Validity of Ultrasound in Patients with Acute Pelvic Pain Related to Suspected Ovarian Torsion

doi:10.4236/ss.2011.26069

Paper Menu >>

Journal Menu >>

Surgical Science, 2011, 2, 326-330

doi:10.4236/ss.2011.26069 Published Online August 2011 (http://www.SciRP.org/journal/ss)

Validity of Ultrasound in Patients with Acute Pelvic Pain

Related to Suspected Ovarian Torsion

Leena Mawaldi1, Charu Gupta1, Hanadi Bakhsh1, Maissa Saadeh2, Mostafa A. Abolfotouh2*

1Department of Obstetrics and G ynaecology, King AbdulAziz Medical City, National Guard Health Affairs,

Riyadh, Saudi Arabia

2King Abdullah International Medical Research Center (KAIMRC), King Saud Bin-Abdulaziz University fo r

Health Sciences, National Guard Health Affairs, Riyadh, Saudi Arabia

*E-mail: [email protected] m

Received April 27, 2011; revised July 2, 2011; accepted July 18, 2011

Abstract

Objective: Ultrasound has been proven to be useful in detecting underlying ovarian pathology. However, its

role in the prediction of ovarian torsion has been controversial. The aim of the study was to assess the valid-

ity of ultrasound in the prediction of ovarian torsion in patients with acute pelvic pain related to clinically

suspected ovarian torsion. Methods: A retrospective observational study was conducted at the Ob/Gyn de-

partment using a 10-year chart review of all female patients older than 11 years of age with highly suspected

ovarian torsion who underwent clinical assessment and ultrasound prior to surgery (n = 62). The sensitivity

and specificity of ultrasound were determined by cross-tabulation of the ultrasound and surgical findings.

Results: Of the suspected cases, 54 (87.1%) were confirmed to be cases of ovarian torsion by surgery. The

majority of the cases were suggestive of ovarian torsion, which was indicated by clinical examination

(77.4%), ultrasound (77.4%), or pathological examination (79%). Almost one-half of the cases (46.8%)

showed a pain score >6; two-thirds (62.9%) presented with vomiting and/or nausea; and more than one-third

(38.7%) presented with leukocytosis. The estimated sensitivity and specificity of ultrasound were 0.74 and

0.0, respectively. The positive predictive value was 0.83. Ultrasound was significantly associated with both

clinical examination (p = 0.039) and pain score (p = 0.008). Conclusion: The diagnosis of ovarian torsion

cannot be exclusively based on ultrasound. Both clinical and sonographical evaluation of acute pelvic pain

should be considered for the diagnosis. A definitive diagnosis remains challenging.

Keywords: Acute Pelvic Pain, Ultrasound, Ovarian Torsion

1. Introduction

Ovarian torsion refers to the twisting of the ovary on its

ligamentous supports, which often results in an imped-

ance of its blood supply. This condition is the fifth most

common gynecological emergency and affects females

of all ages [1]. Expedient diagnosis is important to pre-

serve ovarian function and prevent adverse sequelae [2];

however, the diagnosis can be challenging because the

symptoms are relatively nonspecific. The ovaries were

examined in a large series of patients with surgically

confirmed torsion-associated cysts in 48% and neopla-

sms in 46%; the remainder occurred in normal-appearing

ovaries [3]. Histopathology was benign in over 90% of

patients [3-4]. Whereas anatomic factors usually account

for ovarian torsion in adults, normal ovaries have been

demonstrated in over 50% of ovarian torsion cases in

children under the age of 15 years [5]. In adults, torsion

has also been described following laparoscopic hyster-

ectomy, suggesting that even release of the fulcrum, on

which the ovaries usually twist, does not protect against

torsion [6]. Strenuous exercise or a sudden increase in

abdominal pressure also promotes torsion of the ovary

around the vascular pedicle [7]. The right ovary is more

likely than the left to underg o torsion, suggestin g that the

sigmoid colon may help to prevent torsion [8].

