SEARCH

SEARCH BY CITATION

Keywords:

  • placenta accreta;
  • power Doppler;
  • prenatal diagnosis;
  • receiver–operating characteristics;
  • three-dimensional;
  • ultrasound

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Objective

To assess the role of three-dimensional (3D) power Doppler in the antenatal diagnosis of placenta accreta and compare its diagnostic performance with gray-scale and color Doppler ultrasonography.

Methods

One hundred and seventy pregnant women with persistent placenta previa totalis (after 28 weeks' gestation) were prospectively enrolled into this study. Gray-scale transabdominal ultrasound examination was performed to detect loss of the subendometrial echolucent zone and other abnormalities suggestive of placenta accreta. Color flow mapping was used to scan the whole placenta to detect any newly formed vessels at the serosa–bladder border or the presence of abnormal lacunae. Finally a targeted examination of angioarchitecture in the basal and lateral views of the placenta was carried out using 3D power Doppler. The ultrasound findings were analyzed with reference to the final diagnosis made during Cesarean delivery.

Results

Placenta accreta and its variants (including increta and percreta) were confirmed in 39 patients at the time of Cesarean delivery. Based on receiver–operating characteristics analysis, ‘numerous coherent vessels’ visualized using 3D power Doppler in the basal view was the best single criterion for the diagnosis of placenta accreta, with a sensitivity of 97% and a specificity of 92%. If we considered the presence of at least one criterion to be diagnostic when using each ultrasound technique, then 3D power Doppler would have the best positive predictive value (76%), followed by gray-scale (51%) and color Doppler (47%). The majority of patients with placenta accreta showed multiple characteristic features on ultrasound imaging. In contrast, those patients with a false-positive diagnosis (i.e. the final diagnosis was placenta previa alone) tended to show isolated ultrasound markers of the condition.

Conclusion

3D power Doppler may be useful as a complementary technique for the antenatal diagnosis or exclusion of placenta accreta. Copyright © 2009 ISUOG. Published by John Wiley & Sons, Ltd.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

It is estimated that 140 000 women die of postpartum hemorrhage per year1. Although statistically a rare complication, placenta accreta has now become an important etiology of maternal morbidity and mortality2. Owing to the increasing rate of Cesarean delivery, there has been a 10-fold rise in the incidence of placenta accreta since the 1970s3. In a recent survey, placenta accreta was even found to have become the leading cause of emergency peripartum hysterectomy4.

Clinically, the most significant feature of placenta accreta is the abundant uteroplacental neovascularization, which can lead to life-threatening hemorrhage5–7. However, its antenatal diagnosis is usually based on characteristic findings on gray-scale ultrasound imaging, such as the loss of subendometrial echolucent zone or the presence of abnormal placental lacunae8–10. Several reports have suggested that neither color Doppler nor power Doppler, both of which assess uteroplacental vascularization, add any additional information to gray-scale ultrasound imaging in the diagnosis of placenta accreta10–12. However, much of the literature on this subject either describes single cases or lacks adequate controls (i.e. placenta previa without accreta), making it rather difficult to validate the existing criteria13–15. Despite the modern advances in imaging techniques, no single diagnostic technique affords complete assurance for the presence or absence of placenta accreta1. The diagnosis is most often made during the third stage of labor or on Cesarean delivery.

The aim of this study was to introduce additional criteria for the diagnosis of placenta accreta using three-dimensional (3D) power Doppler complementary to gray-scale and color Doppler techniques, and to compare their diagnostic performance based on receiver–operating characteristics (ROC) curve analysis.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Between December 2000 and September 2007, 170 pregnant women with persistent placenta previa (after 28 weeks' gestation) were prospectively enrolled into this study. The inclusion criteria were: complete imaging using all diagnostic techniques (gray-scale, color Doppler and 3D power Doppler), and full availability of delivery information. Ultrasound examination was performed using a 3D ultrasound system equipped with a 4–8-MHz transabdominal transducer (Voluson 730, GE Medical Systems, Zipf, Austria). The patients were asked about their past histories of uterine surgery, such as prior Cesarean delivery or uterine curettage. Informed consent was obtained before scanning.

For each patient, the whole placenta was scanned in a systematic fashion using both gray-scale ultrasound and color flow mapping. The placenta was imaged with a sufficient bladder volume to clearly visualize the serosa–bladder interface, and the angle of insonation was kept as low as possible. The resistance index of flow within the abnormal lacunae and any newly formed vessels over the serosa–bladder border were measured in at least three different locations to obviate selection bias, with the lowest value being used for analysis. On gray-scale ultrasound imaging, we considered the presence of at least one of the following characteristics to indicate placenta accreta (including its variants, placenta increta and placenta percreta): (1) complete loss of the retroplacental sonolucent zone (Figure 1a); (2) irregular retroplacental sonolucent zone; (3) thinning or disruption of the hyperechoic uterine serosa–bladder interface (Figure 1b); (4) the presence of focal exophytic masses invading the urinary bladder (Figure 1b); and (5) the presence of abnormal placental lacunae (Figure 1c)9, 16, 17. Likewise, the diagnosis of placenta accreta was regarded as positive when any one of these color Doppler criteria was present: (1) diffuse or focal lacunar flow pattern (Figure 1d); (2) sonolucent vascular lakes with turbulent flow typified by high-velocity (peak systolic velocity > 15 cm/s) and low-resistance waveform (Figure 1e); (3) hypervascularity of the uterine–bladder interface with abnormal blood vessels linking the placenta to the bladder (Figure 1f); and (4) markedly dilated vessels over the peripheral sub-placental region18–21.

thumbnail image

Figure 1. Gray-scale and color Doppler images illustrating criteria for the diagnosis of placenta accreta and its variants. (a) Complete loss of the retroplacental sonolucent zone. (b) Focal disruption of hyperechoic bladder mucosa (arrows) and the presence of an exophytic placenta invading the bladder (arrowhead). An abnormal lacuna was also seen in this exophytic placenta. (c) Appearance of ‘moth-eaten’ placental lacunae. (d) Pattern of diffuse lacunar flow visualized by color Doppler mapping. (e) Multiple vascular lakes with high peak systolic velocity (PSV) and low-resistance turbulent flow. In this case, the highest PSV was 74 cm/s and the lowest resistance index was 0.18. (f) Hypervascularity of serosa–bladder interface.

Download figure to PowerPoint

To analyze the angioarchitecture of the lower uterine segment and placenta, we carried out a 3D power Doppler examination targeted to this region. Three to five 3D volumes were obtained in automatic sweeps using a motorized curved-array transducer while the women held their breath. In order to interpret the data consistently, the settings of power, pulse repetition frequency, gain and wall motion filter were maintained for all examinations at 100%, 0.9 kHz, − 5.4 and low2, respectively. Offline image analysis was performed using either the ultrasound platform or a PC-based program (4D View®; GE Medical Systems). Two views were successively evaluated: the lateral view was used to observe the intraplacental vasculature and serosa–bladder complex along the sagittal axis of the maternal pelvis, and the basal view illustrated the serosa–bladder interface in a 90° rotation of the lateral view (observing from the direction of the bladder).

Since the abundant neovascularization of the uteroplacental region is a notable feature of placenta accreta, we further analyzed the patterns of placental vasculature in an attempt to differentiate between placenta previa totalis and placenta accreta. We made the diagnosis of placenta previa without accreta when the following 3D power Doppler criteria were demonstrated: (1) in the basal view, the vessels in the uterine serosa–bladder border were discretely arranged (Figure 2a,b); (2) in the lateral view, both cotyledonal (fetal villous) circulation and intervillous (maternal) circulation were essentially parallel to each other, and perpendicular to the decidual plate; (3) the cotyledonal and intervillous circulation were separately distributed; and (4) the cotyledonal circulation was longer and more apparent than the intervillous circulation (Figure 2c). Conversely, the diagnosis of placenta accreta and its variants was regarded as positive if at least one of the following 3D power Doppler criteria was illustrated: in the lateral view, (1) intraplacental hypervascularity, (2) inseparable cotyledonal and intervillous circulations, and (3) tortuous vascularity with ‘chaotic branching’ (Figure 3a). ‘Chaotic branching’ was defined as vessels growing in an irregular manner, with tortuous courses, varying calibers and complex vessel arrangement, as seen in the tumoral vessels in ovarian malignancies22. Furthermore, the basal view revealed numerous coherent vessels involving the serosa–bladder interface (Figure 3b).

