top of page

COMPLETED: Predicting Patient Deterioration Risks in COMMunities


Principle Investigator – Professor Michael Boniface, University of Southampton

Co-Investigator - Dr Francis Chmiel and Dr Dan Burns, University of Southampton

Clinical Lead – Dr Matt Inada-Kim, Hampshire Hospitals NHS Foundation Trust


What did we learn?

We found evidence for policies and practices supporting safe COVID-19 integrated care pathways in community settings for early identification of deterioration and timely escalation to hospital.

Research validating home oxygen saturations as a marker of clinical deterioration in patients with suspected COVID-19 informed national policies and was critical and rapid evidence necessary to support the COVIDoximetry@home ( under peer review at BMJ Open)

What difference did it make?

  • £500K NHSx RECOxCARE (Remote COVID oximetry care) project for COVID Virtual Wards in collaboration with HHFT, North Hants Primary Care Network, Mid Hants Primary Care Network, South Central Ambulance Service NHS Foundation Trust, WAHSN and University of Southampton IT Innovation Centre.

  • Scaling nationally as the COVID oximetry@home programme, the guidance for people at home to purchase a pulse oximeter published on the BBC News Website and broadcast on BBC News at 10 on Thursday 21st January 2021 (citing the ARC publication)

  • Clinical model was adopted by the World Health Organisation

  • Digital and data analytics approach for remote monitoring of patients in communities was presented by Inada-Kim and Boniface as part of the NHSx@home innovators engagement event 16th Sept 2021.

  • COVID oximetry@home received the overall winner of the prestigious for HSJ Patient Safety Award 2021

  • Service evaluation demonstrated improved patient outcomes

Boniface, M., Burns, D., Duckworth, C., Ahmed, M., Duruiheoma, F., Armitage, H., Ratcliffe, N., Duffy, J., O’Keeffe, C. and Inada-Kim, M., 2022. COVID-19 Oximetry@ home: evaluation of patient outcomes. BMJ Open Quality, 11(1), p.e001584.

What impact has it had on patient and the health system?

  • Remote monitoring of patients in communities is important for improved patient outcomes, hospital attendance/admissions avoidance and resource planning.

  • Remote monitoring is a key NHS transformation within the NHS@Home programme

  • Remote monitoring is being extended from COVID to all Acute Respiratory Infections, and is likely to be extended to other conditions in the future

What's next?

We are working with Hampshire Hospitals and NHS England to transfer the learning from COVID to more general approaches.

This research is being conducted in:

  • HDRUK rapid insight project (Jan22-Sep23) called PHILOSARIP “Predicting Hospital Length of Stay in Acute Respiratory Infections Patients”

  • Wessex ARC PARIEDA project which is tacking “Prediction of Acute Respiratory Infection outcomes prior to Emergency Department Attendance”


Covid Oximetry at Home Toolkit - via AHSN Wessex


Original Project outline

ARC Wessex is supporting research to explore COVID patient risks (deterioration, admission and readmission) in community settings working with Hampshire Hospitals NHS Foundation (HHFT) Trust who are co-leading the development of national pathways linking community, primary and secondary care.

According to leading acute care clinicians (Dr. Inada-Kim - HHFT) working at the forefront of UK’s COVID-19 emergency response and policymaking, two of most pronounced COVID-19 Unmet Medical Care Needs (UMCN) include:

  • UMCN-1) Risk prediction tools on triage and admission to emergency care: Evidence shows that early identification of physiological deterioration risks improves patient outcomes through timely and appropriate interventions, including escalations to higher levels of acute care through hospital admissions and intensive care[1].

  • UMCN-2) Rapid follow up of patients post discharge: There is little evidence to predict the occurrence of COVID19-related complications following discharge, particularly for vulnerable patients with multiple long term conditions at high risk of adverse complication events, and therefore rapid follow up and continuous monitoring of a patients recovery is needed to reduce risk of readmission to hospital.

In addition, consideration of population infection risks resulting from contact and transmission from infected individuals has demanded alternative care delivery models. During the initial phase of the pandemic patients freely made their way to GPs and hospitals increasing infection rates within the general population and the healthcare workforce, leading to policies aimed at reducing contact between infected patients and health care workers (HCWs)[2]. This has driven then need to reimagine care pathways that minimise physical interaction using virtual care (video conferencing, mobile symptom reporting/scores, real-time remote sensing, and surveillance) delivered through telemedicine solutions. Virtual care not only protects the population and HCWs during highly infectious periods of a pandemic but importantly offers significant benefits to patients who can now be treated longer in community settings reducing the number of admissions to hospital, the length of stay and mortality.

PPDRCOMM proposes to undertake research to develop predictive models for early warning detection arising from a COVID-19 infection, capable of running in residential settings such as care homes. Models will use near-patient observation data (e.g., temperature, respiration rate, and blood oxygen levels), patient demographics, and comorbidities from patients in the community who are in the early stages of a COVID-19 infection. The measurements will be collected with high frequency such that machine-learning algorithms will be able to report real-time risk scores of imminent deteriorations. Overall, this models will allow for real-time detection of deterioration earlier than currently possible with conventional techniques. This will help address the clinical need for pre-emptively stopping the severe deterioration of those with a seemingly mild case of COVID-19.

Read the Evaluation

Pre-Print evaluation paper

bottom of page