Women who are pregnant [9-10] or are undergoing

ovarian hyperstimulation during infertility treatment are

at increased risk of ovarian torsion [11]. The overall in-

cidence of torsion in pregnant women was reported as

L. MAWALDI ET AL.327

15% [12]. In association with pregnancy, torsion most

commonly occurred between 10 and 17 weeks of gesta-

tion and during the postpartum period. A much lower

incidence was reported in another series of pregnant

women [13-14]. Torsion was half as common as appen-

dicitis during pregnancy [15]. The clinical presentation

of ovarian torsion is nonspecific, and therefore, it is a

challenge for the clinician to recognize this conditio n and

differentiate it from other etiologies.

The two most common presenting features of ovarian

torsion are acute pelvic pain (83%) and an adnexal mass

(72%) [16]. Other symptoms and findings include nausea

and vomiting (70%), stabbing pain (70%), sudden and

sharp pain in the lower abdomen (59%), pain radiating to

the back, flank, or groin (51%), peritoneal signs (3%),

fever (<2%), leukocytosis, and an increased level of In-

terleukin-6 [17,18] .

Ultrasound can detect adnexal lesions and ovarian

enlargement. An enlarged, heterogeneous-appearing ovary

is the most common ultrasound finding [19]; however,

the presence of normal-appearing ovaries does not rule

out the diagnosis [20]. Doppler ultrasound is also con-

troversial [21] because it shows diminishing or absent

ovarian vessel flow in two-dimensional color [21]. The

aim of this study was to assess the valid ity of ultrasound

in the prediction of ovarian torsion in patients with acute

pelvic pain suggestive of ovarian torsion.

2. Methods

A retrospective observational study was conducted in the

Department of Obstetrics and Gynaecology, King Ab-

dulAziz Medical City, Riyadh, Saudi Arabia. All female

patients (n = 62) older than 11 years of age who pre-

sented with acute pelvic pain with highly suspected

ovarian torsion from January 2000 through December

2009 and for whom surgeries were performed within less

than 6 hours of the ultrasound assessment were in cluded.

Patients underwent a transabdominal scan using 3 - 5

MHz probes and a 5 - 7 MHz endovaginal transducer

with both real-time gray-scale and color Doppler imag-

ing. The data were collected by chart review for all pa-

tients.

All categorical variables (age, marital status, pregnancy,

pain score, vomiting, white blood cell counts, abdomi-

nal/pelvic exam, ultrasound, and surgery) were recorded,

and their frequency distributi ons were measured.

Sensitivity, specificity, and predictive values were de-

termined by cross-tabulation of the results of ultrasound

and the surgical findings of ovarian torsion as the gold

standard. Based on this tabulation, the sensitivity, speci-

ficity and positive predictive values were computed for

ultrasound. The sensitivity of ultrasound diagnosis com-

pared with the surgical diagnosis “gold standard” was

determined by calculating how frequently the correct

ultrasound diagnosis was made in each surgical diagnosis.

The specificity of the ultrasound diagnosis was deter-

mined by calculating how frequently the ultrasound di-

agnosis was not made when the corresponding surgical

diagnosis was not present. Positive predictability indi-

cated how frequently the ultrasound diagnosis correctly

reflected the surgical diagnosis. In addition, the level of

agreement between the ultrasound diagnosis and the sur-

gical diagnosis was determined by calculating the kappa

coefficient.

The data were analyzed using SPSS version 15.0. All

of the statistical tests were considered significant at a

P-value < 0.05

3. Results

Table 1 shows a summary of the sociodemographic and

clinical characteristics of the 62 cases of suspected ovar-

ian torsion. Of these cases, 54 (87.1%) were confirmed

as cases of ovarian torsion by surgery. The majority of

the cases were suggestive of ovarian torsion, as deter-

mined by clinical examination (77.4%), ultrasound

(77.4%), or pathological examination (79%). Almost

one-half of the cases (46.8%) were associated with a pain

score > 6; two-thirds (62.9%) presented with vomiting

and/or nausea; and more than one-third (38.7%) pre-

sented with leukocytosis.

