Maternal severe obesity and sonographic visualization of fetal anatomy

Abstract

Objective

We sought to investigate adequacy of fetal anatomic visualization in pregnant patients with BMI ≥ 50 kg/m² during second trimester fetal anatomic survey.

Methods

This was a retrospective chart review of singleton pregnancies presenting for fetal anatomic survey from 18 0/7 to 21 6/7 weeks gestation between January 2019 and December 2020 at a single tertiary care center. The primary outcome was rate of anatomic survey completion stratified by body mass index (BMI) category, as defined by the World Health Organization. Severe obesity was defined as BMI ≥50 kg/m². Secondary outcomes included scan duration, number of images acquired, and adequacy of visualization of individual organ systems. Logistic regression was used to generate an adjusted odds ratio (aOR) for anatomic survey completion compared to the normal BMI referent group.

Results

Of 2372 anatomic surveys, 9.0% of patients were normal weight, 31.5% of patients were overweight, 54.7% of patients were obese (26.0% class I, 15.9% class II, 12.2% class III), and 4.8% of patients were with severe obesity. Rates of anatomic survey completion were highest in normal weight patients, and lowest in those with severe obesity (67.6%  vs. 4.4%, p < .001). Severe obesity impacted visualization of all organ systems; cardiac structures were most affected compared to normal‐weight individuals (75.1%  vs. 8.7%, p < .001). Compared to patients with normal BMI, there are significantly reduced odds of anatomic survey completion in patients with class I obesity (aOR 0.70, 95% CI 0.50–0.98), class II obesity (aOR 0.42 95% CI 0.30–0.61), class III obesity (aOR 0.13, 95% CI 0.08–0.19), and severe obesity (aOR 0.02, 95% CI 0.01–0.05). The median exam duration and number of images acquired decreased with increasing BMI (ranging between 33.5–28.4 min and 86–63 images).

Conclusions

Patients with severe obesity have extremely low rates of anatomic survey completion at the first attempt. The inability to adequately evaluate fetal anatomy in this patient population is especially concerning due to the higher baseline risk of fetal anomalies with increasing obesity. The healthcare disparity demonstrated in this data justifies further research on advancements in practice standards and technology in obstetric imaging.

1. INTRODUCTION

Obesity rates continue to rise across all age groups, with a prevalence of 39.7% among 20–39 year‐old reproductive age patients in the United States in 2017–2018 [1]. It is well established that obesity presents challenges in prenatal ultrasonography, with evidence reporting poor detection of soft markers for aneuploidy and suboptimal visualization in patients with obesity [2, 3, 4]. This becomes especially impactful as pregnancies affected by obesity are at increased risk for anomalies, specifically central nervous system, cardiovascular, facial, and limb abnormalities [5]. Prior evidence has suggested the difference in completion of anatomic survey differs from 72% in patients with a body mass index (BMI) less than 25 kg/m² to 50% in those with obesity [4]. The data remain limited on how this trend varies within the strata of obesity and there is a paucity of data on patients with BMI ≥ 50 kg/m². As a tertiary care center, we receive referrals to perform anatomic surveys on patients with BMI ≥ 50 kg/m² based on that indication alone, which suggests there is a perception that sonography in this population requires a different level of expertise, though there is limited evidence on anatomic survey completion in this patient population. Our study sought to further describe the rate of anatomic survey completion in patients with BMI ≥ 50 kg/m².

2. METHODS

This study was a retrospective review evaluating all singleton pregnancies presenting for fetal anatomic survey from January 2019 to December 2020 at a single tertiary care institution from 18 0/7 to 22 6/7 weeks gestation. Institutional Review Board approval was obtained through the Medical University of South Carolina (IRB # Pro00111469). Patients undergoing amniocentesis were excluded. Standard anatomic survey examinations were performed in accordance with the American Institute of Ultrasound in Medicine (AIUM) guidelines. The primary outcome was the rate of anatomic survey completion stratified by BMI category. BMI at the first prenatal visit was recorded by chart extraction. Weight classes were defined by the World Health Organization, and stratified as normal weight (BMI 18.5–24.9 kg/m²), overweight (BMI 25.0–29.9 kg/m²), class I obesity (BMI 30.0–34.9 kg/m²), class II obesity (BMI 35.0–39.9 kg/m²), class III obesity (BMI 40.0–49.9 kg/m²) [6]. In our study, we employed the term severe obesity to identify patients with a BMI ≥ 50.0 kg/m². Secondary outcomes included completed visualization of specific organ systems, scan duration, and number of images acquired.

