Predictive Capacity of First Trimester Diagnosis of Placenta Previa

Abstract

Objective

First trimester transabdominal ultrasound is sometimes used to diagnose placenta previa and counsel patients accordingly. We aimed to determine the predictive capacity of a first trimester transabdominal ultrasonographic placenta previa diagnosis for persistence to the second trimester.

Study Design

Retrospective cohort study of patients with singleton pregnancies and first trimester transabdominal ultrasonographic placenta previa diagnoses from January to December 2022. The primary outcome was the predictive capacity of a first trimester transabdominal ultrasound diagnosis of placenta previa for placenta previa persistence into the second trimester. Secondary outcomes included the predictive capacity of a first trimester transabdominal ultrasound for placenta previa persistence to delivery and risk factors associated with placenta previa persistence. Chi-square and Student’s t test were used to determine statistical significance, and a multivariable logistic regression determined the strength of associations.

Results

Of the 185 patients with a first trimester transabdominal ultrasound diagnosis of placenta previa, 159 (86.0%) resolved by the second trimester resulting in a predictive capacity for persistence to the second trimester of 14.0%. Moreover,182 (98.4%) resolved by delivery, resulting in a predictive capacity for persistence to delivery of 1.6%. Among the 27 patients with a prior cesarean delivery, the predictive capacity of a first trimester placenta previa diagnosis for persistence to the second trimester was 22.2% and to delivery was 7.4%. Advanced maternal age and posterior placenta in the first trimester were risk factors for previa persistence to the second trimester, while prior cesarean delivery and reproductive assistance were not.

Conclusion

First trimester transabdominal ultrasonographic placenta previa diagnosis has a poor predictive capacity for placenta previa persistence to the second trimester and even lower for persistence to delivery. Counseling patients regarding placenta previa diagnosis in the first trimester may result in unnecessary patient anxiety and activity restrictions.

Introduction

Placenta previa refers to the presence of placental tissue that extends over the internal cervical os. Placenta previa leads to increased maternal and fetal morbidity including the need for cesarean delivery, potential for antepartum bleeding, preterm birth, and postpartum hemorrhage.^1,2 The prevalence of placenta previa at delivery is reported to be 4 to 5 per 1000 births³ Over the years, the incidence of placenta previa has increased, largely due to the increased presence of associated risk factors such as cesarean deliveries, uterine procedures, assisted reproductive technology, endometriosis, and increased maternal age at delivery.^{4,5,14–16,6–13} Placenta previa is most often diagnosed at the time of the midtrimester anatomical scan between 18 and 22 weeks’ gestation.^17–20 Notably, over 90 percent of midtrimester placenta previa will resolve before delivery.²¹

With the recent increase in utilization of the first trimester ultrasound evaluation, it has become more common to diagnose placenta previa even earlier in pregnancy.^17,22–24 Although the diagnosis is preferentially made by transvaginal ultrasound (TVUS), the majority of first trimester diagnoses will be made via transabdominal ultrasound (TAUS).^20,25–29 Numerous studies have investigated the clinical utility of a first trimester TVUS diagnosis of placenta previa and found high rates of false positives with a low predictive capacity.^30–33 Despite this poor predictive capacity of a first trimester TVUS diagnosis of placenta previa, often an even less optimal first trimester TAUS diagnosis of placenta previa is communicated to patients, which may result in unnecessary patient anxiety and activity restrictions.

Our objective was to characterize the predictive capacity of a first trimester TAUS placenta previa diagnosis for persistence to the second trimester and until delivery. Given the findings for TVUS, we hypothesize that TAUS first trimester diagnosis of placenta previa will likewise have a poor predictive capacity.

Methods

We conducted a retrospective cohort study of singleton pregnancies with first trimester placenta previa diagnosed during a first trimester transabdominal ultrasound between 12w0d and 13w6d at a single academic center from January to December 2022. Only patients who delivered at the institution were included. At our institution, all patients are recommended to have a first trimester ultrasound for pregnancy dating and nuchal translucency. All ultrasonographic evaluations are conducted by accredited fetal sonographers and formally read by maternal-fetal medicine physicians and radiologists. For the evaluation, TAUS is routinely used, and a diagnosis of placenta previa alone does not trigger the utilization of TVUS. TVUS during the evaluation is reserved for suspected fetal anomalies or concern for placenta accreta spectrum. All records were queried from GE Healthcare Viewpoint 6 Ultrasound Reporting Software for Women’s Health to determine our cohort. We included all singleton patients with a first trimester TAUS diagnosis of placenta previa. We excluded patients delivering at an outside hospital, those without a second trimester ultrasound in our system, and pregnancies ending in a termination or fetal loss < 24-weeks’ gestation.

