基于超声特征的早期稽留流产预测模型构建分析

孔文翠; 张瑶佳; 刘明松

doi:10.16766/j.cnki.issn.1674-4152.003472

基于超声特征的早期稽留流产预测模型构建分析

doi: 10.16766/j.cnki.issn.1674-4152.003472

湖州市妇幼保健院超声科，浙江湖州 313000

基金项目:

浙江省医药卫生科技计划项目 2023KY1183

详细信息

通讯作者:
孔文翠，E-mail：Kongwc1080@163.com

中图分类号: R714.21 R445.1
计量
- 文章访问数: 160
- HTML全文浏览量: 119
- PDF下载量: 6
- 被引次数: 0
出版历程
- 收稿日期: 2023-10-18
- 网络出版日期: 2024-05-29

Analysis on the construction of early missed abortion prediction model based on ultrasonic characteristics

Department of Ultrasound, Huzhou Maternal and Child Health Hospital, Huzhou, Zhejiang 313000, China

摘要

摘要: 目的超声特征可用于早期稽留流产的预测，但目前尚无研究说明经阴道超声特征预测早期不同孕周稽留流产的价值，因此本研究建立基于超声特征的早期稽留流产的预测模型，并分析其预测效能。方法纳入湖州市妇幼保健院2021年1月—2022年2月收治的经阴道超声诊断为宫内单胎妊娠孕妇251名为研究对象，通过超声每周测量患者孕囊(GS)、卵黄囊(YS)直径、顶臀长(CRL)和胚胎心率(HR)等超声特征至妊娠10周末，随访并记录妊娠11~12周251例患者稽留流产发生率及GS、YS、CRL、HR水平。采用SPSS 22.0统计学软件中logistic回归数据包筛选早期稽留流产的影响因素，并构建logistic回归预测模型，绘制ROC曲线分析预测效能。结果根据孕妇是否发生早期稽留流产分为流产组(43例)和继续妊娠组(208例)。多因素logistic回归结果显示：临床特征中睾酮(OR=0.452)、雌二醇(OR=1.422)、孕酮(OR=1.442)均为早期稽留流产的预测因子(P < 0.05)；基于临床特征的预测模型预测早期稽留流产的AUC为0.815，妊娠6周超声特征预测模型为0.801，7周超声特征预测模型为0.926，8周超声特征预测模型为0.906，9周超声特征预测模型为0.883，10周超声特征预测模型为0.924。Delong分析显示，妊娠7周及10周超声特征预测模型的曲线下面积高于基于临床特征的预测模型(P < 0.05)。结论基于6~10周超声特征建立的预测模型可用于妊娠11~12周稽留流产的预测，但仅基于妊娠7周及10超声特征构建的预测模型的预测价值高于基于临床特征的预测模型，其余孕周超声特征预测模型预测价值与基于临床特征的预测模型预测价值相当，其中YS是预测早期稽留流产的最强超声特征指标。
- 早期稽留流产 /
- 超声特征 /
- 妊娠 /
- 预测模型
Abstract: Objective Ultrasound features can be used to predict early missed abortion, but there is no research to explain the value of transvaginal ultrasound features in predicting early missed abortion in different gestational weeks, so this study established a prediction model of early missed abortion based on ultrasound features and analyzed its prediction efficiency. Methods A total of 251 pregnant women diagnosed as intrauterine singleton pregnancy by transvaginal ultrasound from January 2021 to February 2022 in Huzhou Maternal and Child Health Hospital were included as the research objects. The gestational sac (GS), yolk sac (YS) diameter, crown-rump length (CRL) and fetal heart rate (HR) were detected by the ultrasound until the 10th week of pregnancy. Two hundred and fifty-one patients were followed up and recorded the incidence of missed abortion in 11-12 weeks and the levels of GS, YS, CRL and HR. Logistic regression data package in SPSS 22.0 software was used to screen the influencing factors of early missed abortion, and a logistic regression prediction model was constructed to draw ROC curve to analyze the prediction efficiency. Results Pregnant women were divided into abortion group (43 cases) and continued pregnancy group (208 cases) according to whether the patients had early missed abortion. The results of multivariate logistic regression showed that testosterone (OR=0.452), estradiol (OR=1.422) and progesterone (OR=1.442) were all predictors of early missed abortion (P < 0.05). The prediction model based on clinical features predicted the area under the curve of early missed abortion was 0.815, the prediction model of 6-week ultrasound characteristics was 0.801, the prediction model of 7-week ultrasound characteristics was 0.926, the prediction model of 8-week ultrasound characteristics was 0.906, the prediction model of 9-week ultrasound characteristics was 0.883 and the prediction model of 10-week ultrasound characteristics was 0.924. According to Delong analysis, the area under the curve of the ultrasonic feature prediction model at the 7th and 10th weeks of pregnancy is higher than that of the prediction model based on clinical features (P < 0.05). Conclusion The prediction model based on ultrasound characteristics of 6-10 weeks can be used to predict missed abortion in 11-12 weeks of pregnancy, but the prediction value of the prediction model based on ultrasound characteristics of 7-week and 10-week of pregnancy are higher than that based on clinical characteristics, and the prediction value of other pregnancy ultrasound characteristics prediction models is equivalent to that based on clinical characteristics, among which YS is the strongest ultrasound characteristic index to predict early missed abortion.
- Early missed abortion /
- Ultrasonic characteristics /
- Pregnancy /
- Prediction model