thumbnail image

Figure 2. Three-dimensional power Doppler patterns of placenta previa totalis without accreta. The finding of hypervascularity over the serosa–bladder interface, as in this lateral view (a), will generally suggest a diagnosis of placenta accreta. However, when observed in a basal view (b), this hypervascularity consisted of a group of discretely distributed vessels (arrowhead). The arrows in (a) and (b) indicate the same group of vessels. The final diagnosis was confirmed to be placenta previa without accreta. (c) Transparent rendering (in a different patient from (a) and (b)) showing that both intervillous and cotyledonal circulation were essentially parallel to each other and perpendicular to the decidual plate. The cotyledonal circulation was generally longer and more prominent than the intervillous circulation. c, cotyledonal circulation; d, decidual plate; f, fetal (chorionic) plate; i, intervillous circulation; LT, left uterine artery; uc, umbilical cord.

Download figure to PowerPoint

thumbnail image

Figure 3. Three-dimensional power Doppler patterns of placenta accreta in a patient at 38 weeks' gestation. (a) Lateral view showing extreme hypervascularity and vascular detours with chaotic branching. The cotyledonal and intervillous circulations cannot be further discriminated. (b) Basal view showing numerous dilated and coherent vessels fused into a huge vascular complex involving the whole placental base (arrowheads).

Download figure to PowerPoint

All the pregnancies enrolled in this study were delivered by Cesarean section either at our hospital or elsewhere with full availability of information on the delivery. Definitive diagnosis of placenta accreta was made at delivery when the myometrium was seen to be invaded by the placenta, and the pathological examination of the removed uterus showed the villi attached to the myometrium without intervening decidua (accreta), invading into the myometrium (increta) or reaching the serosa (percreta) (Figure 4). To analyze the performance of the diagnostic techniques, ROC analysis was used. The software package SigmaPlot v. 8.0 for Windows (Systat Software Inc., San Jose, CA, USA) was used for statistical analysis and graph plotting.

thumbnail image

Figure 4. Photograph taken during laparotomy in the same patient as in Figure 3b showing numerous engorged and coherent vessels over the uterine serosa. The urinary bladder is not seen in this image as it was below the incision site. Placenta percreta was confirmed after pathological inspection.

Download figure to PowerPoint

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

During the 7-year study period, a total of 170 women who met the inclusion criteria were enrolled into the study. The mean age of our patients was 32.1 ± 4.7 years at the time of diagnosis. Among them, 72 patients had had at least one previous Cesarean section, and 46 patients had a past history of other uterine surgery (such as uterine curettage or myomectomy). Twenty-one patients had at least one episode of antepartum hemorrhage during the current pregnancy.

In this series, placenta accreta and its variants (including increta and percreta) were confirmed in 39 patients at the time of Cesarean delivery. The mean gestational age at sonographic diagnosis of placenta accreta and delivery was 30.7 ± 2.2 and 34.3 ± 1.7 weeks, respectively. Among these patients, all but one had a past history of prior Cesarean delivery. Cesarean hysterectomy was performed in 37 patients in order to prevent massive postpartum hemorrhage. The remaining two patients underwent focal resection of myometrium invaded by placenta, followed by additional hemostatic techniques such as square suture or B-Lynch techniques. In those patients who underwent Cesarean hysterectomy, two sustained bladder injury and received primary repair during the operation. The final pathological analysis revealed six cases of placenta accreta, 24 cases of placenta increta and nine cases of placenta percreta. The definitive diagnosis of placenta previa without accreta was made in 131 out of the 170 patients. Cesarean hysterectomy was performed in two of these patients owing to uncontrolled bleeding from the implantation site after placental separation.

