Human Reproduction, Vol.28, No.6 pp.

1569– 1579, 2013

Advanced Access publication on March 15, 2013 doi:10.1093/humrep/det065

ORIGINAL ARTICLE Gynaecology

Ultrasound for diagnosing acute

salpingitis: a prospective observational
diagnostic study
G. Romosan 1,*, C. Bjartling 1, L. Skoog 2, and L. Valentin 1

Downloaded from https://academic.oup.com/humrep/article-abstract/28/6/1569/604182 by guest on 22 May 2019

1
Department of Obstetrics and Gynaecology, Skåne University Hospital Malmö, Lund University, Malmö 205 02, Sweden
2
Department of Clinical Pathology Malmö, Laboratory Medicine, Malmö 205 02, Sweden

*Correspondence address. Tel: +4640336948; Fax: +4640962600; E-mail: gina.romosan@gmail.com

Submitted on October 31, 2012; resubmitted on January 8, 2013; accepted on February 18, 2013

study question: What are the diagnostic benefits of using ultrasound in patients with a clinical suspicion of acute salpingitis and signs
of pelvic inflammatory disease (PID)?
summary answer: In patients with a clinical suspicion of acute salpingitis, the absence of bilateral adnexal masses at ultrasound
decreases the odds of mild-to-severe acute salpingitis about five times, while the presence of bilateral adnexal masses increases the odds
about five times.
what is known already: PID is difficult to diagnose because the symptoms are often subtle and mild. The diagnosis is usually
based on clinical findings, and these are unspecific. The sensitivity and specificity of ultrasound with regard to salpingitis have been reported
in one study (n ¼ 30) of appropriate design, where most patients had severe salpingitis (i.e. pyosalpinx) or tubo-ovarian abscess.
study design, size, duration: This diagnostic test study included 52 patients fulfilling the clinical criteria of PID. Patients were
recruited between October 1999 and August 2008.
participants/materials, setting, methods: The patients underwent a standardized transvaginal gray scale and Doppler
ultrasound examination by one experienced sonologist (index test) before diagnostic laparoscopy by a laparoscopist blinded to the
ultrasound results. The final diagnosis was determined by laparoscopy, histology of the endometrium and other histology where relevant
(reference standard).
main results and the role of chance: Of the 52 patients, 23 (44%) had a final diagnosis unrelated to genital infection,
while the other 29 had cervicitis (n ¼ 3), endometritis (n ¼ 9) or salpingitis (n ¼ 17; mild n ¼ 4, moderate n ¼ 8, severe, i.e. pyosalpinx
n ¼ 5). Bilateral adnexal masses and bilateral masses lying adjacent to the ovary were seen more often on ultrasound in patients with
salpingitis than with other diagnoses (bilateral adnexal masses: 82 versus 17%, i.e. 14/17 versus 6/35, P ¼ 0.000, positive likelihood ratio
4.8, negative likelihood ratio 0.22; bilateral masses adjacent to ovary: 65 versus 17%, i.e.11/17 versus 6/35, P ¼ 0.001, positive likelihood
ratio 3.8, negative likelihood ratio 0.42). In cases of salpingitis, the masses lying adjacent to the ovaries were on average 2– 3 cm in diameter,
solid (n ¼ 14), unilocular cystic (n ¼ 4), multilocular cystic (n ¼ 3) or multilocular solid (n ¼ 1), with thick walls and well vascularized at
colour Doppler. In no case were the cogwheel sign or incomplete septae seen. All 13 cases of moderate or severe salpingitis were diagnosed
with ultrasound (detection rate 100%, 95% confidence interval 78–100%) compared with 1 of 4 cases of mild salpingitis. Three of six cases of
appendicitis, and two of two ovarian cysts were correctly diagnosed with ultrasound, and one case of adnexal torsion was suspected and then
verified at laparoscopy.
limitations, reasons for caution: The sample size is small. This is explained by difficulties with patient recruitment. There
are few cases of mild salpingitis, which means that we cannot estimate with any precision the ability of ultrasound to detect very early
salpingitis. The proportion of cases with salpingitis of different grade affects the sensitivity and specificity of ultrasound, and the sensitivity
and specificity that we report here are applicable only to patient populations similar to ours.
wider implications of the findings: The information provided by transvaginal ultrasound is likely to be of help when
deciding whether or not to proceed with diagnostic laparoscopy in patients with symptoms and signs suggesting PID and, if laparoscopy
is not performed, to select treatment and plan follow-up.