HTML全文

表 1 继续妊娠组与流产组一般资料比较

Table 1. Comparison of general data between continued pregnancy group and abortion group

项目	类别	流产组(n=43)	继续妊娠组(n=208)	统计量	P值
年龄[M(P₂₅, P₇₅)，岁]		25.3(22.2, 34.7)	26.9(22.0，37.1)	1.310^a	0.196
BMI(x±s)		25.34±4.96	26.96±4.33	1.991^b	0.051
文化程度[例(%)]	初中及以下	12(27.91)	34(16.35)	3.523^c	0.172
	高中	14(32.56)	89(42.79)
	大学及以上	17(39.53)	85(40.87)
长期吸烟[例(%)]		10(23.26)	39(18.75)	0.460^c	0.497
长期饮酒[例(%)]		3(6.98)	19(9.13)	0.207^c	0.649
流产史[例(%)]	人工流产	1(2.33)	6(2.88)	0.041^c	0.839
	自然流产	1(2.33)	3(1.44)	0.177^c	0.674
叶酸使用[例(%)]		21(48.84)	78(37.50)	1.918^c	0.166
剖宫产史[例(%)]		4(9.30)	30(14.42)	0.798^c	0.372
患者铅汞暴露史[例(%)]		1(2.33)	3(1.44)	0.177^c	0.674
高电磁设备暴露史[例(%)]		12(27.91)	45(21.63)	0.799^c	0.371
日常运动时间[M(P₂₅, P₇₅), min]		45(8, 63)	47(13, 70)	1.312^a	0.196
合并症[例(%)]	子宫肌瘤	3(6.98)	6(2.88)	1.726^c	0.189
	卵巢肿瘤	4(9.30)	8(3.85)	2.330^c	0.127
宫颈感染[例(%)]		16(37.21)	96(46.15)	1.154^c	0.283
性激素(x±s)	FSH(U/L)	7.81±0.80	7.78±0.46	0.238^b	0.813
	LH(U/L)	13.82±1.72	14.11±1.66	1.012^b	0.315
	PRL(μg/L)	14.87±1.61	15.84±1.74	3.546^b	0.001
	T(mmol/L)	4.76±1.12	4.14±0.89	3.414^b	0.001
	E₂(ng/L)	47.18±2.18	48.96±2.87	4.594^b	＜0.001
	P(μg/L)	18.46±2.52	19.65±2.21	2.876^b	0.006
注：^a为Z值，^b为t值，^c为χ²值；FSH为卵泡刺激素，LH为黄体生成素，PRL为催乳素，T为睾酮，P为孕酮，E₂为雌二醇。

