Development and Validation of a Risk Prediction Model of One-year Survive Rate in Vancomycin-Associated Acute Kidney Injury

In the intensive care units of hospitals around the world, doctors often turn to a powerful antibiotic called vancomycin to fight severe bacterial infections that other medicines cannot stop. While this drug is a vital tool for saving lives, it carries a hidden danger: it can sometimes damage the kidneys, a condition known as vancomycin-associated acute kidney injury. When the kidneys fail, patients face a much harder road to recovery, requiring longer hospital stays and more complex treatments. For years, researchers have known that certain factors, like a patient's age or the severity of their illness, make kidney injury more likely to happen. However, a critical question remained unanswered: once a patient has already developed this kidney injury while on the drug, what are the chances they will survive the next year? Until now, there has been no reliable way to predict this long-term outcome, leaving doctors to guess at the future of their most vulnerable patients.

To answer this difficult question, a team of researchers from hospitals in Henan Province and Zhengzhou University turned to a massive digital archive of medical records. They analyzed data from nearly 13,000 patients who had received vancomycin in an intensive care unit and subsequently developed kidney injury. By splitting this group into two sets—one to build a new prediction tool and another to test it—they looked for patterns in the data that could separate those who would survive from those who would not. The researchers examined a wide range of information, from basic details like age and body weight to complex measurements of how the body was functioning, such as blood pressure, heart rate, and the levels of various chemicals in the blood. They were looking for specific signs that appeared consistently in the patients who passed away within a year compared to those who lived.

The study identified eight specific factors that acted as strong warning signs for a lower chance of survival one year later. These included the patient's age, their body mass index, the level of urea nitrogen in their blood, their breathing rate, whether they had an active infection, the severity of their kidney injury, the presence of cancer, and the level of lactate in their blood. The researchers found that older patients, those with higher levels of kidney stress markers, those with active infections or cancer, and those with faster breathing rates faced a significantly higher risk of not surviving the year. In contrast, a higher body mass index appeared to be slightly protective, while lower levels of kidney stress markers were associated with better outcomes.

Using these eight factors, the team constructed a simple scoring system, which they presented as a visual chart known as a nomogram. This tool allows a doctor to take a patient's specific measurements, find the corresponding points for each factor on the chart, and add them up to get a total score. That total score translates directly into a predicted probability of surviving one year. To ensure this tool was accurate, the researchers tested it on a separate group of patients who had not been used to create the model. The results showed that the tool could correctly distinguish between survivors and non-survivors with a high degree of reliability, performing well in both the initial group and the test group. The model also showed that it was clinically useful, meaning it could help doctors make better decisions about treatment and care planning.

This work represents the first time a specific risk scoring system has been created to predict the one-year survival rate for patients who develop kidney injury while being treated with vancomycin. While the study was based on data from a single medical center and relied on past records, the findings offer a concrete new way to look at patient outcomes. By identifying the specific combination of factors that signal a poor prognosis, the model gives medical teams a clearer picture of the road ahead. It does not change the biology of the disease, but it provides a structured way to understand the risks, potentially helping doctors intervene earlier or manage expectations more effectively for patients facing this serious complication.