Regarding 3D power Doppler as an additional tool to diagnose placenta accreta, based on the criteria described above, placenta accreta or its variants (n = 39) was diagnosed by demonstrating numerous coherent vessels involving the placental base (n = 38, Figure 3b), hypervascularity (n = 35, Figure 3a) or inseparable cotyledonal/intervillous circulation (n = 21, Figure 3a). In contrast to those cases with placenta previa alone, the intervillous (maternal) circulation was more prominent than the cotyledonal (fetal villous) circulation in placenta accreta (Figure 5). In some cases parts of the intervillous circulation even merged together or with abnormal lacunae to develop a giant aneurysm connected to the serosa–bladder complex (Figure 5a,b). Moreover, those cases with exophytic placentae also showed extension of their neovascularization out of the serosa and into the bladder (Figure 5c,d).

thumbnail image

Figure 5. Additional three-dimensional (3D) power Doppler manifestations of placenta accreta. (a) During serial examinations, part of the intervillous circulation and abnormal lacunae merged together to develop a large aneurysm (arrowhead) in a patient with placenta increta at 35 weeks' gestation. (b) Basal view showing numerous dilated vessels fused and extended to the whole placental base. These coherent vessels, along with the huge aneurysm in (a), developed an extensively vascular complex over the serosa–bladder border. (c) In another patient at 31 weeks' gestation, 3D surface rendering depicted abnormal placental lacunae and an exophytic placenta (arrow). (d) The neovascularization of this exophytic placenta (arrow) also extended out of the uterine serosa and into the bladder (arrow). A small aneurysm was also found (arrowhead) to connect with the serosa–bladder complex. In both patients, the intervillous circulation (i) was more conspicuous than was the cotyledonal circulation (c). d, decidual plate; uc, umbilical cord.

Download figure to PowerPoint

To evaluate and compare the diagnostic efficacy of the different ultrasound techniques, we carried out a systematic review of the ultrasound records of the patients included in the study (Table 1). Based on ROC analysis (Figure 6), ‘numerous coherent vessels’ visualized in the basal view on 3D power Doppler was the best single criterion for the diagnosis of placenta accreta, with a sensitivity of 97% and a specificity of 92%. The criteria of ‘loss of retroplacental echolucent zone’ and ‘abnormal lacunae’, which have previously been used in the diagnosis of placenta accreta9, 21, 23, 24, were both less effective in their diagnostic performance. Furthermore, despite their nearly perfect specificities, both ‘disrupted bladder mucosa’ and ‘exophytic placenta invading the bladder’ had low sensitivities for detecting placenta accreta (18% and 10%, respectively), and hence could not be effective screening tests for placenta accreta. In addition, the discrete classifiers based on 3D power Doppler generally had the shortest distance to the perfect classification performance (upper left corner) on the ROC graph, followed by color Doppler and then gray-scale criteria. Therefore, any individual criterion of 3D power Doppler generally provided a better diagnostic performance than did color Doppler and gray-scale ultrasound. If we assume that the presence of at least one criterion could be diagnostic, 3D power Doppler still showed the best diagnostic performance for placenta accreta, but there was little difference between the performances of the gray-scale and the color Doppler criteria.

thumbnail image

Figure 6. Receiver–operating characteristics analysis of the discrete classifiers for placenta accreta. The upper left corner indicates an ideally perfect prediction method (i.e. 100% sensitivity and 0% false-positive rate). The number beneath each classifier relates to the criterion numbers given in Table 1 (●, gray-scale criteria; □, color Doppler criteria; equation image, three-dimensional power Doppler (3D-PD) criteria). The hexagons represent the integrated assessment (i.e. the presence of any one criterion) for the gray-scale (G), color Doppler (C), and 3D-PD techniques.

