Early Detection of Imperforate Hymen Prevents Morbidity From Delays in Diagnosis
http://www.100md.com
《小儿科》
the Division of Emergency Medicine, Department of Pediatrics, Children's Hospital of Philadelphia, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania
ABSTRACT
Objective. Although it is detectable at all ages through inspection of the external genitalia, imperforate hymen (IH) is a diagnosis that is missed commonly. We hypothesized that children with late diagnoses (predefined as 8 years of age, chosen to reflect the timing of normal menarche) would be more likely to be symptomatic, undergo more diagnostic testing, and lack appropriate documentation in their medical records, compared with children with earlier diagnoses (ie, <8 years of age).
Methods. All patients with IH were identified through searches of 3 hospital databases with International Classification of Diseases, Ninth Revision, codes. The medical records of eligible subjects were reviewed by a single, blinded researcher. Comparisons were made between children diagnosed at younger versus older ages.
Results. A bimodal distribution of age at diagnosis was demonstrated; 43% (n = 10) of girls were diagnosed at <8 years of age, and 57% (n = 13) were diagnosed at 8 years of age. Among older girls, 100% were symptomatic (abdominal pain and/or urinary symptoms; duration of symptoms: 1–120 days), whereas 90% of cases in the younger group were detected incidentally. Documentation was lacking for breast development (77%), pubic hair development (69%), and menstrual history (46%) among the older girls. Older children were more likely to present symptomatically (odds ratio: 42.0; 95% confidence interval: 3.1–1965.7) and to undergo ancillary testing (odds ratio: 20.3; 95% confidence interval: 1.6–983.1).
Conclusions. Two distinct populations of girls with IH exist, ie, those diagnosed without symptoms at a young age and those not diagnosed until >8 years of age. By incorporating an examination of the external genitalia into their routine practice, clinicians caring for children can prevent the significant delays in diagnosis, misdiagnosis, and morbidity associated with the latter group.
Key Words: imperforate hymen hydrocolpos hydrometrocolpos hematocolpos hematometrocolpos
Abbreviations: IH, imperforate hymen
Imperforate hymen (IH) is a rare congenital anomaly in which the hymenal membrane occludes outflow from the female genital tract. This obstruction results in an accumulation of uterine and vaginal secretions. Large amounts of retained fluid have been reported.1,2 The presentations of patients with IH can be variable. There are reports of cases detected incidentally1,3,4 or with symptoms of abdominal1,5–9 or back10–12 pain, constipation,13,14 dysuria or acute urinary retention,14–20 peritonitis,21,22 or primary amenorrhea,23 as well as a report of acquired IH associated with sexual abuse.24
Menarche is a relatively late event in the pubertal process, and most girls have attained Tanner stage 3 at the time of menarche. A discrepancy between a teen's advanced stage of pubertal development and the absence of menarche is a key clinical clue to the possibility of IH. Despite the ease of diagnosis at all ages with simple inspection of the external genitalia, the detection of IH can be delayed or missed, resulting in preventable morbidity and unnecessary diagnostic testing.
Previous works regarding IH consisted of case reports or small case series and therefore were limited in the ability to allow definitive conclusions. To our knowledge, no previous work systematically investigated a population of children with IH. The objective of this study was to describe a population of patients with IH who were treated at a large, tertiary care, referral children's hospital. Specifically, we aimed to characterize the presenting features and diagnostic evaluations performed and to assess the frequency with which the menstrual history, pubertal staging, and hymenal examination results were documented in the medical evaluation. We hypothesized that children with late diagnoses (predefined as 8 years of age, chosen to reflect the timing of normal menarche) would be more likely to be symptomatic, undergo more diagnostic testing, and lack appropriate documentation in their medical records, compared with children with earlier diagnoses (ie, <8 years of age).