At our institution, each individual structure is given one of four labels: normal (AIUM imaging guidelines), visualized (structure visualized and without gross abnormality, but not to the AIUM standard), suboptimal (unable to adequately visualize the structure to determine if normal or abnormal), or not visualized [7]. Structures were considered incompletely assessed if marked as “suboptimal” or “not‐visualized.” Image acquisition was performed by registered diagnostic medical sonographers on GE Voluson E10 and interpreted centrally by faculty trained in Maternal Fetal Medicine, utilizing Viewpoint‐6 software. All sites are accredited by AIUM.

At our institution, completion of anatomic survey is first attempted with utilization of the highest frequency RAB6‐d probe; if additional penetration is required, a lower frequency C 1‐6‐d probe is employed. The RIC5‐9‐RS transvaginal probe is used as needed transvaginally or in the maternal umbilicus to image structures in the mid uterus. Each probe has a setting for increased penetration to use for these exams, for example: low frequency, XBeam CRI 3, SRI 4, Line filter low, Line density normal, Frame Filter 3, OTI normal, Enhance 2, Reject 15. Additionally, with significant central adiposity, it is common for two sonographers to partner to complete visualization so that one can assist in positioning the panniculus for optimization and/or in the setting of fatigue. Our institution does not standardize the clinical scenarios in which to enroll these practices. When anatomic survey is incomplete, the survey is reattempted 4 weeks after the initial survey, and then again 4 weeks later if needed. Three‐dimensional imaging is not routinely acquired at the time of mid‐trimester anatomic survey at our institution.

Six major organ systems were evaluated, including intracranial, spine, face, heart, abdominal/genitourinary (GU), and extremities. Intracranial components included lateral ventricles, choroid plexus, midline falx, cavum septum pellucidum, cerebellum, and cisterna magna. Spinal components included cervical, thoracic, lumbar, and sacral spine. Facial structures included nose and lips. Cardiac components included the following views: four‐chamber, right ventricular outflow tract (RVOT), left ventricular outflow tract (LVOT), three vessel (3VV), and three vessel trachea view (3VT). Abdominal/GU structures include cord insertion, stomach, kidneys, bladder, and genitals. Extremities included arms and legs. Our institution evaluates anatomic structure beyond what is listed above, but for the study purposes, anatomic structures were limited to those outlined above. Scan duration (minutes) and number of images acquired was also obtained from each anatomic survey. Duration of imaging was obtained by noting the time of the first image to time of the last image acquired in Viewpoint software.

Statistical analysis was performed via Stata Statistical Software, Release 16 (StataCorp LLC). Categorical variables were analyzed via chi‐square test. Continuous variables were analyzed using Kruskal–Wallis test, as data were non‐parametric. Logistic regression was used to generate an adjusted odds ratio (aOR) for anatomic survey completion compared to the normal BMI referent group, adjusting for covariates including fetal position, gestational age at anatomic survey. A p value of < 0.05 was considered statistically significant for all analyses.

3. RESULTS

Of 2372 anatomic surveys performed during the study period, 213 (9.0%) of patients were normal weight, 748 (31.5%) were overweight, 1297 (54.7%) were obese (631 [26.6%] class I, 377 [15.9%] class II, 289 [12.2%] class III), and 114 (4.8%) had severe obesity (Figure 1). The median BMI with interquartile range (IQR) for each class were as follows: normal 23.8 (IQR 22.7–24.5), overweight 27.8 (IQR 26.6–28.9), class I obesity 32.4 (31.3–33.6), class II obesity 37.3 (IQR 36.0–38.6), class III obesity 43.1 (IQR 41.3–46.0), and severe obesity 54.8 (IQR 52.3–58.7). Of patients with severe obesity, the overall range was from 50.0 to 101.26 kg/m², with many patients (n = 31) with BMI greater than 60 kg/m². There was no difference in gestational age at anatomic survey between BMI classes (p = 0.57). The average gestational age at detailed anatomic survey ranged from 19.1 to 19.3 weeks' gestation for all BMI classes (p = 0.57).

FIGURE 1.

BMI categories of study population. BMI, body mass index.

Table 1 depicts anatomic survey completion on initial attempt by BMI category. The overall rate of anatomic survey completion at initial anatomic survey attempt was 53%. Anatomic survey completion rates significantly differed between groups, ranging from 69.8% to 4.4% with severe obesity. Image acquisition for total anatomic visualization progressively decreased with increasing BMI (p < 0.001), Table 1.

TABLE 1.

Completion of anatomy by BMI category and anatomic systems.

