OBJECTIVE: To determine the time period after sexual assault of a child that specimens may yield evidence using DNA amplification. Secondary questions included the comparative laboratory yields of body swabs versus other specimens, and the correlation between physical findings and laboratory results.
PATIENTS AND METHODS: Data from evidence-collection kits from children 13 years and younger were reviewed. Kits were screened for evidence using traditional methods, and DNA testing was performed for positive specimens. Laboratory data were compared with historical information.
RESULTS: There were 277 evidence-collection kits analyzed; 151 were collected from children younger than 10; 222 kits (80%) had 1 or more positive laboratory screening test, of which 56 (20%) tested positive by DNA. The time interval to collection was <24 hours for 30 of the 56 positive kits (68% positives with a documented time interval), and 24 (43% of all positive kits) were positive only by nonbody specimens. The majority of children with DNA were aged 10 or older, but kits from 14 children younger than 10 also had a positive DNA result, of which 5 were positive by a body swab collected between 7 and 95 hours after assault. Although body swabs were important sources of evidence for older children, they were significantly less likely than nonbody specimens to yield DNA among children younger than 10 (P = .002). There was no correlation between physical findings and laboratory evidence.
CONCLUSIONS: Body samples should be considered for children beyond 24 hours after assault, although the yield is limited. Physical examination findings do not predict yield of forensic laboratory tests.
WHAT'S KNOWN ON THIS SUBJECT:
The American Academy of Pediatrics recommends that forensic evidence collection be considered for up to 72 hours after sexual assault. Data on child victims reveal that the yield is low beyond 24 hours, particularly for specimens collected from children's bodies.
WHAT THIS STUDY ADDS:
Review of forensic laboratory results including DNA amplification indicates that collection of body swabs from children beyond 24 hours after assault may yield evidence. Most children with positive laboratory evidence have normal or nonspecific physical findings.
National data for 2008 reveal that among ∼758 000 reported child maltreatment victims, 9.1% were sexually abused. Just over half of sexually abused children were younger than 12.1 For children who present to a medical facility after a recent episode of sexual assault, standard care includes collection of body swabs and other specimens (such as clothing) in the hope of identifying an assailant. Forensic evidence-collection kits (“rape kits”) contain receptacles and instructions for collection of cotton swab specimens from the mouth, vagina/penis, and anus; scrapings or swabs from under the victim's nails; combed and pulled hair specimens; as well as miscellaneous items such as stained clothing and debris. The American Academy of Pediatrics recommends that forensic evidence collection be considered for up to 72 hours after sexual assault and in cases of acute injury.2 The 72-hour timeframe is supported by studies of adult women after consensual intercourse and studies of adult rape victims.3,–,5 The increased availability of sensitive DNA amplification methods in recent years has enabled detection of foreign DNA in adult women victims even when cytological tests were negative for spermatozoa, which has prompted some jurisdictions to request evidence collection beyond 72 hours after sexual assault.6
Data on the yield of evidence collection in child sexual assault victims is limited. Christian et al7 reported that among 273 children younger than 10 who underwent forensic evidence collection after sexual assault, no swabs were positive for blood, semen, or sperm after 13 hours after assault, and 64% of the evidence recovered was found on clothing or linens. In their study of 39 adolescent and 41 prepubertal sexual assault victims, Young et al8 found evidence of semen on swabs collected from 13 adolescents but none of the prepubertal children. Additional evidence was recovered from clothing and linens for 3 prepubertal children. None of the 39 kits collected beyond 24 hours yielded any evidence. Both investigations concerned the use of pre-DNA amplification laboratory methods, however, and their applicability to modern child sexual assault cases is unclear. In a more recent analysis of forensic findings that included some DNA methods it was concluded that positive examination findings, age older than 10, and pubertal Tanner stage were predictive of positive forensic evidence.9 Clothing in the latter study was very likely to be positive but was collected in only a minority of cases.
The primary purpose of our investigation was to determine the time period after sexual assault of a child that specimens may yield positive laboratory evidence using DNA amplification techniques. Our secondary questions were to examine the correlations of specimen source and physical examination findings to the likelihood of finding positive laboratory results. A better understanding of these questions will enable practitioners to limit collection of specimens that are found to have low forensic yields and can help reassure victims and their caregivers about the importance of other specimens that prove to have greater evidentiary potential.
This investigation was a collaborative effort between child abuse pediatricians from 2 Houston medical schools (University of Texas Houston Medical School and Baylor College of Medicine) and the Houston Police Department (HPD) Crime Laboratory. The laboratory is the largest crime laboratory in Houston, and processes ∼70% of evidence-collection kits in the community. Permission to conduct the study was obtained from the human subjects panels for both medical schools and from HPD officials.