下载: 导出CSV

表 2 基于临床特征的早期稽留流产发生的多因素logistic回归分析

Table 2. Multivariate logistic regression analysis of early missed abortion based on clinical features

变量	B	SE	Waldχ²	P值	OR值	95% CI
PRL	0.248	0.130	3.627	0.057	1.282	0.993~1.654
T	-0.794	0.241	10.868	0.001	0.452	0.282~0.725
E₂	0.352	0.108	10.633	0.001	1.422	1.151~1.758
P	0.366	0.137	7.169	0.007	1.442	1.103~1.885

下载: 导出CSV

表 3 临床特征预测早期稽留流产发生的效能分析

Table 3. Analysis of the predictive efficacy of clinical features in predicting the occurrence of early missed abortion

变量	AUC	敏感度	特异度	95% CI
T	0.739	76.00	70.00	0.624~0.822
E₂	0.693	62.00	72.00	0.593~0.781
P	0.742	60.00	76.00	0.640~0.835

下载: 导出CSV

表 4 继续妊娠组与流产组不同孕周GS水平(mm)

Table 4. The levels of GS at different gestational weeks between the continuing pregnancy group and the abortion group(mm)

组别	例数	妊娠5周	妊娠6周	妊娠7周
继续妊娠组	208	11(8, 12)	15(13, 19)	23(20, 26)
流产组	43	10(7, 12)	11(8, 13)	12(10, 20)

组别	例数	妊娠8周	妊娠9周	妊娠10周
继续妊娠组	208	30(27, 34)	36(33, 41)	42(38, 47)
流产组	43	14(11, 20)	22(18, 26)	20(18, 21)

下载: 导出CSV

表 5 流产组与继续妊娠组及流产组不同孕周GS水平回归分析

Table 5. Regression analysis of GS levels at different gestational weeks in aborted group, continued pregnancy group and aborted group

项目	Waldχ²	P值	B	95% CI
组别^a
继续妊娠组	68.642	＜0.001	3.742	2.857~4.628
时间^b
妊娠6周	27.368	＜0.001	0.380	0.237~0.522
妊娠7周	17.592	＜0.001	0.195	0.104~0.286
妊娠8周	34.801	＜0.001	0.202	0.135~0.269
妊娠9周	51.910	＜0.001	0.246	0.179~0.313
妊娠10周	50.000	＜0.001	0.160	0.115~0.204
注：^a表示以流产组为参照；^b表示以妊娠5周为参照。

下载: 导出CSV

表 6 继续妊娠组与流产组不同孕周YS水平比较(mm)

Table 6. Comparison of YS levels at different gestational weeks between continued pregnancy group and abortion group(mm)

组别	例数	妊娠5周	妊娠6周	妊娠7周
继续妊娠组	208	2.1(2.0, 2.4)	2.5(2.2, 2.8)	3.1(2.9, 3.9)
流产组	43	2.2(1.5, 2.5)	2.9(2.1, 3.6)	3.0(2.5，4.0)

组别	例数	妊娠8周	妊娠9周	妊娠10周
继续妊娠组	208	3.3(2.9, 3.6)	3.6(3.3, 3.9)	4.0(3.6, 4.4)
流产组	43	4.0(3.0, 5.1)	3.4(2.0, 5.4)	2.1(2.1, 3.1)

下载: 导出CSV

表 7 流产组与继续妊娠组及流产组不同孕周YS水平回归分析

Table 7. Regression analysis of YS levels at different gestational weeks in aborted group, continued pregnancy group and aborted group

项目	Waldχ²	P值	B	95% CI
组别^a
继续妊娠组	61.754	＜0.001	3.357	2.520~4.194
时间^b
妊娠6周	7.864	0.005	-1.427	-2.424~-0.430
妊娠7周	20.874	＜0.001	-1.616	-2.309~-0.923
妊娠8周	3.182	0.002	-0.617	-1.296~0.061
妊娠9周	5.848	0.016	1.218	0.231~2.205
妊娠10周	24.202	＜0.001	1.632	0.982~2.283
注：^a表示以流产组为参照；^b表示以妊娠5周为参照。

