Sonography‐Assisted Management of Suspected Retained Products of Conception Upon Completion of the Puerperium: A Prospective Single‐Center Study

Abstract

Objectives

Retained products of conception (RPOC) occur in approximately 1% of term deliveries and may result in bleeding, infection, and reproductive complications. However, early postpartum symptoms and ultrasound findings are often nonspecific, making diagnosis challenging. This study evaluated whether transvaginal sonography performed 6 weeks postpartum can effectively stratify RPOC risk in suspected cases and guide follow‐up and management.

Methods

In this prospective study, 116 hemodynamically stable women with risk factors for RPOC and without significant symptoms underwent transvaginal ultrasound 6 weeks postpartum. Findings were classified as high, moderate, or low probability based on the presence of an intrauterine mass, endometrial thickness, and vascularity. Low probability cases required no further evaluation, whereas those in the moderate and high probability groups were referred for diagnostic or operative hysteroscopy, respectively.

Results

Patients were categorized into high (19%), moderate (28%), and low (53%) probability groups. The prevalence of histologically confirmed RPOC in these groups was 50, 12.5, and 0%, respectively. Compared with prior studies that performed earlier postpartum assessments, evaluation at 6 weeks improved diagnostic accuracy, particularly in moderate probability cases, while predictive values in high and low probability groups remained comparable.

Conclusions

Sonographic assessment at 6 weeks postpartum provides reliable risk stratification for RPOC and supports a more conservative management approach in hemodynamically stable women without significant symptoms by deferring imaging and potential interventions until completion of the postpartum period. This approach improves diagnostic accuracy in the moderate probability group while reducing false positives and unnecessary interventions.

Abbreviations

D&C

dilation and curettage

kHz

kilohertz

MHz

megahertz

PPV

positive predictive value

RPOC

retained products of conception

SD

standard deviation

SPSS

statistical package for the social sciences

Retained products of conception (RPOC) refer to persistent trophoblastic tissue within the uterine cavity and occur in approximately 1% of full‐term deliveries, ¹ including cesarean sections. ² RPOC is associated with early and long‐term complications. Early complications include bleeding and infection, while long‐term complications include intrauterine adhesions, Asherman's syndrome, and subfertility, which are important to prevent through appropriate treatment. ³ Diagnosis is challenging because early symptoms such as prolonged postpartum hemorrhage, fever, and abdominal pain are nonspecific, and ultrasound is usually the first‐line imaging modality. However, ultrasound findings of RPOC may overlap with normal postpartum intrauterine contents such as blood clots or necrotic decidua, ⁴ creating diagnostic uncertainty. When ultrasound findings raise suspicion for RPOC, standard management is surgical uterine evacuation, underscoring the importance of accurate assessment to avoid unnecessary interventions. In 2 earlier studies from our group, we identified key diagnostic patterns that directly guided the design of the present study. The first, a retrospective study, demonstrated that gray‐scale ultrasound combined with color Doppler imaging can reliably stratify patients into high, moderate, or low risk for RPOC (Figure 1), with confirmed rates of 62, 32, and 0% in each group, respectively. ⁵ We also found that diagnostic accuracy was markedly improved when ultrasound was performed at the end of the puerperium, as physiological intrauterine contents commonly present in the early postpartum weeks typically resolve spontaneously by this time. ⁶ In a subsequent prospective study, we validated this sonographic classification system and again confirmed its superior performance among the subgroup of patients in whom imaging was performed after the puerperal period. ⁷

Figure 1.

Transvaginal sagittal images of the uterine cavity in patients with suspected retained products of conception, 6 weeks after delivery. A, High probability: an echogenic mass with vascularity. B, Moderate probability: an echogenic mass without vascularity. C, Low probability: no echogenic mass, endometrial thickness <10 mm, no vascularity.

