高级检索
赵钦康, 魏俊妮. 中老年跌倒住院患者治疗效果影响因素分析及预测模型建立[J]. 中国公共卫生, 2023, 39(7): 902-907. DOI: 10.11847/zgggws1139752
引用本文: 赵钦康, 魏俊妮. 中老年跌倒住院患者治疗效果影响因素分析及预测模型建立[J]. 中国公共卫生, 2023, 39(7): 902-907. DOI: 10.11847/zgggws1139752
ZHAO Qinkang, WEI Junni. Treatment outcome of falls injury and its associates in middle-aged and elderly inpatients: medical records-based analysis and prediction model construction[J]. Chinese Journal of Public Health, 2023, 39(7): 902-907. DOI: 10.11847/zgggws1139752
Citation: ZHAO Qinkang, WEI Junni. Treatment outcome of falls injury and its associates in middle-aged and elderly inpatients: medical records-based analysis and prediction model construction[J]. Chinese Journal of Public Health, 2023, 39(7): 902-907. DOI: 10.11847/zgggws1139752

中老年跌倒住院患者治疗效果影响因素分析及预测模型建立

Treatment outcome of falls injury and its associates in middle-aged and elderly inpatients: medical records-based analysis and prediction model construction

  • 摘要:
      目的  了解中老年跌倒住院患者治疗效果的影响因素并建立列线图预测模型,为及时合理有效地对中老年跌倒住院患者进行医疗处置提供参考依据。
      方法  收集山西省汾阳医院2017年1月 — 2022年1月1500例中老年跌倒住院患者的病历资料,应用多因素非条件logistic回归模型分析中老年跌倒住院患者治疗效果的影响因素,并应用R 4.2.0软件建立列线图预测模型,通过受试者工作特征曲线(ROC)下面积(AUC)评价模型的区分度,采用Hosmer-Lemeshow检验和Calibration校准曲线评价模型的拟合度。
      结果  1500例中老年跌倒住院患者中,出院时治愈1017例,治愈率为67.8 %;多因素非条件logistic回归分析结果显示,入院途径为急诊、躯干跌伤、跌伤性质为其他和校正年龄的Charlson共病指数(aCCI) ≥ 6分中老年跌倒住院患者的治疗效果较差,上肢和下肢跌伤、跌伤严重程度为中度和重度中老年跌倒住院患者的治疗效果较好;据此建立的列线图预测模型的AUC = 0.732(95%CI = 0.705 ~ 0.759),预测模型的拟合度良好(χ2 = 12.692,P = 0.123)。
      结论  入院途径、跌伤部位、跌伤性质、跌伤严重程度和aCCI与中老年跌倒住院患者出院时的疗效密切相关,基于上述5个影响因素建立的列线图预测模型拟合度和区分度良好,有助于中老年跌倒住院患者的个体化治疗和针对性治疗。

     

    Abstract:
      Objective  To examine treatment outcome of falls injury and its influencing factors in middle-aged and elderly inpatients and to establish a nomogram prediction model of treatment efficacy for providing evidence to effective treatment of falls injury of the inpatients.
      Methods  The medical records of 1 500 patients aged 45 years and above and hospitalized due to falls injury from January 2017 through January 2022 were collected from Fenyang Hospital of Shanxi province. Unconditional multivariate logistic regression model was used to analyze factors affecting treatment efficacy of falls injury of the inpatients. A nomogram model for predicting falls injury treatment efficacy was established with R4.2.0 software; the discrimination of the established model was evaluated with the area under the receiver operating characteristic curve (AUC) and the goodness of fit of the model was assessed with Hosmer-Lemeshow test and calibration curve.
      Results  Of all the inpatients, 1 017 (67.8%) were cured at the discharges. The results of logistic regression analysis showed that the inpatients with following characteristics were more likely to have poor treatment outcome: with emergency admission to hospital, suffering from torso injury, with other unspecified falls injury, and being assessed as having an age-adjusted Charlson comorbidity index (aCCI) of 6 and greater; while, the inpatients with falls-induced upper or lower limb injury and having moderate or severe falls injury were more likely to have a better treatment outcome. A nomogram prediction model involving 5 variables was constructed, with an AUC of 0.732 (95% confidence interval: 0.705, 0.759) and a good goodness-of-fit (χ2 = 12.692, P = 0.123).
      Conclusion  Admission mode, body part of falls injury, type of falls injury, severity of falls injury, and aCCI are closely related to treatment outcome of falls injury in middle-aged and elderly inpatients. The established nomogram model covering those variables for predicting treatment outcome of falls injury could be helpful for individualized and targeted treatment of falls injury among middle-aged and elderly inpatients.

     

/

返回文章
返回