Prediction model for spontaneous preterm birth less than 32 weeks of gestation in low-risk women with mid-trimester short cervical length: a retrospective cohort study

doi:10.21203/rs.3.rs-4836121/v1

Download PDF

Research Article

Prediction model for spontaneous preterm birth less than 32 weeks of gestation in low-risk women with mid-trimester short cervical length: a retrospective cohort study

https://doi.org/10.21203/rs.3.rs-4836121/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 01 Oct, 2024

Read the published version in BMC Pregnancy and Childbirth →

You are reading this latest preprint version

Background

A short cervix in mid-trimester pregnancy is a risk factor for spontaneous preterm birth. However, there is currently a lack of predictive models and classification systems for predicting spontaneous preterm birth in these patients, especially those without additional risk factors for spontaneous preterm birth.

Methods

A retrospective observational cohort study of low-risk singleton pregnant women with a short cervix (≤ 25 mm) measured by mid-trimester fetal ultrasound scan between 22 and 24 weeks was conducted. A multivariate logistic regression model for spontaneous preterm birth < 32 weeks in low-risk pregnant women with a short cervix was constructed. Moreover, we developed a nomogram to visualize the prediction model and stratified patients into three risk groups (low-, intermediate-, and high-risk groups) based on the total score obtained from the nomogram model.

Results

Between 2020 and 2022, 213 low-risk women with a short cervix in mid-trimester pregnancy were enrolled in the study. Univariate logistic analysis revealed that a high body mass index, a history of three or more miscarriages, multiparity, a short cervical length, leukocytosis, and an elevated C-reactive protein level were associated with spontaneous preterm birth < 32 weeks, but multivariate analysis revealed that multiparity (OR, 3.31; 95% CI, 1.13–9.68), leukocytosis (OR, 3.96; 95% CI, 1.24–12.61) and a short cervical length (OR, 0.88; 95% CI, 0.82–0.94) were independent predictors of sPTB < 32 weeks. The model incorporating these three predictors displayed good discrimination and calibration, and the area under the ROC curve of this model was as high as 0.815 (95% CI, 0.700-0.931). Patients were stratified into low- (195 patients), intermediate- (14 patients) and high-risk (4 patients) groups according to the model, corresponding to patients with scores ≤ 120, 121–146, and > 146, respectively. The predicted probabilities of spontaneous preterm birth < 32 weeks for these groups were 6.38, 40.62, and 71.88%, respectively.

Conclusions

A noninvasive and efficient model to predict the occurrence of spontaneous preterm birth < 32 weeks in low-risk singleton pregnant women with a short cervix and a classification system were constructed in this study and can provide insight into the optimal management strategy for patients with different risk stratifications according to the score chart.

Short cervix

low-risk

spontaneous preterm birth

predictive model

classification system

Preterm birth (PTB) less than 32 weeks of gestation (wks) accounts for approximately 15% of the 15 million births annually worldwide and is associated with approximately 52% of infant deaths, 4% of cerebral palsy, and 8% of neurological developmental delays[1, 2]. Although the rate of PTB in the United States declined from 10.4% in 2007 to 10.09% in 2020, the distribution of PTB < 32 wks did not change[3, 4], which might be one reason why infant deaths decrease were slow. Therefore, the prediction and prevention of PTB < 32 wks is essential for reducing healthcare costs and improving perinatal outcomes.

The causes of PTB include maternal or fetal complications, chorioamnionitis, preterm premature rupture of membranes (PPROM), and cervical insufficiency[5]. In addition to medical indications, cervical insufficiency is one of most common risk factors for spontaneous preterm birth (sPTB). At present, the definitions of cervical insufficiency are not consistent, and cervical insufficiency is usually associated with painless, progressive dilatation of the uterine cervix in mid-trimester pregnancy, which consequently arises from the inability to support pregnancy. A classic obstetric history, such as recurrent mid-trimester pregnancy losses, sPTB or PPROM at < 32 wks, raises the suspicion of cervical insufficiency[6]. Based on a classic history, cervical cerclage can be recommended for these women, as it is beneficial for reducing the risk of sPTB. However, it is extremely difficult to identify how patients without these classic histories can be diagnosed with cervical insufficiency and further provide optimal management.

A short cervix is defined as a transvaginal sonographic cervical length (CL) ≤ 25 mm in mid-trimester pregnancy and is a risk factor for sPTB[7, 8]. Hence, for singleton pregnant women with a short cervix combined with a classic obstetric history, cervical cerclage is strongly indicated[9]. In our previous study, we found that cervical cerclage could significantly improve neonatal outcomes for singleton pregnant women with a short cervix and a history of sPTB; however, conservative treatment was more suitable for those without a history of sPTB[10]. Interestingly, approximately 80% of the asymptomatic women without high-risk factors for sPTB who were occasionally diagnosed with a short cervix in mid-trimester pregnancy did not deliver prematurely; on the other hand, a few women had cervical insufficiency and eventually delivered prematurely[11]. The management and treatment of a short cervix in these low-risk individuals remains an unsolved clinical challenge.