下载: 导出CSV

表 8 继续妊娠组与流产组不同孕周CRL水平比较(mm)

Table 8. Comparison of CRL levels at different gestational weeks between continued pregnancy group and abortion group(mm)

组别	例数	妊娠5周	妊娠6周	妊娠7周
继续妊娠组	208	2.4(1.7, 4.2)	4.9(4.1, 6.3)	11.4(9.5, 13.5)
流产组	43	2.2(1.4, 3.1)	3.9(2.8, 4.3)	5.0(3.5, 8.3)

组别	例数	妊娠8周	妊娠9周	妊娠10周
继续妊娠组	208	18.2(16.6, 19.9)	25.3(23.1, 28.5)	35.7(32.9, 38.0)
流产组	43	5.3(3.8, 12.0)	16.9(8.0, 24.9)	5.3(5.1, 14.9)

下载: 导出CSV

表 9 流产组与继续妊娠组及流产组不同孕周CRL水平回归分析

Table 9. Regression analysis of CRL levels at different gestational weeks in aborted group, continued pregnancy group and aborted group

项目	Waldχ²	P值	B	95% CI
组别^a
继续妊娠组	46.108	＜0.001	2.685	1.910~3.460
时间^b
妊娠6周	2.436	0.119	0.385	-0.098~0.868
妊娠7周	26.761	＜0.001	0.341	0.212~0.470
妊娠8周	30.952	＜0.001	0.182	0.118~0.247
妊娠9周	17.410	＜0.001	0.140	0.074~0.206
妊娠10周	40.580	＜0.001	0.086	0.059~0.112
注：^a表示以流产组为参照；^b表示以妊娠5周为参照。

下载: 导出CSV

表 10 继续妊娠组与流产组不同孕周HR水平比较(次/min)

Table 10. Comparison of HR levels at different gestational weeks between continued pregnancy group and abortion group(beats/min)

组别	例数	妊娠5周	妊娠6周	妊娠7周
继续妊娠组	208	109(88, 120)	114(109, 124)	146(133, 157)
流产组	43	103(93, 114)	105(99，125)	118(102, 149)

组别	例数	妊娠8周	妊娠9周	妊娠10周
继续妊娠组	208	165(163, 175)	170(165, 174)	171(163, 175)
流产组	43	123(85, 161)	165(159, 176)	164(146, 173)

下载: 导出CSV

表 11 流产组与继续妊娠组及流产组不同孕周HR水平回归分析

Table 11. Regression analysis of HR level at different gestational weeks in aborted group, continued pregnancy group and aborted group

项目	Waldχ²	P值	B	95% CI
组别^a
继续妊娠组	38.083	＜0.001	2.380	1.624~3.137
时间^b
妊娠6周	3.245	0.072	0.059	-0.005~0.123
妊娠7周	12.671	＜0.001	0.050	0.023~0.079
妊娠8周	12.171	＜0.001	0.030	0.013~0.047
妊娠9周	8.659	0.003	0.083	0.028~0.138
妊娠10周	5.273	0.022	0.043	0.006~0.080
注：^a表示以流产组为参照；^b表示以妊娠5周为参照。

下载: 导出CSV

表 12 基于超声特征的早期稽留流产发生的多因素logistic回归分析

Table 12. Multivariate logistic regression analysis of early missed abortion based on ultrasonic characteristics

孕周	相关指标	B	SE	Waldχ²	P值	OR值(95% CI)
6	GS	-1.399	0.307	20.726	＜0.001	0.247(0.135~0.451)
	CRL	-0.902	0.507	3.161	0.045	0.406(0.150~0.897)
7	GS	-2.018	0.694	8.448	0.004	0.133(0.034~0.518)
	YS	-0.821	0.324	6.401	0.011	0.440(0.233~0.831)
	CRL	0.180	0.101	3.159	0.046	1.197(1.182~1.460)
	HR	-2.081	0.693	9.017	0.003	0.125(0.032~0.485)
8	GS	-0.193	0.103	3.530	0.060	0.824(0.674~1.008)
	YS	1.298	0.394	10.852	0.001	3.660(1.191~7.922)
	CRL	-0.893	0.192	21.564	＜0.001	0.409(0.281~0.597)
	HR	-0.142	0.026	2.540	0.111	0.867(0.711~0.910)
9	GS	-0.196	0.026	14.010	＜0.001	0.822(0.804~0.955)
	CRL	0.012	0.035	0.125	0.723	1.119(1.045~1.485)
10	GS	-0.125	0.070	3.177	0.075	0.882(0.769~1.013)
	YS	-3.060	0.722	17.974	＜0.001	0.047(0.011~0.193)
	CRL	-0.107	0.098	1.199	0.274	0.899(0.742~1.088)