METHODS
We conducted a retrospective, case-series study of all patients who were diagnosed with and treated for IH at a large, tertiary care, children's hospital during a 13-year period (January 1, 1987, through December 31, 2000). Eligible patients were identified through a search of the hospital's health information management (ie, medical records) database with International Classification of Diseases, Ninth Revision, codes for IH and related terms (hydrocolpos, hydrometrocolpos, hematocolpos, and hematometrocolpos). Additional searches of the urologic procedures database and the operative anesthesia database were performed to supplement the primary search. We used this strategy to optimize the identification of patients with IH. These databases were not mutually exclusive. Because we aimed to describe the presentation of uncomplicated IH, patients with genetic syndromes or coexistent genital anomalies were excluded. The medical records of identified subjects were abstracted for data pertaining to the presentation and diagnostic evaluation. All available components of the hospital medical records were subject to review, including emergency department records, preoperative evaluation findings, operative reports, and inpatient records. Demographic features, presenting symptoms, documentation of Tanner stage, hymenal examination results, menstrual history, preliminary diagnosis, ancillary test findings, operative procedure, and final diagnosis were recorded on a standardized data collection sheet by a single researcher, who was blinded with respect to the study hypothesis.
The sample size was based on identification of all eligible patients. Statistical analyses were conducted with Stata 7.0 software (Stata Corp, College Station, TX). The sample was described with means, SDs, medians, and ranges for continuous variables and frequencies for categorical variables. Between-group comparisons (ie, older versus younger girls) were performed with exact 2 tests.
This study received institutional review board approval. After chart abstraction had been completed, all data were deidentified.
RESULTS
Computerized searches of the databases resulted in the identification of 95 girls. Fourteen patients were excluded because of multiple congenital anomalies or genetic syndromes associated with IH. Of the remaining 81 patients, 58 were excluded because of diagnoses other than IH, most commonly dysfunctional uterine bleeding. There were 23 patients in the final study sample, 74% (n = 17) white, 22% (n = 5) black, and 4% (n = 1) Asian (Table 1). The distribution of age at diagnosis demonstrated a clear bimodal distribution that was even more dramatic than results with the predefined cutoff point of 8 years (Fig 1); 43% of patients (n = 10) were diagnosed at <4 years of age (median age: 1.2 years; range: 9 days to 3.7 years), and the remaining 57% (n = 13) were diagnosed at >10 years of age (median age: 12.3 years; range: 10.9–14.2 years).
In the older group, 100% of patients (n = 13) were symptomatic at the time of diagnosis. In the younger group, almost all patients (90%, n = 9) were asymptomatic. One girl in the younger group was diagnosed during an evaluation because of dysuria at 3 years of age. Although it was unclear whether the presence of an IH predisposed this patient to develop dysuria, we considered this patient to be symptomatic for the purposes of analysis.
In the group of older girls, all presented with abdominal pain (n = 6), urinary symptoms (n = 2), or both abdominal pain and urinary symptoms (n = 5). In the same group, the median duration of symptoms before diagnosis was 15 days (range: 1–120 days). Almost one half of the older girls (46%, n = 6) were given preliminary diagnoses other than IH, including urinary tract infection (n = 2), appendicitis (n = 2), nephrolithiasis (n = 1), and abdominal tumor (n = 1). In the older group, the medical records lacked documentation of breast development in 77% of cases (n = 10), pubic hair development in 69% (n = 9), and menstrual history in 46% (n = 6). Most of the older girls (85%, n = 11) underwent ancillary testing before definitive diagnosis, including blood analyses (n = 10), urinalyses (n = 9), and abdominal plain radiographic evaluations (n = 4). In comparison, the only patient in the younger group to undergo diagnostic testing before definitive diagnosis was the child diagnosed as having vesiculoureteral reflux after a previous urinary tract infection.
Ultrasonographic imaging was performed for 15 of the 23 patients. Interestingly, all 3 of the ultrasonographic evaluations performed in the younger group were to confirm diagnoses of IH that were made first through physical examination; however, 8 of the 12 ultrasonographic evaluations performed in the older group were part of diagnostic evaluations that led to the eventual diagnosis of IH. Of note, there was no documentation of any clinical suspicion of sexual abuse in either of the groups.
Compared with the younger children, the older children were more likely to present symptomatically (odds ratio: 42.0; 95% confidence interval: 3.1–1965.7; P < .001) and undergo laboratory or plain radiographic testing (odds ratio: 20.3; 95% confidence interval: 1.6–983.1; P < .001). All children underwent hymenectomy at or near the time of diagnosis, with the notable exception of 2 young children for whom surgery was delayed for several years.