Table 2 shows that when ultrasound is used for pre-

dicting ovarian torsion, the sensitivity is 74%. That is,

the ultrasound scan correctly diagnosed 74% of ovarian

torsion cases and missed 26% of these cases (false nega-

tives). However, free subjects were misclassified as

ovarian torsion cases (false positives).

Table 3 shows that ultrasound was significantly asso-

ciated with clinical examination (p = 0.039) and pain

score (p = 0.008).

Table 4 shows that positive ultrasound alone is not

predictive of ovarian torsion (PPV = 0.0). Additionally,

examination and pain scores have 100% PPV each, yet

the yield is only 13% and 3.7%, respectively. When re-

lying upon ultrasound and one of these two techniques,

the PPV reaches 87%, and the yield is 74%.

4. Discussion

Ovarian torsion is an uncommon condition; however, it

is the most common gynecological surgical emergency

and has an overall incidence of 2.7% [22]. Awareness of

its clinical and sonographical features may enable

prompt treatment that can spare the ovary. The possible

consequences of delayed diag nosis are serious, including

L. MAWALDI ET AL.

328

Table 1. Sociodemographic and clinical characteristics of

the 62 cases of suspected ovarian torsion.

Demographics N %

Age Group

11 - 18 years 23 37.1

19 and older 39 62.9

Marital status

Married 31 50.0

Single 31 50.0

Pregnant

Yes 10 16.1

No 52 83.9

Pain Score

Less than 6 33 53.2

More than 6 29 46.8

Vomiting & Nausea

Yes 39 62.9

No 23 37.1

White Blood cells

Increased 24 38.7

Normal 38 61.3

Clinical Examination

Positive 48 77.4

Negative 14 22.6

Pathology

Yes 49 79.0

No 13 21.0

Ultrasound

Positive 48 77.4

Negative 14 22.6

Definitive diagnosis

Positive 54 87.1

Negative 8 12.9

ovarian necrosis, peritonitis, and death. Underlying pa-

thology was found in 79% of the cases. Enlargement of

the ovary was the common predisposing factor, including

polycystic ovary, functional cyst, and ovarian hyper-

stimulation syndrome. There were cases of ovarian tor-

sion without ovarian pathology due to an elongated

Table 2. 2 × 2 table of the ultrasound results and the defini-

tive diagnosis of ovarian torsion.

Ovarian Torsion

Ultrasound Positive Negative Total

Positive 40 8 48

Negative 14 0 14

Total 54 8 62

Sensitivity: 40/54 = 74 %; Specificity: 0/8 = 0.0%; PPV: 40/45 = 83.3%;

NPV: 0/14 = 0 %; Kappa: 0.20, P-value: 0.102.

Table 3. Association of the results of ultrasound and a de-

finitive diagnosis of ovarian torsion with patient character-

istics.

Variable Ultrasound Diagnosis

Age 0.61b 0.12b

Marital Status 0.068b 0.71a

Pregnant 0.30b 0.47b

Pain Score 0.008b* 1.00a

Vomiting & nausea 0.90b 0.42b

White blood cells 0.32b 0.47a

Clinical examinat ion 0.039b* 0.86b

Pathology 0.43b 0.22b

*p < 0.05; aFisher’s exact test was applied, bChi-square test was applied.

utero-ovarian ligament in 21% of the cases.

Ultrasound has been proven to be useful in detecting

any underlying ovarian pathology. Lee et al. [21] con-

cluded that identification of the twisted vascular pedicle

through ultrasonogr aphy is suggestive of ovarian tors ion,

and color Doppler sonography could be helpful in pre-

dicting the viability of adnexal structures by depicting

blood flow within the twisted vascular pedicle. Gray-

scale findings typically include asymmetric enlargement,

a solid heterogeneous appearance, and peripheral cystic

areas; however, ultrasound was less capable of deter-

mining ovarian torsion. It shows a sensitivity of 0.74

(resulting in a false reassurance of 26%), and a specific-

ity of 0.0 (resulting in a false warning of 100%), and a

PPV of 0.83 (suggesting it might be useful in the clinical

setting). Pena et al. [23] concluded that abnormal flow

detected by Doppler sonography is highly predictive of

adnexal torsion and is therefore useful in the diagnosis of

ovarian torsion; however, the detection of normal flow

does not necessarily exclude ovarian torsion. Thus, our

results are similar to those of international studies.

