口腔颌面恶性肿瘤患者术后发生便秘风险预测模型构建

doi:10.12016/j.issn.2096-1456.2022.08.005

摘要/Abstract

摘要：

目的了解口腔颌面恶性肿瘤患者术后便秘发生情况与便秘影响因素,构建便秘风险预测模型,为防治术后便秘提供参考。方法回顾性分析2019年6月至2020年6月在中山大学附属口腔医院行口腔颌面恶性肿瘤手术的191例患者资料,采用单因素分析和Logistic多因素回归分析筛选出独立影响因素,建立风险预测列线图,采用ROC曲线下面积对预测模型进行评价。内部和外部均采用C指数验证该模型的准确度。结果 191例患者中有52例（27.23%）术后出现便秘症状。单因素分析显示,术前便秘史、进食能全素、气管切开、吸烟、喝酒、手术时长、出血量、卧床时间、进食匀浆膳、性别、手术修补方式、使用益生菌、癌症T分期、进食量均可能是口腔颌面恶性肿瘤患者术后便秘的影响因素（P<0.05）。多因素分析结果显示,修补方式、卧床时间、性别是口腔颌面恶性肿瘤患者术后便秘的独立危险因素（P<0.05）,修补方式为腓骨瓣、卧床时间长、男性患者术后易出现便秘。将修补方式、卧床时间、性别纳入预测模型中,训练组和验证组的C指数值分别为0.882和0.953;训练组的ROC曲线下面积为0.909（95%CI: 0.850 ~ 0.968）,验证组的ROC曲线下面积为0.893（95%CI: 0.787 ~ 0.999）,列线图显示出良好的鉴别能力。结论修补方式、卧床时间、性别是口腔颌面恶性肿瘤患者术后便秘的独立危险因素,构建的风险预测模型有较好的判别能力。

关键词: 便秘, 术后, 口腔颌面部, 外科, 恶性肿瘤, 风险预测, 模型, 危险因素

Abstract:

Objective To understand the incidence and influencing factors of postoperative constipation in patients with malignant tumors who undergo oral and maxillofacial surgery and construct a constipation risk prediction model to provide a reference for the prevention and treatment of postoperative constipation. Methods The data of 191 patients who underwent oral and maxillofacial malignant tumor surgery at the Affiliated Stomatological Hospital of Sun Yat sen University from June 2019 to June 2020 were analyzed retrospectively. The independent influencing factors were selected via univariate analysis and logistic multivariate regression analysis, a risk prediction nomogram was established, and the prediction model was evaluated by the area under the ROC curve. Both internal and external use the C index to verify the accuracy of the model. Results Among 191 patients, 52 (27.23%) had postoperative constipation. Univariate analysis showed that a preoperative secret history of defecation, total energy intake, tracheotomy, smoking, drinking, operation duration, bleeding volume, bed time, eating homogenate diet, sex, surgical repair method, use of probiotics, T-stage of cancer and food intake may be the influencing factors of postoperative constipation in patients with oral and maxillofacial malignant tumors (P<0.05). Multivariate analysis showed that repair method, bed time and sex were independent risk factors for postoperative constipation in patients with oral and maxillofacial malignant tumors (P<0.05). The repair method was a fibular myocutaneous flap with a long bed time, and male patients were prone to constipation after surgery. The c-index values in the training group and the verification group were 0.882 and 0.953, respectively. The area under the ROC curve of the training group was 0.909 (95%CI: 0.850-0.968), and the area under the ROC curve of the verification group was 0.893 (95%CI: 0.787-0.999). The nomogram showed good discrimination ability. Conclusion The repair method, bed time and sex are independent risk factors for postoperative constipation in patients with oral and maxillofacial malignant tumors. The risk prediction model has good discrimination ability.

Key words: constipation, postoperative, oral and maxillofacial, surgical, malignant tumors, risk prediction, model, risk factor

中图分类号:

朱惠璇, 何杏芳, 黄秋雨, 刘满凤, 林艳彤. 口腔颌面恶性肿瘤患者术后发生便秘风险预测模型构建[J]. 口腔疾病防治, 2022, 30(8): 564-570.

ZHU Huixuan, HE Xingfang, HUANG Qiuyu, LIU Manfeng, LIN Yantong. Establishment of a risk prediction model for postoperative constipation in patients with oral and maxillofacial malignant tumors[J]. Journal of Prevention and Treatment for Stomatological Diseases, 2022, 30(8): 564-570.