DISCUSSION
In this study, the first large case series of patients with IH exclusively, we demonstrated that 2 distinct subpopulations of girls with IH exist, on the basis of the age at diagnosis and the presenting clinical features. One group was composed of younger girls, nearly all of whom were asymptomatic and were diagnosed incidentally through physical examination alone. The other group consisted of older girls, all of whom presented symptomatically. Although we planned in our analyses to compare girls 8 years of age with those <8 years of age (coinciding with the lower limit of normal puberty), a more striking dichotomy in age was observed. All patients in the younger group were <4 years of age, whereas the older group consisted of girls 10 years of age. This finding is consistent with the age distribution reported in a case series that aimed to describe the utility of laparoscopy in the evaluation of vaginal malformations.25
Girls in the older group commonly experienced protracted symptoms before being diagnosed as having IH. Although this might be attributable, in part, to a delay in seeking care by the girl and/or her family, our results indicate that the medical evaluation might also be contributory. Nearly one half of the older girls were assigned an incorrect preliminary diagnosis. In addition, this group was significantly more likely to undergo unnecessary laboratory and radiographic studies, further delaying the eventual diagnosis. Furthermore, we found that a substantial proportion of the older girls had incompletely documented histories and physical examination results and frequently lacked appropriate pubertal evaluations. Specifically, no genital examination was performed for two thirds of the older patients, and there was no breast examination for three fourths. Almost one half of the older patients were not asked about menstrual history.
The diagnosis of IH can be made from a physical examination alone, eliminating the need for expensive laboratory testing or radiographic evaluations. Observation of the hymen and viewing into the vaginal vault are best accomplished by applying gentle labial traction with the child in the supine frog-leg or knee-chest position.26 An uncomfortable speculum examination is not indicated, because mere observation of a bulging mass at the vaginal introitus for an older patient (Fig 2) or, ideally, of a membrane obstructing the view into the vaginal vault for an infant or young child establishes the diagnosis. Our results indicate, however, that some patients with IH experience delays in diagnosis, misdiagnosis, and additional diagnostic evaluations. It can be surmised that, if each older girl had undergone a complete examination of the external genitalia as part of a routine well-child visit early in her life, then none would have had to endure the symptoms and diagnostic evaluations associated with late diagnosis. It is our recommendation, therefore, that the hymenal examination be emphasized as a routine part of newborn and infant health maintenance evaluations. In addition, routine periodic hymenal reexaminations should be performed to document the hymenal changes that may occur with normal development.27,28
Treatment of IH consists of surgical hymenectomy, releasing the obstructing hymenal tissue. The majority of patients in our study were treated surgically at or near the time of diagnosis, although there were 2 notable exceptions of patients who underwent delayed treatment. The optimal timing for this procedure is unclear. However, it is possible that patients with IH may be at higher risk of endometriosis.25 Therefore, treating physicians must weigh the risks of anesthesia for young children against the potential risks resulting from delayed correction.
There are limitations to this study that should be considered in interpreting our results. First, as with all retrospectively conducted investigations, there is a potential for information bias. It is possible that key components of the history and physical examination were actually performed but were not documented in the medical record. This could have led to an overestimation of incomplete pubertal evaluations for older girls. However, the fact that older girls experienced significant delays in diagnosis and underwent additional testing supports the validity of our results. Second, ascertainment bias might have occurred, resulting in incomplete identification of all eligible patients. However, we minimized the potential for this bias by using 3 hospital databases (not mutually exclusive) to ensure inclusion of all patients diagnosed with IH during the study period. Third, although this study of IH is the largest to date, the small sample size led to wide confidence intervals around the odds ratios. Finally, this study was undertaken at a tertiary children's hospital; therefore, the results might not be reflective of all children with IH.
Clinicians caring for children and teens should incorporate inspection of the female genitalia into their routine practices. If clinicians detect IH among young children, they can prevent the morbidity associated with delays in diagnosis, including the development of symptoms and additional diagnostic testing. For older girls presenting with abdominal or urinary symptoms, the complete medical evaluation must include a review of the menstrual history and an examination of the breasts and external genitalia. A high index of suspicion for IH should be maintained, especially if a discrepancy is noted between the Tanner stage of pubertal development and a reported lack of menarche.