The study by Cohen et al. [24] concluded that patients

and surgeons alike should be aware of the difficulty in

L. MAWALDI ET AL.

329

Table 4. Predictive value and yield of different criteria for the diagnosis of ovarian torsion.

Criteria for Diagnosis Positive Negative Total PPV (%) Yield (%)

US only 0 1 1 0.0 0 .0

Exam. only 7 0 7 100.0 13.0

Pain score only 2 0 2 100.0 3.7

US + exam 14 3 17 82.4 25.9

US + pain score 6 1 7 85.7 11.1

US + pain score + exam 20 3 23 87.0 37.0

74%*

Pain + exam 1 0 1 100.0 1.9

All negatives 4 0 4 100.0 7.4

Total 54 8 62

*This figure reflects the yield of ultrasound in addition to examination and\or p ain score.

making accurate preoperative diagnoses of acute gyne-

cologic pathologies in the emergency room. In the pre-

sent study, a pain score > 6 (100.0) and an abdominopel-

vic examination (100.0) are highly predictive of the di-

agnosis. However, the yield of cases was very low (13%

for examination and 3.7% for pain score).

Bouguizane et al. [9] concluded th at clinicians must be

aware of possible adnexal torsion in women with acute

pelvic pain; ultrasound is a useful tool in these situations

(10). In the present stud y, the predictive value of US and

pain score is 85.7%, but the number of cases is very low

(11.1%).

Ignacioa et al. [25] concluded that an ultrasound im-

age can usually be used to make a diagnosis in conjunc-

tion with clinical parameters; however, this is most dif-

ficult in patients with ovarian torsion. In the present

study, the predictive value of US with only examination

is 82.4%, but again, the yield is low (25.9%). However,

when relying upon US in conjunction with either exami-

nation and/or the pain score, the yield reached 74%.

These were the only two variables that h ad positive asso-

ciations with ultrasound (p = 0.008 for the pain score and

p = 0.039 for the examination). Clinical assessment by

nausea-vomiting, pain score, age group, marital status,

and pregnancy status was not statistically significant in

predicting ovarian torsion, although the PPV was high

for leukocytosis (0.92), existing ovarian pathology (0.89),

and nausea-vomiting (0.87).

5. Conclusions

From the collective findings of this study, and consider-

ing its limitations in terms of sample size, it is concluded

that the definitive diagnosis of ovarian torsion remains

challenging. Both clinical and sonographical evaluation

of acute pelvic pain should be considered for the diagno-

sis of ovarian torsion. The diagnosis cannot be exclu-

sively based on ultrasound only, on the presence or ab-

sence of colo r flow Doppler, or even on the morp hologi-

cal findings. Therefore, surgical intervention is recom-

mended in suspicions of a nonviable ovary in order to

decrease the morbidity.

6. Acknowledgements

The study was approved by the research committee of

King Abdullah International Medical Research Center

(KAIMRC), King Saud Bin-Abdulaziz University for

Health Sciences. Special thanks go to Mr. Mahmoud

Salam, the research coordinator at KAIMRC, for helping

in reference formatting. We also thank the editing service

office of KAIMRC for the English language editing of

the manuscript via a specialist English language copy

editor (American Journal Experts).

7. Conflict of Interest

The authors declare no conflicts of interest.