We defined low placentation as either having a placenta previa defined as any portion of the placenta covering the internal cervical os or a low-lying placenta defined as having less than 2 centimeters from the edge of the placenta to the cervix. The primary outcome was the predictive capacity of a first trimester TAUS diagnosis of placenta previa for persistence into the second trimester. Patients with placenta previa in the first trimester with a low-lying placenta in the second trimester were considered to be a resolved previa. The secondary outcomes included the predictive capacity of a first trimester TAUS diagnosis of placenta previa for persistence to delivery and risk factors associated with placenta previa persistence from first trimester to second trimester. Placenta previa persistence to delivery was defined as a finding of placenta previa on admission to labor and delivery and requiring a scheduled cesarean for delivery. Additionally, we performed a sub-group analysis among only patients with a history of prior cesarean delivery.

We used a combination of chi-square, Fisher exact, student’s T, and Mann-Whitney U tests for univariable analysis based on results distribution and sample sizes as statistically appropriate. Multivariable regression tests were used to determine the strengths of association between risk factors associated with placenta previa persistence. Multivariable analysis included prior cesarean delivery, advanced maternal age, assisted reproductive technology, and placental location at first trimester scan. In all analyses, a p-value of less than 0.05 or 95% confidence interval not crossing 1.00 indicated statistical significance. All statistical analyses were performed on STATA IC Version 15. All data was retrospectively abstracted by members of the research team from maternal and neonatal charts within the electronic medical record and then stored in a Research Electronic Data Capture (REDCap) database. The Institutional Review Board approved this study, and a waiver of consent was obtained.

Results

During the study period, there were 3021 patients with a first trimester ultrasound completed; among those, 637 (21.1%) patients had a TAUS diagnosis of low placentation (Figure 1). After exclusions, there were 185 (6.1%) patients with a first trimester TAUS diagnosis of placenta previa of which 159 (86.0%) resolved by the second trimester resulting in a predictive capacity of 14.0% and 182 (98.4%) resolved by delivery resulting in a predictive capacity of 1.6% (Table 2). Among the 27 patients who had a prior cesarean delivery with a first trimester TAUS diagnosis of placenta previa, 21 (77.8%) resolved by the second trimester resulting in a predictive capacity of 22.2% and 25 (92.6%) resolved by delivery resulting in a predictive capacity of 7.4% (Table 2).

Figure 1.

Cohort flowchart highlighting breakdown of study groups.

Table 2.

Summary of Placental Progression

	Resolved Previa at Second Trimester Scan (n = 159, 86.0%)	Persistent Previa at Second Trimester Scan (n = 26, 14.0%)	p-value
Placental Diagnosis at Second Scan, n (%)			<0.01
Resolved Placenta Previa	159 (100.0)	18 (69.2)
Low-Lying Placenta	0 (0.0)	5 (19.2)
Placenta Previa	0 (0.0)	3 (11.5)
	Resolved Previa at Delivery (n = 182, 98.4%)	Persistent Previa at Delivery (n = 2, 1.6%)	p-value
Placental Diagnosis at Delivery, n (%)			<0.01
Resolved Placenta Previa	182 (100.0)	0 (0.0)
Low-Lying Placenta	0 (0.0)	0 (0.0)
Placenta Previa	0 (0.0)	2 (100.0)

Patients that had placenta previa persist to the second trimester were more likely to be of advanced maternal age (resolved 31.5% versus persistent 73.1%) and have used reproductive assistance (resolved 7.6% versus persistent 26.9%) (Table 1). There were no significant differences in nulliparity, BMI at first trimester or delivery, prior uterine surgery, fibroid uterus, uterine anomalies, or medical comorbidities (Table 1).

Table 1.