These cumulative findings directly shaped the design of the present study, the third in this research sequence. Building on insights from the earlier studies, we refined our clinical protocol to intentionally defer transvaginal sonography to approximately 6 weeks postpartum, using this standardized time point to classify patients into high, moderate, or low probability of RPOC. This strategy aims to enhance diagnostic accuracy, reduce false positives, and minimize unnecessary interventions. Accordingly, the objective of the present study was to determine whether transvaginal sonography at 6 weeks postpartum can effectively stratify RPOC risk and guide clinical decision‐making in hemodynamically stable women.

Material and Methods

This prospective study was conducted between March 2021 and November 2023 and included women who delivered in the third trimester at our institution and met predefined inclusion criteria for suspected RPOC. Inclusion criteria comprised risk factors such as manual placental removal, uterine cavity revision, abnormal placental appearance, clinical symptoms (vaginal bleeding, abdominal pain, or fever), or a history of RPOC, as well as abnormal intrauterine ultrasound findings or a hemoglobin decrease of greater than 3 g/dL compared with admission values. Only women fulfilling these criteria were invited for follow‐up evaluation 6 weeks postpartum. All participants were hemodynamically stable and asymptomatic or mildly symptomatic; otherwise, earlier intervention would have been required.

Ultrasound examinations were performed at least 6 weeks postpartum using a 5–9 MHz transvaginal probe (GE E8 or E10, GE Medical Systems, Milwaukee, WI, USA). Sonographic findings were categorized into high, moderate, and low probability groups according to the classification established in our earlier work, ⁷ which defines RPOC likelihood based on endometrial thickness, the presence of an echogenic intrauterine mass, and the presence or absence of vascularity (Table 1).

Table 1.

Sonographic Classification and Management Recommendations for Suspected RPOC at 6 Weeks Postpartum

Low probability	Moderate probability	High probability
No endometrial mass, endometrial thickness <10 mm, complex fluid <10 mm	Endometrial echogenic mass or thick endometrium (≥10 mm) without vascularity or complex fluid (≥10 mm)	Endometrial echogenic mass or thick endometrium (≥10 mm) with vascularity
Recommendation: No follow‐up is required	Recommendation: Diagnostic hysteroscopy	Recommendation: Operative hysteroscopy

All examinations were performed by experienced sonographers (O.L.T., O.S., A.O.) and included grayscale and Doppler imaging. Endometrial thickness was measured in the sagittal plane at its maximal diameter, and vascularity was assessed by color Doppler signals arising from the endometrial mass or thickened endometrium. Low‐flow sensitivity was optimized by setting the pulse repetition frequency to 0.3–0.6 kHz, with color gain adjusted just below the noise threshold. Representative images illustrating the 3 risk categories are presented in Figure 1. All sonographers were familiar with the classification system, which is routinely used and documented in our department. Results were communicated to the patient and the surgical team. Cases classified as low probability required no further follow‐up. Those with moderate probability were referred to diagnostic hysteroscopy without anesthesia, whereas cases with high probability were directed to operative hysteroscopy performed in the operating room, usually under general anesthesia (Table 1). RPOC diagnosis was confirmed histologically following endoscopic removal. In cases referred directly to operative hysteroscopy, removal was performed in the operating room. For patients undergoing diagnostic hysteroscopy, suspected RPOC could be removed during the same procedure in a “see‐and‐treat” setting, depending on operator judgment and lesion characteristics; otherwise, patients were referred for operative hysteroscopy for definitive treatment.

Demographic and obstetric data, delivery details, operative procedures, and pathological results were collected. If no surgical intervention was performed at our center, patients were contacted at least 12 months postpartum to assess clinical follow‐up and obtain external histopathology reports when available. In cases without surgical intervention, RPOC was considered absent only if this follow‐up confirmed the absence of subsequent symptoms or sonographic findings suggestive of RPOC. The study was approved by the institutional ethics committee, and written informed consent was obtained from all participants. Their participation did not influence clinical management, which was conducted according to departmental protocol, and consent was obtained for the use of clinical data for research purposes.

Statistical Analysis

Statistical analysis was conducted using SPSS Statistics version 24 (Chicago, IL, USA). Categorical variables are presented as percentages and continuous variables as mean ± SD or median (range). Differences were analyzed using Student's t‐test, and associations between sonographic findings and RPOC were evaluated with Fisher's exact test or Pearson's chi‐square test. A p‐value of less than .05 was considered statistically significant.