8. References

[1] D. Houry and J. T. Abbott, “Ovarian Torsion: A Fifteen-

Year Review,” Annals of Emergency Medicine, Vol. 38,

No. 2, 2001, pp. 156-159. doi:10.1067/mem.2001.114303

[2] J. H. Becker, J. De Graff and C. M. Vos, “Torsion of the

Ovary: A Known but Frequently Missed Diagnosis,”

European Journal of Emergency Medicine , Vol. 16, No. 3,

2009, pp. 124-126. doi:10.1097/MEJ.0b013e32831cbaf8

[3] M. Varras, A. Tsikini, D. Polyzos, Ch. Samara, G. Had-

jopoulos and Ch. Akrivis, “Uterine Adnexal Torsion:

Pathologic and Gray-Scale Ultrasonographic Findings,”

L. MAWALDI ET AL.

330

Clinical & Experimental Obstetrics & Gynecology, Vol.

31, No. 1, 2004, pp. 34-38.

[4] M. Sommerville, D. A. Grimes, P. P. Koonings and K.

Campbell, “Ovarian Neoplasms and the Risk of Adnexal

Torsion,” American Journal of Obstetrics & Gynecology,

Vol. 164, No. 2, 1991, pp. 577-578.

[5] J. F. Anders and E. C. Powell, “Urgency of Evaluation

and Outcome of Acute Ovarian Torsion in Pediatric Pa-

tients,” Archives of Pediatrics & Adolescent Medicine,

Vol. 159, No. 6, 2005, pp. 532-535.

doi:10.1001/archpedi.159.6.532

[6] R. Mashiach, M. Canis, K. Jardon, G. Mage, J. L. Pouly

and A. Wattiez, “Adenexal Torsion after Laparoscopic

Hysterectomy: Description of Seven Cases,” The Journal

of the American Association of Gynecologic Laparo-

scopists, Vol. 11, No. 3, 2004, pp. 336-339.

doi:10.1016/S1074-3804(05)60046-1

[7] E. D. Littman, J. Rydfors and A. A. Milki, “Exer-

cise-Induced Ovarian Torsion in the Cycle Following

Gonadotrophin Therapy: Case Report,” Oxford Journals:

Human Reproduction, Vol. 18, No. 8, 2003, pp.

1641-1642. doi:10.1093/humrep/deg338

[8] M. Beaunoyer, J. Chapdelaine, S. Bochard and A. Ouimet,

“Asynchronous Bilateral Ovarian Torsion,” Journal of

Pediatric Surgery, Vol. 39, No. 5, 2004, pp. 746-749.

doi:10.1016/j.jpedsurg.2004.01.037

[9] S. Bouguizane, H. Bibi, Y. Farhat, S. Dhifallah, F. Darraji,

S. Hidar, L. Lassoued, A. Chaieb and H. Khairi, “Adnexal

Torsion: A Report of 1 35 Cases,” Journal de Gynécologie,

Obstétrique et Biologie de l a Reproductio n, Vol. 32, No. 6,

2003, pp. 535-540.

[10] C. Born, S. Wirth, A. Stabler and M. Reiser, “Diagnosis

of Adnexal Torsion in the Third Trimester of Pregnancy:

A Case Report,” Abdominal Imaging Journal, Vol. 29,

No. 1, 2004, pp. 123-127.

[11] H. Gorkemli, M. Camus and K. Clasen, “Adnexal Tor-

sion after Gonadotrophin Ovulation Induction for IVF or

ICSI and Its Conservative Treatment,” Archives of Gy-

necology and Obstetrics, Vol. 267, No. 1, 2002, pp. 4-6.

doi:10.1007/s00404-001-0251-x

[12] C. F. Yen, S. L. Lin, W. Murk, C.-J. Wang, C.-L. Lee, Y.

K. Soong, M. Phil and A. Arici, “Risk Analysis of torsion

an d Malignancy for Adnexal Masses during Pregnancy,”

Fertility Sterility Home, Vol. 91, No. 5, 2009, pp. 1895-

1902. doi:10.1016/j.fertnstert.2008.02.014

[13] B. Bromley and B. Benacerraf, “Adnexal Masses during

Pregnancy: Accuracy of Sonographic Diagnosis and

Outcome,” Journal of Ultrasound in Medicine, Vol. 16,

No. 7, 1997, pp. 447-452.

[14] K. M. Schmeler, W. W. Mayo-Smith, J. F. Peipert, S.

