A prognostic model such as the ACS NSQIP Surgical Risk Calculator predicts the likelihood of early mortality or significant complications after surgery ( 5).
For example, the TREAT model (Thoracic Research Evaluation And Treatment model) estimates the risk of a lung nodule being cancer using information most likely available to evaluating surgeons ( 4). Risk prediction models use patient characteristics to estimate the probability that a certain outcome is present or will occur within a defined time period ( 3). Support of human cognition by allowing models to inform decision-making is a scalable way to manage growing data volumes and information complexity ( 2). In addition, information generation in health care is growing very quickly and outstripping the capacity of human cognition to adequately manage. Conversely, biases in the way the data are collected or filtered for use by the model can introduce other types of biases, and so the choice of underlying data and cohort selection are paramount. While not a substitute for clinical experience, they can provide objective data about an individual’s disease risk and avoid some common biases observed in clinical decision making ( 1). Prediction models are designed to assist healthcare professionals and patients with decisions about the use of diagnostic testing, starting or stopping treatments, or making lifestyle changes ( 1).
0 kommentar(er)
