FormalPara What does this study add to the clinical work

Our research indicates a link between incorrect diagnoses of fetal growth restriction (FGR) and adverse long-term neurological outcomes. Improving the accuracy in diagnosing FGR could potentially lessen related complications.

Introduction

Fetal growth restriction (FGR) is a prevalent obstetric condition wherein the fetus fails to achieve its growth potential due to an underlying pathological factor [1]. FGR is typically defined as an estimated fetal weight (EFW) below the 10th percentile for gestational age [2], occasionally accompanied by abnormal Doppler blood flow patterns [3].

FGR is associated with an increased risk of short-term adverse perinatal outcomes such as hypoglycemia and prematurity [4], in addition to long-term morbidities such as neurodevelopment abnormalities [5], respiratory [6], and cardiovascular morbidity [7]. Key factors contributing to FGR include maternal conditions, placental abnormalities, and genetic factors [8]. However, the diagnosis of FGR is far from straightforward and continues to raise challenges. The complexity of accurately identifying true cases of FGR is compounded by the prevalence of false positives, where healthy fetuses are misclassified as growth-restricted. Serial ultra-sonograms are important for the evaluation of FGR, and umbilical artery Doppler velocimetry is used to guide pregnancy management decisions [9].

Unfortunately, the current most common clinical mentioned definition for FGR is inaccurate and may lead at times to falsely diagnosed fetuses that are eventually born appropriate for gestational age (AGA). The phenomenon of false FGR diagnosis seems more common than generally perceived. Monier et al. found that approximately 50% of the neonates diagnosed with FGR in the antepartum period, were eventually born AGA, i.e., above the 10th weight percentile [10]. This is concordant with a previous study, which also documented low detection rates for FGR [11].

Accurate diagnosis of FGR is crucial because it helps to take informed decisions about patient and pregnancy care. False FGR diagnoses can lead to unnecessary interventions such as labor inductions, cesarean deliveries (CD), or premature birth, which can pose additional risks to both the pregnant patients and the fetus. Moreover, the psychological distress caused by a false FGR diagnosis can be substantial, causing anxiety and fear for the expectant individuals.

Previous studies have associated a false FGR diagnosis with short-term adverse perinatal outcomes [12]. This study seeks to investigate the long-term neurological outcomes of offspring in cases where pregnancies were initially misdiagnosed with FGR but ultimately resulted in the delivery of appropriately grown neonates.

Materials and methods

A population-based cohort study was performed comparing the risk for long-term neurological diagnoses according to whether the newborns were falsely diagnosed with FGR or not. This cohort analysis included all singleton deliveries with a birth weight appropriate for gestational age (AGA) which occurred during the years of 1991 to 2020. Missing gestational age, fetuses with congenital malformations, and cases of perinatal deaths were excluded.

The study was conducted at the Soroka University Medical Center (SUMC). This center is the only hospital in southern Israel, serving the entire population of the region totaling over one million inhabitants, thus the study is based on nonselective population data. The study is approved by the institutional review board (in accordance with the Helsinki Declaration). Neurological disorders in accordance with International Classification of Diseases, Ninth Revision (ICD-9) were compared among AGA offspring with and without a false diagnosis of FGR. Follow-up was conducted up to the age of 18 years. The relevant neurological diagnoses were given based on an evaluation performed on offspring’s admission to the hospital or based on the patient’s history. It may have been the main reason for admission or a background disorder. For diagnosing FGR, we used Dollberg and colleagues’ birth weight growth curves. These curves are a commonly employed, population-based standard for evaluating weight percentiles based on the birth weights of live-born infants in Israel [13]. Fetal weights were estimated using sonographic measurements. After birth, we applied these same growth curves to classify the neonate’s weight. When necessary, the timing of delivery was determined based on several factors: gestational age, the severity of suspected FGR, fetal heart rate monitoring, the biophysical profile, and Doppler velocimetry results.

Data were collected from the computerized hospitalization database of SUMC (“Demog-ICD9”) and the computerized perinatal database of the obstetrics and gynecology department. The Demog-ICD9 database includes demographic information and ICD-9 codes for all medical diagnoses made during encounters with SUMC. In order to ensure maximal integrity and accuracy, experienced medical secretaries routinely review the information before entering it into the database. Coding is performed after assessing medical and perinatal records as well as routine hospital documents.

Statistical analysis was performed using SPSS (version 24) and STATA (version 12.0) software. Differences between the groups were assessed using χ2 test, t test, in accordance with the variable type and its distribution. Kaplan–Meier survival curves were used to compare cumulative hospitalization incidences over time, and the differences were analyzed using the log-rank test. To establish an association with cumulative hospitalization incidence, while controlling for potential confounders, we used a multivariate Cox proportional hazards model to account for repeated occurrence of mothers and the dependence among siblings in the cohort. P-values < 0.05 were considered statistically significant.

Results

During the study period, 324,620 AGA infants met the inclusion criteria. 3249 of them (1%) were falsely classified as FGR in the antepartum period. Table 1 demonstrates the maternal and obstetrical outcomes, according to false FGR diagnosis, or not. Pregnancies with false FGR diagnosis demonstrated higher rates of hypertensive disorders (14.7% vs. 4.3% p < 0.001), preterm delivery (36.4% vs. 6.6%, p < 0.001), and CD (33.9% vs. 13.0%, p < 0.001). The fetal weights in the false FGR group followed a normal distribution but were consistently lower than those of the non-false FGR newborns across all percentiles. The mean gestational age for the group falsely diagnosed with FGR was 36.7 ± 2.47.