The existing models for predicting sPTB include maternal clinical characteristics, serum biomarkers, and sonographic examination results[7, 12, 13]. However, none of these models have shown enough accuracy as a noninvasive and efficient tool for predicting sPTB, especially for patients who deliver < 32 wks. More importantly, none of these tools involve individualized stratification of sPTB < 32 wks in low-risk women with a short cervix, which may be advantageous in clinical management and early decision-making.

Hence, the aim of the present study was 2-fold: (1) to evaluate clinical, laboratory and sonographic data to assess their potential predictive value and (2) to construct a nomogram model and classification system to predict the individual probability of sPTB < 32 wks for low-risk women with a short cervix in mid-trimester pregnancy.

Patients

This retrospective observational cohort study was performed from January 2020 to December 2022 at Women’s Hospital, School of Medicine, Zhejiang University, China. In our hospital, all pregnant women undergo mid-trimester fetal ultrasound scan between 22 and 24 weeks of gestation, and the transabdominal cervical length (CL) is simultaneously measured. Among women with CL ≤ 25 mm measured by transabdominal ultrasonography, a diagnosis with a short cervix is further identified by transvaginal ultrasonography. Singleton pregnant women with a short cervix without a history of recurrent mid-trimester pregnancy losses, sPTB or PPROM at < 32 wks were considered low-risk individuals and were enrolled in the study. The exclusion criteria included multiple pregnancy, PPROM, uterine malformation, uterine scar, the presence of fetal anomalies, intrauterine fetal demise, and symptomatic women that defined as the presence of symptoms such as persistent contractions or vaginal bleeding. Patients with clinically induced PTB due to conditions such as placenta previa, placental abruption, severe pregnancy-induced hypertension/preeclampsia, clinical symptoms or laboratory findings suggestive of chorioamnionitis were considered ineligible for this study. We further excluded low-risk women who underwent cervical cerclage according to the decision of the attending physician. All patients underwent routine prenatal examinations and ultimately gave birth at our hospital. During this period, 54,702 deliveries occurred in our hospital, and 878 women were diagnosed with a short cervix (1.6%) in mid-trimester pregnancy, including 213 low-risk individuals (0.4%). According to the weeks of delivery, the patients were divided into the sPTB < 32 wks (n = 21) and control groups (n = 192).

Data Collection

The following variables were assessed to identify predictors of sPTB < 32 wks: age, body mass index (BMI) at the time of CL measurement, assisted reproductive technology (ART), a history of ≥ 3 miscarriages, multiparity, smoking, alcohol consumption, medical history, obstetric complications, gestational age at the time of cervical length measurement, CL measurement, maternal blood assay results, sPTB prevention treatments (vaginal progesterone), gestational week at delivery, type of delivery, and neonatal outcomes. CL was assessed by vaginal ultrasonography from the internal os to the external os in the midsagittal plane, with the entire cervical canal visible on a grayscale sonographic image[14]. Additionally, Roman and Hadar separately proposed that cervical changes can be categorized into 5 stages through sonographic and physical examination before 24 wks in women with a short cervix[15]. The women identified as stage 1–3 with closed internal cervical os were included in our study. We excluded patients with a CL < 15 mm and visible dilation with membranes apparent in the cervical canal (stage 4) or membranes protruding into the vagina (stage 5) because these patients usually undergo cervical cerclage in our hospital. In total, 156 (73.2%) patients were classified as Stage 1, 53 (24.9%) as Stage 2, and 4 (1.9%) as Stage 3. Maternal blood analysis was performed to determine the white blood cell (WBC) count (the Natt and Herrick method using a Neubauer hemocytometer) and C-reactive protein (CRP) level (a turbidimetric method using a Beckman Coulter AU5800/AU680 autoanalyzer) between 22 and 24 wks. The kit-recommended standard WBC and CRP reference intervals were used to determine whether leukocytosis (defined as > 10×10⁹/L) or an elevated CRP level (defined as > 5 mg/L) was present. All enrolled subjects with a short cervix were observed or prescribed vaginal progesterone for sPTB prevention according to the decision of the attending physician. Vaginal progesterone therapy included the use of progesterone soft capsules twice daily (in the morning and at night) (each pill contained 100 mg of progesterone; Besins Manufacturing Belgium, France) until > 34 wks.

Statistical analysis

Continuous data are expressed as the mean ± standard deviation (SD) or median (interquartile range), and categorical data are expressed as numbers (proportions). Significant differences in the continuous and categorical variables between the sPTB < 32 wks and control groups were compared by univariate analysis (t-test, χ² test or rank sum test).

