The University of Pittsburgh used automated machine learning to develop a model that predicts surgical complications.
A team at the University of Pittsburgh used machine learning to create a model that predicts a patient’s likelihood of developing postoperative problems. The scientists reported that the program accurately detected risk variables in more than 60,000 of his surgeries, and he published the results in JAMA Surgery.
Automated machine learning predicts the risk of patients encountering complications such as readmissions, surgical site infections, and adverse transfusion reactions. Models with over 300 variables. In addition to the patient’s age, gender and medical history, information about the operation and the hospital where the operation will be performed is also taken into account.
According to the study authors, this kind of predictive tool could benefit medicine in several ways. First, it could help medical professionals determine which patients are most likely to have problems and need more in-depth monitoring and follow-up care. Additionally, hospitals may be able to identify problem-prone areas and take steps to enhance care there.
This model was created using information from operations at the University of Pittsburgh Medical Center. Still, the researchers argue that it could be adapted to work at other institutions as well. However, they caution that more research is needed to improve and validate the model before it is widely used.
Machine learning is being used more and more in healthcare, with researchers around the world creating useful models for everything from image analysis to predicting patient outcomes. Some experts believe this will be an important tool for increasing the effectiveness and efficiency of healthcare, especially as the amount of data available increases.
However, machine learning models have certain limitations. For example, if the data generated is not representative of the majority of the population, the data can be biased and produce conclusions that are difficult to understand and act on.
The University of Pittsburgh model researchers say they are optimistic about the model’s potential significance, despite its inherent difficulties. In their study, they said, “We believe that the introduction of this model into the clinical setting will improve patient outcomes, increase hospital efficiency, and improve resource allocation.” .