Data were derived from a retrospective review of case information and laboratory results from forensic evidence-collection kits collected from children 13 or younger. All kits from children 13 or younger that were processed by the laboratory between January 1, 2007, and December 31, 2008, were analyzed. The laboratory did not process all kits, but only those in which evidence was sought for investigative or legal purposes. Laboratory methods for processing of kits remained constant during the study period. Separate police reports also were consulted when data points were missing from the kits (such as time of assault or victim age).
Kits were processed according to standard laboratory protocol. Body swabs were tested for semen by acid phosphatase assay, and the presence of sperm/semen was confirmed with microscopy and/or prostate-specific (p30) antigen assay. Specimens suspected of containing blood were screened by using standard color tests (phenolphthalein and hematrace). Nonbody specimens were inspected visually and with the aid of an alternate light source (Omnichrome Omniprint 1000, 450 nm [Omnichrome, Carlsbad, CA] or Mini Crime Scope MCS-400, 455 nm [Horiba Scientific, Edison, NJ]) for stains. Identified stains were tested for evidence of sperm/semen and blood using the tests used for body swabs. Confirmative DNA testing was performed for specimens with positive microscopy, p30, or color test result. Specimens thought to contain other biological evidence (eg, saliva) were tested directly by DNA analysis. For kits with multiple positive results on screening tests, only the strongest or most probative specimens were tested further for DNA. In cases involving multiple assailants, all specimens with positive screens were tested for DNA.
Data collected from forensic evidence kits is stored in a secure database at the crime laboratory according to routine laboratory practice. For the purposes of this investigation, laboratory data were recorded as positive or not positive (ie, negative or indeterminate). The source of each sample was recorded. Laboratory data were compared with historical and physical examination information recorded on standard evidence-collection documents; specifically, the type of sexual contact, the time since most recent sexual assault, the presence of acute anogenital trauma, and the facility where evidence collection was performed were abstracted. The retrospective nature of this investigation precluded an accurate assessment of additional data points such as whether the child had bathed or changed clothes before evidence collection. Because information regarding subjects' pubertal stage was inconsistently recorded on kit documents, this information was not abstracted. Multiple evidence-collection kits collected from the same child during the study period were counted as separate cases.
Anogenital examination findings recorded on kit documents were classified using a standard table.10 Forensic photographs are not maintained by the HPD crime laboratory and, therefore, were not available for analysis. For the purposes of this investigation, only acute anogenital injuries were considered. Anogenital findings were analyzed separately by 2 investigators for evidence of recent penetrating trauma, and a consensus opinion was sought from all investigators for discrepant cases.
There were 290 total kits abstracted, of which 277 met study criteria. Of the 13 excluded cases, 7 contained insufficient data for study analysis, and the remainder were collected from subjects older than 13. One child had 2 kits during the study period. Among the included cases, 228 (82%) were female, and 153 (55%) were younger than 10. The majority of cases were from 1 of 2 children's hospitals serving the Houston area and the child advocacy center located in Houston. There were 244 kits (88%) collected in the year 2000 or later; the oldest kit was collected in 1988. Data on the numbers of kits that were stored but not processed during the study period were not available.
The time from assault to evidence collection was within 24 hours for 111 (40%) kits, between 25 and 48 hours for 24 (9%) kits, between 49 and 72 hours for 9 (3%) kits, and beyond 72 hours for 8 (3%) kits. The time interval for 125 (45%) kits was unknown; of these, 85 involved subjects younger than 10. Invasive body sampling (defined as oral, vaginal, and anal swabs) was performed for 253 kits. For 8 kits, the only invasive or intimate body specimens were oral swabs (7 kits) or pubic hair combings (1 child). Only objects from the crime scene (clothing and/or bedding, and a Ziploc bag in 1 case) were submitted for laboratory analysis for 24 cases.
There were 222 kits (80%) that had 1 or more positive laboratory nonDNA test that included alternate light source fluorescence. Yields of all nonDNA laboratory tests and their comparisons to DNA results are listed in Table 1. Specimens from 56 kits (20%) tested positive by DNA (53 girls). Two of the 56 kits with a positive DNA result did not have a nonDNA laboratory test; swabs from around the mouth and from the hands were the only specimens available for 1 child, and only objects from the crime scene were available for another. Positive DNA was found for 35 of 162 kits with a positive alternate light source test (22%), for 9 of 39 kits with a positive test for blood (phenolphthalein and/or hematrace; 23%), and for 39 of 130 kits with a positive test for semen/sperm (acid phosphatase, microscopy and/or p30; 30%). For 1 12-year-old girl with positive screens for blood and semen/sperm by all testing methods from body and nonbody specimens, DNA was recovered only from debris and clothing specimens. Another 3 kits with positive screens for semen/sperm by all modalities had no positive DNA test.
