The accuracy of the rectovaginal examination in detecting cul-de-sac disease in patients under general anaesthesia

K.G. Dragisic1, L.A. Padilla2 and M.P. Milad1,3

1 Department of Obstetrics and Gynecology, Northwestern University School of Medicine, Chicago, Illinois and 2 Department of Obstetrics and Gynecology, University of New Mexico, Santa Fe, New Mexico, USA

3 To whom correspondence should be addressed at: Northwestern University Medical School 333 East Superior Street, Room 183 Chicago, IL 60611, USA. Email: mmilad{at}nmh.org


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
BACKGROUND: The rectovaginal examination is frequently used as an adjunct to a bimanual examination. The accuracy of rectovaginal examination in detecting cul-de-sac disease under ideal circumstances of the operating room was studied. METHODS: Fifty-two attending physicians and 30 residents were selected on experience and immediate availability to evaluate the presence of uterosacral nodularity and external rectal compression on 140 women undergoing general anaesthesia, followed by laparoscopy or laparotomy, at an academic medical centre. Physicians were masked to the indications for surgery and procedure ranging from diagnostic laparoscopy to laparotomy for suspected malignancy. RESULTS: Mean patient body mass index (BMI) was 26.1 kg/m2 and nearly one-third had had previous surgery. Based on the surgical findings, cul-de-sac disease was common, including uterosacral nodularity (5.8%) and rectal compression (10.1%). Both sensitivity and positive predictive value of the rectovaginal examination for detecting uterosacral nodularity were zero, whereas specificity approached 95%. Accuracy in detecting rectal compression was somewhat better with a sensitivity of 34%, specificity 96.7% and positive predictive value 55.6%. Neither examiner-years of experience nor the laterality of the dominant or examining hand affected the accuracy of the rectovaginal examination. CONCLUSIONS: The rectovaginal examination has marked limitations despite the controlled circumstances of the operating room including general anaesthesia, an empty bladder and ideal patient positioning. As suspected, the specificity of the rectal examination is high due to the low prevalence of disease. However, the sensitivity of the rectovaginal examination is very low, limiting its capacity as a screening test.

Key words: cul-de-sac disease/endometriosis/rectovaginal examination/uterosacral nodularity


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The rectovaginal examination is considered to be an important part of a pelvic examination (Seidel et al., 1995Go). When compared with the bimanual examination, the rectovaginal examination is thought to allow the examiner to reach almost 2.5 cm higher in the pelvis, enabling a better evaluation of pelvic viscera. On average, the examining finger extends 7–8 cm and has access to most of the rectal mucosa, which is 11 cm in length. The rectovaginal examination permits palpation of the rectovaginal septum for thickness and masses (Seidel et al., 1995Go; Mishell et al., 1997Go) and of the uterosacral ligaments for thickening or nodularity (Mishell et al., 1997Go).

Although the rectovaginal examination theoretically may provide additional information during a routine pelvic examination, little data exists validating the ability of the rectovaginal examination to detect pelvic disease. The purpose of our study was to estimate the accuracy of the rectovaginal examination to detect cul-de-sac disease under the ideal circumstances of the operating room.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
This prospective cohort study was conducted at Prentice Women’s Hospital of Northwestern Memorial Hospital, Chicago, Illinois, USA from March 1997 to March 1998 with institutional review board approval. Women presenting for laparoscopy or laparotomy consented to have a bimanual examination and rectovaginal examination under anaesthesia. The subjects had many indications for surgery, from diagnostic or sterilization procedures to laparotomy for suspected pelvic malignancy. There was no requirement for pre-operative bowel preparation.

Examiners included attending gynaecologists and gynaecology residents who were not aware of the patients’ symptoms or indications for surgery. When a patient was brought to the operating room, examiners were selected by immediate availability and by experience. No single physician accounted for more than 9% of total examinations. Examiners were blinded to the preoperative diagnosis, indications for surgery and surgical findings.

Pelvic examination conditions were standardized. Once general anaesthesia was attained, the patient was placed in the dorsal lithotomy position using Allen stirrups (Allen Medical Systems, Bedford Heights, OH, USA). The bladder was emptied by straight catheterization. In addition to a bimanual examination, each examiner performed a rectovaginal examination and was asked to record perceived findings, including the presence of external rectal compression and uterosacral nodularity.

Post-operatively, surgeons completed forms that described surgical findings, which entailed the same variables assessed during the rectovaginal examination. Demographic examiner information was also recorded including gender, dominant hand, and years of postgraduate training.

