Scholarlink

Background <p>Schizophrenia represents a significant global public health burden, with considerable impact on morbidity and mortality. This study sought to construct a prognostic model to estimate mortality risk in individuals diagnosed with schizophrenia, employing a Cox proportional hazards framework with time-dependent variables.</p> Methods <p>Participants were recruited from a cohort of individuals with schizophrenia in a less-developed area of Southwest China and followed for a median of 6.4 years. A time-dependent Cox model based on the Anderson-Gill method was established to predict survival, followed by comprehensive model evaluation and graphical presentation.</p> Results <p>A total of 1,009 participants were enrolled, with 86 deaths recorded during follow-up. The model identified female sex (HR = 0.599, 95%CI= [0.387–0.928]) and hospitalization (HR = 0.499, 95%CI= [0.251–0.992]) as protective factors. Higher social function scores were associated with reduced mortality risk (HR = 0.889, 95%CI= [0.793–0.995]). Conversely, poor sleep quality (HR = 3.824, 95%CI= [1.482–9.867]) and abnormal electrocardiographic findings (HR = 2.185, 95%CI= [1.236–3.862]) increased mortality risk. Each additional year of age at initial diagnosis corresponded to an elevated risk of death (HR = 1.044, 95%CI= [1.027–1.060]). Model performance was robust, with a concordance index (C-index) of 0.820. Time-dependent AUC values at 1, 3, and 5 years were 0.814, 0.848, and 0.831, respectively.</p> Conclusions <p>Mortality risk among individuals with schizophrenia was significantly associated with sex, hospitalization status, social functioning, age at diagnosis, sleep quality, and electrocardiographic abnormalities. The model demonstrated strong predictive accuracy, and the accompanying nomogram together with the Excel-based scoring system offers a practical tool for rapid clinical risk assessment, supporting timely implementation of targeted interventions to improve patient outcomes.</p> Clinical trial number <p>Not applicable.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Prognostic prediction in schizophrenia using the Anderson-Gill time-dependent Cox model: a cohort study in a less-developed region of Southwest China

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月30日

查看全文
  • Weiqi Tang,
  • Mingsong Li,
  • Zhongxin Hou,
  • Ensheng Pu,
  • Nan Cheng,
  • Li Deng,
  • Shifan He,
  • Dejue Men,
  • Jinglin Huang,
  • Yi Li,
  • Qian Wang,
  • Jianzhong Yin,
  • Qiong Meng,
  • Teng Zhang

摘要

Background

Schizophrenia represents a significant global public health burden, with considerable impact on morbidity and mortality. This study sought to construct a prognostic model to estimate mortality risk in individuals diagnosed with schizophrenia, employing a Cox proportional hazards framework with time-dependent variables.

Methods

Participants were recruited from a cohort of individuals with schizophrenia in a less-developed area of Southwest China and followed for a median of 6.4 years. A time-dependent Cox model based on the Anderson-Gill method was established to predict survival, followed by comprehensive model evaluation and graphical presentation.

Results

A total of 1,009 participants were enrolled, with 86 deaths recorded during follow-up. The model identified female sex (HR = 0.599, 95%CI= [0.387–0.928]) and hospitalization (HR = 0.499, 95%CI= [0.251–0.992]) as protective factors. Higher social function scores were associated with reduced mortality risk (HR = 0.889, 95%CI= [0.793–0.995]). Conversely, poor sleep quality (HR = 3.824, 95%CI= [1.482–9.867]) and abnormal electrocardiographic findings (HR = 2.185, 95%CI= [1.236–3.862]) increased mortality risk. Each additional year of age at initial diagnosis corresponded to an elevated risk of death (HR = 1.044, 95%CI= [1.027–1.060]). Model performance was robust, with a concordance index (C-index) of 0.820. Time-dependent AUC values at 1, 3, and 5 years were 0.814, 0.848, and 0.831, respectively.

Conclusions

Mortality risk among individuals with schizophrenia was significantly associated with sex, hospitalization status, social functioning, age at diagnosis, sleep quality, and electrocardiographic abnormalities. The model demonstrated strong predictive accuracy, and the accompanying nomogram together with the Excel-based scoring system offers a practical tool for rapid clinical risk assessment, supporting timely implementation of targeted interventions to improve patient outcomes.

Clinical trial number

Not applicable.