Among the kits with positive tests for DNA, 28 (50%) concerned a disclosure of penile-vaginal contact, 4 (7%) a disclosure of penile-anal contact, 3 (5%) a disclosure of both penile-vaginal and penile-anal contact, 5 (9%) a disclosure of other forms of sexual contact (cunnilingus, fondling, and/or fellatio), and no sexual contact was specified for another 16 (29%) kits. Among children with positive DNA from a body swab, 30 (91% of children in this group) were female.
The majority of children with a positive DNA result were 10 or older, but 14 kits taken from children younger than 10 (9% of children in this age group) also had a positive DNA result (Fig 1). Of the 56 kits with a positive DNA result, 24 (43%) were positive only by specimens taken from sources other than the child's body (Fig 2). Of the 14 kits from children younger than 10 that had a positive DNA result, 13 included invasive body specimens, and 5 were positive by a swab taken from the child's body. Among these 5 kits, 2 also had positive DNA evidence from nonbody specimens, and nonbody specimens were not collected for the remaining 3 kits. Overall, among children younger than 10 with positive DNA evidence, the source of the positive evidence was more likely to have been a nonbody specimen than a body swab (P = .002). When body swabs were analyzed separately, positive results for DNA were uncommon except among children age 12 and 13, for whom the vagina was the most common source of evidence (Fig 3). Overall, vaginal specimens had the highest yield of positive DNA (21 of 200 kits), followed by penile (2 of 33), and anal (3 of 173). Fingernail swabs and scrapings yielded only 1 positive DNA result among 60 kits that contained these specimens. Hair combings (from the head in 114 kits and pubis in 69 kits) and oral specimens (145 kits) yielded no positive foreign DNA.
The time to evidence collection for kits testing positive for DNA was within 24 hours for 30 (54%) kits, between 25 and 48 hours for 9 (16%) kits, between 49 and 72 hours for 3 (5%) kits, and between 73 and 96 hours for 2 (4%) kits. The time to evidence collection was unknown for 12 (21%) kits that had a positive DNA result. When only kits with known time intervals were considered, 68% (30 of 44) were collected within 24 hours. Among the 7 kits containing specimens collected >96 hours after assault, none had a positive sample for DNA.
For the 5 kits taken from children younger than 10 who had a body swab specimen that tested positive for DNA, the time intervals to evidence collection were as follows: 7 hours (perioral and hands); 35 hours (vaginal and anal swabs); 66 hours (pubic hair from anus); 95 hours (fingernail swab); and an unknown time interval (penile swab).
Acute anogenital findings were classified differently by first and second reviewers for 87 cases, but agreement was achieved for all cases after group discussion among study authors that focused on consistent application of the standard criteria proposed by Adams.10 The majority of children with a positive DNA result had normal, nonspecific, or indeterminate acute anogenital findings according to written kit documentation. Kits from 23% of cases with documented normal or nonspecific anogenital findings tested positive for DNA versus only 13% of kits from cases with diagnostic physical findings, but this difference was not statistically significant (P = .187). When clinical findings were compared with the source of positive DNA specimens, 23 of 28 (82%) kits in the normal/nonspecific group with a body swab tested yielded a positive result, versus all 6 kits from children with indeterminate or diagnostic findings (P = .627). Of the cases that did not include sufficient information to make a determination regarding the presence of acute anogenital trauma, 16% yielded a positive DNA result (Table 2).
Our results support the findings of previous investigators who found that the majority of children with positive biological evidence undergo examination within 24 hours of assault, and that a significant proportion of evidence is collected from objects such as linens and clothing rather than from the children themselves.7,–,9 Recovery of laboratory evidence is particularly rare in children younger than 10, for whom the majority of evidence was recovered from nonbody specimens. However, kits from 5 children younger than 10 years had a positive DNA test from a body swab collected between 7 and 95 hours, 1 of which included positive DNA from invasive body swabs collected 35 hours after the child's molestation. Two of the 5 kits also had positive DNA tests from nonbody sources, and nonbody specimens were not available for testing for the remaining 3 kits. It is conceivable that positive specimens taken from inside the child's body may be regarded as stronger proof of sexual contact than positive biological evidence recovered from nonbody objects, although this subject has not been investigated to our knowledge. Given the potentially higher probative value of a positive laboratory result from an invasive body swab than from a nonbody specimen, our data suggest that invasive body samples should be considered for prepubertal children beyond the 24-hour limit proposed by previous investigators, although the yield will be limited.