ACKNOWLEDGMENTS
This work was presented in part at the Ambulatory Pediatrics Association Regional Conference; March 5, 2002; Philadelphia, PA.
We thank Gina Ruggierio for her assistance with data management.
FOOTNOTES
Accepted Sep 1, 2004.
No conflict of interest declared.
REFERENCES
Bogen DL, Gehris, RP, Bellinger MF. Picture of the month. Arch Pediatr Adolesc Med. 2000;154 :959 –960
Loscalzo IL, Catapano M, Loscalzo J, Sama A. Imperforate hymen with bilateral hydronephrosis: an unusual emergency department diagnosis. J Emerg Med. 1995;13 :337 –339
Henderson KC. Hydrometrocolpos in a newborn. Am J Dis Child. 1975;129 :1190 –1191
Kahn R, Duncan B, Bowes W. Spontaneous opening of congenital imperforate hymen. J Pediatr. 1975;87 :768 –770
Ward A, Maher P. Hematocolpos: an unusual presentation. Br J Clin Pract. 1979;33 :83 –84
Dickson CA, Saad S, Tesar JD. Imperforate hymen with hematocolpos. Ann Emerg Med. 1985;14 :467 –469
Maffeo DA, Santagata P, Santagata M. Hematocopos by imperforated hymen: case report. Clin Exp Obstet Gynecol. 1999;26 :230 –231
Schneider K, Hong J, Fong J, Sanders CG. Hematocolpos as an easily overlooked diagnosis. Curr Opin Pediatr. 1999;11 :249 –252
McIvor RA. Haematocolpos, once-in-a-lifetime cause of recurrent abdominal pain. Arch Emerg Med. 1990;7 :51 –57
London NJ, Sefton GK. Hematocolpos: an unusual cause of sciatica in an adolescent girl. Spine. 1996;21 :1381 –1382
Buick RG, Chowdhary SK. Backache: a rare diagnosis and unusual complication. Pediatr Surg Int. 1999;15 :586 –587
Letts M, Haasbeek J. Hematocolpos as a cause of back pain in premenarchal adolescents. J Pediatr Orthop. 1990;10 :731 –732
Isenhour JL, Hanley M, Marx JA. Hematocolpometra manifesting as constipation in the young female. Acad Emerg Med. 1999;7 :752 –753
Kitapci F, Avsar AG, Senses DA. A girl with constipation and acute urinary retention. Eur J Pediatr. 1999;158 :337 –338
Hall DJ. An unusual case of urinary retention due to imperforate hymen. J Accid Emerg Med. 1999;16 :232 –233
Peterson-Sweeney KL, Stevens J. 13-year-old female with imperforate hymen. Nurse Pract. 1996;21 :90 –94
Nisanian AC. Hematocolpometra presenting as urinary retention: a case report. J Reprod Med. 1993;38 :57 –64
Wort SJ, Heman-Ackah C, Davies A. Acute urinary retention in the young female. Br J Urol. 1995;76 :659 –670
Yu T, Lin M. Acute urinary retention in two patients with imperforate hymen. Scand J Urol Nephrol. 1993;27 :543 –544
Hammond Y. Haematocolpos. Nurs Times. 1970;66 :1131 –1132
Bakos O, Berglund L. Imperforate hymen and ruptured hematosalpinx: a case report with a review of the literature. J Adolesc Health. 1999;24 :226 –228
Gazit E, Frand M, Mashiah S, Rotem Y. Imperforate hymen causing pyocolpos in an infant. Clin Pediatr. 1975;14 :414 –415
Siegberg R, Tenhunen A, Ylostalo P. Diagnosis of mucocolpos and hematocolpos by ultrasound: two case reports. J Clin Ultrasound. 1985;13 :421 –423
Philbois O, Guye E, Richard O, et al. Role of laparoscopy in vaginal malformation. Surg Endosc. 2004;18 :87 –91
Christian C, Decker JM. Prepubertal genital examination. In: Henretig FM, King C, eds. Textbook of Pediatric Emergency Procedures. Philadelphia, PA: Lippincott Williams & Wilkins; 1997:961–970
Berenson AB, Grady J. A longitudinal study of hymenal development from 3 to 9 years of age. Pediatrics. 2002;140 :600 –607
Berenson AB. A longitudinal study of hymenal morphology in the first three years of life. Pediatrics. 1995;95 :490 –496(Jill C. Posner, MD, MSCE )
ABSTRACT
Objective. Although it is detectable at all ages through inspection of the external genitalia, imperforate hymen (IH) is a diagnosis that is missed commonly. We hypothesized that children with late diagnoses (predefined as 8 years of age, chosen to reflect the timing of normal menarche) would be more likely to be symptomatic, undergo more diagnostic testing, and lack appropriate documentation in their medical records, compared with children with earlier diagnoses (ie, <8 years of age).