& The Author 2013. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved.
For Permissions, please email: journals.permissions@oup.com
1570
study funding/competing interest(s): This work was supported by funds administered by Malmö University Hospital and
two Swedish governmental grants (ALF-medel and Landstingsfinansierad Regional Forskning). The authors have no conflict of interest.
Key words: salpingitis / ultrasonography / Doppler ultrasonography / emergency medicine / sensitivity/specificity
Introduction
Pelvic inflammatory disease (PID) is difficult to diagnose because the
symptoms are often subtle and mild. Because there are no precise
tests for PID, a diagnosis is usually based on clinical findings, but clinical
diagnosis of PID is hampered by the lack of specificity of signs and
symptoms. On physical examination, pelvic and abdominal tenderness,
abnormal cervical secretions and fever are findings associated with PID
(CDC, 2010). However, in a study by Jacobson and Westrom (1969),
only 65% of patients with a clinical diagnosis of PID had salpingitis con-
firmed when laparoscopy was performed, 23% had normal findings
and in the remaining 12% laparoscopy revealed pathologic conditions
unrelated to PID (acute appendicitis, ectopic pregnancy, pelvic endo-
metriosis and several other pelvic disorders) (Jacobson and Westrom,
1969). Palpable adnexal fullness or mass is a common finding in
women with salpingitis and is related to the severity of inflammation
as determined by laparoscopy. However, palpable adnexal fullness
or mass is also reported in some women with normal findings on
laparoscopy (Jacobson and Westrom, 1969).
Transvaginal ultrasound has become increasingly common as an aid
to establish a correct diagnosis in women with acute pelvic pain
(Okaro and Valentin, 2004). Ultrasound findings suggestive of pyosal-
pinx have been described, i.e. a pear-shaped fluid-filled structure with
thick walls, presence of incomplete septae and cogwheel sign (Timor-
Tritsch et al., 1998). Moreover, increased vascularisation as deter-
mined by Doppler ultrasound has been reported in cases of
inflammation-induced hyperemia in the tubes (Alatas et al., 1996;
Molander et al., 2001).
The aim of this study was to (i) describe ultrasound findings in cases
of acute mild, moderate and severe salpingitis verified by laparoscopy
and (ii) to estimate the sensitivity and specificity of transvaginal ultra-
sound for diagnosing acute salpingitis in patients with clinical signs
of PID.
Methods
This is a prospective observational study. Consecutive patients scheduled
for diagnostic laparoscopy at the department of Obstetrics and Gynaecol-
ogy, Skåne University Hospital, Malmö, Sweden, because of a clinical sus-
picion of acute salpingitis were eligible for inclusion. The clinical
examination in the emergency room upon which the doctor based his/
her clinical diagnosis included speculum examination, wet smear and gy-
naecological palpation. Presenting symptoms and clinical findings were
documented prospectively in a research protocol by the doctor in the
emergency room. In addition, blood was drawn for analysis of C-reactive
protein (CRP) and body temperature was measured. Salpingitis was sus-
pected if the following criteria were fulfilled: acute abdominal pain for
1 – 14 days, cervical motion tenderness and adnexal tenderness at pelvic
bimanual examination, negative pregnancy test and at least one of the fol-
lowing three signs: pathological saline prepared vaginal wet smear (more
Romosan et al.
leucocytes than epithelial cells in the absence of clue-cells and inflamma-
tory vaginitis) or pathological discharge at speculum examination, elevated
CRP, oral temperature .38.08C. These criteria are similar to the criteria
of PID of the Centers for disease control and prevention, USA (CDC,
2010). According to both the policy of our department and our research
Downloaded from https://academic.oup.com/humrep/article-abstract/28/6/1569/604182 by guest on 22 May 2019
protocol, all patients fulfilling these criteria should undergo diagnostic
laparoscopy.
Tests for Neisseria gonorrhoeae (N. gonorrhoeae) and Chlamydia trachoma-
tis (C. trachomatis) were performed. Neisseria gonorrhoeae was detected by
culture from a charcoal-treated cotton swab which was sent to the labora-
tory in Stuart’s transport medium. First void urine together with cervical
samples, or vaginal swabs together with urine, were collected to diagnose
C. trachomatis using PCR (Roche Molecular Diagnostics, Pleasanton, CA,
USA, or m2000, Abbott Molecular, Inc., Des Plaines, IL, USA).
Our inclusion criteria were age ≥18 years (the age of legal consent),
signed informed consent, no unequivocal alternative diagnosis to PID on
the basis of clinical evaluation and no ongoing treatment with antibiotics
or anti-inflammatory drugs, but antibiotic therapy started ,24 h before
the laparoscopy was accepted. Exclusion criteria were violation of the
study protocol or treatment with antibiotics .24 h before laparoscopy.
After a decision to perform laparoscopy had been taken by the phys-
ician in the emergency room, and after the patient had consented to par-
ticipate in the study, a standardized transvaginal gray scale and colour and
spectral Doppler ultrasound examination was performed by a gynaecolo-
gist with more than 10 years of experience in gynecological ultrasound
(LV). The ultrasound examiner knew that the eligibility criteria of the
study were fulfilled but did not have access to any other clinical informa-
tion. The ultrasound system used was a Sequoia 512 (Siemens Medical
Solutions, Inc., Ultrasound Division, Mountain View, CA, USA) equipped
with a 5 – 8 MHz transvaginal transducer. All women were examined in
the lithotomy position with an empty bladder. The uterus and adnexa
were scanned systematically following the research protocol. The pres-
ence/absence of fluid in the endometrial cavity, the cervical canal and
the pouch of Douglas was noted. The ovaries were described with
regard to the presence of corpus luteum, polycystic appearance and any
pathological intra-ovarian lesions. Any adnexal lesions were noted, and
the gray scale morphology of any such lesion was described using the ter-
minology of the International Ovarian Tumor Analysis group (Timmerman
et al., 2000). Subjective assessment, i.e. pattern recognition was also used
for evaluation of ultrasound findings (Valentin, 1999; Valentin, 2004) with
ultrasound signs reported to be specific for pyosalpinx (Timor-Tritsch
et al., 1998, Valentin, 2004) being searched for. Three orthogonal dia-
meters of any mass were measured. We report the size of masses as
the mean of three orthogonal diameters. The dynamic and interactive
nature of transvaginal ultrasound was made full use of for pain mapping
and for estimating the mobility of organs and lesions. After the gray
scale ultrasound examination had been completed, the ultrasound
system was switched into the colour Doppler mode, and the colour
content of the endometrium, myometrium and any adnexal mass sus-
pected to be a diseased tube was estimated subjectively by the ultrasound
examiner on a visual analog scale graded from 0 to 100. Standardized
colour Doppler settings were used (frequency 6 MHz; power Doppler
gain 50; dynamic range 10 dB; edge 1; persistence 2; colour map 1; gate
2; filter 3). Finally, blood flow velocities were measured in the uterine
Ultrasound for diagnosing acute salpingitis
arteries, in the tubal arteries at the place where the tube leaves the uterus
(Kirchler et al., 1992) and in the wall of any adnexal mass. Angle correction
was used when measuring blood flow velocities in the uterine arteries. For
the other vessels angle correction was not used, but the highest achievable
Doppler shift signals were sought for each vessel (Valentin et al., 1994).
Ultrasound images were documented on hard copies, on video tape or
electronically. Based on the subjective evaluation of the ultrasound find-
ings, the ultrasound examiner suggested a likely diagnosis. The ultrasound
results were entered into a research protocol. The ultrasound examiner
played no role in the clinical management of the patient, and all staff man-
aging the patient including the laparoscopist was blinded to the ultrasound
results.
After the ultrasound examination, laparoscopy was performed by
gynaecologists with different levels of expertise (senior registrar or consult-
ant). Immediately before the laparoscopy with the patient anaesthetizsed,
any intrauterine contraceptive device was removed, and an endometrial
sample was taken using both an EndoretteTM outpatient endometrial sam-
pling device (Medscand AB, Malmö, Sweden) and a currette. The Endor-
ette is a sterile device with a polyethylene piston which slides within a
straight but flexible polypropylene sheath with four lateral holes near its
tip. Its length is 285 mm and its outer diameter is 2.6 mm. The endomet-
rial samples were analyzed by one dedicated pathologist (L.S.) with the
specific aim of confirming or excluding endometritis. At laparoscopy, cul-
tures were taken from the tubal fimbriae and from fluid in the pouch of
Douglas for analysis of C. trachomatis and N. gonorrhoeae. The laparoscopy
was documented on video tape or on DVD. Immediately after the lapar-
oscopy, the laparoscopist described the laparoscopic findings in a standar-
dized research protocol.
The final diagnosis was determined by the authors on the basis of the
results of endometrial histology and findings at laparoscopy. A diagnosis
of acute salpingitis was made if the laparoscopic criteria of mild, moderate
or severe salpingitis as suggested by Hager et al. (1983) were fulfilled. The
minimal criteria for a diagnosis of salpingitis were tubal redness, tubal
edema and pus or exudate from the tubal fimbriae provoked by manipu-
lation (salpingitis grade 1, i.e. mild salpingitis). If, in addition to the minimal
criteria, pus was present spontaneously and the tubes were fixed and
closed, we classified the condition as salpingitis grade 2 (i.e. moderate sal-
pingitis). If there was a pyosalpinx or a tubo-ovarian abscess, the condition
was classified as salpingitis grade 3 (i.e. severe salpingitis) (Hager et al.,
1983). The diagnosis of endometritis was made when there were neu-
trophlic microabscesses plus infiltration and destruction of glandular epi-
thelium in the endometrial sample (acute endometritis) or infiltration of
plasma cells, histiocytes, lymphocytes and lymphoid follicles (chronic endo-
metritis) or both but no signs of salpingitis at laparoscopy (Sellors et al.,
1991; Blaustein and Kurman, 2002). A final diagnosis of cervicitis was
made if there were clinical signs of cervicitis but neither the criteria of
endometritis nor those of salpingitis were fulfilled.
We collected information regarding parity, gynecological history and use
of contraceptives from the patient records retrospectively.
Statistical analysis
Statistical calculations were undertaken using the Statistical Package for the
Social Sciences (SPSS, Inc., Chicago, Illinois, USA, version 16.0 or 17.0).
The statistical significance of a difference in unpaired proportions was
determined using the x 2 test or Fisher’s exact test as appropriate, and
the statistical significance of a difference in continuous unpaired data was
determined using the Mann– Whitney test. Exact 95% confidence intervals
(CI) for sensitivity and specificity were calculated. When the sensitivity was
100%, positive and negative likelihood ratios were calculated by adding 0.5
to all four fields in the four-field table. A P-value ,0.05 was considered
statistically significant.
1571
Ethical approval
The Ethics Committee of Lund University, Sweden, approved the study
protocol. Written informed consent was obtained from all participants,
after the nature of the procedures had been fully explained.
Results
Recruitment was from October 1999 until August 2008. A total of 85
patients scheduled for diagnostic laparoscopy because of a clinical sus-
picion of salpingitis consented to be included in the study, but 33 had
Downloaded from https://academic.oup.com/humrep/article-abstract/28/6/1569/604182 by guest on 22 May 2019
to be excluded: 5 because they were treated with antibiotics for
.48 h before the laparoscopy, 1 because the laparoscopy was
carried out 13 days after the ultrasound examination, 11 because
the planned laparoscopy was cancelled and 16 because the ultrasound
examiner was not available to perform the scan before the laparos-
copy. Thus, 52 patients were included.
Of the 52 patients included, 8 received antibiotics during the 24 h
preceding the ultrasound examination and the laparoscopy (three
cases with a final diagnosis of salpingitis or pyosalpinx and five cases
with a final diagnosis of pelvic pain with unknown etiology). The
median time between the ultrasound examination and the diagnostic
laparoscopy was 4.9 h (range 0.5 to 48 h).
The final diagnosis of the 52 patients included was salpingitis (17/
52, 32.7%), endometritis (9/52, 17.3%), cervicitis (3/52, 5.8%) and
other (23/52, 44.2%). Four of the 17 cases of salpingitis were mild,
8 were moderate and 5 were severe (i.e. pyosalpinx). No patient
had a tubo-ovarian abscess. The other diagnoses were pelvic pain
with unknown etiology (7/52, 13.5%), appendicitis (6/52, 11.5%),
peritoneal or ovarian endometriosis (2/52, 3.8%), ovarian cyst
(2/52, 3.8%), urinary tract infection (2/52, 3.8%), adnexal torsion
(2/52, 3.8%), i.e. one case of torsion of a hydrosalpinx and one of
torsion of a Morgagni hydatid, mucinous cystadenocarcinoma in the
appendix (1/52, 1.9%) and Crohn’s disease (1/52, 1.9%).
Demographic background data, symptoms, findings at clinical exam-
ination, results of cultures/PCR and endometrial histology are shown
in Table I. No patient was diagnosed with N. gonorrhoeae. More
patients with salpingitis and endometritis than with other diagnoses
complained of discharge (17/25 versus 8/26, P ¼ 0.012), but there
were no other obvious differences in either demographic background
data, symptoms or clinical findings between women with laparoscop-
ically confirmed salpingitis and those with other diagnoses. Patients
with salpingitis had the highest CRP values (median 75, range 15 –
204 versus 42, 8– 374, P ¼ 0.016).
Some ultrasound results are shown in Tables II and III. There was no
obvious difference in the presence of or the amount of fluid in the
pouch of Douglas between patients with salpingitis, endometritis
and other diagnoses. Polycystic appearance of the ovaries was not
more common in patients with salpingitis than in other patients.
Bilateral adnexal masses and bilateral adnexal masses lying adjacent
to the ovary were seen more often at ultrasound examination in
patients with salpingitis than with other diagnoses (14/17 versus
6/35, P ¼ 0.000; 11/17 versus 6/35, P ¼ 0.001). The colour score
of the endometrium was highest in patients with endometritis, while
pulsatility index (PI) values were lowest and time averaged
maximum velocities (TAMXV) were highest in the uterine and tubal
1572 Romosan et al.