A robust and well-calibrated multivariate logistic regression model was established to predict the risk of sPTB < 32 wks using backward elimination. Additionally, we evaluated the performance of the prediction model through its discrimination and calibration abilities. To assess discrimination, a receiver operating characteristic (ROC) curve was drawn, and the area under the ROC curve (AUC) was calculated. With respect to the evaluation of calibration, we utilized a calibration curve to visually compare the consistency between the predicted and observed probabilities, and the Hosmer–Lemeshow test was also employed.

For convenient clinical application, we developed a nomogram to visualize the prediction model. Based on the total score obtained from the nomogram, patients were stratified into three risk groups (low-, intermediate-, and high-risk groups). In addition, we calculated the positive likelihood ratio (LR+), negative likelihood ratio (LR-), and rate of sPTB < 32 wks for each risk group. All analyses were performed using R version 4.0.3 (http://cran.r-project.org/mirrors.html) and SPSS version 26.0 (Armonk, NY, USA). A two-tailed P value < 0.05 was considered to indicate statistical significance.

Clinical characteristics

Among 878 women who were diagnosed with a short cervix between 22 and 24 weeks by transvaginal ultrasonography, 213 were low-risk patients who met the inclusion criteria (see Additional Fig. 1). Among these patients, 42 (19.7%) delivered with sPTB, and the incidence of sPTB < 32 wks was 9.9%. Table 1 summarized the maternal clinical characteristics and perinatal outcomes. Maternal clinical characteristics did not differ between the two groups, except for BMI at the time of short cervix diagnosis, multiparity, and with a history of three or more miscarriages. The results showed that the women with sPTB < 32 wks were usually had a higher BMI at the time of short cervix diagnosis (24.6 ± 4.1 vs. 21.5 ± 3.4 kg/m², P < 0.001), had a history of three or more miscarriages (14.3% vs. 2.6%, P = 0.034), and multiparity (47.6% vs. 21.9%, P = 0.009) than those without sPTB. Leukocytosis (76.2% vs. 46.4%, P = 0.009) and an elevated CRP level (42.9% vs. 19.8%, P = 0.040) were observed in women with sPTB < 32 wks compared to those without sPTB. More importantly, CL measurements revealed that women with sPTB < 32 wks had a significantly shorter CL than women without (10.8 ± 8.7 vs. 18.2 ± 5.6 mm, respectively, P < 0.001). Additionally, the fetuses in the group with sPTB < 32 wks had a significantly lower birth weight, lower Apgar score at 1 minute and 5 minute, and higher rates of neonatal intensive care unit admission than did those in the control group (P < 0.001).

Table 1

Clinical characteristics and perinatal outcomes of women in the sPTB < 32 wks and control groups.
Variables	sPTB < 32 wks		P value
	Yes	No
	N = 21 (9.9%)	N = 192 (90.1%)
Age, years	30.7 ± 4.9	29.7 ± 4.2	0.350
BMI at time of short cervix diagnosis, kg/m²	24.6 ± 4.1	21.5 ± 3.4	< 0.001
ART, No. (%)	0 (0.0)	21 (10.9)	0.226
A history of ≥ 3 miscarriages, No. (%)	3 (14.3)	5 (2.6)	0.034
Multiparity, No. (%)	10 (47.6)	42 (21.9)	0.009
Smoking, No. (%)	1 (4.8)	7 (3.6)	0.570
Alcohol consumption, No. (%)	1 (4.8)	5 (2.6)	0.460
History of cervical surgery, No. (%)	1 (4.8)	5 (2.6)	0.491
History of D&C, No. (%)	13 (61.9)	99 (51.6)	0.104
Hypothyroidism, No. (%)	0 (0.0)	16 (8.3)	0.347
GDM, No. (%)	3 (14.3)	31 (16.1)	0.999
ICP, No. (%)	0 (0.0)	8 (4.2)	0.999
Gestational hypertension or chronic hypertension, No. (%)	2 (9.5)	9 (4.7)	0.666
Gestation age at time of short cervix diagnosis, wk	23.3 ± 0.8	23.3 ± 0.8	0.974
Cervical length, mm	10.8 ± 8.7	18.2 ± 5.6	< 0.001
Leukocytosis at time of short cervix diagnosis, No. (%)	16 (76.2)	89 (46.4)	0.009
Elevated CRP level at time of short cervix diagnosis, No. (%)	9 (42.9)	38 (19.8)	0.040
Vaginal progesterone use, No. (%)	19 (90.5)	170 (88.5)	0.790
Gestational week at delivery, wk	28.8 ± 2.0	38.8 ± 2.0	< 0.001
Type of delivery, No. (%)			0.112
Vaginal delivery	19 (90.5)	141 (73.5)
Cesarean section	2 (9.5)	51 (26.6)
Birth weight, g	1099.5 ± 366.5	3165.8 ± 493.5	< 0.001
Male fetus, No. (%)	10 (47.6)	100 (52.1)	0.819
Apgar score at 1 min, median (IQR)	7 (2–8)	10 (10–10)	< 0.001
Apgar score at 5 min, median (IQR)	9 (5–9)	10 (10–10)	< 0.001
NICU admission rate, No. (%)	20 (95.2)	36 (18.8)	< 0.001
Data are presented as the mean ± standard deviation or median (interquartile range) for continuous variables and as number (%) for categorical variables. Abbreviations: sPTB, spontaneous preterm birth; BMI, body-Mass Index; ART, assisted reproductive technology; D&C, dilation and curettage; GDM, gestational diabetes mellitus; ICP: intrahepatic cholestasis of pregnancy; CRP, C-reactive protein; IQR, interquartile range; NICU, neonatal intensive care unit.