Download figure to PowerPoint

Table 1. Comparison of gray-scale, color Doppler and three-dimensional (3D) power Doppler criteria in cases of placenta accreta and placenta previa in all patients and in the subgroup with previous Cesarean section
CriterionAll patientsPatients with previous Cesarean section
Placenta accreta* (n = 39)Placenta previa (n = 131)Sens (%)Spec (%)PPV (%)NPV (%)Placenta accreta* (n = 38)Placenta previa (n = 34)Sens (%)Spec (%)PPV (%)NPV (%)
  • *

    Or its variants, placenta increta and placenta percreta. NPV, negative predictive value; PPV, positive predictive value; PSV, peak systolic velocity; Sens, sensitivity; Spec, specificity,

Gray-scale criteria
 1. Loss of the retroplacental sonolucent zone1764495748517444888173
 2. Irregular retroplacental sonolucent zone19204985498518247949061
 3. Thinning or disruption of hyperechoic serosa–bladder interface701810010080701810010052
 4. Presence of focal exophytic masses invading the urinary bladder41109980794110978049
 5. Abnormal placental lacunae21205485518620453888372
 At least one gray-scale criterion37369573519836895768293
Color Doppler criteria
 6. Diffuse or focal lacunar flow pattern29157489669228474888875
 7. Vascular lakes with turbulent flow (PSV > 15 cm/s)2155496818820253949164
 8. Hypervascularity of serosa–bladder interface30287779529229876767874
 9. Markedly dilated vessels over peripheral subplacental zone78189447797518855948
 At least one color Doppler criterion364092694797351192687689
3D power Doppler sonography criteria
 10. Numerous coherent vessels involving the whole uterine serosa–bladder junction (basal view)38119792779937497889097
 11. Hypervascularity (lateral view)35159089709634389919289
 12. Inseparable cotyledonal and intervillous circulations, chaotic branching, detour vessels (lateral view)213549888882005210010065
 At least one 3D power Doppler criterion391910085761003851008588100

Likewise, to evaluate the usefulness of either diagnosing or ruling out placenta accreta, the positive predictive value (PPV) and negative predictive value (NPV) of these diagnostic measures were calculated. As shown in Table 1, ‘thinning or disruption of the hyperechoic serosa–bladder interface (i.e. bladder mucosa)’ had a PPV of 100%, indicating that the presence of this finding could confidently confirm the diagnosis. If we considered the presence of at least one criterion to be diagnostic, then 3D power Doppler gave the best PPV (76%), followed by gray-scale (51%) and color Doppler (47%). The criterion ‘coherent vessels in the placental base’ (detected on 3D power Doppler) had the highest NPV (99%). However, if the absence of any sign was used to reject the diagnosis of placenta accreta, all of the three ultrasound techniques gave a similar NPV.

To adjust for possible confounding effects from previous Cesarean deliveries, we further compared the ultrasound criteria in only those patients who had undergone a previous Cesarean delivery (Table 1). However, the sensitivities and specificities of the criteria for diagnosing placenta accreta were nearly unchanged. Surprisingly, the PPVs of most of the criteria for diagnosing accreta were increased, indicating a better reliability for detecting placenta accreta in this subset of the study population. It also showed that in this subset 3D power Doppler reached a very high level of accuracy for diagnosing placenta accreta (PPV ranging from 90 to 100% for criteria 10–12 in Table 1), even better than the diagnostic performance in the study population as a whole.

To assess the influence of single markers vs. multiple markers in the decision-making process of the differential diagnosis, we reviewed the numbers of positive ultrasound criteria in those patients with placenta accreta. Notably, the majority of patients with placenta accreta showed multiple ultrasound criteria of placenta accreta on gray-scale, color Doppler and 3D power Doppler imaging (Table 2). In contrast, those patients with a false-positive diagnosis (i.e. final diagnosis was placenta previa) tended to show single isolated ultrasound markers.

Table 2. Comparison of the number of positive ultrasound criteria for placenta accreta in patients proved to have placenta accreta or its variants, or placenta previa without placenta accreta
Ultrasound methodPlacenta accreta (n = 39)Placenta previa (n = 131)
Single criterionMultiple criteriaSingle criterionMultiple criteria
  1. Data are expressed as n (%). 3D-PD, three-dimensional power Doppler.