Methods. All patients with IH were identified through searches of 3 hospital databases with International Classification of Diseases, Ninth Revision, codes. The medical records of eligible subjects were reviewed by a single, blinded researcher. Comparisons were made between children diagnosed at younger versus older ages.
Results. A bimodal distribution of age at diagnosis was demonstrated; 43% (n = 10) of girls were diagnosed at <8 years of age, and 57% (n = 13) were diagnosed at 8 years of age. Among older girls, 100% were symptomatic (abdominal pain and/or urinary symptoms; duration of symptoms: 1–120 days), whereas 90% of cases in the younger group were detected incidentally. Documentation was lacking for breast development (77%), pubic hair development (69%), and menstrual history (46%) among the older girls. Older children were more likely to present symptomatically (odds ratio: 42.0; 95% confidence interval: 3.1–1965.7) and to undergo ancillary testing (odds ratio: 20.3; 95% confidence interval: 1.6–983.1).
Conclusions. Two distinct populations of girls with IH exist, ie, those diagnosed without symptoms at a young age and those not diagnosed until >8 years of age. By incorporating an examination of the external genitalia into their routine practice, clinicians caring for children can prevent the significant delays in diagnosis, misdiagnosis, and morbidity associated with the latter group.
Key Words: imperforate hymen hydrocolpos hydrometrocolpos hematocolpos hematometrocolpos
Abbreviations: IH, imperforate hymen
Imperforate hymen (IH) is a rare congenital anomaly in which the hymenal membrane occludes outflow from the female genital tract. This obstruction results in an accumulation of uterine and vaginal secretions. Large amounts of retained fluid have been reported.1,2 The presentations of patients with IH can be variable. There are reports of cases detected incidentally1,3,4 or with symptoms of abdominal1,5–9 or back10–12 pain, constipation,13,14 dysuria or acute urinary retention,14–20 peritonitis,21,22 or primary amenorrhea,23 as well as a report of acquired IH associated with sexual abuse.24
Menarche is a relatively late event in the pubertal process, and most girls have attained Tanner stage 3 at the time of menarche. A discrepancy between a teen's advanced stage of pubertal development and the absence of menarche is a key clinical clue to the possibility of IH. Despite the ease of diagnosis at all ages with simple inspection of the external genitalia, the detection of IH can be delayed or missed, resulting in preventable morbidity and unnecessary diagnostic testing.
Previous works regarding IH consisted of case reports or small case series and therefore were limited in the ability to allow definitive conclusions. To our knowledge, no previous work systematically investigated a population of children with IH. The objective of this study was to describe a population of patients with IH who were treated at a large, tertiary care, referral children's hospital. Specifically, we aimed to characterize the presenting features and diagnostic evaluations performed and to assess the frequency with which the menstrual history, pubertal staging, and hymenal examination results were documented in the medical evaluation. We hypothesized that children with late diagnoses (predefined as 8 years of age, chosen to reflect the timing of normal menarche) would be more likely to be symptomatic, undergo more diagnostic testing, and lack appropriate documentation in their medical records, compared with children with earlier diagnoses (ie, <8 years of age).