Table 1 Maternal and obstetrical outcomes, according to false FGR diagnosis
Full size table

These infants exhibited a higher rate of long-term neurological morbidities as compared to those without a false FGR diagnosis (OR 1.431, 95% CI 1.278–1.608; p < 0.001, Table 2). Specifically, movement disorders (4.2% vs. 2.9%, p < 0.01), cerebral palsy (0.4% vs. 0.2%, p < 0.002), and developmental disorders (2.2% vs. 0.8%, p < 0.001) were each significantly more common in the false FGR diagnosis group.

Table 2 Long term neurological outcomes of offspring according to false diagnosis of FGR
Full size table

The cumulative incidence of long-term neurological morbidity was significantly higher among false FGR diagnosis (Kaplan–Meier survival curve Log rank p < 0.001, Fig. 1). In the Cox proportional hazard regression model, an independent association was found between false FGR diagnosis and long-term neurological morbidity, after adjusting for gestational age, birth year (since our study included a long range of years), CD and hypertensive disorders as possible confounders (adjusted HR 1.086, 95% CI 1.003–1.177, P = 0.043).

Fig. 1
figure 1

Cumulative incidence of hospitalization with neurological morbidity in offspring according to whether falsely classified as FGR or not (Log Rank P value < 0.001)

Full size image

Discussion

Based on the findings of our study, it has been observed that fetuses suspected to be growth-restricted who subsequently had a normal birth weight were at an increased risk of experiencing long-term neurological morbidity during childhood. This includes a heightened likelihood of conditions such as movement disorders, developmental impairments, and cerebral palsy.

Prior research on this subject has shown that false diagnoses of FGR have been linked to higher rates of early-term deliveries, preterm CD, and increased risk for short-term adverse neonatal outcomes. Consistent with our findings, a study assessing the consequences of misdiagnosed FGR on the mode of delivery indicated a tenfold increase in the rates of preterm CD (12.7% versus 1.2%) [14]. Gabbay-Benziv et al., additionally found that these offspring were at increased risk for transient tachypnea of the newborn, mechanical ventilation, hypoglycemia, and NICU admissions [12].

Our study specifically focuses on the less-explored area of long-term neurological impacts resulting from a misdiagnosed FGR. Analyzing our results, we suggest a possible explanation for the difference in the observed outcomes. The elevated prevalence of long-term neurological complications may be connected to iatrogenic early deliveries, as physicians might be inclined to induce preterm or early-term labor and perform CD in cases of suspected FGR, potentially overlooking the iatrogenic risks [10]. Preterm birth is an established risk factor associated with long-term adverse neurological outcomes in offspring. The immature brain is susceptible to both the primary brain damage (such as white matter injury, germinal intraventricular hemorrhage, and cerebellar hemorrhage) and the secondary consequences of incomplete maturation [15]. Furthermore, there is also evidence to suggest that even early-term deliveries are associated with long-term adverse outcomes [16,17,18,19] specifically neurodevelopmental impairment compared with full-term birth [20]. In a study conducted by Baumfeld et al., it was demonstrated that CD, in contrast to vaginal delivery, emerged as an independent risk factor for pediatric neurological hospitalizations [21]. Notably, the results in our study retained their significance even after accounting for confounders such as gestational age at birth, hypertensive disorders, CD, and birth year. A possible reason for the increased morbidity is that even though these fetuses reached the “AGA threshold,” they might have still experienced restricted growth in utero. These infants could potentially have a similarly unfavorable prognosis as those with lower birth weights [22].

This interpretation aligns with our observation that the fetal birth weights in the false FGR group were consistently smaller than those in the non-false FGR group across all percentiles, despite a normal distribution. However, this difference might also be attributed to an earlier gestational age, considering that all misclassified cases in the group were born AGA and that gestational age was accounted for in the multivariable analysis.

Our study has several limitations. First, it only includes diagnoses of hospitalized children, potentially excluding those with mild-to-moderate neurological issues not requiring hospitalization. A potential solution is a follow-up study integrating community medical records with hospital data. Secondly, retrospective studies may have coding errors, despite skilled review. Thirdly, immigration and potential loss to follow-up are considerations, but their effects likely apply similarly to both groups.

Our study’s major strength stems from the fact that SUMC is the sole tertiary hospital in the southern region of Israel. That, combined with free essential health insurance provided to each citizen of Israel, makes it safe to assume that if a woman gave birth to a child in SUMC, the child would reach SUMC when in need of major medical assistance.

In conclusion, our study shows an independent association between misdiagnosed FGR and an elevated risk of long-term neurological health issues in offspring. It is possible that these increased neurological morbidities may be a consequence of iatrogenic deliveries at a lower gestational age and prompt the question of the adequacy of ultrasound measurements for FGR detection. Additional research may help develop a more effective method for predicting FGR, particularly in the third trimester, through wider application of an individualized growth curve and the implementation of routine third-trimester screening for FGR [23, 24].