天津医药 ›› 2026, Vol. 54 ›› Issue (11): 1209-1214.doi: 10.11958/20260406

• 临床研究 • 上一篇    下一篇

膝骨关节炎患者全膝关节置换术后恶心呕吐的影响因素预测模型建立与验证

钟杰林(), 蔡迎, 金成浩, 苏啸天, 刘博, 李栩   

  1. 1 天津市泰达医院骨四科 (邮编 300450)
    2 天津市泰达医院麻醉科 (邮编 300450)
  • 收稿日期:2026-02-04 修回日期:2026-04-08 出版日期:2026-11-15 发布日期:2026-11-10
  • 作者简介:钟杰林(1979),男,副主任医师,主要从事膝关节炎方面研究。E-mail:liusf909@126.com

Establishment and validation of a nomogram for predicting postoperative nausea and vomiting after total knee arthroplasty in patients with knee osteoarthritis

ZHONG Jielin(), CAI Ying, JIN Chenghao, SU Xiaotian, LIU Bo, LI Xu   

  1. 1 Department of Orthopedics 4, TEDA Hospital, Tianjin 300450, China
    2 Department of Anesthesiology, TEDA Hospital, Tianjin 300450, China
  • Received:2026-02-04 Revised:2026-04-08 Published:2026-11-15 Online:2026-11-10

摘要:

目的 探讨膝骨关节炎(KOA)患者行全膝关节置换术(TKA)术后恶心呕吐(PONV)的危险因素,构建列线图模型并验证其预测价值。方法 回顾性分析330例行TKA的KOA患者临床资料,按7∶3分为建模组(231例)和验证组(99例)。比较2组患者基线临床资料,以术后24 h内是否发生PONV将建模组分为PONV组(54例)和非PONV组(177例),进行单因素分析和二元Logistic回归多因素分析。根据多因素分析结果构建列线图预测模型,通过受试者工作特征(ROC)曲线、Hosmer-Lemeshow检验、校准曲线进行模型内部验证,采用决策曲线分析(DCA)与临床影响曲线(CIC)评估模型临床效用。结果 与非PONV组比较,PONV组女性、非吸烟状态、晕动病史、PONV史及术后使用阿片类药物占比高,年龄低(P<0.05);女性(OR=4.559,95%CI:1.897~10.957)、非吸烟状态(OR=2.499,95%CI:1.053~5.928)、晕动病史(OR=5.474,95%CI:2.141~13.998)、PONV史(OR=4.715,95%CI:1.264~17.588)、术后使用阿片类药物(OR=3.270,95%CI:1.431~7.475)为术后PONV独立危险因素,年龄大(OR=0.835,95%CI:0.778~0.895)为保护因素。基于上述独立影响因素构建列线图预测模型,建模组曲线下面积(AUC)为0.872(95%CI:0.814~0.929),敏感度为79.63%,特异度为84.18%。验证组AUC为0.849(95%CI:0.757~0.942),提示模型区分度良好。Hosmer-Lemeshow检验χ2=9.355(P=0.313),模型拟合优度良好;DCA曲线显示高风险阈值在0~0.8范围内干预可获正向收益;CIC曲线提示模型预测病例数与实际发病情况契合度良好。结论 KOA患者TKA术后PONV列线图风险预测模型具有良好的区分度、拟合优度及临床实用价值,可有效评估PONV发生风险。

关键词: 骨关节炎,膝, 关节成形术,置换,膝, 手术后恶心呕吐, 危险因素, 列线图

Abstract:

Objective To explore the risk factors for postoperative nausea and vomiting (PONV) in patients with knee osteoarthritis (KOA) undergoing total knee arthroplasty (TKA), and to establish and validate a clinically applicable nomogram for predicting PONV risk. Methods A retrospective analysis was performed on clinical data from 330 KOA patients who underwent TKA between January 2022 and December 2024. Patients were randomly divided into the training cohort (n = 231) and the validation cohort (n = 99) at a ratio of 7∶3. Baseline clinical data were compared between the two groups. The modeling group was divided into the PONV group (54 cases) and the non-PONV group (177 cases) according to the occurrence of PONV within 24 hours postoperatively. Univariate analysis and binary Logistic multivariate analysis were performed. A nomogram prediction model was constructed accordingly. Model performance was evaluated using receiver operating characteristic (ROC) curves, Hosmer-Lemeshow tests and calibration curves. Clinical utility was further assessed by decision curve analysis (DCA) and clinical impact curve (CIC). Results Compared with the non-PONV group, in the PONV group, the proportions of females, non-smoking status, history of motion sickness, history of PONV and postoperative use of opioids were higher, and patient age was lower (P<0.05). Multivariate analysis identified that female (OR = 4.559, 95%CI: 1.897-10.957), non-smoking status (OR = 2.499, 95%CI: 1.053-5.928), history of motion sickness (OR = 5.474, 95%CI: 2.141-13.998), history of PONV (OR = 4.715, 95%CI: 1.264-17.588) and postoperative use of opioid drugs (OR =3.270, 95%CI: 1.431-7.475) were independent risk factors for postoperative PONV, whereas older age was a protective factor (OR = 0.835, 95%CI: 0.778-0.895,P < 0.05). A nomogram prediction model was constructed based on the above independent influencing factors. The area under the ROC curve (AUC) of the modeling group was 0.872 (95%CI: 0.814-0.929) with sensitivity of 79.63%, and a specificity of 84.18%. The AUC of the validation group was 0.849 (95%CI: 0.757-0.942), suggesting that the model had good discrimination. The Hosmer-Lemeshow test (χ² = 9.355, P = 0.313) confirmed satisfactory goodness of fit. DCA demonstrated a positive net benefit within a threshold probability range of 0 to 0.8, and CIC showed good agreement between predicted and actual PONV cases. Conclusion The established nomogram for predicting PONV after TKA in KOA patients demonstrates favorable discrimination, calibration and clinical applicability. It can be used to reliably identify the risk of PONV occurrence.

Key words: osteoarthritis, knee, arthroplasty, replacement, knee, postoperative nausea and vomiting, risk factors, nomograms

中图分类号: