is still a suspicion of an ectopic pregnancy, consider a vaginal examination to assess for cervical excitation or pelvic tenderness. If either of these is present, together with a positive pregnancy test, refer the woman immediately to a dedicated early pregnancy assessment service.
1
2
Number
Research recommendation
RR 3 Research should be undertaken to design and validate a decision tool for evaluating signs, symptoms and risk factors for correctly identifying ectopic pregnancy
3
6.2
Ultrasound for determining a viable intrauterine
4
pregnancy
5
Review question
6
What is the diagnostic value of ultrasound for determining a viable intrauterine pregnancy?
7
Introduction
8
The application of ultrasound is well established and important in the assessment and evaluation of
9
early pregnancy events and complications. Its use in early pregnancy assessment may be routine but
10
practices vary considerably. Although high resolution transvaginal ultrasound has been widely
11
adopted, the limitations of accuracy in defining ultra-small structures, such as a fetal heart at early
12
gestations, are acknowledged. The aim of this review was to identify the point at which viability of a
13
pregnancy can be definitively confirmed using ultrasound. This threshold also represents the point at
14
which miscarriage can be definitively diagnosed.
15
Description of included studies
16
Fifteen studies were included in this review (Abaid et al., 2007; Abdallah et al., 2011; Bree et al.,
17
1989; Brown et al., 1990; Cacciatore et al., 1990; de Crespigny, 1988; Ferrazzi et al., 1993; Goldstein,
18
1992; Hassan et al., 2009; Levi et al., 1988; Levi et al., 1990; Pennell et al., 1991; Rempen, 1990;
19
Rowling et al., 1999; Steinkampf et al., 1997).
20
The included studies consist of nine prospective observational studies (Abdallah et al., 2011; Bree et
21
al., 1989; Cacciatore et al., 1990; de Crespigny, 1988; Goldstein, 1992; Hassan et al., 2009; Pennell
22
et al., 1991; Rempen, 1990; Rowling et al., 1999), five retrospective observational studies (Abaid et
23
al., 2007; Ferrazzi et al., 1993; Levi et al., 1988; Levi et al., 1990; Steinkampf et al., 1997) and one
24
partially retrospective observational study (Brown et al., 1990).
25
The studies were conducted in the UK (Abdallah et al., 2011; Hassan et al., 2009), Germany
26
(Rempen, 1990), Italy (Ferrazzi et al., 1993), Finland (Cacciatore et al., 1990), the USA (Abaid et al.,
27
2007; Bree et al., 1989; Brown et al., 1990; Goldstein, 1992; Pennell et al., 1991; Rowling et al., 1999;
28
Steinkampf et al., 1997), Canada (Levi et al., 1988; Levi et al., 1990), and Australia (de Crespigny,
29
1988).
30
All included studies evaluated the use of transvaginal ultrasound for visualising fetal cardiac activity in
31
intrauterine pregnancies, and stratified their findings by gestational age, crown-rump length or
32
gestation sac size. Two studies additionally compared the performance of transabdominal ultrasound
33
in visualising cardiac activity (Ferrazzi et al., 1993; Pennell et al., 1991).
34
Evidence profile
1
Table 6.2 GRADE summary of findings for evaluation of ultrasound for determining a viable intrauterine
2
pregnancy3
Number of studies Type of ultrasound scan (transvaginal (TVU) or transabdominal (TAU)) Number of women scanned for fetal cardiac activity, and stratified by fetal size/ age (Total study participants)Threshold at which 100% of fetuses that later proved to be viable can be identified
Quality
Visualisation of cardiac activity by crown-rump length / mm
1 study (Rempen, 1990) 5-MHz TVU 292 (363) 3 High 1 study (Pennell et al., 1991) 5-MHz/3.5-MHz TAU 5-MHz/7.5-MHz TVU 175 (175) TAU: 9 TVU: 5 High 1 study (Hassan et al., 2009) TVU 1174 (1174) 6.0 Moderate 1 study (Abaid et al., 2007) 8-MHz TVU 179 (179) 3.5 Moderate 1 study (Brown et al., 1990) 5-MHz TVU 375 (375) 5 Moderate 1 study (Abdallah et al., 2011) 6-12-MHz TVU 24 (1060) 5.3 Low 1 study (Levi et al., 1990) 6.5-MHz TVU 71 (71) 4.0 Low 1 study (Goldstein, 1992) 5-MHz/7.5-MHz TVU 96 (96) 4 Low
Visualisation of cardiac activity by gestation sac diameter / mm
1 study (Rempen, 1990) 5-MHz TVU 354 (363) 18.3 * High 1 study (Abdallah et al., 2011) 6-12-MHz TVU 183# (1060) 21 Moderate 266# (1060) 21 1 study (Bree et al., 1989) 7-MHz TVU 53 (53) > 9 Moderate
Pain and bleeding in early pregnancy
72
1 study (Rowling et al., 1999) 9-5-MHz TVU 39 (39) 13 Low 1 study (Levi et al., 1988) 6.5-MHz TVU 35 (62) 16 Very low 1 study (de Crespigny, 1988) 5-MHz TVU 353 (353) > 12 Low 1 study (Steinkampf et al., 1997) 5-MHz TVU 82 (82) 19** Very low 1 study (Cacciatore et al., 1990) 5-MHz/6.5-MHz TVU 20 (22) > 18*** Very lowVisualisation of cardiac activity by gestational or menstrual age (days)
1 study (Rempen, 1990) 5-MHz TVU 252 (363) 46 (menstrual age) High 1 study (Bree et al., 1989) 7-MHz TVU 53 (53) > 40 (gestational age) Moderate 1 study (Steinkampf et al., 1997) 5-MHz TVU 82 (82) 45.5** (gestational age) Very low 1 study (Ferrazzi et al., 1993) 5-MHz TAU/ 5-MHz TVU 76 (598) TAU: 37 TVU: 35 (menstrual age) Very low 1 study (Cacciatore et al., 1990) 5-MHz/6.5-MHz TVU 20 (22) > 43*** (gestational age) Very low
* Chorionic cavity diameter
1
** Point of 99% probability of visualisation
2
*** Point of “reliable detection”
3
# 183 scans showed a gestation sac without a visible embryo or yolk sac. 266 scans showed a gestation sac with a yolk sac,
4
but without a visible embryo.
5
Evidence to recommendations
6
Relative value placed on the outcomes considered
7
The outcomes for this review were the thresholds of gestational age, CRL or gestation sac diameter
8
at which 100% of fetuses that later proved to be viable had cardiac activity visible on ultrasound. The
9
group noted that, even when women could be completely certain about the date of intercourse or last
10
menstrual period, variation in the menstrual cycle and rate of fetal growth might result in inaccurate
11
estimates of gestational age. Therefore, the GDG felt that the use of a gestational age threshold alone
12
was not appropriate for the determination of viability. From their own clinical experience, the group felt
13
that, where it was possible to measure, the crown-rump length would provide the most accurate
14
estimate of development; therefore, measurement of mean gestational sac diameter was only
15
recommended in cases where a fetal pole could not be identified.
16
Consideration of clinical benefits and harms
1
The GDG felt that it was appropriate to set thresholds for the determination of a viable intrauterine
2
pregnancy; however, considering the consequences of misdiagnosing a viable intrauterine pregnancy
3
as a miscarriage, they felt that the thresholds should be based on the most conservative findings
4
reported in the studies. The GDG were aware of recent work by Pexsters et al. (2011) that
5
documented the potential for considerable intra- and inter-observer variation in measurements of CRL
6
and mean gestational sac diameter. They also noted that there was additional potential for variation in
7
measurements linked to the quality of scanning equipment and the skill level of the sonographer. In
8
light of these considerations and the findings from the included studies, the GDG determined that fetal
9
non-viability should not be diagnosed based on the absence of a heartbeat in fetuses with a CRL of
10
less than or equal to 6.0 mm or a mean gestational sac diameter of less than or equal to 25.0 mm, as
11
measured on a single transvaginal ultrasound. They felt that at such small sizes, it would not be
12
possible to determine whether a miscarriage had truly occurred or whether the embryo was simply too
13
small for there to be a visible heartbeat. Therefore, up to and including these thresholds, all women
14
should have a repeat scan to confirm the findings of the initial scan. The GDG discussed what would
15
be an appropriate interval between scans, balancing the fact that sufficient time would need to pass to
16
be able to confirm the diagnosis with the fact that women might understandably want an answer as
17
soon as possible. They felt that, for embryos with a measurable crown-rump length, an interval of 7
18
days would be sufficient to definitively demonstrate viability. However, they felt that a longer interval
19
would be required for women in whom only a gestation sac could be measured, and therefore
20
recommended that in such cases, the scan was performed after 14 days.
21
Due to the significant consequences of misclassifying a viable pregnancy as a miscarriage, the GDG
22
felt that it was reasonable to recommend that, in the absence of a visible heartbeat above these
23
thresholds (i.e. CRL > 6 mm, mean gestational sac diameter > 25 mm), all sonographers should seek
24
a second opinion before definitively diagnosing a non-viable pregnancy. However, they also realised
25
that some women might instead wish to wait and have a second confirmatory scan at a later date, and
26
felt that this was a reasonable alternative.
27
Based on their clinical experience, and evidence from another review comparing transvaginal and
28
transabdominal ultrasound for diagnosing ectopic pregnancy (see section 5.3 below), the GDG
29
decided that transvaginal ultrasound would generally be the optimal mode of scanning. However, they
30
recognised that, in some circumstances, a transvaginal ultrasound might not be appropriate or
31
acceptable to women, and therefore a transabdominal scan could be offered as an alternative. Based
32
on evidence from Pennell et al., (1991) the GDG decided that a higher CRL threshold of > 10 mm
33
should be used for diagnosing miscarriage on the basis of a single transabdominal ultrasound scan.
34
However, they felt it was safe to use the same mean gestational sac diameter threshold of > 25 mm,
35
because the larger size of the gestational sac results in almost equivalent accuracy with
36
transabdominal ultrasound.
37
Consideration of health benefits and resource uses
38
The group did not feel that recommending a repeat ultrasound scan would add significantly to the
39
case load and resource use, because in practice this often happens anyway. However, they did feel
40
that cost-effectiveness should be a component of any research conducted in this area, and therefore
41
incorporated it into a research recommendation. Given the number of women requesting scans in
42
early pregnancy and the associated service and cost implications, the group felt that it was important
43
that research be done in this area, to determine the timing and frequency of ultrasound examinations
44
that would maximise improvements in diagnosis, outcomes and women’s experience.
45
Quality of evidence
46
The quality of evidence ranged from high to very low quality and the GDG felt that, in conjunction with
47
their clinical experience, it was appropriate to make recommendations based on the findings.
48
Information giving and psychological support
49
As discussed above, although the GDG felt that transvaginal scanning was the optimal mode of
50
scanning, they understood that some women would choose to have a transabdominal scan instead.
51
The group fully supported this choice; however, they did feel that women should be given information
52
about the potential limitations of transabdominal scanning so that they could make a fully informed
53
choice.