Clinical prediction models for COVID-19: development, international validation and use

For good care of COVID-19 patients, it is crucial to be able to estimate the expected course of their disease. For example: What is the probability that a patient will require intensive care (ICU)? What is the chance of death? Research In this study, a prediction model (https://mdmerasmusmc.shinyapps.io/COPE/ ) has been developed that provides answers to these questions upon admission to hospital. Data from almost 6,000 patients from four Dutch hospitals were used for this. Six easily measurable predictive factors have been used in the models, so that the model is easy to use. The model provided accurate predictions for more than 3,000 Dutch and 2,000 American patients in the second wave of the pandemic. Results Interviews with healthcare providers, patients and relatives showed that they see the value of these types of models when making decisions about optimal care for the patient. The models must then be available electronically and the predictions must be clearly communicated. View the scientific publication with the results of this research.