A new study is exploring the journey of hospital patients who have tested positive for COVID-19.
Researchers will analyse what staffing and resources were needed for patients’ care. This real world data will create a model to help hospitals plan care for patients,
The study will use de-identified data from all patients who have tested positive for COVID-19 at Guy’s and St Thomas’. The team estimate that this could be around 1,600 patients.
The model could allow clinicians to enter clinical data from a patient on admission to the hospital, and predict how long they are likely to stay in hospital, and what they may need while they are there. This would allow hospitals to plan a few days in advance to make sure the resources are there when needed.
The team are starting to process the clinical data. First they will strip out all data that could identify a patient, such as their name, NHS number or address. The remaining data will then form an ‘aggregated’ dataset which represents the patient population as a whole. The dataset will not contain information about any individual case. Researchers at Imperial College London will then analyse the aggregated data and create the model.
Dr Irum Kotadia is a Clinical Research Fellow working on the project. She said: “At the beginning of the pandemic, many of our research team were on COVID-19 wards. We noticed the challenge of predicting this new disease and having resources in place. We want to create a model that will predict what resources we will need, how many ventilators or members of staff we will require for the patients coming in, and for how long. This way we could predict any shortages and make sure the resources are available early to ensure patients receive the best possible care.”
“We hope that the model will be available in a year or so, so that we can use this for COVID-19 patients. We also hope to adapt the model beyond the pandemic, to meet the needs of other periods of high hospital demand, such as seasonal flu.”
The team have worked with patient and public involvement (PPI) groups to design of the study. The groups have also had an input on patient information about consent and confidentiality, ensuring it is easy to understand.