Prediction model of sPTB < 32 wks for low-risk women with a short cervix in mid-trimester pregnancy

The results of the univariate and multivariate logistic regression models are shown in Table 2. Univariate analysis revealed that a high BMI, a history of three or more miscarriages, multiparity, a short CL, leukocytosis and an elevated CRP level were associated with sPTB < 32 wks. Multivariate logistic regression analysis revealed that multiparity (odds ratio [OR], 3.31; 95% confidence interval [CI], 1.13–9.68; P = 0.028), leukocytosis (OR, 3.96; 95% CI, 1.24–12.61; P = 0.020) and a short CL (OR, 0.88; 95% CI, 0.82–0.94; P < 0.001) were independent predictors of sPTB < 32 wks. Furthermore, a nomogram was constructed based on statistically significant predictors (P < 0.05), which can be applied for calculating the probability of sPTB < 32 wks in low-risk women with a short cervix (Fig. 1). The nomogram illustrated that CL had the largest contribution to predicting sPTB < 32 wks, and the cutoff value was 15.5 mm. More importantly, the nomogram using the significant statistical variables, i.e., multiparity, leukocytosis and a short CL, had excellent diagnostic performance, with a sensitivity of 81.0% and a specificity of 76.6% at the optimal cutoff value, and had an AUC as high as 0.815 (95% CI, 0.700-0.931; Fig. 2A). Figure 2B shows the calibration plots of the validation cohorts. When plotting the probabilities of sPTB < 32 wks predicted by the model compared with the actual probabilities, the calibration curve was close to the dotted line.

Table 2

Univariate and multivariate logistic regression models for predicting sPTB at < 32 wks.
Variables	Univariate analysis		Multivariate analysis
Variables	OR (95% CI)	P value	OR (95% CI)	P value
BMI, kg/m2	1.23 (1.01–1.38)	< 0.001
A history of ≥ 3 miscarriages
Yes	6.23 (1.37–8.24)	0.034
No	Referent
Multiparity
Yes	3.25 (1.29–8.17)	0.009	3.31 (1.13–9.68)	0.028
No	Referent		Referent
Cervical length, mm	0.86 (0.81–0.92)	< 0.001	0.88 (0.82–0.94)	< 0.001
Leukocytosis
Yes	3.70 (1.30–0.52)	0.009	3.96 (1.24–12.61)	0.020
No	Referent		Referent
Elevated CRP level
Yes	3.04 (1.19–7.74)	0.040
No	Referent
Data are presented as OR (95% CI). Abbreviations: OR, odds ratio; CI, confidence interval; BMI, body-Mass Index; CRP, C-reactive protein.

Establishing a classification system

As shown in Table 3, scores of three independent variables were assigned according to the established nomogram. We stratified the patients into three risk groups based on the probability predicted by this logistic regression model: the low-risk (≤ 30%), intermediate-risk (30%-50%) and high-risk (> 50%) groups, corresponding to women with scores ≤ 120, 121–146, and > 146, respectively. The model classified 195/213 women (91%) into the low-risk group; the predicted probability of sPTB < 32 wks was 6.38%, and the actual rate was 6.15% (12/195). The model classified 4/213 women (2%) into the high-risk group. In this group, the predicted probability of sPTB < 32 wks was 71.88%, and the actual rate was 75% (3/4). The remaining patients (7%, 14/213) were assigned to the intermediate-risk group. In this group, the predicted probability of sPTB < 32 wk was 40.62%, and the actual rate was 42.86% (6/14). Taken together, our data suggest that women with a short cervix who are at high-risk group could have a greater rate of sPTB < 32 wks, and further evaluation is needed to determine whether cervical cerclage may benefit these women.