Results

Data from 116 patients were analyzed, with a mean maternal age of 31.5 years (range: 19–42 years). Of these, 100 patients (86%) had vaginal deliveries, including 12 vacuum‐assisted births, and 16 (14%) underwent cesarean section. At the time of ultrasound, 80% were asymptomatic. Among symptomatic patients, vaginal bleeding was most common (88%), followed by fever (8%) and abdominal pain (4%). The incidence of symptoms varied significantly by RPOC probability group: 52% in the high probability group, 34% in the moderate probability group, and only 2% in the low probability group, representing a statistically significant difference.

Ultrasound examinations were performed at the end of the puerperium. Based on the sonographic findings, 22 patients (19%) were classified as high probability, 32 (28%) as moderate probability, and 62 (53%) as low probability for RPOC. Out of the 116 patients, 46 (40%) underwent uterine evacuation procedures, 44 via hysteroscopy and 2 via suction curettage. This included 21 patients (95%) from the high probability group, 24 (75%) from the moderate group, and 1 patient (1.2%) from the low probability group.

Histopathological confirmation of RPOC was obtained in 15 cases (13% overall): 11 in the high probability group and 4 in the moderate probability group, with no confirmed cases in the low probability group. These findings correspond to a statistically significant gradient in RPOC prevalence across the groups, with rates of 50% in the high probability group, 12.5% in the moderate probability group, and 0% in the low probability group (p < .05) (Figure 2). All women who did not undergo surgical intervention at our center or elsewhere remained asymptomatic during follow‐up, based on telephone contact conducted at least 12 months postpartum, and their most recent ultrasound examination demonstrated no sonographic findings suggestive of RPOC.

Figure 2.

Flow chart summarizing the detection rate of retained products of conception in low, moderate, and high probability groups. p < .0001, Pearson's chi‐square test.

Discussion

In contrast to our previous management approach, in which sonographic evaluation for suspected RPOC and classification into high, moderate, or low probability could be performed at any time after delivery, depending on clinical presentation and referral timing, the current study implemented a different approach. When less than 6 weeks had elapsed since delivery, patients were scheduled for a follow‐up visit at the end of the puerperium, regardless of the initial ultrasound findings. The classification and corresponding treatment recommendations were determined only at that stage. This approach is appropriate for hemodynamically stable women without significant symptoms, who represent most cases, as RPOC is often asymptomatic and detected on routine ultrasound. ⁸ Setting 6 weeks postpartum as the time point for follow‐up was guided by both clinical and methodological considerations. First, our previous studies ⁵ , ⁷ have demonstrated that postponing the evaluation until the completion of the puerperal period significantly reduces the false‐positive rate in diagnosing RPOC, as other intrauterine contents that may mimic residual tissue, such as blood clots, debris, and necrotic decidua, are typically expelled spontaneously by that time. ⁶ , ⁹ Second, genuine retained placental tissue itself may also be spontaneously expelled in certain cases, and allowing adequate time for this natural resolution is therefore beneficial. A 6‐week interval has been consistently reported as a safe and appropriate timeframe in this regard. ¹⁰ , ¹¹ Another consideration is that hysteroscopy is widely regarded as the optimal approach for both diagnosis and treatment of RPOC. ¹² , ¹³ Performing this procedure after the puerperium, when uterine involution has occurred and the cervical canal is closed, facilitates a technically easier and safer intervention. In addition, conducting a follow‐up examination at the end of the puerperium is a practical approach, as it corresponds with the standard post‐delivery follow‐up that is generally recommended at this point.