METHODS
We conducted a retrospective, case-series study of all patients who were diagnosed with and treated for IH at a large, tertiary care, children's hospital during a 13-year period (January 1, 1987, through December 31, 2000). Eligible patients were identified through a search of the hospital's health information management (ie, medical records) database with International Classification of Diseases, Ninth Revision, codes for IH and related terms (hydrocolpos, hydrometrocolpos, hematocolpos, and hematometrocolpos). Additional searches of the urologic procedures database and the operative anesthesia database were performed to supplement the primary search. We used this strategy to optimize the identification of patients with IH. These databases were not mutually exclusive. Because we aimed to describe the presentation of uncomplicated IH, patients with genetic syndromes or coexistent genital anomalies were excluded. The medical records of identified subjects were abstracted for data pertaining to the presentation and diagnostic evaluation. All available components of the hospital medical records were subject to review, including emergency department records, preoperative evaluation findings, operative reports, and inpatient records. Demographic features, presenting symptoms, documentation of Tanner stage, hymenal examination results, menstrual history, preliminary diagnosis, ancillary test findings, operative procedure, and final diagnosis were recorded on a standardized data collection sheet by a single researcher, who was blinded with respect to the study hypothesis.
The sample size was based on identification of all eligible patients. Statistical analyses were conducted with Stata 7.0 software (Stata Corp, College Station, TX). The sample was described with means, SDs, medians, and ranges for continuous variables and frequencies for categorical variables. Between-group comparisons (ie, older versus younger girls) were performed with exact 2 tests.
This study received institutional review board approval. After chart abstraction had been completed, all data were deidentified.
RESULTS
Computerized searches of the databases resulted in the identification of 95 girls. Fourteen patients were excluded because of multiple congenital anomalies or genetic syndromes associated with IH. Of the remaining 81 patients, 58 were excluded because of diagnoses other than IH, most commonly dysfunctional uterine bleeding. There were 23 patients in the final study sample, 74% (n = 17) white, 22% (n = 5) black, and 4% (n = 1) Asian (Table 1). The distribution of age at diagnosis demonstrated a clear bimodal distribution that was even more dramatic than results with the predefined cutoff point of 8 years (Fig 1); 43% of patients (n = 10) were diagnosed at <4 years of age (median age: 1.2 years; range: 9 days to 3.7 years), and the remaining 57% (n = 13) were diagnosed at >10 years of age (median age: 12.3 years; range: 10.9–14.2 years).
In the older group, 100% of patients (n = 13) were symptomatic at the time of diagnosis. In the younger group, almost all patients (90%, n = 9) were asymptomatic. One girl in the younger group was diagnosed during an evaluation because of dysuria at 3 years of age. Although it was unclear whether the presence of an IH predisposed this patient to develop dysuria, we considered this patient to be symptomatic for the purposes of analysis.
In the group of older girls, all presented with abdominal pain (n = 6), urinary symptoms (n = 2), or both abdominal pain and urinary symptoms (n = 5). In the same group, the median duration of symptoms before diagnosis was 15 days (range: 1–120 days). Almost one half of the older girls (46%, n = 6) were given preliminary diagnoses other than IH, including urinary tract infection (n = 2), appendicitis (n = 2), nephrolithiasis (n = 1), and abdominal tumor (n = 1). In the older group, the medical records lacked documentation of breast development in 77% of cases (n = 10), pubic hair development in 69% (n = 9), and menstrual history in 46% (n = 6). Most of the older girls (85%, n = 11) underwent ancillary testing before definitive diagnosis, including blood analyses (n = 10), urinalyses (n = 9), and abdominal plain radiographic evaluations (n = 4). In comparison, the only patient in the younger group to undergo diagnostic testing before definitive diagnosis was the child diagnosed as having vesiculoureteral reflux after a previous urinary tract infection.
Ultrasonographic imaging was performed for 15 of the 23 patients. Interestingly, all 3 of the ultrasonographic evaluations performed in the younger group were to confirm diagnoses of IH that were made first through physical examination; however, 8 of the 12 ultrasonographic evaluations performed in the older group were part of diagnostic evaluations that led to the eventual diagnosis of IH. Of note, there was no documentation of any clinical suspicion of sexual abuse in either of the groups.
Compared with the younger children, the older children were more likely to present symptomatically (odds ratio: 42.0; 95% confidence interval: 3.1–1965.7; P < .001) and undergo laboratory or plain radiographic testing (odds ratio: 20.3; 95% confidence interval: 1.6–983.1; P < .001). All children underwent hymenectomy at or near the time of diagnosis, with the notable exception of 2 young children for whom surgery was delayed for several years.