Table 3. Risk estimate and stratification of patients into subgroups.
Total points	Number of patients (n = 213)	Number of observed sPTB < 32 wks (n = 21)	Actual rate of sPTB < 32 wks	Predicted probability of sPTB < 32 wks	LR+ (95% CI)	LR- (95% CI)	Risk group	Number of patients (n = 213)	Number of observed sPTB < 32 wks (n = 21)	Actual rate of sPTB < 32 wks	Predicted probability of sPTB < 32 wks
0–80	157	5	3.18%	≤ 10%	1.92 (1.68, 2.16)	0.28 (0.25, 0.32)	Low	195	12	6.15%	6.38%
91–104	31	5	16.13%	10% − 20%	4.51 (3.87, 5.23)	0.47 (0.45, 0.50)
105–120	7	2	28.57%	20% − 30%	8.90 (8.25, 9.59)	0.54 (0.52, 0.57)
121–133	6	2	33.33%	30% − 40%	12.6 (11.02, 14.17)	0.60 (0.59, 0.62)	Intermediate	14	6	42.86%	40.62%
134–146	8	4	50.00%	40% − 50%	13.41 (13.41, 1.79)	0.67 (0.65, 0.70)
147–158	0	0	/	50% − 60%	21.31 (12.8, 13.71)	0.76 (0.68, 0.87)	High	4	3	75.00%	71.88%
159–171	1	1	100.00%	60% − 70%	27.43 (27.43, 7.43)	0.86 (0.86, 0.86)
> 172	3	2	66.67%	> 70%	18.29 (18.29, 8.29)	0.91 (0.91, 0.91)
sPTB, spontaneous preterm birth; LR+, positive likelihood ratio; LR−, negative likelihood ratio; CI, confidence interval.

Our study revealed that 9.9% (21/213) low-risk women with mid-trimester short cervix who delivered < 32 wks received only observation or were prescribed vaginal progesterone treatment, suggesting that building a predictive model is important for sPTB prediction and prevention. Furthermore, we incorporated maternal clinical characteristics, sonographic assessment of cervical length and maternal blood results to construct a noninvasive prediction model and a simple scoring system to predict the individual probability of sPTB < 32 wks for low-risk women with a short cervix in mid-trimester pregnancy. This innovative sPTB prediction model can potentially provide better predictive capacity and management strategies in clinical practice.

In women with singleton pregnancies without a history of sPTB, the incidence of CL ≤ 25 mm was 0.81-2%[16–18]. Our population was assessed using vaginal ultrasonography during the mid-trimester fetal ultrasound scan between 22 and 24 wks, which revealed that the prevalence of CL ≤ 25 mm was 1.6% and 0.4% in the low-risk population. Among the low-risk subjects, 42 delivered with sPTB, and 21 delivered before 32 wks. Hadar et al. reported that the total rate of sPTB before 32 weeks was 9.9% in asymptomatic singleton pregnant women¹¹, but some women enrolled in the study underwent cervical cerclage, which might have led to bias in the actual incidence of sPTB.

A history of sPTB is a strong high-risk factor for sPTB. Accumulating evidence indicates that transvaginal sonographic CL measurement in mid-trimester pregnancy is also a key clinical tool for predicting the occurrence of sPTB. Therefore, the American College of Obstetricians and Gynecologists (ACOG) guideline recommend routine transvaginal CL screening from 16 to 24 weeks for women with singleton pregnancies and a history of sPTB. However, differences exist regarding further screening and management for singleton pregnant women with a short cervix without high-risk factors for sPTB[5, 9, 19–21]. Our data showed that 0.4% of the low-risk subjects had a CL ≤ 25 mm. These individuals underwent observation or received vaginal progesterone treatment, and 9.9% ultimately delivered prematurely < 32 wks; consequently, a predictive model for low-risk women with a short cervix will be constructed to allow for better management and interventions to prevent or delay sPTB.

Our study defined predictors, including multiparity, leukocytosis and a short CL, that are associated with the occurrence of sPTB < 32 wks. A short CL (< 15.5 mm) was the most significant independent predictor for sPTB. In a study of unselected pregnant women between 22 and 24 wks conducted by Heath et al., only 1.7% had a CL < 15 mm, but they accounted for approximately 58% of those with sPTB < 32 wks[22]. Additionally, multiparity, as a predictor, can underscore the complex relationship between parity and sPTB. Our study revealed that low-risk women who are multiparous are at high risk for sPTB < 32 wks. A retrospective cohort study by Koullali et al reported that the likelihood of sPTB occurring in nulliparous women was greater than that in multiparous women in their second pregnancy (OR, 2.15; 95% CI, 1.98–2.3)[23]. This conclusion contradicts what was observed in our study, which may explain the status of the population. Our study focused on low-risk women and might hypothesize that greater parity may increase the risk of sPTB < 32 wks due to a damaged cervix, which is consistent with the finding of Koullali that when analyzing multiparous patients with a parity > 4, the risk of sPTB < 32 wks increased with increasing parity (OR, 1.26; 95% CI, 1.13–1.41)[23]. Multiple studies have corroborated that elevated levels of inflammatory biomarkers are associated with an increased risk of sPTB. Various inflammatory biomarkers, such as matrix metalloproteinases, WBC, and CRP, have been studied in both the amniotic fluid and maternal serum for an association with the occurrence of sPTB[24–26]. Our study supports this reasoning, but the mechanism by which it impacts sPTB needs further exploration.