图/表 7

表1 口腔颌面恶性肿瘤患者术后便秘影响因素的单因素分析

Table 1 Univariate analysis of the factors influencing postoperative constipation in patients with oral and maxillofacial malignancies

	No constipation group(n=139)	Constipation group(n=52)	χ²/t	P
Sex
Male	72	40	9.949	0.002
Female	67	12
Smoke
No	103	28	7.206	0.007
Yes	36	24
Drink
No	117	37	4.107	0.043
Yes	22	15
Constipation before surgery
No	133	45	4.989	0.026
Yes	6	7
Intake of total nutrients
No	38	2	11.235	0.001
Yes	101	50
Tracheotomy
No	116	16	49.202	<0.001
Yes	23	36
Operation duration/h	3.17 ± 2.81	6.75 ± 3.18	-7.555	<0.001
Amount of bleeding/mL	193.60 ± 273.00	472.02 ± 404.19	-4.591	<0.001
Time in bed/day	2.34 ± 2.39	7.31 ± 3.13	-11.648	<0.001
Eat homogenate meals/(sub/day)	3.04 ± 0.48	2.65 ± 0.62	4.079	<0.001
Food-intake/mL	1 471.94 ± 267.58	1 694.12 ± 214.86	-5.895	<0.001
Surgical repair method
Anterolateral thigh flap	8	17	54.685	<0.001
Fibula flap	1	8
Forearm flap	13	10
Other flap	12	4
No	105	13
Use of probiotics
No	95	11	34.124	<0.001
Yes	44	41
The T stage of cancer pain
1-2	119	21	39.550	<0.001
3-4	20	31

表2 口腔颌面恶性肿瘤患者术后便秘影响因素的二元logistic回归分析

Table 2 Binary logistic regression analysis of influencing factors of postoperative constipation in patients with oral and maxillofacial malignant tumors

Variable	β	SE	Wald	P	OR	95%CI
Sex	-1.595	0.643	6.162	0.013	0.203	0.058-0.715
Time in bed	0.608	0.153	15.822	<0.001	1.837	1.361-2.479
Surgical repair method			8.297	0.081
Anterolateral thigh flap	1.128	1.296	0.757	0.384	3.090	0.243-39.210
Fibula flap	4.263	1.646	6.709	0.010	71.011	2.821-1 787.267
Forearm flap	0.430	1.284	0.112	0.738	1.537	0.124-19.038
Other flap	0.607	1.482	0.168	0.682	1.835	0.101-33.502
Constants	-18.967	13 820.377	0.000	0.999	0.000

图1 预测便秘的列线图

Figure 1 Nomogram for predicting constipation Gender (1= male, 2= female), fibula flap (0= nonfibula flap, 1= fibula flap), time in bed by days (1 to 15 days)

图2 训练组的列线图校准曲线

Figure 2 Calibration curve for the nomogram training group The standard curve is a straight line passing through the origin of the coordinate axis with a slope of 1. The closer the predicted calibration curve is to the standard curve, the better the predictive power of the nomogram

图3 验证组的列线图校准曲线

Figure 3 Calibration curve of the nomogram validation group The calibrated curve is close to the diagonal, that is, the gap between the predicted value and the real value is small, and the prediction performance of the model is good

图4 训练组口腔颌面恶性肿瘤患者术后便秘风险预测的ROC曲线

Figure 4 ROC curve of predicting the risk of postoperative constipation in patients with oral and maxillofacial malignant tumors in the training group On the ROC curve, the point near the upper left of the coordinate map is the high cutoff value for both sensitivity and specificity. The sensitivity is drawn as the vertical coordinates and (1-specificity) as the horizontal coordinates, and the larger the area under the curve is, the higher the diagnostic accuracy.

图5 验证组口腔颌面恶性肿瘤患者术后便秘风险预测的ROC曲线

Figure 5 ROC curve of predicting the risk of postoperative constipation in patients with oral and maxillofacial malignant tumors in the validation group On the ROC curve, the point near the top left of the coordinate map has high cutoff values for both sensitivity and specificity, showing the excellent discrimination effect of this model for patients with constipation