DISCUSSION
In this study, the first large case series of patients with IH exclusively, we demonstrated that 2 distinct subpopulations of girls with IH exist, on the basis of the age at diagnosis and the presenting clinical features. One group was composed of younger girls, nearly all of whom were asymptomatic and were diagnosed incidentally through physical examination alone. The other group consisted of older girls, all of whom presented symptomatically. Although we planned in our analyses to compare girls 8 years of age with those <8 years of age (coinciding with the lower limit of normal puberty), a more striking dichotomy in age was observed. All patients in the younger group were <4 years of age, whereas the older group consisted of girls 10 years of age. This finding is consistent with the age distribution reported in a case series that aimed to describe the utility of laparoscopy in the evaluation of vaginal malformations.25
Girls in the older group commonly experienced protracted symptoms before being diagnosed as having IH. Although this might be attributable, in part, to a delay in seeking care by the girl and/or her family, our results indicate that the medical evaluation might also be contributory. Nearly one half of the older girls were assigned an incorrect preliminary diagnosis. In addition, this group was significantly more likely to undergo unnecessary laboratory and radiographic studies, further delaying the eventual diagnosis. Furthermore, we found that a substantial proportion of the older girls had incompletely documented histories and physical examination results and frequently lacked appropriate pubertal evaluations. Specifically, no genital examination was performed for two thirds of the older patients, and there was no breast examination for three fourths. Almost one half of the older patients were not asked about menstrual history.
The diagnosis of IH can be made from a physical examination alone, eliminating the need for expensive laboratory testing or radiographic evaluations. Observation of the hymen and viewing into the vaginal vault are best accomplished by applying gentle labial traction with the child in the supine frog-leg or knee-chest position.26 An uncomfortable speculum examination is not indicated, because mere observation of a bulging mass at the vaginal introitus for an older patient (Fig 2) or, ideally, of a membrane obstructing the view into the vaginal vault for an infant or young child establishes the diagnosis. Our results indicate, however, that some patients with IH experience delays in diagnosis, misdiagnosis, and additional diagnostic evaluations. It can be surmised that, if each older girl had undergone a complete examination of the external genitalia as part of a routine well-child visit early in her life, then none would have had to endure the symptoms and diagnostic evaluations associated with late diagnosis. It is our recommendation, therefore, that the hymenal examination be emphasized as a routine part of newborn and infant health maintenance evaluations. In addition, routine periodic hymenal reexaminations should be performed to document the hymenal changes that may occur with normal development.27,28
Treatment of IH consists of surgical hymenectomy, releasing the obstructing hymenal tissue. The majority of patients in our study were treated surgically at or near the time of diagnosis, although there were 2 notable exceptions of patients who underwent delayed treatment. The optimal timing for this procedure is unclear. However, it is possible that patients with IH may be at higher risk of endometriosis.25 Therefore, treating physicians must weigh the risks of anesthesia for young children against the potential risks resulting from delayed correction.
There are limitations to this study that should be considered in interpreting our results. First, as with all retrospectively conducted investigations, there is a potential for information bias. It is possible that key components of the history and physical examination were actually performed but were not documented in the medical record. This could have led to an overestimation of incomplete pubertal evaluations for older girls. However, the fact that older girls experienced significant delays in diagnosis and underwent additional testing supports the validity of our results. Second, ascertainment bias might have occurred, resulting in incomplete identification of all eligible patients. However, we minimized the potential for this bias by using 3 hospital databases (not mutually exclusive) to ensure inclusion of all patients diagnosed with IH during the study period. Third, although this study of IH is the largest to date, the small sample size led to wide confidence intervals around the odds ratios. Finally, this study was undertaken at a tertiary children's hospital; therefore, the results might not be reflective of all children with IH.
Clinicians caring for children and teens should incorporate inspection of the female genitalia into their routine practices. If clinicians detect IH among young children, they can prevent the morbidity associated with delays in diagnosis, including the development of symptoms and additional diagnostic testing. For older girls presenting with abdominal or urinary symptoms, the complete medical evaluation must include a review of the menstrual history and an examination of the breasts and external genitalia. A high index of suspicion for IH should be maintained, especially if a discrepancy is noted between the Tanner stage of pubertal development and a reported lack of menarche.
ACKNOWLEDGMENTS
This work was presented in part at the Ambulatory Pediatrics Association Regional Conference; March 5, 2002; Philadelphia, PA.