Weitzen, M. D. Manuel and M. E. Gordinier, “Adnexal

Masses in Pregnancy: Surgery Compared with Observa-

tion,” The American College of Obstetricians and Gyne-

cologists, Vol. 105, No. 5, 2005, pp. 1098-1103.

[15] T. R. Johanson Jr. and J. D. Woodruff, “Surgical Emer-

gencies of the Uterine Adnexae during Pregnancy,” In-

ternational Journal of Gynecology & Obstetrics, Vol. 24,

No. 5, 1986, pp. 331-335.

doi:10.1016/0020-7292(86)90150-5

[16] A. I. Bayer and A. K. Wiskind, “Adnexal Torsion: Can

the Adnexa Be Saved?” American Journal of Obstetrics

& Gynecology, Vol. 171, No. 6, 1994, pp. 1506-1510.

[17] A. B. Pinto, V. S. Ratts, D. B. Williams, S. L. Keller and R.

R. Odem, “Reduction of Ovarian Torsion 1 Week after

Embryo Transfer in a Patient with Bilateral Hype- rstimu-

lated Ovaries,” Fertility Sterility Home, Vol. 76, No. 2,

2001, pp. 403-406. doi:10.1016/S0015-0282(01)01910-0

[18] S. B. Cohen, A. Wattiez, D. Stockheim, D. S. Seidman, A.

L. Lidor, S. Mashiach and M. Goldenberg, “The Accu-

racy of Serum Interleukin-6 and Tumour Necrosis Factor

as Markers for Ovarian Torsion,” Oxford Journals: Hu-

man Reproduction, Vol. 16, No. 10, 2001, pp. 2195-2197.

doi:10.1093/humrep/16.10.2195

[19] S. Servaes, D. Zurakowski, M. R. Laufer, N. Feins and J.

S. Chow, “Sonographic findings of ovarian torsion in

children,” Journal of Pediatric Radiology, Vol. 37, No. 5,

2007, pp. 446-451.

doi:10.1007/s00247-007-0429-x

[20] M. Ben-Ami, Y. Perlitz and S. Haddad, “The Effective-

ness of Spectral and Color Doppler in Predicting Ovarian

Torsion. A Prospective Study,” European Journal of Ob-

stetrics & Gynecology and Reproductive Biology, Vol.

104, No. 1, 2002, pp. 64-66.

doi:10.1016/S0301-2115(02)00056-8

[21] E .J. Lee, H. C. Kwon, H. J. Joo, J. H. Suh and A. C.

Fleischer, “Diagnosis of Ovarian Torsion with Color

Doppler Sonography: Depiction of Twisted Vascular

Pedicle,” Journal of Ultrasound in Medicine, Vol. 17, No.

2, 1998, pp. 83-89.

[22] L. T. Hibbard, “Adnexal Torsion,” American Journal of

Obstetrics & Gynecology, Vol. 152, No. 4, 1985, pp.

456-461.

[23] J. E. Pena, D. Ufberg, N. Coony and A. L. Denis, “Use-

fulness of Doppler Sonography in the Diagnosis of Ovar-

ian Torsion,” Fertility Sterility Home, Vol. 73, No. 5,

2000, pp. 1047-1050.

doi:10.1016/S0015-0282(00)00487-8

[24] S. B. Cohen, B. Weisz, D. S. Seidman, S. Mashiach, A. L.

Lidor and M. Goldenberg, “Accuracy of the Preoperative

Diagnosis in 100 Emergency Laparoscopies Performed

Due to Acute Abdomen in Non-pregnant Women,” The

Journal of the American A ssociation of Gynecolog ic Lapa-

roscopist s, Vol. 8, No. 1, 2001, pp. 92-94.

doi:10.1016/S1074-3804(05)60555-5

[25] E. A. Ignacio and M. C. Hill, “Ultrasound of the Acute

Female Pelvis,” Ultrasound Quarterly, Vol. 19, No. 2,

2003, pp. 86-98.

doi:10.1097/00013644-200306000-00004