue, transections, perforations, and notches. It was found that only 2.5% of physical exam findings were unique to the group with a history of penetration.
Kellog et al. studied a cohort in the United States with definitive evidence of previous vaginal penetration. In the study of 26 pregnant adolescents who reported sexual abuse, 22 participants (64%) had normal or nonspecific genital examination findings, eight (22%) had inconclusive findings, four (8%) had suggestive findings, and two (6%) had findings of definite evidence of vaginal penetration .
In one large cohort of 2384 children in the United States, 957 girls reported penetration from sexual abuse. Of these 957 girls, only 61 (6%) had abnormal genital examination findings . The study’s parameters for abnormal examination included: “acute trauma, transections of the hymen that extended to the base of the hymen, scarring, sexually transmitted diseases, and positive forensics” .
In a retrospective case review in the United States of 213 girls who reported vaginal penetration/contact from sexual abuse, there was a normal genital exam in 59 cases (28%), non-specific findings in 104 cases (49%), and suspicious findings in 20 cases (9%) . Size of hymenal opening was also measured on the study group (7.7 ± 2.6 mm) and compared to published data on non-abused children of the same age (6.9 ± 2.2 mm), and no significant difference was found in mean size .
Hymenal opening size was measured in a United States case-control study of 189 girls with, and 197 girls without, a history of penile or digital penetration from sexual abuse . The former group had a larger mean transverse hymen diameter than the latter when examined in the knee-chest position but not supine position; hymenal orifice size also increased with age.
Healing of injuries to the hymen was reviewed in two studies, both in the contexts of sexual abuse allegations in the United States [27, 28]. In a study of 75 girls with a history of a vaginal penetration or trauma, injuries to the hymen were found in 37 cases (49.3%), significant genital findings (i.e., a transection of the hymen) were found in 15 girls (20%), and in the remaining 80%, there was no increase in the hymenal diameter or irregularity or narrowing of the hymen . In a study of 113 prepubertal girls and 126 pubertal girls with previous penetration that reported on healing of injuries to the hymen, it was found that hymenal injuries in both groups healed rapidly and often left little or no evidence of previous trauma .
A 1978 study in the United States reported on the accuracy of physicians in confirming virginity through hymen examination . Two gynecologists inspected the hymens of a cohort of 28 self-declared virgins. The physicians reported that examination of the hymen confirmed virginity in 16 cases (58%), was inconclusive in nine cases (31%), and uncertain in three cases (11%) .
In a study of forensic physicians in Turkey, 66% of respondents reported that their findings from at least one virginity exam contradicted a recent virginity exam of the same patient .
Lastly, a study examined physician knowledge of hymen anatomy . In 2005, 137 United States pediatric chief residents were asked to identify the hymen on a photograph of pediatric anatomy; 64% were able to correctly identify the structure .
Quality of the ten studies reporting on the medical relevance of virginity testing was assessed according to USPSTF guidelines and is reported in Table 1. (Additional information regarding the USPSTF grading criteria is provided in the Additional file 1). The level of evidence ranged from level II-2 to level III. Seven studies were level II-2 [22,23,24,25,26,27,28], one study was level II-3 , and two studies were level III [30, 31]. The internal validity of the studies reporting on medical relevance ranged from good to poor. Two studies had good internal validity [22, 26], six had fair internal validity [24, 25, 27, 28, 30, 31], and two had poor interval validity [23, 29].
Impact on examinee
Eight studies provided evidence on the effects of virginity testing on the examinee [4, 6, 14, 30, 32,33,34,35]. The study characteristics and key findings are summarized in Table 2. Included studies provided data on the experiences of the examinee and those who worked directly with the examinee (such as doctors, social workers, police officers, and lawyers). Six studies [4, 6, 14, 32, 33, 35] provided data from interviews and focus group discussions and two studies [30, 34] provided survey data from healthcare professionals. Three themes were constructed from the data: physical harm, psychological harm, and social harm. Themes are presented in Table 3, and expanded on below with relevant study findings and rationale of theme selection.