Data were analysed using a Microsoft Excel database (Microsoft Corp, Seattle, WA, USA), EpiInfo2000 (CDC, Atlanta, GA, USA), and SPSS 10.1 (SPSS, Inc., Chicago, IL, USA). Parametric continuous variables were assessed using the Student’s t-test and non-parametric analysis with Mann–Whitney U-test. Sensitivity, specificity, and positive and negative predictive value were expressed as proportions and analysed with {chi}2-test.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
A total of 270 rectovaginal examinations were performed on 140 patients with a mean (± SEM) age of 39.3 ± 0.9 years, and a minority over age 50 (n = 14, 10%). Patients had a variety of indications for surgery ranging from diagnostic laparoscopy to laparotomy for suspected malignancy. The mean body mass index (BMI) was 26.1 kg/m2 and nearly one-third had prior surgeries (Table I).


View this table:
[in this window]
[in a new window]
 
Table I. Patient characteristics and surgical findings
 
Of examiners, 63.4% were attending gynaecologists and 36.6% were gynaecology residents, and 46.3% were female. A total of 44% of examiners were right-handed, and 42.7% of examiners performed right-handed rectovaginal examinations. Mean years of postgraduate experience were 8.62 ± 9.2.

Cul-de-sac disease was a frequent intraoperative finding among study patients, including uterosacral nodularity (5.8%) and rectal compression (10.1%). Both the sensitivity and the positive predictive value for detecting uterosacral nodularity were 0% (Table II). The specificity of the rectovaginal examination for uterosacral nodularity was 95.3% and the negative predictive value was 94%. The sensitivity of the rectovaginal examination for detecting rectal compression was 34.5%, with 96.7% specificity and 55.6 % positive predictive value and the negative predictive value was 93% (Table III).


View this table:
[in this window]
[in a new window]
 
Table II. Accuracy of surgical findings for uterosacral nodularity
 

View this table:
[in this window]
[in a new window]
 
Table III. Accuracy of surgical findings for rectal compression
 
Examiners participating in the study included attending physicians (n = 52) and residents (n = 30). Residents and attending physicians did not differ in their ability to detect uterosacral nodularity (P = 0.13) or rectal compression (P = 0.26). Similarly, examiner gender did not influence the ability to detect uterosacral nodularity (P = 0.3). However, female examiners were more frequently correct in identifying rectal compression than were their male counterparts (P = 0.048).

Handedness does not appear to play a role in the accuracy of the rectovaginal examination. Right-handed examinations by right-handed physicians were not more likely than left-handed examinations by left-handed physicians to be accurate in the detection of uterosacral nodularity (P = 0.65) or rectal compression (P = 0.11). There also does not appear to be an association between use of dominant hand and accuracy: use of the dominant hand in the rectovaginal examination was not more accurate in the detection of uterosacral nodularity (Right hand, P = 0.81; Left hand, P = 0.94), or rectal compression (Right hand, P = 0.88; Left hand, P = 0.97).

Receiver operating characteristic (ROC) curves are often used to study the diagnostic accuracy of a given test and to depict the pattern of sensitivities and specificities observed when the performance of the test is evaluated at several different diagnostic thresholds. Figure 1 depicts the receiver operating characteristic curve for uterosacral nodularity and rectal compression. The area under the curve for nodularity was 0.458 and for rectal compression 0.495. The overall diagnostic performance of the rectovaginal examination is poor since the line approaches the rising diagonal.



View larger version (11K):
[in this window]
[in a new window]
 
Figure 1. Receiver Operator Characteristics Curve for the rectovaginal examination and its ability to correctly detect rectal compression and uterosacral nodularity.

 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
When discussing the role of a given test it is important to distinguish between a screening and confirmatory test. Screening tests are employed to evaluate groups of individuals for subclinical disease. For a test to be useful at screening it must have a high sensitivity; this is in contrast to a confirmatory test, which must have high specificity. The rectovaginal examination has been used for both functions. However, supporting evidence for these practices is scarce for these indications.

Authors (Padilla et al., 2000Go) have questioned the function of the pelvic examination as a primary screening test for asymptomatic adnexal disease. Similarly, our study questions the use of the rectovaginal examination as a screening test for asymptomatic cul-de-sac disease. The routine use of the digital rectal examination for colorectal cancer has been shown to be an ineffective screening test. As a result, both the American Cancer Society and the US Preventive Services Task Force do not recommend routine digital rectal examinationss in their guidelines for colorectal cancer screening (US Preventive Services Task Force, 1996Go).