	Normal (BMI 18.5–24.9) n = 213	Overweight (BMI 25.0–29.9) n = 748	Class I obesity (BMI 30.0–34.9) n = 631	Class II obesity (BMI 35.0–39.9) n = 377	Class III obesity (BMI 40.0–49.9) n = 289	Severe obesity (BMI ≥ 50) n = 114	Total n = 2372	p value
Total	144 (67.6)	522 (69.8)	368 (58.3)	174 (46.2)	60 (20.1)	5 (4.39)	1273 (53.6)	< 0.001
Extremities	212 (99.5)	745 (99.6)	629 (99.6)	376 (99.7)	285 (98.6)	105 (92.1)	2352 (99.1)	< 0.001
Spine	186 (87.3)	675 (90.2)	551 (87.3)	290 (76.9)	192 (66.4)	55 (48.3)	1949 (82.1)	< 0.001
GU/Abdomen	203 (95.3)	714 (95.5)	606 (96.0)	354 (93.9)	249 (86.2)	84 (73.7)	2210 (93.2)	< 0.001
Cardiac	160 (75.1)	577 (77.1)	410 (64.9)	207 (54.9)	82 (28.4)	10 (8.7)	1446 (60.9)	< 0.001
Intracranial	209 (98.1)	725 (96.9)	613 (97.2)	361 (95.8)	262 (90.7)	81 (71.1)	2251 (94.9)	< 0.001
Face	201 (94.4)	701 (93.7)	582 (92.2)	310 (82.2)	206 (71.2)	51 (44.7)	2051 (86.5)	<0.001

Note: Data are n (%).

Abbreviation: BMI, body mass index in kg/m².

Completion rates of specific organ systems in the first attempt by BMI class are also shown in Table 1 and Figure 2. There was a significant decrease in adequate visualization of all organ systems assessed with increasing BMI. Cardiac and spinal structures were least likely to be visualized overall, with completion on first attempt in 60.9% and 82.1% of the study population, respectively. When stratified by BMI class, there was a significant decrease in the rate of completion of cardiac structures with increasing BMI class: normal 75.1%, overweight 77.1%, class I obesity 64.9%, class II obesity 54.9%, class III obesity 28.4%, and severe obesity 8.7% (p value <0.001). When divided by BMI class, adequate visualization of the spine differed between groups, ranging from 87.3% with a normal BMI to 48.3% for severe obesity (p value < 0.001).

FIGURE 2.

Anatomic survey completion rates by BMI category. BMI, body mass index; GU, genitourinary.

Logistic regression analysis is presented in Table 2. There are significantly reduced odds of anatomic survey completion in all classes of obesity, including severe obesity with an aOR 0.02, 95% CI 0.01–0.05, compared to normal BMI on initial anatomic survey. For each organ system, odds of completion for severe obesity compared to normal BMI is as follows: extremities (aOR 0.05, 95% CI 0.01–0.40), spine (aOR 0.13, 95% CI 0.07–0.22), genitourinary (aOR 0.14, 95% CI 0.06–0.30), cardiac (aOR 0.03, 95% CI 0.01–0.06), intracranial (aOR 0.05, 95% CI 0.02–0.14), face (aOR 0.04, 95% CI 0.02–0.09).

TABLE 2.

Odds of anatomic survey completion by BMI category.

	Overweight	Class I obesity	Class II obesity	Class III obesity	Severe obesity
	(BMI 25.0–29.9)	(BMI 30.0–34.9)	(BMI 35.0–39.9)	(BMI 40.0–49.9)	(BMI ≥ 50)
	aOR (95% CI)	aOR (95% CI)	aOR (95% CI)	aOR (95% CI)	aOR (95% CI)
Total	1.16 (0.83–1.62)	0.70 (0.50–0.98)	0.42 (0.30–0.61)	0.13 (0.08–0.19)	0.02 (0.01–0.05)
Extremities	1.01 (0.10–9.8)	1.26 (0.11–14.10)	1.67 (0.10–27.06)	0.29 (0.03–2.63)	0.05 (0.01–0.40)
Spine	1.38 (0.86–2.21)	1.02 (0.64–1.62)	0.50 (0.31–0.81)	0.30 (0.18–0.48)	0.13 (0.07–0.22)
GU	1.04 (0.50–2.14)	1.17 (0.55–2.49)	0.79 (0.37–1.70)	0.30 (0.15–0.62)	0.14 (0.06–0.30)
Cardiac	1.17 (0.82–1.68)	0.64 (0.45–0.92)	0.41 (0.28–0.60)	0.13 (0.09–0.20)	0.03 (0.01–0.06)
Intracranial	0.65 (0.22–1.91)	0.71 (0.24–2.15)	0.43 (0.14–1.32)	0.19 (0.07–0.55)	0.05 (0.02–0.14)
Nose/Lips	0.92 (0.48–1.78)	0.75 (0.39–1.45)	0.28 (0.15–0.54)	0.15 (0.08–0.29)	0.04 (0.02–0.09)

Note: Referent group—normal BMI. aOR adjusted for fetal position, gestational age at anatomic survey.