Table I Demographic background data, symptoms and findings at clinical examination.

Final diagnosis
..................................................................................................................................
Salpingitis Endometritis Cervicitis Othersa Total (n 5 52)
(n 5 17) (n 5 9) (n 5 3) (n 5 23)
.............................................................................................................................................................................................
Hours between scan and laparoscopy 4.7 (0.5–48) 3.7 (1.5–24) 6.2 (2–8) 6.4 (1–48) 4.9 (0.5–48)
Antibiotic treatment before laparoscopyb 3/17 (17.6) 0/9 0/3 5/23 (21.7) 8/52 (15.4)
Age (years) 28 (20–50) 27 (21–30) 27 (20– 28) 28 (17–48) 28 (17– 50)
Nullipara 7/17 (41.2) 6/9 (66.7) 2/3 (66.7) 10/23 (43.5) 25/52 (48.1)
Current IUD 4/17 (23.5) 2/9 (22.2) 1/3 (33.3) 7/23 (30.4) 14/52 (26.9)

Downloaded from https://academic.oup.com/humrep/article-abstract/28/6/1569/604182 by guest on 22 May 2019

Current contraceptive pill 6/17 (35.3) 2/7 (28.6) 1/3 (33.3) 7/22 (31.8) 16/49 (32.7)
Previous salpingitis 2/17 (11.8) 0/8 0/3 2/21 (9.5) 4/49 (8.2)
Previous ectopic pregnancy 1/17(5.9) 0/9 0/3 0/23 1/52 (1.9)
Undergone appendectomy 0/17 1/8 (12.5) 1/3 (33.3) 2/23 (8.7) 4/51 (7.8)
Symptoms
Pelvic pain 17/17 (100) 9/9 (100) 3/3 (100) 23/23 (100) 52/52 (100)
Discharge 11/16 (68.8) 6/9 (66.7) 1/3 (33.3) 7/23 (30.4) 25/52 (48.1)
Bleeding 6/17 (35.3) 3/9 (33.3) 1/3 (33.3) 2/23 (8.7) 12/52 (23.1)
Clinical findings
General condition affected 9/16 (56.2) 7/9 (77.8) 2/2 (100) 18/23 (78.3) 36/50 (72.0)
Pathological discharge at speculum 12/17 (70.6) 9/9 (100) 3/3 (100) 19/23(83.6) 43/52 (82.7)
examination
Cervical tenderness 12/16 (75.0) 7/9 (77.8) 1/3 (33.3) 15/21 (71.4) 35/49 (71.4)
Uterine tenderness 14/17 (82.4) 9/9 (100) 3/3 (100) 20/23 (87.0) 46/52 (88.5)
Adnexal tenderness
Unilateral 3/17 (17.6) 1/9 (11.1) 3/3 (100) 10/23 (43.5) 17/52 (32.7)
Bilateral 14/17 (82.4) 8/9 (88.9) 0/3 13/23 (56.5) 35/52 (67.3)
Palpable pelvic mass 6/16 (37.5) 1/9 (11.1) 2/3 (66.7) 2/23 (8.7) 11/51 (21.6)
Temperature ≥38 C8 10/17 (58.8) 2/9 (22.2) 1/3 (33.3) 11/23 (47.8) 24/52 (46.2)
CRP .8 17/17 (100) 7/9 (77.7) 2/3 (66.7) 20/23 (87.0) 46/52 (88.4)
CRP 75 (15–204) 42 (8–145) 10 (8–47) 51 (8–374) 52 (8–374)
Chlamydia trachomatis positivec 6/17 (35.3) 2/9 (22.2) 0/3 0/22 8/51 (15.4)
Neisseria gonorrhoeae positivec 0/13 0/7 0/0 0/16 0/36
Endometrial histology compatible with 11/16 (68.8)d 9/9 (100) 0/2e 0/20f 20/47 (42.6)
endometritis

Results are shown as the median (range) or n (%). IUD, intrauterine device; CRP, C-reactive protein.
a
Pelvic pain with unknown etiology (n ¼ 7), appendicitis (n ¼ 6), peritoneal endometriosis (n ¼ 2), ovarian cyst (n ¼ 2), adnexal torsion (n ¼ 2), urinary tract infection (n ¼ 2),
Crohn’s disease (n ¼ 1), cystadenocarcinoma of the appendix (n ¼ 1).
b
Always ,24 h before laparoscopy.
c
Positive in cervix, urine or abdominal fluid.
d
In one case no endometrial sampling was performed, in two cases the endometrial biopsy yielded insufficient material, in three cases there were no signs of endometritis in the
endometrial sample but the laparoscopy findings clearly fulfilled the criteria of salpingitis (mild in two cases and moderate in one case).
e
In one case endometrial sampling was not performed.
f
In three cases endometrial sampling was not performed.

arteries in patients with salpingitis. However, there was substantial
overlap in Doppler results between patients with different diagnoses.
The diagnoses suggested by the ultrasound examiner are shown in
Table IV. The sensitivity with regard to acute salpingitis of subjective
interpretation of the ultrasound findings by the ultrasound examiner
was 82% (14/17, 95% CI 57 –96%), the specificity was 77% (27/35,
95% CI 60 –90%), the positive likelihood ratio was 3.6 and the nega-
tive likelihood ratio was 0.23. The sensitivity and specificity of
ultrasound findings of bilateral masses lying adjacent to the ovary
were 65% (11/17, 95% CI 38 –86%) and 83% (29/35, 95% CI 66–
93%), and the positive and negative likelihood ratios were 3.8 and
0.42, respectively. The corresponding figures for bilateral adnexal
masses were 82% (14/17, 95% CI 57 –96%), 83% (29/35, 95% CI
66– 93%), 4.8 and 0.22. Bilateral adnexal masses were found in all
13 patients with moderate or severe salpingitis but in only one of
four patients with mild salpingitis. Thus, the sensitivity and specificity
Ultrasound for diagnosing acute salpingitis 1573

Table II Characteristics of ultrasound examination and ultrasound findings.

Final diagnosis
.............................................................................................................................
Salpingitis (n 5 17) Endometritis (n 5 9) Othersa (n 5 26) Total (n 5 52)
.............................................................................................................................................................................................
Cycle day 14.8 (2–35) 19.7 (1–37) 16.6 (2–60)b 16.6 (1–60)
Fluid in the
Cervical canal 3/17 (17.2) 5/9 (55.6) 2/25 (8.0) 10/51 (19.6)
Endometrial cavity 5/17 (29.4) 3/9 (33.3) 0/26 8/52 (15.4)
Pouch of Douglas 14/17 (82.4) 7/9 (77.8) 16/26 (61.5) 37/52 (71.2)
Amount of fluid in the pouch of Douglas (mm) 10.0 (0–22) 8.8 (0–21) 8.5 (0–29) 9.0 (0–29)

Downloaded from https://academic.oup.com/humrep/article-abstract/28/6/1569/604182 by guest on 22 May 2019

Polycystic appearance of ovaryc
Right ovary 4/15 (26.7) 2/9 (22.2) 5/26 (19.2) 11/50 (22.0)
Left ovary 2/15 (13.3) 3/8 (37.5) 1/24 (4.2) 6/47 (12.8)
Bilateral masses adjacent to the ovary 11/17 (64.7) 3/9 (33.3) 3/26 (11.5) 17/52 (38.5)
Bilateral adnexal masses 14/17 (82.4) 3/9 (33.3) 3/26 (11.5) 20/52 (38.5)
Unilateral mass adjacent to the ovary 0/17 0/9 6/26 (23.1) 6/52 (11.5)
Unilateral ovarian cyst 0/17 0/9 2/26 (7.7) 2/52 (3.8)
No adnexal mass 3/17 (17.6) 6/9 (66.6) 15/26 (57.7) 24/52 (46.2)
Bilateral masses adjacent to the ovary
Both with colour score ≥70 5/10 (50)d 1/2 (50)e 1/3 (33.3) 7/15 (46.7)
One with colour score ≥70 1/10 (10) d
0/2 e
2/3 (77.8) 3/15 (20)
None with colour score ≥70 4/10 (40)d 1/2 (50)e 0/3 5/15 (33.3)
Bilateral adnexal masses
Both with colour score ≥70 5/12 (41.7)d,f 1/2e (50) 1/3 (33.3) 7/17 (41.2)
One with colour score ≥70 3/12 (25) d,f
0/2 e
2/3 (77.8) 5/17 (29.4)
None with colour score ≥70 4/12 (33.3)d,f 1/2e (50) 0/3 5/17 (29.4)

Results are shown as the median (range) or n (%).

a
Pelvic pain with unknown etiology (n ¼ 7), appendicitis (n ¼ 6), cervicitis (n ¼ 3), peritoneal endometriosis (n ¼ 2), ovarian cyst (n ¼ 2), adnexal torsion (n ¼ 2), urinary tract
infection (n ¼ 2), Crohn’s disease (n ¼ 1), cystadenocarcinoma of the appendix (n ¼ 1).
b
One woman in perimenopause (adnexal torsion) is not included in the calculations.
c
Polycystic appearance was defined as an ovary containing ≥10 follicles 2–10 mm in diameter.
d
The information on colour score was missing for the biggest mass in one patient with bilateral solid masses adjacent to the ovary.
e
The information on colour score was missing for both masses in one patient with bilateral solid masses adjacent to the ovary.
f
The information on colour score was missing for the smallest mass in one patient with bilateral multilocular-solid masses without a clearly discernable ovary.

of bilateral adnexal masses with regard to moderate or severe salpin-
gitis were 100% (13/13, 95% CI 78 –100%) and 82% (32/39, 95% CI
67 –91%) and the positive and negative likelihood ratios were 5.1 and
0.04, respectively.
Ultrasound findings in patients with mild
salpingitis (red, swollen tubes and pus or
exudate from the tubal fimbriae provoked
only by manipulation at laparoscopy)
The only patient with mild salpingitis that was diagnosed with ultra-
sound had bilateral solid masses with diameters of 19 and 23 mm, re-
spectively, lying adjacent to the ovary. The colour scores were 14 and
65, peak systolic velocities (PSVs) were 4 and 23 cm/s, TAMXV were
2 cm/s and 10 cm/s and PI were 1.60 and 2.03 in the smallest and
largest masses, respectively. In the remaining three patients with
mild salpingitis, no adnexal masses were seen with ultrasound.
Ultrasound findings in patients with
moderate salpingitis (red, swollen, fixed
and closed tubes, pus present spontaneously
at laparoscopy)
Seven of the eight patients with moderate salpingitis had bilateral
masses lying adjacent to the ovary, and the other patient had a
mass lying adjacent to the ovary on one side and a mass with no dis-
cernable ovary on the other side. Ten of the 15 masses lying adjacent
to the ovary were solid, 4 were unilocular cystic (sausage-shaped,
thick-walled unilocular structure with echogenic fluid inside, no cog-
wheel sign, no incomplete septae) and 1 was multilocular solid (thick-
walled roundish lesion filled with echogenic fluid, no cogwheel sign, no
incomplete septae). The mass without a discernable ovary was solid.
The size of the smallest mass lying adjacent to the ovary was median
18 mm (15– 23) and that of the largest mass 28 mm (24 –32). Median
colour scores in the smallest and largest masses were 75 (45 –99) and
92 (5–97), median PSVs were 17 cm/s (9–26) and 15 cm/s (8–47),
1574 Romosan et al.

Table III Results of Doppler ultrasound examination of the uterus and uterine and tubal arteries.

Final diagnosis
.....................................................................................................................................................
Salpingitis (n 5 17) Endometritis (n 5 9) Othersa (n 5 26) Total (n 5 52)
.............................................................................................................................................................................................
Colour score
Endometrium 20 (0– 99) 51 (3–92) 19 (0– 72) 22 (0–99)
Myometrium 51 (8– 98) 34 (19– 89) 55 (8– 93) 54 (8–98)
Spectral Doppler results
Right uterine artery
PSV (cm/s) 61 (17 –115) 40 (25– 51) 51 (20–97) 51 (17– 115)

Downloaded from https://academic.oup.com/humrep/article-abstract/28/6/1569/604182 by guest on 22 May 2019

TAMX, cm/s 20 (6– 46) 11 (4–19) 18 (8– 50) 17 (4–50)
PI 2.40 (1.47 –4.99) 3.00 (1.85– 5.89) 2.81 (1.31–4.89) 2.74 (1.31– 5.89)
Left uterine artery
PSV (cm/s) 59 (34 –127) 56 (37– 60) 50 (17–168) 52 (17– 168)
TAMXV, cm/s 26 (7– 50) 16 (7–20) 14 (5– 53) 18 (5–53)
PI 2.21 (1.29 –5.05) 2.67 (1.66– 4.14) 3.04 (1.51–5.02) 2.66 (1.29– 5.05)
Right tubal artery
PSV (cm/s) 21 (5– 57) 16 (10– 37) 24 (5– 46) 21 (5–57)
TAMXV (cm/s) 11 (2– 26) 7 (3–12) 8 (2– 65) 8 (2–65)
PI 1.71 (1.18 –3.53) 1.98 (1.65– 3.70) 2.08 (0.87–3.65) 2.04 (0.87– 3.70)
Left tubal artery
PSV (cm/s) 27 (5– 49) 26 (6–36) 18 (7– 35) 21(5–49)
TAMXV (cm/s) 13 (3– 26) 6 (2–15) 7 (2– 18) 7 (2–26)
PI 1.40 (1.02 –4.17) 2.65 (1.42– 4.59) 2.51(1.01–4.27) 2.29 (1.01– 4.59)

Results are shown as the median (range). PSV, peak systolic velocity; TAMXV, time averaged maximum velocity; PI, pulsatility index.
a
Pelvic pain with unknown etiology (n ¼ 7), appendicitis (n ¼ 6), cervicitis (n ¼ 3), peritoneal endometriosis (n ¼ 2), ovarian cyst (n ¼ 2), adnexal torsion (n ¼ 2), urinary tract
infection (n ¼ 2), Crohn’s disease (n ¼ 1) and cystadenocarcinoma of the appendix (n ¼ 1).

median TAMXV were 9 cm/s (5–16) and 9 cm/s (4 –27) and median
PI were 1.15 (0.78–2.00) and 1.16 (0.80– 1.46). The mass without a
discernable ovary measured 60 mm, had a colour score of 18, a PSV of
17 cm/s, a TAMXV of 12 cm/s and a PI of 0.63.
Ultrasound images from patients with laparoscopically confirmed
moderate salpingitis are shown in Figs 1–3.
Ultrasound findings in patients with severe
salpingitis (pyosalpinx at laparoscopy)
Three of the five patients with severe salpingitis had bilateral masses
lying adjacent to the ovary: three of the six masses were solid and
the other three were roundish or elongated multilocular thick-walled
cystic structures containing echogenic fluid without cogwheel sign or
incomplete septae. Two patients had bilateral adnexal masses
without a discernable ovary: two were solid masses and two were
multilocular-solid masses with thick walls and septae and with cyst
locules containing homogenously echogenic fluid or fluid with echo-
genicity similar to what is seen in haemorrhagic corpora lutea (Valen-
tin, 2004). Neither cogwheel sign nor incomplete septae were seen.
The median sizes of the smallest and largest adnexal masses lying ad-
jacent to the ovary were 21 (19 –30) and 33 (27– 46) mm, the median
colour scores were 87 (29 –95) and 76 (56 –97), the median PSVs
were 20 cm/s (12 –21) and 14 cm/s (13 –37), the median TAMXV
were 8 cm/s (7 –14) and 6 cm/s (6– 22) and the median PI were
0.98 (0.69–2.06) and 2.06 (1.07– 2.07), respectively. The median
diameter of the smallest mass without a discernable ovary was
32 mm (30 –34) and that of the largest was 48 (40 –60) mm. The
median colour scores were 71 and 73 (65 –81), the median PSVs
were 35 and 30 cm/s (18–41), the median TAMXV were 10 and
15 cm/s (8– 21) and the median PI were 3.22 and 1.68 (1.30–2.07)
(velocimetry results were missing for one of the smallest masses),
respectively.
Ultrasound findings in patients
with endometritis
In six (66.6%) of the nine patients with a final diagnosis of endometri-
tis, no adnexal masses were seen at ultrasound examination, while bi-
lateral solid masses lying adjacent to the ovary were described in three
patients. The median size of the smallest of the bilateral masses was
16 mm (15 –16) and that of the largest was 19 mm (16 –23). The
colour content of the masses lying adjacent to the ovary in women
with endometritis was lower than in patients with confirmed salpin-
gitis, while the PI values were higher; the median colour scores in
the smallest and largest masses in patients with endometritis being
51 (20– 81) and 54 (20– 88) and the median PI being 1.63 (0.97–
2.31) and 1.31 (1.00– 1.63). PSV and TAMXV in masses lying adjacent
Ultrasound for diagnosing acute salpingitis 1575

