While there are numerous models that predict risks for patients in hospital with COVID-19 using only the information collected when the patient arrives at hospital, there remain few reliable models that are able to update their predictions as new clinical information is collected.
New research, published in BMJ Open, by Martin Wiegand, Brian Tom, Robert Goudie and several clinicians and scientists at Addenbrooke’s Hospital, focused on developing a risk prediction model for patients admitted to hospital with COVID-19 that updates as new clinical information is collected during a hospital visit. We used de-identified data extracted from the hospital medical records at Addenbrooke’s Hospital, Cambridge, to develop our model. Addenbrooke’s is one of only six European hospitals to reach the highest digital maturity, meaning we have been able to use detailed hospital data. We included demographic data, observations, vital signs, laboratory tests, comorbidities and information about interventions and treatments.
We were able to develop a model founded on sound statistical underpinnings including: accounting properly for patients who leave hospital because they recovered or required transfer to another hospital for more specialised treatment; and acknowledging that e.g. extra blood tests may be requested by doctors for sicker patients.
The resulting prediction model has the potential to provide physicians with an assessment of a patient’s evolving prognosis throughout the course of active hospital treatment. An online calculator is available http://shiny.mrc-bsu.cam.ac.uk/apps/covid19mortalityrisk/. However, the model requires external validation to assess its performance in other hospitals and for recent variants of SARS-CoV-2.
Dr Robert Goudie, Senior Investigator Statistician at the MRC Biostatistics Unit, said:
This work demonstrates the potential for de-identified hospital medical records to improve our understanding of the short-term prognosis of patients in hospital with COVID-19. The framework developed as part of this work also has considerable potential to be adapted for other conditions beyond COVID-19.”
Read the full paper at https://doi.org/10.1136/bmjopen-2021-060026
Martin Wiegand, first author, was funded by NIHR Cambridge Biomedical Research Centre, to carry out this research. The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care.
This work uses data provided by patients and collected by the NHS as part of their care and support and would not have been possible without access to this data. The NIHR recognises and values the role of patient data, securely accessed and stored, both in underpinning and leading to improvements in research and care. www.nihr.ac.uk/patientdata