A novel visual dynamic nomogram to online predict the risk of unfavorable outcome in elderly aSAH patients after endovascular coiling: A retrospective study

Wei Lu, Yulan Tong, Cheng Zhang, Xiang Lan, Liang Xiang, Chen Chen, LeHeng Guo, YaJie Shan, Xuemei Li, Zheng Zhao, Xiding Pan, Zhihong Zhao, Jianjun ZouFrontiers in Neuroscience 2023 · DOI

This study introduces the first web-based, dynamic nomogram designed to predict 6-month recovery outcomes for elderly patients suffering from aneurysmal subarachnoid hemorrhage (aSAH) who have undergone endovascular coiling. By distilling complex clinical data into four key predictors, the researchers provide a tool that allows clinicians to move beyond "gut feeling" toward personalized, data-driven risk stratification.

TL;DR

Researchers have developed a highly accurate (AUC 0.882) online tool to predict whether elderly patients will have an "unfavorable outcome" six months after brain aneurysm treatment. By using just four clinical variables—age, ventilator use, white blood cell count, and cerebral infarction—doctors can now identify high-risk patients at the bedside and tailor their care accordingly.

What Problem Does This Paper Address?

Aneurysmal subarachnoid hemorrhage (aSAH)—a life-threatening type of stroke caused by a ruptured brain aneurysm—is particularly devastating for the elderly. While endovascular coiling (a minimally invasive procedure to "plug" the aneurysm) is often preferred over open surgery, roughly 18% to 49% of these patients still face poor long-term outcomes.

Existing prognostic tools, like the Hunt-Hess or Fisher scales, are often criticized for being too subjective or for "dichotomizing" data (e.g., treating age 64 and 66 as fundamentally different categories). Furthermore, most models are built for the general population, not specifically for the unique physiological challenges of elderly patients. There is a critical need for a precise, user-friendly tool that can help clinicians and families make informed decisions about post-operative care.

Methods

The researchers conducted a retrospective study of 209 elderly patients (age ≥ 65) treated between 2016 and 2021. They defined an "unfavorable outcome" as a score of 3 or higher on the modified Rankin Scale (mRS) at the six-month mark.

To build the model, they started with 31 potential variables and used forward stepwise regression to identify the most significant independent predictors. Unlike traditional static nomograms (which are essentially paper charts), the authors developed a dynamic nomogram—a web-based interface built using the R "Shiny" package.

The model's performance was rigorously tested using:

  • Area Under the Curve (AUC): To measure discriminative power.
  • Hosmer–Lemeshow test: To check "calibration" (how well predicted risks match actual outcomes).
  • Decision Curve Analysis (DCA) and Clinical Impact Curves (CIC): To determine the "net benefit" of using the tool in a real-world clinical setting.

Datasets

The study utilized a single-center dataset from the People’s Hospital of Hunan Province, China.

  • Cohort size: 209 patients.
  • Demographics: Median age of 70; 36.4% experienced unfavorable outcomes.
  • Inclusion: Patients ≥ 65 years old with aSAH treated specifically via endovascular coiling.

Key Insights and Ideas

  • The Power of Four: Despite the complexity of brain bleeds, only four variables were needed to achieve high predictive accuracy: Age, Ventilator use, White Blood Cell (WBC) count, and Complicated Cerebral Infarction (CCI).
  • Inflammation as a Signal: The inclusion of WBC count highlights the critical role of the systemic inflammatory response in aSAH. A higher WBC count at admission was a strong indicator of poor recovery, likely reflecting the severity of the initial brain injury.
  • Ventilator Use as a Proxy: Needing a ventilator wasn't just a respiratory metric; it served as a powerful proxy for overall neurological depression and physical frailty.
  • Dynamic vs. Static: By hosting the model online, the researchers removed the "manual calculation" barrier that often prevents paper-based nomograms from being used in busy ICUs.

Key Results

  • High Accuracy: The nomogram achieved an AUC of 0.882 (95% CI: 0.835–0.929), outperforming many generalized models.
  • Strong Reliability: The Hosmer–Lemeshow test ($p = 0.3717$) indicated excellent calibration, meaning the model's predictions were highly consistent with observed results.
  • Clinical Utility: DCA showed that using the nomogram to guide treatment decisions provided a significant "net benefit" across almost all risk thresholds (5% to 95%).
  • Accessibility: The authors provided a live web link and a QR code, allowing any clinician with a smartphone to calculate a patient's risk in seconds.

Limitations and Future Work

The authors acknowledge several hurdles before widespread adoption:

  1. Single-Center Bias: The data comes from one hospital in China. To prove the model works globally, it needs external validation using datasets from different regions and ethnicities.
  2. Sample Size: While 209 patients provided statistically significant results, larger prospective studies would strengthen the model's robustness.
  3. Retrospective Design: Because the study looked at past data, it may be subject to selection bias. A prospective trial (tracking patients in real-time) is the logical next step to confirm these findings.

Summary updated

Abstract

This study is the first visual dynamic online nomogram that accurately predicts the risk of 6-month unfavorable outcome in elderly aSAH patients after endovascular coiling. Clinicians can effectively improve interventions by taking targeted interventions based on the scores of different items on the nomogram for each variable.
BibTeX
@inproceedings{lu2023novel,
  author = {Wei Lu and Yulan Tong and Cheng Zhang and Xiang Lan and Liang Xiang and Chen Chen and LeHeng Guo and YaJie Shan and Xuemei Li and Zheng Zhao and Xiding Pan and Zhihong Zhao and Jianjun Zou and ZhiHong Zhao and JianJun Zou},
  title = {{A novel visual dynamic nomogram to online predict the risk of unfavorable outcome in elderly aSAH patients after endovascular coiling: A retrospective study}},
  booktitle = {Frontiers in Neuroscience},
  year = {2023},
  doi = {10.3389/fnins.2022.1037895},
}
Figure