Abbreviations: aOR, adjusted odds ratio; BMI, body mass index; CI, Confidence interval.

Ultrasound duration and number of images are presented in Table 3. The median scan duration for the total cohort was 34.44 min (IQR 26.87–50.45). The median exam duration significantly decreased with increasing BMI category (p < 0.001). The median duration ranged from 33.48 min (IQR 26.33–44.82) for normal weight to 28.38 min (IQR 24.32–34.50) for severe obesity, p < 0.001. The median number of images acquired significantly decreased with increasing BMI category (p < 0.001) and ranged from 86 (IQR 75–101) for normal weight to 63 (IQR 54–73) for severe obesity.

TABLE 3.

Scan duration and number of images.

	N	Scan duration, minutes	Number of images acquired
Normal (BMI 18.5–24.9)	213	33.48 (26.33–44.82)	86 (75–101)
Overweight (BMI 25.0–29.9)	748	32.48 (26.47–43.87)	84 (74–97)
Class I obesity (BMI 30.0–34.9)	631	33.00 (26.07–43.45)	79 (69–90)
Class II obesity (BMI 35.0–39.9)	377	34.72 (27.02–44.15)	76 (67–87)
Class III obesity (BMI 40.0–49.9)	289	32.25 (25.67–40.40)	70 (60–83)
Severe obesity (BMI ≥ 50)	114	28.38 (24.32–34.50)	63 (54–73)
p value		<0.001	<0.001

Note: All data are median (IQR).

Abbreviations: BMI, body mass index; IQR, interquartile range.

4. DISCUSSION

Our study demonstrates that patients with severe obesity (BMI ≥ 50.0 kg/m²) have anatomic survey completion rates of approximately 5% at initial attempt, and significantly reduced odds of complete visualization compared to normal weight counterparts. When evaluating specific organ systems, visualization of all systems was significantly affected by increasing BMI. Cardiac and central nervous system structures are greatly affected by increasing BMI, which is of particular concern given that anomalies in these systems have a known association with obesity [5].

While evidence exists that ultrasound detection falls with increasing BMI, it has not been well established whether this trend eventually plateaus. Our study found that the rate of anatomic survey completion on initial attempt continues to decline as BMI rises within the strata of class III obesity. We detected similar rates of completion for normal BMI compared to other studies; however, prior studies have had relatively small numbers of patients with BMI ≥50 kg/m² [4]. This is a particularly important patient population to assess, as obesity rates continue to rise in the United States, with recent epidemiological studies finding that in the United States, 42% of adults have obesity and 9% have a BMI ≥40 kg/m² [8].

In pregnancies complicated by obesity, there is an increased risk of fetal anomalies, particularly neural tube defects (NTD) and cardiac abnormalities [9]. One systematic review found significantly increased odds of NTD for patients with obesity to be 1.87 [5]. Another meta‐analysis evaluating the prevalence of NTD found a linear relationship between NTD and increasing BMI, with OR reported as 1.22 for overweight, 1.70 for obesity, and 3.11 for patients with severe obesity compared to normal BMI [10].

When focusing on cardiac anomalies, there is an increasing prevalence of congenital heart defects when comparing normal BMI to obesity, even when controlled for pregestational diabetes (OR 1.39) [11]. The trend was most notable for increased risk of specific anomalies such as transposition of great arteries, tetralogy of Fallot, hypoplastic left heart, and right ventricular outflow tract obstruction [11]. While some cardiac anomalies can be identified with a standard four‐chamber view of the heart, additional views are often required for accurate prenatal diagnosis, further highlighting the importance of a thorough and completed anatomic survey for patients with obesity. Additionally, BMI has a linear relationship with pregestational diabetes [12]. Patients with pregestational diabetes have significantly increased rates of congenital anomalies, emphasizing the importance of ultrasound detection in these cohorts [13, 14, 15].