In the last case, a multilocular-solid mass was suggested by the ultra-

Table IV Diagnoses suggested by the ultrasound sound examiner to be either torsion of a diseased tube, a pyosalpinx
examiner on the basis of subjective evaluation
or, very unlikely, a malignancy. This mass was the only one manifesting
of ultrasound findings.
the cogwheel sign (Fig. 4). The correct diagnosis was torsion of a
True diagnosis hydrosalpinx.
.............................................................. In 3 of the 26 patients, bilateral small diffusely delineated-solid
Ultrasound Salpingitis Endometritis Others masses lying adjacent to the ovary were seen at ultrasound examin-
diagnosis (n 5 17) (n 5 9) (n 5 26)
........................................................................................ ation. In two of these cases, the ultrasound examiner suggested a diag-
Salpingitis 14 4 4a nosis of acute salpingitis, but no pathology was judged to be present at
Appendicitis 0 0 4b laparoscopy, and the final diagnosis was pain of unknown etiology (the
Normal 2 3 10c sizes of the smallest masses were 15 and 26 mm, colour scores were
ultrasound 70 and 12, PSVs were 13 and 22 cm/s, TAMXV were 5 and 9 cm/s

Downloaded from https://academic.oup.com/humrep/article-abstract/28/6/1569/604182 by guest on 22 May 2019

findings
No diagnosis 1 2 5d
suggested
Ovarian cyst 0 0 2e
Adnexal mass of 0 0 1f
unclear origin
a to the hydatid of Morgagni was solid and had a size of 12 mm, a colour
Final diagnosis was pain of unknown etiology (n ¼ 2), cervicitis (n ¼ 1) and
appendicitis (n ¼ 1).
b
Final diagnosis was appendicitis (n ¼ 3) and peritoneal endometriosis (n ¼ 1).
c
Final diagnosis was pelvic pain of unknown etiology (n ¼ 4), cervicitis (n ¼ 2),
urinary tract infection (n ¼ 2), appendicitis (n ¼ 1) and cystadenocarcinoma of the
appendix (n ¼ 1).
d
Final diagnosis was pelvic pain of unknown etiology (n ¼ 1), appendicitis (n ¼ 1),
Crohn’s disease (n ¼ 1) and adnexal torsion (n ¼ 2); in one case of torsion, torsion
of a diseased tube was one of the three possible diagnoses suggested by the
ultrasound examiner.
e
Final diagnosis was ovarian cyst in both cases.
f
The final diagnosis was peritoneal endometriosis (no mass was confirmed at
laparoscopy).
to the ovary were similar in women with salpingits and endometritis,
the median PSV and TAMVX in women with endometritis being
14 cm/s (8–16) and 6 cm/s (3 –10) in the smallest mass and
14 cm/s (7 –21) and 6 cm/s (3–13) in the largest mass.
Ultrasound findings in patients with a final
diagnosis other than salpingitis or
endometritis
In 15 of the 26 patients with a final diagnosis other than salpingitis or
endometritis, no masses were detected at ultrasound examination.
The final diagnosis in these 15 cases were pain of unknown etiology
(n ¼ 4), cervicitis (n ¼ 3), appendicitis (n ¼ 3), urinary tract infection
(n ¼ 2), peritoneal endometriosis (n ¼ 1), Crohn’s disease (n ¼ 1)
and cystadenocarcinoma of the appendix (n ¼ 1).
In 2 of the 26 patients, unilateral ovarian lesions were seen at ultra-
sound examination and suggested by the ultrasound examiner to be a
hemorrhagic corpus luteum cyst and a 50-mm bilocular ovarian cyst.
The two diagnoses of ovarian cyst were confirmed at laparoscopy.
A unilateral mass lying adjacent to the ovary was seen in 6 of the 26
patients. In one of these six cases, no mass was found at laparoscopy
(the final diagnosis was pain of unknown etiology, and in retrospect
the mass seen at ultrasound was probably a bowel loop). In four
cases, the ultrasound image of the mass was judged to be compatible
with acute appendicitis. This diagnosis was confirmed in three of the
four cases. In the fourth case the appendix was judged to be normal
at laparoscopy, and the final diagnosis was peritoneal endometriosis.
and PI were 1.88 and 1.75, while the sizes of the largest masses were
18 and 27 mm, colour scores were 95 and 87, PSVs were 8 and
13 cm/s, TAMXV were 4 and 8 cm/s and PI were 1.58 and 0.88, re-
spectively). In the third case, torsion of a gangrenous hydatid of Mor-
gagni was found at laparoscopy, while no mass was confirmed on the
contralateral side. At ultrasound examination the mass corresponding
score of 22, PSV of 16 cm/s, TAMXV of 9 cm/s and PI of 0.69. The
mass that was not confirmed at laparoscopy was also considered to be
solid, and had a size of 14 mm, a colour score of 70, PSV of 17 cm/s,
TAMXV of 8 cm/s and PI of 1.64.
Discussion
Our results showed that ultrasound findings suggestive of moderate or
severe acute salpingitis were the presence of bilateral adnexal masses
or the presence of bilateral masses lying adjacent to the ovary,
with 100% (13 of 13) and 77% (10 of 13) of the patients with
moderate-to-severe salpingitis confirmed by laparoscopy manifesting
these signs compared with 25% (1 of 4) of the patients with mild sal-
pingitis and 17% (6 of 35) of the patients with other diagnoses. Most
masses lying adjacent to the ovaries in cases of salpingitis were solid,
fewer were thick-walled elongated unilocular cystic masses or round-
ish or elongated multilocular or multilocular-solid masses containing
echogenic fluid but not manifesting the cogwheel sign or incomplete
septae. Most of these lesions were 2–3 cm in size and well vascular-
ized at colour Doppler ultrasound examination.
The strength of our study is that it contributes information to an
area where scientific evidence is very scarce. To the best of our
knowledge, the ultrasound findings in cases of mild or moderate sal-
pingitis, but not distinguishing between the two, verified by laparos-
copy have been described in only one publication. It includes six
cases of mild or moderate salpingitis and shows ultrasound images
of two of these cases (Molander et al., 2001). We are aware of
only two published studies reporting on the sensitivity and specificity
of ultrasound with regard to laparoscopically confirmed acute salpin-
gitis (Tukeva et al., 1999; Molander et al., 2001). Both include
mainly patients with pyosalpinx or tubo-ovarian abscess, and only
one of them has a design appropriate for estimating sensitivity and
specificity (Table V) (Tukeva et al., 1999). Details on published
studies reporting on the sensitivity and specificity of ultrasound
for diagnosing ‘upper genital tract infection’ are shown in Table VI
(Cacciatore et al., 1992; Boardman et al., 1997).
Our study has limitations. First, the sample size is small. This is
explained mainly by difficulties with patient recruitment (doctors in
1576 Romosan et al.

Downloaded from https://academic.oup.com/humrep/article-abstract/28/6/1569/604182 by guest on 22 May 2019

Figure 1 Ultrasound images from one patient with moderate acute salpingitis verified by laparosocopy. (A) and (B) show gray scale ultrasound
images of the left tube, (C) and (D) show colour Doppler images of the same tube. The lesion is a sausage-shaped thick-walled unilocular cystic
structure with a very small amount of echogenic fluid inside. We interpret the white oval ring in (B) as the mucosa of the inflamed tube. As seen
in (C) and (D), the tube is extremely well vascularized in Doppler ultrasound examination. Please note the ring of colour in (D). We interpret
this as rich vascularisation surrounding a transverse section through the inflamed tube. We observed this finding also in other cases of moderate sal-
pingitis, see Figs 2 and 3.

Figure 2 Ultrasound images of moderate acute salpingitis verified by laparosocopy in a second patient. (A) A sausage-shaped solid structure cor-
responding to the inflamed tube. (B) The rich vascularisation of the same structure and rings of colour are discernable, see also Figs 1 and 3.

the emergency ward forgetting to recruit patients, patients declining to
undergo diagnostic laparoscopy). According to our hospital statistics,
245 patients had a diagnosis of acute salpingitis either as inpatients
or outpatients in our hospital during the study period. Our study
sample includes only 17 of these. Secondly, there were few cases
of mild salpingitis in our study sample. This means that we cannot es-
timate with any precision the ability of ultrasound to detect very early
salpingitis. Clearly, the proportion of cases with salpingitis of different
grade affects the sensitivity and specificity of ultrasound, and the sen-
sitivity and specificity that we report here are generalizable only to
patient populations similar to our study population. Moreover, one
can expect results similar to ours only if ultrasound is carried out
by experienced ultrasound examiners using high-end ultrasound
systems. The small number of patients with mild salpingitis may
reflect either that doctors did not recommend laparoscopy to all
patients fulfilling our eligibility criteria (which they should have done
both according to the policy of our department and our study proto-
col), that a higher proportion of patients with mild salpingitis than
moderate or severe salpingitis declined to participate in the study,
or that few patients with mild salpingitis fulfilled our eligibility criteria.
A third limitation, and one that we share with other studies trying to
estimate the sensitivity and specificity of ultrasound with regard to
Ultrasound for diagnosing acute salpingitis
Figure 3 Ultrasound images of moderate acute salpingitis verified by laparosocopy in a third patient. (A) A unilocular sausage-shaped thick-walled
structure corresponding to the inflamed tube with a very small amount of echogenic fluid inside. (B) A transverse section through the same tube.
(C) The rich vascularisation of the same tube. Please note the ring of colour surrounding the transverse section of the tube, see also Figs 1 and 2.
Figure 4 Ultrasound images of a hydrosalpinx that has undergone
torsion. This is a 3-cm multilocular solid mass lying adjacent to the
ovary. Swollen mucosal folds protrude into the fluid-filled lumen of
the mass. This is the only mass in our series manifesting the cogwheel
sign. The ultrasound examiner suggested three possible diagnoses:
torsion of a diseased tube, pyosalpinx, or, very unlikely, a malignancy.
1577
Downloaded from https://academic.oup.com/humrep/article-abstract/28/6/1569/604182 by guest on 22 May 2019
acute salpingitis, is the lack of an obvious gold standard. Salpingitis may
be present in the absence of histological signs of endometritis (Sellors
et al., 1991). In our opinion, the best gold standard is diagnostic lapar-
oscopy. Still, diagnostic laparoscopy is not ideal. Sellors et al. (1991)
found laparoscopy to have low sensitivity (58%) with regard to histo-
logically confirmed salpingitis, and Molander et al. (2003) reported
poor intra-and inter-observer agreement with regard to PID when
six gynaecologists evaluated laparoscopic images of the female pelvis.
The ultrasound image of early salpingitis has been described by
Molander et al. (2001) as a solid mass with high colour content at
power Doppler examination located close to the ovary. Our results
agree fairly well. However, neither we nor Boardman et al. (1997)
were able to confirm the findings of Cacciatore et al. (1992) that
fluid in the pouch of Douglas and polycystic appearance of the
ovaries are reliable signs of ‘upper genital tract infection’. Even
though our results confirm that inflamed tubes are richly vascularized
at Doppler examination (Tinkanen and Kujansuu, 1992; Alatas et al.,
1996; Molander et al., 2001), all our Doppler results overlapped too
much between patients with different diagnoses for them to be clinic-
ally useful. This is in accordance with the findings in the case–control
study by Molander et al. (2001) where ultrasound images of acute sal-
pingitis (cases) were compared with those of hydrosalpinx (controls).
Possibly, the inability of Doppler ultrasound to distinguish
upper genital tract infections from other conditions in the pelvis is
1578 Romosan et al.

explained by Doppler measurements not being sufficiently precise or

..........................................................................................................................................................................................................................................................
accurate.
Two studies (Tukeva et al., 1999; Molander et al., 2001) reported

Number of

peritonitis
cases with
very high sensitivity and specificity of ultrasound for diagnosing PID.
The higher sensitivity in these studies than in ours is likely to be
explained by a much higher prevalence of pyosalpinx and tubo-ovarian

0
?
abscess in the other studies (Table V). The sensitivity and specificity
reported by Molander et al. (2001) are not applicable to patients
with clinical signs of acute PID, because they calculated sensitivity
Number of
cases with

using patients with laparoscopically confirmed PID (cases) and specifi-

abscess

city using patients with hydrosalpinx (control group).

Our finding that 44% (23/52) of the patients with clinical signs of
8

Downloaded from https://academic.oup.com/humrep/article-abstract/28/6/1569/604182 by guest on 22 May 2019

PID had normal laparoscopy findings or a final diagnosis unrelated
to PID agrees with those of others, the corresponding figures in
Number of cases with

other studies being 30% (9/30) (Tukeva et al., 1999), 35%

salpingitis grade 3

(282/814) (Jacobson and Westrom, 1969), 39% (13/33) (Molander

et al., 2000) and 33% (10/30) (Molander et al., 2001). In our study
(pyosalpinx)

as well as in others, common diagnoses unrelated to PID in patients

with clinical signs of PID were appendicitis (Jacobson and Westrom,
Table V Summary of studies reporting the sensitivity and specificity of ultrasound in diagnosing salpingitis.

1969; Molander et al., 2000) adnexal torsion (Tukeva et al., 1999;

6

Molander et al., 2000), ovarian cysts (Jacobson and Westrom, 1969;

Tukeva et al., 1999; Molander et al., 2000) and endometriosis
Number of cases

(Jacobson and Westrom, 1969; Tukeva et al., 1999; Molander et al.,

with salpingitis

2000).
grade 1 or 2

Our results support that ultrasound is likely to be helpful when

managing patients with symptoms and clinical signs of acute PID,
because symptoms and signs of PID overlap with those of several
5

12

diagnoses unrelated to genital infection. Ultrasound signs of affected

tubes changed the odds of salpingitis only moderately (about five
Specificity, %

90 (18/20)

83 (29/35)

times) (Jaeschke et al., 1994), but even a small change in odds

78 (7/9)

may be helpful in patients presenting a diagnostic dilemma (Jaeschke

et al., 1994). Moreover, in the absence of bilateral adnexal masses,
the odds of moderate or severe salpingitis decreased .20-fold,
while mild salpingitis could not be excluded. This information pro-
Sensitivity, %

81 (17/21)

100 (20/20)

82 (14/17)

vided by transvaginal ultrasound is likely to be of help when deciding

on whether or not to proceed with diagnostic laparoscopy in
patients with symptoms and signs suggesting PID and, if laparoscopy
is not performed, for selecting treatment and planning follow-up.
Results presented for the current study are those for bilateral adnexal masses.