参考文献 20

[1]	Nagao T, Warnakulasuriya S. Screening for oral cancer: future prospects, research and policy development for Asia[J]. Oral Oncol, 2020, 105: 104632. doi: 10.1016/j.oraloncology.2020.104632. DOI URL
[2]	王远鑫. 膳食干预对颅脑外伤患者术后便秘的影响[J]. 中国医学创新, 2021, 18(10): 101-104. doi: 10.3969/j.issn.1674-4985.2021.10.024. DOI
	Wang YX. Effect of dietary intervention on postoperative constipation in BI patients[J]. Chin Med Innovation, 2021, 18 (10): 101-104. doi: 10.3969/j.issn.1674-4985.2021.10.024. DOI
[3]	龚惠华, 曾兴, 章玲, 等. 腹部振颤闪罐法治疗胸腰椎骨折术后便秘的护理研究[J]. 全科护理, 2021, 19(11): 1497-1499. doi: 10.12104/j.issn.1674-4748.2021.11.015. DOI
	Gong HH, Zeng X, Zhang L, et al. Nursing study of constipation after thoracolumbar fractures[J]. General Care, 2021, 19 (11): 1497-1499. doi: 10.12104/j.issn.1674-4748.2021.11.015. DOI
[4]	Iovino P, Chiarioni G, Bilancio G, et al. New onset of constipation during long-term physical inactivity: a proof-of-concept study on the immobility-induced bowel changes[J]. PLoS One, 2013, 8(8): e72608. doi: 10.1371/journal.pone.0072608. DOI URL
[5]	Jehn P, Korn P, Gellrich NC, et al. Fulfillment of patients' information needs during oral cancer treatment and its association with posttherapeutic quality of Life[J]. Qual Life Res, 2021, 30(1): 169-180. doi: 10.1007/s11136-020-02616-0. DOI URL
[6]	Bharucha AE, Lacy BE. Mechanisms, Evaluation, and Management of Chronic Constipation[J]. Gastroenterology, 2020, 158(5): 1232-1249.e3. doi: 10.1053/j.gastro.2019.12.034. DOI PMID
[7]	Erichsén E, Milberg A, Jaarsma T, et al. Constipation in specialized palliative care: factors related to constipation when applying different definitions[J]. Support Care Cancer, 2016, 24(2): 691-698. doi: 10.1007/s00520-015-2831-5. DOI PMID
[8]	高永祥, 张晋昕. Logistic回归分析的样本量确定[J]. 循证医学, 2018, 18(2): 122-124. doi: 10.12019/j.issn.1671-5144.2018.02.015. DOI
	Gao YX, Zhang JX. Sample size was determined for the logistic regression analysis[J]. Evidence-Based Medicine, 2018, 18(2): 122-124. doi: 10.12019/j.issn.1671-5144.2018.02.015. DOI
[9]	孙亚清. Logistic回归样本量确定所需自变量事件数的模拟研究[D]. 广州: 南方医科大学, 2016.
	Sun YQ. Simulation study of Logistic regression sample size to determine the number of required independent variable events required[D]. Guangdong: Southern Medical University, 2016.
[10]	刘劼, 严冬青, 刘健. 膳食纤维―人类的健康卫士[J]. 现代食品, 2020, 1(1): 113-115. doi: 10.16736/j.cnki.cn41-1434/ts.2020.01.043. DOI
	Liu J, Yan DQ, Liu J. Dietary fiber- -human health guard[J]. Modern food, 2020, 1 (1): 113-115. doi: 10.16736/j.cnki.cn41-1434/ts.2020.01.043. DOI
[11]	Van Gijn DR, D'souza J, King W, et al. Free flap head and neck reconstruction with an emphasis on postoperative care[J]. Facial Plast Surg, 2018, 34(6): 597-604. doi: 10.1055/s-0038-1676076. DOI URL
[12]	Li Y, Tong WD, Qian Y. Effect of physical activity on the association between dietary fiber and constipation: evidence from the National health and nutrition examination survey 2005-2010[J]. J Neurogastroenterol Motil, 2021, 27(1): 97-107. doi: 10.5056/jnm20051. DOI URL
[13]	Feuvrier D, Sagawa Y Jr, Béliard S, et al. Long-term donor-site morbidity after vascularized free fibula flap harvesting: clinical and gait analysis[J]. J Plast Reconstr Aesthet Surg, 2016, 69(2): 262-269. doi: 10.1016/j.bjps.2015.10.007. DOI PMID
[14]	Serra J, Pohl D, Azpiroz F, et al. European society of neurogastroenterology and motility guidelines on functional constipation in adults[J]. Neurogastroenterol Motil, 2020, 32(2): e13762. doi: 10.1111/nmo.13762. DOI
[15]	Yurtdaş G, Acar-Tek N, Akbulut G, et al. Risk factors for constipation in adults: a Cross-Sectional study[J]. J Am Coll Nutr, 2020, 39(8): 713-719. doi: 10.1080/07315724.2020.1727380. DOI URL
[16]	姚克青, 郭志利, 代忠, 等. 营养支持小组对肿瘤患者经鼻饲管喂养并发症预防的临床实践[J]. 肿瘤综合治疗电子杂志, 2019, 5(4): 40-44. doi: 10.12151/JMCM.2019.04-09. DOI
	Yao KQ, Guo ZL, Dai Z, et al. Clinical practice of nasal feeding tube feeding complications in oncology patients[J]. Elect J Comprehen Treat Oncol, 2019, 5(4): 40-44. doi: 10.12151/JMCM.2019.04-09. DOI
[17]	Dzierżanowski T, Ciałkowska-Rysz A. Behavioral risk factors of constipation in palliative care patients[J]. Support Care Cancer, 2015, 23(6): 1787-1793. doi: 10.1007/s00520-014-2495-6. DOI PMID
[18]	Lee EJ. Correlations among pain, depressive symptoms, constipation, and serotonin levels in smokers and non smokers[J]. Perspect Psychiatr Care, 2020, 56(4): 864-870. doi: 10.1111/ppc.12503. DOI URL
[19]	Yang F, Wei J, Shen M, et al. Integrated analyses of the gut microbiota, intestinal permeability, and serum metabolome phenotype in rats with alcohol withdrawal syndrome[J]. Appl Environ Microbiol, 2021, 87(18): e0083421. doi: 10.1128/AEM.00834-21. DOI
[20]	Chen Y, Du H, Wei BH, et al. Development and validation of risk-stratification delirium prediction model for critically ill patients: a prospective, observational, single-center study[J]. Medicine (Baltimore), 2017, 96(29): e7543. doi: 10.1097/MD.0000000000007543. DOI URL