Prior literature suggests that completion of the anatomic survey occurs later in gestation for patients with obesity. One study found the average gestational age of completion was 20.6 weeks for normal weight and 23.9 weeks for BMI ≥ 40 kg/m² [16]. Dashe et al. highlighted the risk of a missed postnatal diagnosis, and found that patients with obesity had at least a 20% difference in antenatal detection of fetal anomalies compared to normal BMI [17]. A sequela of suboptimal visualization at the first anatomic survey attempt is that patients with obesity are more likely to experience delayed diagnoses/completion of anatomy into viable gestational ages.

This delay has implications on reproductive autonomy in the era of increasingly restrictive abortion policies; literature suggests that termination for fetal anomalies occurs at later gestations compared to abortion for other indications [18, 19]. Late diagnosis is limiting, as it can occur at gestational ages where termination is limited by geography, gestational age, legal definitions of lethality, and financial barriers.

The results of this study also have implications for resource utilization. Not only did Gupta et al. identify a later gestational age for anatomic survey completion in higher obesity classes, but the number of ultrasounds required to complete the anatomic survey increases with advancing BMI [16]. Although our study did not evaluate this specific metric, individual ultrasound units can assess this to direct resource utilization such as staffing, use of transvaginal ultrasound for additional views, and scheduling.

In addition to the diagnostic challenges of patients with obesity, prenatal ultrasound examinations have implications for work‐related musculoskeletal injury. Up to 90% of sonographers report scanning in pain [20]. In one study, sonographers were asked to rate factors that contribute to pain; 60.8% reported applying transducer pressure as a very aggravating factor [20]. Prolonged scanning duration and lack of breaks to allow muscle recovery have also been cited as contributors to repetitive strain injury [21].

The trend of scan duration for our facility did not follow relationships published in prior studies [2]. Although there was a statistically significant difference in duration of scan, the median difference between normal weight and BMI ≥ 50 kg/m² was a median of only 5 min (Table 3). This may highlight a perception of futility for anatomic survey completion in patients with severe obesity, as well as the potential for implicit bias. Focus on muscle strain and repetitive workplace injury are highlighted in our ultrasound unit, and sonographers are encouraged to acknowledge when strain becomes a factor. In addition, patients with obesity are instructed to follow up for serial growth in our institution where anatomic completion can be attempted again; these factors may contribute to the trends identified in our study. A confounding factor at our institution is the integration of sonography students, which increases scan duration but cannot be accounted for in the analysis. Students are rarely assigned to participate in examinations on patients with high BMI due to the technically challenging nature of these scans.

Strengths of this study include the size of the population and the number of patients with class III obesity, specifically BMI ≥ 50 kg/m² (n = 114), and of those, 31 patients had a BMI ≥ 60 kg/m². Because our institution images patients with BMI ≥ 50 kg/m² with relative frequency, we are positioned to describe the limitations with current technology and need for further innovation. Additionally, these findings provide resources for institutions who encounter this patient population less frequently. Our study is limited by its retrospective nature at a single facility. Another limitation is that some anatomic surveys included transvaginal ultrasound for visualization of the cervix or anatomic survey completion. The use of transvaginal imaging was not documented in a way that could be efficiently quantified in this cohort. Moreover, the number of postnatal anomalies that were not identified antenatally is unknown within this cohort. As a referral center for prenatal sonography, patients often undergo their ultrasounds in our facility and then return to prenatal care with their primary OBGYNs. Thus, information on postnatal diagnosis would be significantly incomplete.

These findings call for innovation and standardization in prenatal imaging for patients with increasing BMI. Several studies have demonstrated that first trimester ultrasound (with use of transvaginal ultrasound as an adjunct) increases the success of anatomic survey completions in patients with obesity [22, 23]. At this time, there is no standardized recommendation for ultrasound practices (specifically, timing, use of transvaginal, probe settings) in pregnancies complicated by obesity from ultrasound or obstetrics governing bodies; this may be necessary given increasing rates of obesity in the US. There is also a need for innovation; a recent study randomized patients with class III obesity to use of a panniculus retractor for anatomy ultrasound [24]. This study did not find a statistically significant improvement in completion of anatomical survey components with or without the panniculus retractor [24]. This suggests that mechanical adjustments may be only a single consideration when addressing approaches to improving visualization, and advanced imaging technology may be required. A trial was recently published that demonstrated the improved visualization for all BMI classes with high performance probes in non‐obstetric abdominal sonography; this is may represent a promising research angle for its potential application to obstetric ultrasound [25].

In summary, our study identified that patients with BMI ≥ 50 kg/m² have extremely low rates of fetal anatomic survey completion, with visualization of cardiac and central nervous systems most affected. These findings provide justification to prompt further research methods and innovation with a goal to improve fetal anomaly detection in pregnancies complicated by obesity.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.