Advice
Expert comments
Expert comments
Comments on this technology were invited from clinical experts working in the field and relevant patient organisations. The comments received are individual opinions and do not represent NICE's view.
All 4 experts were familiar with machine-learning algorithms for prediction or the principles of the use of clinical prediction tools in acutely unwell patients.
Level of innovation
All 4 experts thought that SYNE‑COV is novel. The innovative aspects of the technology were using data collected across time points to predict potential risks, short calculation time, and using artificial intelligence for a real-time risk estimate. One expert added that using machine learning to refine the prediction model was innovative.
Potential patient impact
The potential benefits identified by 4 experts included improved prediction of COVID‑19 outcomes and earlier interventions that may lead to improved patient care. One expert thought collating provision of information about patients' prognosis in their hospital stay would help shared decision making for treatment at an early stage. One expert noted that the clinical algorithm has been used in the research setting as a part of the regulatory process and has not yet been used in clinical practice. Therefore, there is uncertainty around its potential benefits to clinical practice. The company noted that the technology has been deployed in the NHS and is starting to be used in live prediction at Chelsea and Westminster hospital. Another expert said that the technology was developed based on a limited patient data set from early in the pandemic. This means it is unclear how the model would perform its prediction as the pandemic evolves (that is, the emergent variants of SARS‑CoV‑2).
Potential system impact
The main system benefits identified by experts were improving the management of COVID‑19. This could be resource releasing by shortening people's hospital stays and reducing the need for expensive care when patients are admitted to the intensive care unit. Two experts thought that earlier identification of people at risk of deterioration and needing mechanical ventilation could potentially allow an earlier transfer to a critical-care setting or avoid escalating their care in hospitals. There might be a significant cost saving. Two experts thought it was too early to say what the system benefits would be because of lack of evidence. The experts said that the software was designed specifically for COVID‑19, and it could change the current pathway for managing COVID‑19 if evidence supports its benefits in clinical practice.
Three experts thought using SYNE‑COV would need changes in hospital IT systems to enable clinicians to see the prediction scores in hospitals. The experts agreed that training is needed to use the technology. Overreliance on the SYNE‑COV results was identified as a main risk, which may lead to inappropriate clinical management. Other potential risks identified by the experts including failure to identify people at a high risk of deterioration, misinterpretation of the SYNE‑COV results, delayed or inaccuracy of data and the technology being used inexperienced or unqualified clinicians.
General comments
The experts said that currently NEWS scores are standard care to predict deterioration and clinical observation. There are some scoring systems that are used in intensive care units to estimate the severity of diseases and risk of death such as APACHE II SOFA. One expert noted that SYNE‑COV is not intended for use in children. The experts agreed that evidence is important to understand SYNE‑COV's clinical and cost effectiveness in the NHS.