Despite the high number of kits with a positive screening laboratory test (80%), only 20% of the study population was confirmed positive by DNA testing. The alternate light source proved to be a particularly nonspecific screen, consistent with previous reports regarding its low specificity11,12; tests for blood and semen/sperm seem to be moderately specific, with 9 of 39 and 39 of 130 confirmed in our series, respectively.
Another important result of our investigation was the high proportion of cases of children with normal or nonspecific anogenital findings who had DNA evidence from a body swab (23 of 28 tested). Previous research has indicated a positive correlation between diagnostic physical findings and laboratory evidence9; although we found a trend toward positive DNA evidence among children with diagnostic findings, it was not significant. Our results are consistent with the reported yield of laboratory tests for sexually transmitted diseases among children with disclosures of genital contact but normal or nonspecific physical findings.13 The preponderance of normal anogenital findings among children presenting for sexual assault has been convincingly established.14,–,16 Reasons put forth to explain this phenomenon include the capacity for rapid healing of mucosal tissues and the inability of young children to accurately describe the details of their assault. Our results indicate that collection of forensic specimens after a disclosure of recent assault is appropriate even when physical findings are normal or nonspecific.
There are several limitations to this study. Like the other investigations of this subject, our ability to draw conclusions is limited by our retrospective design. A potential confounder in our study was the fact that the time interval from assault to evidence collection was unknown for 45% of cases; our data on the yield of laboratory testing within the various time intervals must therefore be interpreted with caution. However, the proportion of kits testing positive for DNA within 24 hours remained significantly greater even if it is theoretically assumed that the 12 positive kits with undocumented time intervals were collected after 24 hours (P = .031). Because data regarding bathing and changing clothes before examination was inconsistently recorded, we chose to forego analysis of these data points, and therefore we cannot comment on the possible effect on specimen yield that these actions might have had. The retrospective design is also problematic when comparing laboratory results to examination findings. The fact that the majority of kits included in our study were completed at 1 of 2 children's hospitals or a child advocacy center potentially mitigates the problem, although 18% of the kits (51 of 277) did not include sufficient documentation to determine whether there was visible trauma. We also cannot know whether all available specimens were collected in every case, and therefore the true contribution of clothing and linens cannot be known. Controlling for these factors in a prospective design would be a challenging but worthwhile endeavor given the implications of the results to protecting children.
The majority of children with DNA-confirmed biological evidence present to a medical facility within 24 hours of assault. DNA recovery from body swabs among children younger than 10 is possible after 24 hours, although it occurs infrequently. The presence of visible acute anogenital trauma on examination does not predict recovery of DNA evidence. Nonbody specimens are more likely to yield positive DNA evidence than body swabs.
We thank Drs Irma Rios and Laura Ghan of the HPD Crime Laboratory for assistance with this investigation.
- Accepted April 7, 2011.
- Address correspondence to Rebecca Girardet, MD, University of Texas Medical School at Houston, 6410 Fannin St, Suite 1425A, Houston, TX 77030. E-mail:
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
COMPANION PAPERS: Companions to this article can be found on pages 221, 227, and 374 and online at www.pediatrics.org/cgi/doi/10.1542/peds.2010-3288, www.pediatrics.org/cgi/doi/10.1542/peds.2010-3498, and www.pediatrics.org/cgi/doi/10.1542/peds.2011-1455.
- HPD —
- Houston Police Department
US Department of Health and Human Services. Administration for Children and Families. Administration on Children, Youth and Families, Children's Bureau. Child maltreatment 2008. Available at: www.acf.hhs.gov/programs/cb/stats_research/index.htm#can. Accessed June 19, 2010
- Kellogg N
US Department of Justice. Office on Violence Against Women. President's DNA initiative; sexual assault medical forensic examinations. Available at: http://samfe.dna.gov/operational_issues/considerations_collecting_evidence/prompt. Accessed May 5, 2009
- Christian CW,
- Lavelle JM,
- De Jong AR,
- Loiselle J,
- Brenner L,
- Joffe M
- Girardet RG,
- Lahoti S,
- Howard LA,
- et al
- Adams JA,
- Harper K,
- Knudson S,
- Revilla J
- Anderst J,
- Kellogg N,
- Jung I
- Copyright © 2011 by the American Academy of Pediatrics