Our study demonstrates that among the low probability group, characterized by negative ultrasound findings, no cases developed clinical or sonographic findings suggestive of RPOC during follow‐up, indicating a very low likelihood of clinically significant retained tissue. This finding is consistent with the results reported by Durfee et al ¹⁴ and with our previous studies, ⁵ , ⁷ and suggests that the absence of ultrasound features indicative of retained products, as defined in the classification, makes the presence of RPOC unlikely. The current study demonstrates that the positive predictive value (PPV) for the high probability group, defined by the presence of intrauterine sonographic findings with vascularity, was 50%. In this group, 4 additional women underwent hysteroscopy, which revealed the presence of RPOC. In 1 case, operative hysteroscopy was not pursued due to a technical limitation identified during the procedure and because the patient had no desire for future pregnancies. Three other patients underwent hysteroscopy in private medical settings and, according to their reports, RPOC were identified, but the corresponding pathology reports could not be obtained. In the absence of histopathological confirmation, these cases were not included in the count of confirmed residuals. Therefore, the actual number of residual cases is likely higher, and the estimated rate of RPOC among the high probability group ranges between 50 and 68%. This result, too, is consistent with our previous prospective study. ⁷ The only patient in the high probability group who did not undergo hysteroscopy had a corpus luteum identified on ultrasound, and a follow‐up ultrasound after the subsequent menstrual cycle was therefore considered appropriate. At the follow‐up examination, the intrauterine finding had resolved spontaneously, rendering hysteroscopy unnecessary. The moderate probability group, characterized by avascular material within the uterine cavity, appears to be the most susceptible to false positive diagnoses, likely due to the challenge of distinguishing avascular RPOC from blood clots or intrauterine debris. In this group, implementation of the updated protocol evaluated in the current study, which involved ultrasound assessment and clinical categorization at the end of the postpartum period, led to a notable improvement in the PPV, increasing from 2% in our previous study ⁷ where most ultrasound assessments were performed during early puerperium, to 12.5%. In the moderate probability group, 23 women underwent hysteroscopy, and 1 underwent dilation and curettage (D&C) due to bleeding. Among the 24 interventions, 4 yielded positive histological findings consistent with RPOC. In 1 additional case, the hysteroscopy was performed in a private setting, where retained tissue was reportedly observed; however, the pathology report could not be obtained, and this case was therefore not included among the confirmed RPOC cases. One hysteroscopy revealed an intrauterine polyp, representing one of the differential diagnoses for RPOC. In 14 hysteroscopies, the intrauterine appearance was consistent with retained tissue, similar to the ultrasound findings; however, pathological examination did not confirm the presence of gestational tissue. These results underscore the inherent limitations of both ultrasound and hysteroscopy in accurately diagnosing RPOC within this group. Despite improvement over previous studies, the relatively low PPV in the moderate probability group (12.5%) may support extending conservative follow‐up beyond 6 weeks postpartum, as intrauterine contents not representing RPOC might be expelled spontaneously even at later postpartum stages, potentially reducing the need for surgical intervention. This issue remains to be addressed in future prospective studies.

Limitations

This study has several limitations. Not all participants underwent surgical management; thus, histopathological confirmation was unavailable for most patients classified as negative for RPOC. This approach reflects routine clinical practice, in which invasive diagnostic procedures are not routinely performed in clinically stable patients with low sonographic suspicion. As a result, a uniform reference standard was not applied across all groups, potentially introducing partial verification bias. This limitation is inherent to most studies addressing this clinical question. It remains possible that a small proportion had residual tissue, although these cases were likely clinically insignificant and resolved spontaneously. Additionally, pathology reports were unavailable for 4 patients who had hysteroscopy, potentially slightly underestimating PPV in the high and moderate probability groups. Nevertheless, statistical significance and study conclusions were unaffected by this.

Despite these limitations, the present study builds on previous research from our institution, which guided the establishment of a standardized 6‐week postpartum ultrasound protocol. The current findings further support the effectiveness of this modified protocol in clinical practice, confirming the positive impact of the changes implemented based on earlier research.

In summary, the present study suggests that the proposed sonographic classification system may assist in the management of patients with suspected RPOC following delivery. Our findings support the use of a conservative management approach for clinically stable women in all risk categories until the end of the puerperium. This approach appears to be both safe and effective, allowing reliable prediction of RPOC while reducing false positives and unnecessary interventions. Additional experience is required to validate our results and improve diagnostic accuracy, particularly in the high probability group.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.