下载: 导出CSV

表 13 超声特征预测早期稽留流产发生的效能分析

Table 13. Efficiency analysis of ultrasonic characteristics in predicting early missed abortion

孕周	相关指标	敏感度(%)	特异度(%)	95% CI
6	GS	66.00	85.00	0.801(0.718~0.880)
	CRL	50.00	70.00	0.537(0.485~0.685)
7	GS	76.00	90.00	0.800(0.615~0.896)
	YS	90.00	80.00	0.912(0.840~0.960)
	CRL	55.00	86.00	0.632(0.517~0.852)
	HR	62.00	82.00	0.746(0.649~0.828)
8	YS	84.00	88.00	0.892(0.815~0.946)
	CRL	58.00	84.00	0.678(0.594~0.783)
	HR	56.00	86.00	0.693(0.600~0.787)
9	GS	72.00	96.00	0.883(0.797~0.935)
10	YS	92.00	84.00	0.924(0.802~0.969)

下载: 导出CSV

表 14 基于临床特征的预测模型与基于超声特征的预测模型预测早期稽留流产效能比较

Table 14. Comparison of predictive models based on clinical features and ultrasonic features in predicting early missed abortion

预测模型	敏感度(%)	特异度(%)	AUC的95% CI
6周超声特征预测模型	66.00	85.00	0.801(0.709~0.874)
7周超声特征预测模型	90.00	80.00	0.926(0.855~0.969)
8周超声特征预测模型	76.00	90.00	0.906(0.831~0.955)
9周超声特征预测模型	72.00	90.00	0.883(0.804~0.939)
10周超声特征预测模型	92.00	84.00	0.924(0.854~0.968)
基于临床特征的预测模型	70.00	86.00	0.815(0.725~0.886)

下载: 导出CSV

参考文献(16)