Other authors have questioned the utility of the rectal examination as part of a pelvic examination. A recent study retrospectively evaluated the diagnostic yield of rectal examinations as part of routine pelvic examinations and concluded that in young patients the use of the rectal examination as part of a routine pelvic examination was not warranted (Campbell and Shaughnessy, 1998Go). Similarly, while evidence from a randomized clinical trial (Mandel et al., 1993Go) indicates that annual faecal occult blood testing in persons >50 years, reduces colorectal cancer mortality, the rectovaginal examination alone does not contribute to detection rates for colon cancer.

When a patient presents with symptoms consistent with a disease state, a targeted confirmatory test is often employed. A confirmatory test should have high specificity. Some argue that the rectovaginal examination is a necessary confirmatory test to evaluate patients with suspected disease, such as endometriosis. The majority of endometrial implants are located in the dependent parts of the female pelvis, including the ovaries, cul-de-sac, and the uterosacral and broad ligaments, with ~10–15% of women with advanced disease having lesions involving the rectosigmoid (Adamson, 1990Go). On pelvic exam, signs of endometriosis have been reported as cul-de-sac nodularity; a fixed, retroverted uterus (Muse, 1998Go; Luciano and Pitkin, 1984Go); and an obstructed pouch of Douglas (Muse, 1998Go). Matorras et al. studied the clinical signs of endometriosis in infertile women and correlated their physical examination with surgical findings (Matorras et al., 1996Go). They enrolled 602 consecutive infertile women with either abnormal physical findings or after six controlled ovarian stimulation cycles. A pelvic examination was first performed followed by laparoscopy. Of the 174 infertile women with laparoscopically confirmed endometriosis, 6.3% had uterosacral nodularity, 19% had a retroverted uterus and 5.8% had an obstructed pouch of Douglas by physical examination. The sensitivity of the rectovaginal examination findings of uterosacral nodularity and an obstructed cul-de-sac for endometriosis was low, 6.3 and 5.8% respectively. As with our study, the specificity for detection of both was quite high. They found a positive predictive value for the detection of uterosacral nodularity was 100%.

The accuracy of the rectovaginal examination in confirming a suspected disease state may depend upon the severity of the disease. Koninckx et al. found that a bimanual examination with positive findings for painful nodularities correlated strongly and significantly with the presence of deep endometriosis, cystic ovarian endometriosis and severe cul-de-sac adhesions (Koninckx et al., 1996Go). Painful nodularity in the cul-de-sac was not entirely specific for deep endometriosis or an endometrioma.

Additionally, the utilization of adjunct studies improved accuracy. Koninckx et al found that the sensitivity of pelvic nodularity during menses for a deep endometriotic lesion, a cystic ovarian endometrioma or severe pelvic adhesions was increased from 79.2 to 87.5% by utilizing a CA-125 concentration >35 IU/ml (Koninckx et al., 1996Go).

Some have suggested that the accuracy of an examination for endometriosis may depend upon the location of the disease. One study found that focal tenderness had the highest sensitivity and specificity in the uterosacral ligaments (Ripps and Martin, 1992Go). Eskenazi et al. found that a positive examination for endometriosis, which included uterosacral pain or nodularity, fixed or painful adnexal mass, and cul-de-sac pain or nodularity, correctly classified the patient’s endometriosis disease status 44% of the time (Eskenazi et al., 2001Go). Only six (38%) of 16 patients with non-ovarian endometriosis were correctly diagnosed preoperatively by examination, compared with 100% with ovarian endometriosis. The authors concluded that examination was poorly predictive of non-ovarian endometriosis, even after ultrasound and patient symptomatology were incorporated.

Our study does have limitations. Despite the optimal circumstances with the abdomen lax, bladder empty and lower extremities abducted, the rectovaginal examination was insufficient to diagnose cul-de-sac disease. We did not analyse the accuracy of the examination for different disease severities, at a certain time of the menstrual cycle, or for rectal compression due to ovarian and non-ovarian disease. We did not ask examiners to record the size of uterosacral nodules, and thus we are not able to conclude if the accuracy of the rectovaginal examination varies with the size of the nodule. The small number of positive findings such as uterosacral nodularity also limits the validity of the study. The high number of false positives suggests that there may have been some degree of bias for examiners; they were aware they were partaking in a study and thus may have been more likely to find abnormalities. We also did not incorporate the use of adjunct pre-operative testing such as CA-125 or ultrasound. Additionally, some may suggest that conscious patients can help to improve examination accuracy by identifying tender areas. Chapron et al. recently reported their experience in the conscious patient using the vaginal digital examination to detect cul-de-sac endometriosis (Chapron et al., 2002Go). Among 102 women with surgically diagnosed deeply infiltrating endometriosis, vaginal examination was a poor indicator for detecting painful nodules (33%) and painful induration (50%). The authors concluded that routine clinical examination is not sufficient to diagnose deeply infiltrative endometriosis.