Gray-scale11 (28)26 (67)26 (20)10 (8)
Color Doppler8 (21)28 (72)28 (21)12 (9)
3D-PD4 (10)35 (90)11 (8)8 (6)

3D power Doppler alone yielded the fewest false-positive diagnoses (19 cases, vs. 36 cases by gray-scale and 40 cases by color Doppler criteria). Most of the false-positive diagnoses that arose from gray-scale criteria were due to the ‘irregularity or absence of retroplacental echolucent zone’ (Figure 7a) and ‘abnormal lacunae’ (Figure 7b). The ‘absent or irregular retroplacental echolucency’ in these non-accreta patients was probably due to a defective decidua with marked thinning of the myometrium. This may be due to a poor healing process after previous Cesarean delivery. The criteria of ‘hypervascularity’, ‘abnormal lacunae’ (Figure 7c,d) and ‘coherent basal vessels’ (Figure 7e,f) led to most of the false-positive diagnoses on color Doppler and 3D power Doppler imaging. Based on observations made during surgery, these manifestations in placenta previa totalis were most likely the counterpart of placental variations due to neovascularization following previous Cesarean delivery. According to our observations, placenta percreta could be further distinguished by the presence of an exophytic placenta invading the bladder or the disruption of the hyperechoic bladder mucosa. However, the sensitivity of neither sign was high. Placenta increta and accreta were otherwise unlikely to be further differentiated by these diagnostic systems.

thumbnail image

Figure 7. Ultrasound images showing false-positive diagnoses using the gray-scale, color Doppler and three-dimensional power Doppler diagnostic criteria. All these cases were confirmed to be placenta previa totalis without any evidence of placenta accreta. The images show loss of retroplacental echolucent zone (arrows) (a); focal absence of retroplacental echolucent zone (arrow) and abnormal lacunae (b); interface hypervascularity (c); hypervascularity and abnormal lacunar flow pattern (d); a hypervascularized interface with a high peak systolic velocity (171 cm/s) and pulsatile flow with a varicose vein within the bladder mucosa (inset) in the same area (e); and confluent vessels in the placental base (f). LUTA, left uterine artery; m, myometrium; RUTA, right uterine artery.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Antenatal detection of placenta accreta in patients with placenta previa is extremely important, since undiagnosed the condition may lead to maternal death or severe morbidity owing to lack of adequate preparation at the time of delivery. If placenta accreta is strongly suspected in the antenatal period, it may be prudent to complete delivery without removing the placenta and then proceed with hysterectomy immediately2, 25. Such patients should be counseled appropriately and prepared for Cesarean hysterectomy1. This may include anesthesia consultation, ureter stent placement, rich sources of blood products, cell-saver technology if available and prophylactic balloon occlusion of the hypogastric or common iliac arteries26, 27. Scheduled Cesarean section should be considered at 34–35 weeks' gestation to prevent unexpected bleeding before term28.

As a screening tool to detect placenta accreta, the prerequisite of each individual ultrasound criterion is a high sensitivity to detect more true positives in women at risk. A low sensitivity but high specificity test (e.g. ‘disrupted bladder mucosa’) would result in a number of undiagnosed patients, despite the high level of confidence in recognizing true negative cases (placenta previa). Since an ROC graph can depict the relative trade-offs between benefit (true positives; sensitivity) and error cost (false positives; 1 − specificity), it therefore provides us with valuable information for the decision-making involved in diagnosis and management. As depicted in Figure 6, all discrete classifiers in this ROC graph indeed produced few false positives, implying that erroneous hysterectomy (error cost) would seldom occur using the ultrasound criteria listed in this paper. However, it also illustrates a wide range of sensitivities for these different ultrasound criteria, indicating the need for further optimization of current diagnostic methods.