[1]	郑峥, 顾向应, 刘欣燕, 等. 早期妊娠稽留流产治疗专家共识[J]. 中国实用妇科与产科杂志, 2020, 36(1): 70-73. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGSF202012011.htm ZHENG Z, GU X Y, LIU X Y, et al. Expert consensus on treatment of missed abortion in early pregnancy[J]. Chinese Journal of Practical Gynecology and Obstetrics, 2020, 36(1): 70-73. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGSF202012011.htm
[2]	王红钰, 马恩萍, 涂国凤. 稽留流产孕早期预测模型的构建与验证[J]. 中国计划生育和妇产科, 2020, 12(5): 43-47. doi: 10.3969/j.issn.1674-4020.2020.05.10 WANG H Y, MA E P, TU G F. Construction and verification of early prediction model of missed abortion[J]. Chinese Journal of Family Planning & Gynecotokology, 2020, 12(5): 43-47. doi: 10.3969/j.issn.1674-4020.2020.05.10
[3]	LIU X, HAN Y, WANG X, et al. Serum anti-Müllerian hormone levels are associated with early miscarriage in the IVF/ICSI fresh cycle[J]. BMC Pregnancy Childbirth, 2022, 22(1): 279. doi: 10.1186/s12884-022-04591-5
[4]	GHAEDI B, CHENG W, AMERI S, et al. Performance of single serum progesterone in the evaluation of symptomatic first-trimester pregnant patients: a systematic review and meta-analysis[J]. CJEM, 2022, 24(6): 611-621. doi: 10.1007/s43678-022-00332-x
[5]	JIANG W Z, YANG X L, LUO J R. Risk factors for missed abortion: retrospective analysis of a single institution ' s experience[J]. Reprod Biol Endocrinol, 2022, 20(1): 115. doi: 10.1186/s12958-022-00987-2
[6]	DEVILBISS E A, MUMFORD S L, SJIAARDA L A, et al. Prediction of pregnancy loss by early first trimester ultrasound characteristics[J]. Am J Obstet Gynecol, 2020, 223(2): 242. e1-242.522. doi: 10.1016/j.ajog.2020.02.025
[7]	YUAN G, LV B, DU X, et al. Prediction model for missed abortion of patients treated with IVF-ET based on XGBoost: a retrospective study[J]. Peer J, 2023, 11(1): e14762. DOI: 10.7717/peerj.14762.
[8]	张悦, 陈红波, 刘羽. 新生育政策下不同高龄段孕产妇安全性分析[J]. 中华全科医学, 2023, 21(4): 629-631, 636. doi: 10.16766/j.cnki.issn.1674-4152.002947?viewType=HTML ZHNAG Y, CHEN H B, LIU Y. Analysis on the safety of different aged pregnant women under the new birth policy[J]. Chinese Journal of General Practice, 2023, 21(4): 629-631, 636. doi: 10.16766/j.cnki.issn.1674-4152.002947?viewType=HTML
[9]	王欣宓, 孙丽芳, 王立芳, 等. 孕早期无症状稽留流产的识别[J]. 中国计划生育和妇产科, 2022, 14(7): 71-74. doi: 10.3969/j.issn.1674-4020.2022.07.17 WANG X F, SUN L F, WANG L F, et al. Recognition of asymptomatic missed abortion in first trimester[J]. Chinese Journal of Family Planning & Gynecotokology, 2022, 14(7): 71-74. doi: 10.3969/j.issn.1674-4020.2022.07.17
[10]	DEVILBISS E A, MUMFORD S L, SJAARDA L A, et al. Prediction of pregnancy loss by early first trimester ultrasound characteristics[J]. Am J Obstet Gynecol, 2020, 223(2): 2421-2422. doi: 10.1016/j.ajog.2020.02.025
[11]	OUYANG Y, QIN J, LIN G, et al. Reference intervals of gestational sac, yolk sac, embryonic length, embryonic heart rate at 6-10 weeks after in vitro fertilization-embryo transfer[J]. BMC Pregnancy Childbirth, 2020, 20(1): 533. doi: 10.1186/s12884-020-03186-2
[12]	DETTI L, FRANCILLON L, CHRISTIANSEN M E, et al. Early pregnancy ultrasound measurements and prediction of first trimester pregnancy loss: a logistic model[J]. Sci Rep, 2020, 10(1): 1545. doi: 10.1038/s41598-020-58114-3
[13]	DETTI L, ROMAN R A, GOEDECKE P J, et al. Pilot study establishing a nomogram of yolk sac growth during the first trimester of pregnancy[J]. J Obstet Gynaecol Res, 2020, 46(2): 223-228. doi: 10.1111/jog.14173
[14]	PHILLIPS C H, BENSON C B, DERFEE S M, et al. "Pseudogestational sac" and other 1980s-era concepts in early frst-trimester ultrasound: are they still relevant today?[J]. J Ultrasound Med, 2020, 39(8): 1547-1551. doi: 10.1002/jum.15243
[15]	PERUMUMAL N, OHUMA E O, PRENTICE A M, et al. Implications for quantifying early life growth trajectories of term-born infants using INTERGROWTH-21st newborn size standards at birth in conjunction with World Health Organization child growth standards in the postnatal period[J]. Paediatr Perinat Epidemiol, 2022, 36(6): 839-850. doi: 10.1111/ppe.12880
[16]	TAN S, PAN P, YANG Z, et al. Study on the correlation between the ultrasound phenotype and copy number variation of abnormal embryo in spontaneous abortion[J]. J Obstet Gynaecol Res, 2021, 47(11): 3779-3788. doi: 10.1111/jog.14987