The rectovaginal examination has marked limitations despite the controlled circumstances of the operating room, including general anaesthesia, an empty bladder and ideal position. As suspected, the specificity of the rectal examination is high due to the low prevalence of disease. However, the sensitivity of the rectovaginal examination is very low limiting its capacity as a screening test for these indications. It is possible that paraclinical studies such as rectal ultrasound (Schröder et al., 1997Go) or magnetic resonance imaging (Kinkel et al., 1999Go) may be more sensitive and should be studied in this context. Also, our study was conducted in patients who were unable to communicate with us and who had muscular relaxation. The inability to detect tender areas in anaesthetized patients limits the applicability of our study to general office practice.


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Adamson, G.D. (1990) Diagnosis and clinical presentation of endometriosis. Am. J. Obstet. Gynecol., 162, 568–569.[ISI][Medline]

Campbell, K.A. and Shaughnessy, A.F. (1998) Diagnostic utility of the digital rectal examination as part of the routine pelvic examination. J. Fam. Prac., 46, 165–167.[ISI][Medline]

Chapron, C., Dubuisson, J.B., Pansini, V., Vieria, M., Fauconnier, A., Barakat, H. and Dousset, B. (2002) Routine clinical examination is not sufficient for diagnosing and locating deeply infiltrating endometriosis. J. Am. Assoc. Gynecol. Laparosc., 9, 115–119.[ISI][Medline]

Eskenazi, B., Warner, M., Bonsignore, L., Olive, D., Samuels, S. and Vercellini, P. (2001) Validation study of nonsurgical diagnosis of endometriosis. Fertil. Steril., 76, 929–935.[CrossRef][ISI][Medline]

Kinkel, K., Chapron, C., Balleyguier, C., Fritel, X., Dubuisson,.B. and Moreau, J.F. (1999) Magnetic resonance imaging characteristics of deep endometriosis. Hum. Reprod., 14, 1080–1086.[Abstract/Free Full Text]

Koninckx, P.R., Meuleman C., Oosterlynck D., Cornillie F.J. (1996) Diagnosis of deep endometriosis by clinical examination during menstruation and plasma CA-125 concentration. Fertil. Steril., 65, 280–287.[ISI][Medline]

Luciano, A.A. and Pitkin R.M. (1984) Endometriosis: approaches to diagnosis and treatment. Surg. Ann., 16, 297–312.[Medline]

Mandel, J.S., Bond, J.H., Church, T.R., Snover, D.C. Bradley, G.M., Schuman, L.M. and Ederer, F. (1993) Reducing mortality from colorectal cancer by screening for fecal occult blood. N. Engl. J. Med., 328, 1365–1371.[Abstract/Free Full Text]

Matorras, R., Rodriguez, F., Pijoan, J.I., Soto, E., Perez, C., Ramon, O. and Rodriguez-Escudero, F. (1996) Are there any clinical signs and symptoms that are related to endometriosis in infertile women? Am. J. Obstet. Gynecol., 174, 620–623.[ISI][Medline]

Mishell, D.R., Stenchever, M.A., Droegemueller, W. and Herbst, A.L. (1997) Comprehensive Gynecology. 3rd edn., Mosby-Year Book, St. Louis, MO, USA.

Muse, K. (1998) Clinical manifestations and classification of endometriosis. Clin. Obstet. Gynecol., 31, 813–822.

Padilla, L.A., Radosevich, D.M. and Milad, M.P. (2000) Accuracy of the rectovaginal examination in detecting adnexal masses. Obstet. Gynecol., 96, 593–598.[Abstract/Free Full Text]

Ripps, B.A. and Martin, D.C. (1992) Correlation of focal pelvic tenderness with implant dimension and stage of endometriosis. J. Reprod. Med., 37, 620–624.[ISI][Medline]

Schröder, J., Löhnert, M., Doniec, J.M. and Dohrmann, P. (1997) Endoluminal ultrasound diagnosis and operative management of rectal endometriosis. Dis. Colon Rectum, 40, 614–617.[ISI][Medline]

Seidel, H.M., Ball, J.W., Dains, J.E. and Benedict, G.W.(1995) Mosby’s Guide to Physical Examination. 3rd edn, Mosby-Year Book, St. Louis, MO., USA.

US Preventive Services Task Force. (1996) Guide to clinical preventive services. 2nd edn, Williams & Wilkins, Baltimore, MD, USA.

Submitted on February 10, 2003; accepted on May 7, 2003.