We thank Gina Ruggierio for her assistance with data management.
FOOTNOTES
Accepted Sep 1, 2004.
No conflict of interest declared.
REFERENCES
Bogen DL, Gehris, RP, Bellinger MF. Picture of the month. Arch Pediatr Adolesc Med. 2000;154 :959 –960
Loscalzo IL, Catapano M, Loscalzo J, Sama A. Imperforate hymen with bilateral hydronephrosis: an unusual emergency department diagnosis. J Emerg Med. 1995;13 :337 –339
Henderson KC. Hydrometrocolpos in a newborn. Am J Dis Child. 1975;129 :1190 –1191
Kahn R, Duncan B, Bowes W. Spontaneous opening of congenital imperforate hymen. J Pediatr. 1975;87 :768 –770
Ward A, Maher P. Hematocolpos: an unusual presentation. Br J Clin Pract. 1979;33 :83 –84
Dickson CA, Saad S, Tesar JD. Imperforate hymen with hematocolpos. Ann Emerg Med. 1985;14 :467 –469
Maffeo DA, Santagata P, Santagata M. Hematocopos by imperforated hymen: case report. Clin Exp Obstet Gynecol. 1999;26 :230 –231
Schneider K, Hong J, Fong J, Sanders CG. Hematocolpos as an easily overlooked diagnosis. Curr Opin Pediatr. 1999;11 :249 –252
McIvor RA. Haematocolpos, once-in-a-lifetime cause of recurrent abdominal pain. Arch Emerg Med. 1990;7 :51 –57
London NJ, Sefton GK. Hematocolpos: an unusual cause of sciatica in an adolescent girl. Spine. 1996;21 :1381 –1382
Buick RG, Chowdhary SK. Backache: a rare diagnosis and unusual complication. Pediatr Surg Int. 1999;15 :586 –587
Letts M, Haasbeek J. Hematocolpos as a cause of back pain in premenarchal adolescents. J Pediatr Orthop. 1990;10 :731 –732
Isenhour JL, Hanley M, Marx JA. Hematocolpometra manifesting as constipation in the young female. Acad Emerg Med. 1999;7 :752 –753
Kitapci F, Avsar AG, Senses DA. A girl with constipation and acute urinary retention. Eur J Pediatr. 1999;158 :337 –338
Hall DJ. An unusual case of urinary retention due to imperforate hymen. J Accid Emerg Med. 1999;16 :232 –233
Peterson-Sweeney KL, Stevens J. 13-year-old female with imperforate hymen. Nurse Pract. 1996;21 :90 –94
Nisanian AC. Hematocolpometra presenting as urinary retention: a case report. J Reprod Med. 1993;38 :57 –64
Wort SJ, Heman-Ackah C, Davies A. Acute urinary retention in the young female. Br J Urol. 1995;76 :659 –670
Yu T, Lin M. Acute urinary retention in two patients with imperforate hymen. Scand J Urol Nephrol. 1993;27 :543 –544
Hammond Y. Haematocolpos. Nurs Times. 1970;66 :1131 –1132
Bakos O, Berglund L. Imperforate hymen and ruptured hematosalpinx: a case report with a review of the literature. J Adolesc Health. 1999;24 :226 –228
Gazit E, Frand M, Mashiah S, Rotem Y. Imperforate hymen causing pyocolpos in an infant. Clin Pediatr. 1975;14 :414 –415
Siegberg R, Tenhunen A, Ylostalo P. Diagnosis of mucocolpos and hematocolpos by ultrasound: two case reports. J Clin Ultrasound. 1985;13 :421 –423
Philbois O, Guye E, Richard O, et al. Role of laparoscopy in vaginal malformation. Surg Endosc. 2004;18 :87 –91
Christian C, Decker JM. Prepubertal genital examination. In: Henretig FM, King C, eds. Textbook of Pediatric Emergency Procedures. Philadelphia, PA: Lippincott Williams & Wilkins; 1997:961–970
Berenson AB, Grady J. A longitudinal study of hymenal development from 3 to 9 years of age. Pediatrics. 2002;140 :600 –607
Berenson AB. A longitudinal study of hymenal morphology in the first three years of life. Pediatrics. 1995;95 :490 –496(Jill C. Posner, MD, MSCE )