The bootstrap-corrected concordance index of our model was 0.815 (95% CI, 0.700-0.931), which indicates good predictive power and calibration. Cobo et al. included gestational age, CL, amniotic fluid glucose and interleukin (IL)-6 in a model for 358 women with a diagnosis of sPTB and intact membranes before 34 wks, and the ROC of the model for predicting the risk of sPTB within 7 days was 0.86 (95% CI, 0.77–0.95)[27]. Similarly, the ROC of the model developed by Watson et al. combined quantitative fetal fibronectin and CL in 689 asymptomatic high-risk women to predict sPTB < 34 wks was 0.76[28]. Compared with those of previous models, the main advantages of our model is that it focuses on low-risk pregnant women with a short cervix and excludes patients who undergo cervical cerclage, which could actually reflect the correct occurrence of sPTB. Additionally, our model is noninvasive and can be used clinically without the need for additional tests, which is cost-effective and acceptable.

In our study, risk group stratification using our score chart could clinically and noninvasively help clinicians more efficiently manage low-risk women with a short cervix in mid-trimester pregnancy. Table 3 shows the different clinical scenarios stratified by risk group. A total of 91.5% (195/213) of the patients were included in the low-risk group, and the actual rate of sPTB < 32 wks was 6.15%. Thus, the management of these patients could be reasonably limited to unnecessary interventions with a low-intensity intervention when necessary, such as observation. For the high-risk group, the actual rate of sPTB < 32 wks was 75%, and aggressive interventions could be a better choice to reduce the risk of sPTB and decrease the rate of perinatal deaths[29]. Nevertheless, evidence for the benefit of vaginal pessary use in select high-risk patients is limited. A randomized clinical trial conducted by Hoffman et al. in singleton pregnant women with a CL < 20 mm (n = 544) demonstrated that cervical pessary did not decrease the risk of sPTB and was associated with a higher rate of fetal mortality[19]. Thus, women in the high-risk group might require surgical interventions, such as cervical cerclage. Moreover, in the intermediate-risk group, the actual rate of sPTB < 32 wks was 42.86%. A meta-analysis indicated that in singleton pregnant women with a mid-trimester short cervix (n = 974), vaginal progesterone decreased the risk of sPTB < 32 wks[30]. The management of these intermediate-risk patients could be determined according to patients’ wishes, and the use of vaginal progesterone may be a better choice.

There is currently a lack of standardized management protocols for low-risk women with a short cervix in mid-trimester pregnancy with different individualized risk stratifications. In our study, we built a model for predicting the occurrence of sPTB in these women, further developed a classification system and provided optional management strategies. The validation of the nomogram and classification system remains essential for the scientific community. Additionally, studies exploring the efficacy of interventions in various risk groups based on the classification system are needed. These findings could help clinicians better identify, counsel and treat patients with various risk stratifications.

Our study has several limitations. First, the relatively small number of patients in sPTB < 32 wks group was a limitation of this retrospective cohort study, which may have confounded our data. Second, patients with visible dilation of membranes apparent in the internal cervical os were excluded from our study. This selection bias may have affected the performance of the model. Third, although we performed a rigorous internal validation, the model requires external validation using independent data sets to confirm its generalizability. Finally, one potential limitation in improving the predictive ability of a model is the lack of specific blood biomarkers. Future multicenter studies with larger sample sizes will be required before the model can be routinely used.

In conclusion, we constructed a noninvasive and efficient predictive model to predict the occurrence of sPTB < 32 wks in women with mid-trimester short cervix without additional risk factors for sPTB. The patients were divided into low-, intermediate-, and high-risk groups according to the score chart for risk stratification. Our nomogram and classification system may be useful for targeted interventions for women with a short cervix and without a history of sPTB, as unnecessary interventions should be reasonably limited to low-risk patients, and cervical cerclage might be considered for high-risk patients. Further prospective validation studies using independent populations will be conducted to evaluate the performance of the model.

sPTB: Spontaneous preterm birth

CL: Cervical length

PTB: Preterm birth

Wks: Weeks of gestation

PPROM: Preterm premature rupture of membrane

BMI: Body mass index

ART: Assisted reproductive technology

WBC: White blood cell

CRP: C-reactive protein

SD: Standard deviation

ROC: Receiver operating characteristic

AUC: Area under the ROC curve

LR+: Positive likelihood ratio

LR-: Negative likelihood ratio

OR: Odds ratio

CI: Confidence interval

ACOG: American College of Obstetricians and Gynecologists

D&C: Dilation and curettage

GDM: Gestational diabetes mellitus

ICP: Intrahepatic cholestasis of pregnancy

IQR: Interquartile range

NICU: Neonatal intensive care unit

Ethics approval and consent to participate: This study was approved by the Ethics Committee of Women’s Hospital, School of Medicine, Zhejiang University (IRB-20220338-R).