In addition, ultrasound may reveal diagnoses unrelated to PID, e.g.

Laparoscopy

Laparoscopy

Laparoscopy
standard

appendicitis, ovarian cysts including endometriomas and torsion of

Gold

the adnexa.
Salpingitis or

Salpingitis or
End-point

Salpingitis
abscess

abscess

Authors’ roles
Controls were patients with hydrosalpinx.

G.R. recruited patients, performed the statistical analysis, drafted the

manuscript and approved the final version submitted for publication.
Case– controla
observational

observational

C.B. recruited patients, revised the manuscript critically for important

Prospective

Prospective

intellectual content and approved the final version submitted for pub-
Design

lication. L.S. examined all endometrial biopsies, revised the manuscript

critically for important intellectual content and approved the final
version submitted for publication. L.V. designed the study, performed
Tukeva et al.

et al. (2001)

all the ultrasound examinations, participated in the statistical analysis

Molander

Current
Study

(1999)

studyb

and interpretation of the results, revised the manuscript critically for

important intellectual content and approved the final version submit-
b
a

ted for publication.

Ultrasound for diagnosing acute salpingitis
Table VI Summary of studies reporting the sensitivity and specificity of ultrasound in diagnosing ‘upper genital tract
infection’.
Study Design End-point
.............................................................................................................................................................................................
Cacciatore Prospective Plasma cell
et al. (1992) observational endometritis
Boardman et al. Prospective ‘Upper genital
(1997) observational tract infection’
Funding
This work was supported by funds administered by Malmö University
Hospital; and two Swedish government grants (ALF-medel and Land-
stingsfinansierad Regional Forskning).
Conflict of interest
None declared.
References
Alatas C, Aksoy E, Akarsu C, Yakin K, Bahceci M. Hemodynamic
assessment in pelvic inflammatory disease by transvaginal color Doppler
ultrasonography. Eur J Obstet Gynecol Reprod Biol 1996;70:75–78.
Blaustein A, Kurman RJ. Blaustein’s Pathology of the Female Genital Tract,
5th edn. New York: Springer, 2002.
Boardman LA, Peipert JF, Brody JM, Cooper AS, Sung J. Endovaginal
sonography for the diagnosis of upper genital tract infection. Obstet
Gynecol 1997;90:54 – 57.
Cacciatore B, Leminen A, Ingman-Friberg S, Ylostalo P, Paavonen J.
Transvaginal sonographic findings in ambulatory patients with suspected
pelvic inflammatory disease. Obstet Gynecol 1992;80:912–916.
CDC. Sexually Transmitted Diseases Treatment Guidelines. Centers for
Disease Control and Prevention, 2010,63 – 67. http://www.cdc.gov/
std/treatment/2010/std-treatment-2010-rr5912.pdf (1 March 2013,
date last accessed).
Hager WD, Eschenbach DA, Spence MR, Sweet RL. Criteria for diagnosis
and grading of salpingitis. Obstet Gynecol 1983;61:113 – 114.
Jacobson L, Westrom L. Objectivized diagnosis of acute pelvic
inflammatory disease. Diagnostic and prognostic value of routine
laparoscopy. Am J Obstet Gynecol 1969;105:1088 – 1098.
Jaeschke R, Guyatt GH, Sackett DL. Users’ guides to the medical
literature. III. How to use an article about a diagnostic test. B. What
are the results and will they help me in caring for my patients? The
Evidence-Based Medicine Working Group. JAMA 1994;271:703– 707.
Kirchler HC, Kolle D, Schwegel P. Changes in tubal blood flow in evaluating
ectopic pregnancy. Ultrasound Obstet Gynecol 1992;2:283–288.
1579
Gold standard Sensitivity, % Specificity, % Number of cases with
salpingitis grade 1 or 2,
pyosalpinx or abscess
Endometrial 85 (11/13) 100 (38/38) ?
histology
Laparoscopy 32 (6/19) 97 (35/36) ?
(n ¼ 28)
or
Endometrial
Downloaded from https://academic.oup.com/humrep/article-abstract/28/6/1569/604182 by guest on 22 May 2019
histology and/or
Microbiology
(n ¼ 27)
Molander P, Cacciatore B, Sjoberg J, Paavonen J. Laparoscopic
management of suspected acute pelvic inflammatory disease. J Am
Assoc Gynecol Laparosc 2000;7:107– 110.
Molander P, Sjoberg J, Paavonen J, Cacciatore B. Transvaginal
power Doppler findings in laparoscopically proven acute pelvic
inflammatory disease. Ultrasound Obstet Gynecol 2001;17:
233 – 238.
Molander P, Finne P, Sjoberg J, Sellors J, Paavonen J. Observer agreement
with laparoscopic diagnosis of pelvic inflammatory disease using
photographs. Obstet Gynecol 2003;101:875– 880.
Okaro E, Valentin L. The role of ultrasound in the management of women
with acute and chronic pelvic pain. Best Pract Res Clin Obstet Gynaecol
2004;18:105 – 123.
Sellors J, Mahony J, Goldsmith C, Rath D, Mander R, Hunter B, Taylor C,
Groves D, Richardson H, Chernesky M. The accuracy of clinical findings
and laparoscopy in pelvic inflammatory disease. Am J Obstet Gynecol
1991;164:113 – 120.
Timmerman D, Valentin L, Bourne TH, Collins WP, Verrelst H, Vergote I.
Terms, definitions and measurements to describe the sonographic
features of adnexal tumors: a consensus opinion from the
International Ovarian Tumor Analysis (IOTA) Group. Ultrasound
Obstet Gynecol 2000;16:500– 505.
Timor-Tritsch IE, Lerner JP, Monteagudo A, Murphy KE, Heller DS.
Transvaginal sonographic markers of tubal inflammatory disease.
Ultrasound Obstet Gynecol 1998;12:56 – 66.
Tinkanen H, Kujansuu E. Doppler ultrasound studies in pelvic inflammatory
disease. Gynecol Obstet Invest 1992;34:240– 242.
Tukeva TA, Aronen HJ, Karjalainen PT, Molander P, Paavonen T,
Paavonen J. MR imaging in pelvic inflammatory disease: comparison
with laparoscopy and US. Radiology 1999;210:209– 216.
Valentin L. Pattern recognition of pelvic masses by gray-scale ultrasound
imaging: the contribution of Doppler ultrasound. Ultrasound Obstet
Gynecol 1999;14:338 – 347.
Valentin L. Use of morphology to characterize and manage
common adnexal masses. Best Pract Res Clin Obstet Gynaecol 2004;
18:71– 89.
Valentin L, Sladkevicius P, Marsal K. Limited contribution of Doppler
velocimetry to the differential diagnosis of extrauterine pelvic tumors.
Obstet Gynecol 1994;83:425– 433.