Given the unique clinical management of placenta accreta, the PPV and NPV of these diagnostic measures are as important as the sensitivity and specificity. Practically, the PPV reported for each criterion in this study indicates its reliability to correctly diagnose placenta accreta; in other words, the confidence with which clinicians can proceed straight to hysterectomy without removing the placenta. In contrast, the NPV represents the precision of diagnosis of placenta previa, relating to the confidence with which clinicians can remove the placenta without concerns of severe bleeding. For example, as a single diagnostic criterion, ‘hypervascularity of serosa–bladder interface’ alone (by color Doppler) has a high sensitivity (77%) yet an unsatisfactory PPV (52%). Hence, despite detecting more true positives, its reliability in confirming placenta accreta is low and thus it might lead to more unnecessary hysterectomies. Indeed, there is a report in the literature describing two cases of unnecessary hysterectomy that resulted from overdiagnosis9.

Most authors calculate the sensitivity, specificity, PPV and NPV by a combined assessment of all of the sub-criteria using gray-scale ultrasound or color Doppler10, 16, 19. Such combined assessment would sometimes result in a dilemma of clinical judgment. For example, if a patient presented with ‘abnormal placental lacunae’ yet without ‘disruption of retroplacental echolucent zone’, the diagnosis would present a dilemma without considering the PPV and NPV of each individual criterion. In this study, the PPV of ‘abnormal lacunae’ was only 51%; that is, the clinician only has around 50% confidence that it relates to a true-positive diagnosis. In contrast, as the NPV of ‘disruption of retroplacental echolucency’ was 85%, the clinician would have greater reason to reject the diagnosis of placenta accreta and remove the placenta directly without significant risk of bleeding.

We acknowledge that there are some potential pitfalls in analyzing our results. Firstly, modern ultrasound machines allow visualization of the residual myometrium beneath the invasive placenta at a greater resolution, thus the prevalence of ‘loss of retroplacental echolucency’ is not as high in our series as in those previously reported. Secondly, most of the patients in this series were referred from other hospitals. Indeed, general obstetricians tend to refer the severe cases with more apparent findings. Therefore the sensitivity, specificity, PPV and NPV may not be comparable with those of other studies. Lastly, it is important to remember that 3D power Doppler does have certain limitations including those that result from post-processing of the signal, such as the threshold and transparency settings, as well as those related to the relative weighting of the power Doppler and gray-scale signals when the data are displayed as a three-dimensionally rendered ‘glass body’ image.

According to the present study, 3D power Doppler ultrasonograpy can be used as a complementary technique for making or excluding the diagnosis of placenta accreta. We have also used the results of our study to produce a flow chart for the differential diagnosis between placenta accreta and placenta previa (Figure 8). In brief, placenta accreta can be confidently diagnosed if multiple ultrasound characteristics of the condition are present. When placenta accreta is suspected because of only a single ultrasound criterion, a comparison of the PPV of the positive criterion and the NPV of the negative criteria should be made. If the former is significantly greater than the latter, the clinical suspicion and management should tend toward the diagnosis of placenta accreta.

thumbnail image

Figure 8. A proposed flow chart for the differential diagnosis between placenta accreta and placenta previa. MS-AFP, maternal serum alpha-fetoprotein.

Download figure to PowerPoint

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

The author sincerely appreciates the great technical assistance provided by Dr Christian Grabner, Dr Guenther Hinterleitner and Dr Lu Xing from GE Healthcare. This research was supported by a grant from National Science Council of Taiwan (NSC 96-2314-B-002-114).

SUPPORTING INFORMATION ON THE INTERNET

The following supporting information may be found in the online version of this article:

Video files correspond to the following figures:

Videoclip S1, Figure 2b; Videoclip S2, Figure 2c; Videoclip S3, Figure 3; Videoclip S4, Figure 5a; Videoclip S5, Figure 5d.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
FilenameFormatSizeDescription
uog_6284_sm_videoclips1.mpg324KVideoclip S1, Figure 2b
uog_6284_sm_videoclips2.mpg496KVideoclip S2, Figure 2c
uog_6284_sm_videoclips3.mpg474KVideoclip S3, Figure 3
uog_6284_sm_videoclips4.mpg816KVideoclip S4, Figure 5a
uog_6284_sm_videoclips5.mpg738KVideoclip S5, Figure 5d

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.