Consent for publication: All subjects signed a written informed consent and consented for publication.

Availability of data and materials: Data is provided within the manuscript or supplementary information files.

Competing interests: The authors report no conflicts of interest.

Funding: This work was supported by the National Key Research and Development Program of China under Grant 2022YFC2704700 and the Zhejiang Provincial Natural Science Foundation under Grants LY24H040003 and LTGC23H180001.

Author contiributions: LBH and ZYM conceived and designed the study. LBH and HXX contributed to the study design and the drafting of the manuscript. HXX wrote the original draft. LBQ and HYH contributed to the data analysis independently. HXX, WMN and LN helped with the data acquisition and prepared figures 1-2 and tables 1-3. All the authors have read the draft manuscript and made significant intellectual contributions to the final version. All authors approved the final manuscript.

Acknowledgments: We thank the pregnant women for their participation in our study.

Barfield WD: Public Health Implications of Very Preterm Birth. Clin Perinatol 2018, 45(3):565-577.
Walani SR: Global burden of preterm birth. Int J Gynaecol Obstet 2020, 150(1):31-33.
Arias E, Xu J: United States Life Tables, 2019. National vital statistics reports : from the Centers for Disease Control and Prevention, National Center for Health Statistics, National Vital Statistics System 2022, 70(19):1-59.
Ravelli ACJ, Eskes M, van der Post JAM, Abu-Hanna A, de Groot CJM: Decreasing trend in preterm birth and perinatal mortality, do disparities also decline? BMC public health 2020, 20(1):783.
Conde-Agudelo A, Romero R: Does vaginal progesterone prevent recurrent preterm birth in women with a singleton gestation and a history of spontaneous preterm birth? Evidence from a systematic review and meta-analysis. Am J Obstet Gynecol 2022, 227(3):440-461.e442.
Brown R, Gagnon R, Delisle MF: No. 373-Cervical Insufficiency and Cervical Cerclage. J Obstet Gynaecol Can 2019, 41(2):233-247.
Iams JD, Goldenberg RL, Meis PJ, Mercer BM, Moawad A, Das A, Thom E, McNellis D, Copper RL, Johnson F et al: The length of the cervix and the risk of spontaneous premature delivery. National Institute of Child Health and Human Development Maternal Fetal Medicine Unit Network. The New England journal of medicine 1996, 334(9):567-572.
Wulff CB, Rode L, Rosthøj S, Hoseth E, Petersen OB, Tabor A: Transvaginal sonographic cervical length in first and second trimesters in a low-risk population: a prospective study. Ultrasound Obstet Gynecol 2018, 51(5):604-613.
ACOG Practice Bulletin No.142: Cerclage for the management of cervical insufficiency. Obstet Gynecol 2014, 123(2 Pt 1):372-379.
Huang X, Chen R, Li B: Analysis of maternal and neonatal outcomes using cervical cerclage or conservative treatment in singleton gestations with a sonographic short cervix. Medicine (Baltimore) 2021, 100(18):e25767.
Gavra-Shlissel H, Porat S, Tadmor O, Bitton E, Lipschuetz M, Kabiri D: Validating staging criteria for sonographic and physical examination for cervical changes at <24 weeks' gestation to predict preterm birth. Am J Obstet Gynecol MFM 2024, 6(5s):101211.
Mitrogiannis I, Evangelou E, Efthymiou A, Kanavos T, Birbas E, Makrydimas G, Papatheodorou S: Risk factors for preterm birth: an umbrella review of meta-analyses of observational studies. BMC Med 2023, 21(1):494.
Lucaroni F, Morciano L, Rizzo G, F DA, Buonuomo E, Palombi L, Arduini D: Biomarkers for predicting spontaneous preterm birth: an umbrella systematic review. J Matern Fetal Neonatal Med 2018, 31(6):726-734.
Iams JD, Grobman WA, Lozitska A, Spong CY, Saade G, Mercer BM, Tita AT, Rouse DJ, Sorokin Y, Wapner RJ et al: Adherence to criteria for transvaginal ultrasound imaging and measurement of cervical length. Am J Obstet Gynecol 2013, 209(4):365.e361-365.
Roman A, Gulersen M, Boelig RC, Berghella V: Proposed staging criteria for sonographic and physical exam for cervical changes at <24 weeks gestation to predict preterm birth. Am J Obstet Gynecol MFM 2023, 5(1):100753.
Orzechowski KM, Boelig RC, Baxter JK, Berghella V: A universal transvaginal cervical length screening program for preterm birth prevention. Obstet Gynecol 2014, 124(3):520-525.
Hessami K, D'Alberti E, Mascio DD, Berghella V: Universal cervical length screening and risk of spontaneous preterm birth: a systematic review and meta-analysis. Am J Obstet Gynecol MFM 2024, 6(5s):101343.
Temming LA, Durst JK, Tuuli MG, Stout MJ, Dicke JM, Macones GA, Cahill AG: Universal cervical length screening: implementation and outcomes. Am J Obstet Gynecol 2016, 214(4):523.e521-523.e528.
Hoffman MK, Clifton RG, Biggio JR, Saade GR, Ugwu LG, Longo M, Bousleiman SZ, Clark K, Grobman WA, Frey HA et al: Cervical Pessary for Prevention of Preterm Birth in Individuals With a Short Cervix: The TOPS Randomized Clinical Trial. Jama 2023, 330(4):340-348.
Care A, Nevitt SJ, Medley N, Donegan S, Good L, Hampson L, Tudur Smith C, Alfirevic Z: Interventions to prevent spontaneous preterm birth in women with singleton pregnancy who are at high risk: systematic review and network meta-analysis. Bmj 2022, 376:e064547.
Saccone G, Maruotti GM, Giudicepietro A, Martinelli P: Effect of Cervical Pessary on Spontaneous Preterm Birth in Women With Singleton Pregnancies and Short Cervical Length: A Randomized Clinical Trial. Jama 2017, 318(23):2317-2324.
Heath VC, Southall TR, Souka AP, Elisseou A, Nicolaides KH: Cervical length at 23 weeks of gestation: prediction of spontaneous preterm delivery. Ultrasound Obstet Gynecol 1998, 12(5):312-317.
Koullali B, van Zijl MD, Kazemier BM, Oudijk MA, Mol BWJ, Pajkrt E, Ravelli ACJ: The association between parity and spontaneous preterm birth: a population based study. BMC Pregnancy Childbirth 2020, 20(1):233.
Chang Y, Li W, Shen Y, Li S, Chen X: Association between interleukin-6 and preterm birth: a meta-analysis. Ann Med 2023, 55(2):2284384.
Gudicha DW, Romero R, Gomez-Lopez N, Galaz J, Bhatti G, Done B, Jung E, Gallo DM, Bosco M, Suksai M et al: The amniotic fluid proteome predicts imminent preterm delivery in asymptomatic women with a short cervix. Sci Rep 2022, 12(1):11781.
Wei SQ, Fraser W, Luo ZC: Inflammatory cytokines and spontaneous preterm birth in asymptomatic women: a systematic review. Obstet Gynecol 2010, 116(2 Pt 1):393-401.
Cobo T, Aldecoa V, Figueras F, Herranz A, Ferrero S, Izquierdo M, Murillo C, Amoedo R, Rueda C, Bosch J et al: Development and validation of a multivariable prediction model of spontaneous preterm delivery and microbial invasion of the amniotic cavity in women with preterm labor. Am J Obstet Gynecol 2020, 223(3):421.e421-421.e414.
Watson HA, Seed PT, Carter J, Hezelgrave NL, Kuhrt K, Tribe RM, Shennan AH: Development and validation of predictive models for QUiPP App v.2: tool for predicting preterm birth in asymptomatic high-risk women. Ultrasound Obstet Gynecol 2020, 55(3):348-356.
Ridout AE, Ross G, Seed PT, Hezelgrave NL, Tribe RM, Shennan AH: Predicting spontaneous preterm birth in asymptomatic high-risk women with cervical cerclage. Ultrasound Obstet Gynecol 2023, 61(5):617-623.
Romero R, Conde-Agudelo A, Da Fonseca E, O'Brien JM, Cetingoz E, Creasy GW, Hassan SS, Nicolaides KH: Vaginal progesterone for preventing preterm birth and adverse perinatal outcomes in singleton gestations with a short cervix: a meta-analysis of individual patient data. Am J Obstet Gynecol 2018, 218(2):161-180.

No competing interests reported.

Additionalfig1.png

Download PDF

Journal Publication

published 01 Oct, 2024

Read the published version in BMC Pregnancy and Childbirth →

Editorial decision: Revision requested
22 Aug, 2024
Reviews received at journal
22 Aug, 2024
Reviews received at journal
21 Aug, 2024
Reviews received at journal
21 Aug, 2024
Reviewers agreed at journal
21 Aug, 2024
Reviewers agreed at journal
15 Aug, 2024
Reviewers agreed at journal
12 Aug, 2024
Reviewers agreed at journal
11 Aug, 2024
Reviewers invited by journal
07 Aug, 2024
Editor invited by journal
06 Aug, 2024
Editor assigned by journal
06 Aug, 2024
Submission checks completed at journal
05 Aug, 2024
First submitted to journal
31 Jul, 2024

You are reading this latest preprint version

Prediction model for spontaneous preterm birth less than 32 weeks of gestation in low-risk women with mid-trimester short cervical length: a retrospective cohort study

Status:

Journal Publication

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

Introduction

Materials and methods

Patients

Data Collection

Statistical analysis

Results

Clinical characteristics

Establishing a classification system

Discussion

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1