Maternal Demographics and Characteristics

	Resolved Previa at Second Trimester Scan (n = 159, 86.0%)	Persistent Previa at Second Trimester Scan (n = 26, 14.0%)	p-Value
Maternal Age, mean +/− SD	32.4 +/− 4.8	35.5 +/− 4.6	<0.01
Advanced Maternal Age, n (%)	50 (31.5)	19 (73.1)	<0.01
Nulliparity, n (%)	78 (49.1)	15 (57.7)	0.41
BMI at First Trimester, mean +/− SD	25.5 +/− 4.9	24.4 +/− 4.2	0.29
BMI at Delivery, mean +/− SD	30.5 +/− 5.0	28.9 +/− 4.5	0.12
Prior Cesarean, n (%)	21 (13.2)	6 (23.1)	0.19
Prior Uterine Surgery, n (%)*	26 (16.4)	5 (19.2)	0.72
Fibroid Uterus, n (%)	14 (8.8)	2 (7.7)	0.85
Uterine Anomaly, n (%)	7 (4.4)	1 (3.9)	0.90
IVF or Reproductive Assistance, n (%)	12 (7.6)	7 (26.9)	<0.01
Diabetes, n (%)**	19 (12.0)	4 (15.4)	0.62
Hypertensive Disorder, n (%)***	38 (23.9)	3 (11.5)	0.16
Suspected Fetal Growth Restriction, n (%)	4 (2.5)	2 (7.7)	0.17
Autoimmune disease, n (%)	10 (6.30)	2 (7.7)	0.79
Triage/Emergency Visit for Bleeding, n (%)	20 (12.6)	6 (23.1)	0.15
Antepartum Admission for Bleeding, n (%)	2 (1.3)	3 (11.5)	<0.01
Gestational Age at Delivery, mean +/− SD	39.1 +/− 1.6	38.0 +/− 2.0	0.23
Gestational Age at First Trimester Scan, mean +/− SD	12.7 +/− 0.5	12.5 +/− 0.5	0.12
Placental Location at First Trimester, n (%)			0.02
Anterior	72 (45.3)	5 (19.2)
Posterior	87 (54.7)	20 (80.8)
Gestational Age at Second Trimester Scan, mean +/− SD	19.9 +/− 1.5	19.2 +/− 1.1	0.07

^*Uterine Surgery included dilation and curettage or evacuation, diagnostic or operative hysteroscopy, endometrial ablation, or laparoscopic or abdominal myomectomy

^**Diabetes includes all gestational and pre-gestational diabetes

^***Hypertensive Disorders includes gestational and chronic hypertension and any pre-eclampsia spectrum disease

Given the known risk factors associated with placenta previa persistence, we assessed the same risk factors for second trimester persistence in a multivariable regression including prior cesarean delivery, advanced maternal age, reproductive assistance, and placental location at first trimester scan. Advanced maternal age (aOR 5.07, 95% CI 1.88–13.68) and posterior placenta in the first trimester (aOR 4.02, 95% CI 1.33–12.15) were risk factors for previa persistence to the second trimester. Prior cesarean delivery (aOR 1.94, 95% CI 0.62–6.13) and reproductive assistance (aOR 2.64, 95% CI 0.77–9.02) were not associated with increased risk (Table 3).

Table 3.

Risk Factors for Second Trimester Previa Persistence

	Unadjusted OR (95% CI)	Adjusted OR* (95% CI)
Prior Cesarean Delivery	1.97 (0.71–5.47)	1.94 (0.62–6.13)
Advanced Maternal Age	5.92 (2.34–14.98)	5.07 (1.88–13.68)
IVF or Reproductive Assistance	4.51 (1.58–12.86)	2.64 (0.77–9.02)
Posterior Placenta at First Trimester**	3.33 (1.19–9.32)	4.02 (1.33–12.15)

^*Multivariable regression including prior cesarean delivery, advanced maternal age, reproductive assistance, and placental location at first trimester scan

^**Reference for odds ratio was comparing to anterior placenta at first trimester scan

Discussion

Our primary outcome analysis reveals that first trimester TAUS diagnosis of placenta previa has a poor predictive capacity for persistence to the second trimester and even worse predictive capacity for persistence to delivery. We performed a sub-group analysis on patients with a history of prior cesarean delivery given findings in the literature regarding increased association of previa persistence with prior cesarean delivery.^4,5,34–36 In a high-risk cohort of patients with a history of prior cesarean, the predictive capacity for persistence of a first trimester TAUS diagnosis of placenta previa to the second trimester or delivery was higher, but remained poor.

Findings that help predict previa persistence from second trimester to delivery include higher gestational age at most recent sonogram depicting previa, prior cesarean delivery, degree of extension over cervical os, placental thickness at midtrimester scan, reproductive assistance, and posterior placentation.^{18,30,37–43} Not surprisingly, patients with persistent previa at second trimester scan were more likely to be advanced maternal age and have used reproductive assistance. Although there was a trend towards prior cesarean delivery and placenta previa persistence to the second trimester, we were likely underpowered to show a statistical difference.

Despite literature suggesting anterior placentation is associated with increased likelihood of placenta previa persistence, our findings show posterior placentation was associated with persistence to the second trimester. There are two leading hypotheses for placenta previa resolution: 1) as the lower uterine segment develops throughout gestation, this uterine remodeling pulls the placenta away from the lower uterine segment, and 2) preferential migration of the placenta towards a more cephalad position since the lower uterine segment and internal cervical os provide less vascular support for a pregnancy. With these hypotheses in mind, posterior placentation does not have significant biologic plausibility to be associated with persistence and given the overall low likelihood of placenta previa persistence to delivery in this cohort, it is more likely that our finding of posterior placentation and persistence is a chance occurrence.

Multiple studies have aimed to determine the clinical utility of a TVUS first trimester diagnosis of placenta previa. Three of the earliest studies found that first trimester TVUS diagnosis of placenta previa had a predictive capacity for persistence to term of 5.2% (4 out of 77), 5.1% (8 out of 156), and 7.6% (8 out of 105).^30–32 These findings suggest a very high rate of false positives. Taipale, et al. also noted that the placenta extending greater than 15 millimeters over the internal os would have an increased association with previa persistence. Another study aimed to determine the extent of overlapping of the placental edge that would best predict placenta previa at term and found that placental edge overlap greater than 23 millimeters has an 83.3% sensitivity and 86.1% specificity.³³ Regardless of the poor predictive capacity of the optimal TVUS diagnosis of first trimester placenta previa, some institutions continue reporting and counseling based on findings from the inferior TAUS imaging modality. Our findings represent a novel investigation of the clinical utility of a TAUS first trimester placenta previa diagnosis in predicting persistence to second trimester or delivery.

Our findings provide context for counseling after placenta previa is diagnosed via first trimester TAUS imaging. Namely, it allows providers the ability to reassure patients on the high likelihood of false positive. Taking into consideration the poor predictive capacity of a first trimester TAUS diagnosis of placenta previa, providers may opt to not disclose the findings in otherwise low risk pregnancies to lessen the possible maternal anxiety that may result. Specifically, one study assessed maternal psychological outcomes associated with the FTAS scan and found that 99% of patients prefer to know about pregnancy abnormalities as early as possible, however, their study did not specifically address placental abnormalities and instead focused primarily on fetal structural anomalies.⁴⁴

While our findings here provide information for counseling, additional studies are needed to further characterize this relationship. Specifically, it is not known whether a second trimester resolved placenta previa is associated with adverse outcomes such as bleeding at delivery, uterine atony, or need for cesarean delivery. Similar studies to these have been conducted for midtrimester placenta previas that resolved by time of delivery and found an association with uterine atony and postpartum hemorrhage.^45–47 Additionally, studies are needed to evaluate patient attitudes regarding first trimester TAUS placenta previa diagnosis and the resulting psychological impact of receiving this information early with a higher risk of a false positive result. While the previously mentioned study looking at psychological outcomes with first trimester ultrasound suggests patients desire earlier information, these findings are not generalizable to the discovery of placenta previa given that unlike fetal anomalies, placenta previa may resolve on its own. While there have not been any studies to date looking at maternal anxiety in the setting of placenta previa diagnosis, a 2021 systematic review demonstrated that among women hospitalized for obstetric complications, risk of anxiety and depression were doubled, suggesting a strong link between diagnosis of a complication and development of a maternal mood disorder.⁴⁸ Beyond understanding the way patients approach first trimester diagnosis of placenta previa, it remains unclear whether there is any clinical benefit from having this information since, regardless of findings, routine screening is performed during the second trimester anatomic survey. One benefit of early diagnosis of placenta previa may be risk stratification and optimized management in the presence of first trimester or early second trimester vaginal bleeding.

There are multiple strengths of this study. Since we performed this study at a single academic center over a short time interval, there were minimal changes in sonographers performing the scans, no deviations in the standardized scanning protocols, and consistent prenatal management. Because a single electronic medical record was utilized for the study, we had complete records for each patient included. Additionally, there were minimal demographic differences noted throughout our study population.

Even with complete records at a single institution, as a retrospective study, we are limited by errors in chart review and residual confounding from unmeasured variables. Despite our cohort being larger than any of the prior first trimester TVUS placenta previa studies, we remain underpowered to comment on our secondary outcomes assessing risk factors for persistence to the second trimester such as prior cesarean delivery or reproductive assistance. With respect to risk factors for previa persistence, we do not have information on the placental thickness or the degree of extension over the internal cervical os. Lastly, as a single institution, the generalizability of our study findings may be limited by our patient population and ultrasound protocols.

Conclusions

Our findings suggest that a first trimester TAUS diagnosis of placenta previa is a poor predictor of persistence to second trimester, both among patients with theoretical risk factors for persistence (prior cesarean or reproductive assistance) and those without. Counseling patients regarding placenta previa diagnosis in the first trimester may unnecessarily result in patient anxiety and activity restrictions. Further studies are needed to determine whether patients prefer to be counseled about a placenta previa diagnosis in the first trimester, and whether resolution of placenta previa by the second trimester is associated with adverse maternal or fetal outcomes.

Key Points.

• First trimester transabdominal placenta previa diagnosis has poor predictive capacity

• Counseling regarding first trimester placenta previa may result in unnecessary patient anxiety

• Studies needed to see whether patients prefer placenta previa disclosure in the first trimester