Physical harm of virginity examinees was reported in four studies [6, 14, 32, 33]. In a study of virginity testing in Palestine, a social worker present during her client’s virginity exam reported that, “the process was very painful, she was crying, screaming, holding my hands” . Undergoing virginity exams also caused two of the social worker’s clients to become suicidal, with one reported attempted suicide . When one examinee failed her virginity test, she was told that she would need to, “search for a way to save herself from the deadly consequences that awaited her” . In a report of virginity testing in Iran, one medical examiner reported an exam that lead to death, “I told her that her hymen was not intact, and she said that she had done nothing. Then I heard she had committed suicide” . A report on India’s two-finger test describes doctors who harmed the examinee during the test by aggravating existing injuries . In South Africa, a report was made to Childline, a helpline that offers rape counseling, of an examinee’s relatives breaking both of her arms after she failed a virginity test .
Psychological harm was reported in five studies [6, 14, 30, 32, 34]. In a study of virginity testing in Palestine, focus group discussions revealed that women who underwent virginity testing were:
extremely fearful of and indeed felt terrorized by [the experience]. … Their feelings of fear and invasion were manifested in a variety of ways: by their refusal to sit on the examination chair, through crying, screaming, pushing, freezing-up, being silent, fainting, etc. 
One social worker described virginity exams as torture: “… I also felt it is so unfair to be sexually abused and then [have to] go through such a vicious process of torture” . In depth interviews of medical professionals who performed virginity testing in Iran revealed that the virginity test resulted in the psychological distress of the examinee, causing “rejection, suicide, depression, and weakened self-confidence” . A survey of forensic physicians in Turkey found that 93% of 118 respondents agreed that virginity tests are psychologically traumatic for the patient, 64% believed they were a violation of privacy, and 60% believed they result in loss of examinee’s self-esteem . Interviews from a report on India’s two-finger test documented the fear and re-traumatization the examination causes . In a study of 101 nurses and midwives in Turkey, 90% indicated they were opposed to virginity testing, and when asked why, nearly half agreed that they were opposed because the examinations are being done against the examinee’s will . Sixty-two percent of the nurses and midwives also agreed that a forced virginity exam may result in severe negative effects such as anxiety, depression, isolation from society, a dysfunctional sex life, guilt, worsened self-respect, and fear of death .
Social harm was reported in three studies [4, 6, 35]. Leclerc-Madlala reported in a study in South Africa that those who fail virginity tests are often expected to pay a fine for tainting the community . They are also excluded from certain employment opportunities, illustrated by one factory owner who stated that her various franchises throughout KwaZulu-Natal and the Eastern Cape had tested and selectively employed over 4000 virgins, which she believed was a service to the community and state . Social harms of positive virginity tests were also noted. In a focus group interview of 14 girls who were planning to attend a virginity testing event, the girls reported that their primary concern was that being “certified” a virgin would result in brothers, friends, or neighbors raping them. They spoke of previous cases in which this had occurred in their community. Rape in this context was reported most likely to occur as a gang rape by several boys who needed to “teach her a lesson” and show her “what men are all about” . A study in Palestine detailed one examinee’s fear of adverse social consequences . She was afraid that a failed virginity test would result in loss of honor and social condemnation. The examinee stated, “I wanted to do [the examination]. I wanted to know if I lost my honor. I paid to learn that I lost my honor” . In another study of South Africa’s KwaZulu-Natal province by Leclerc-Madlala, those who failed a virginity test were subject to name-calling and social exclusion . One respondent referred to a girl who failed as a "rotten potato" who must be kept away from the ‘virgin girls’ because she will surely “spoil the bunch.” Another respondent noted that being in close proximity to a girl who failed the test would, “cause the flowers of the nation to wilt” .
Quality of the eight studies reporting on the impact of virginity testing on examinee was assessed according to USPSTF guidelines and is reported in Table 2. All eight studies were level III evidence [4, 6, 14, 30, 32,33,34,35]. The internal validity of the studies reporting on impact on examinee ranged from fair to poor. Four studies had fair internal validity [6, 14, 30, 34] and four had poor internal validity [4, 32, 33, 35].