FLU-CATs : Evaluation and refinement of pandemic influenza community assessment tools
Real time refinement and validation of criteria and tools used in primary care to aid hospital referral decisions for patients of all ages in the event of surge during an influenza pandemic.
Sponsor: University of Liverpool
CI: Prof Malcolm Semple
Approval Date: 20 March 2020
When a pandemic occurs, health care capacity both in the community and hospitals can be overwhelmed. When this happens, doctors need to make difficult decisions about who should be admitted to hospital and who can safely be allowed to stay at home. This process is called triage. Triage tools should help doctors identify which people are most likely to benefit from treatments only available in hospital and which people can safely be managed at home. This study will develop processes that will test how parts of a GP's questions and assessment of children and adults with flu like illness can predict: who can safely be kept at home; who needs hospital admission; who needs high dependency or intensive care; and who are most likely to die. The study uses the GP's routine electronic record to capture routine healthcare information and links to the hospital record if the patient is admitted to hospital with COVID-19. These records will be looked at by researchers and will be anonymous (will not identify the patient). The study will develop technology that allows records from about 600 GP surgeries across the UK to be studied automatically every week. The technology is set up in advance of a pandemic by working closely with the GPs that will use these tools. It takes about six months to develop the information technology processes and six months to test the processes. In the event of a pandemic, the processes that have been developed will be used to refine triage tools and check that they are "fit-for-purpose" in readiness for use if needed during a severe pandemic.
Design: A prospective analysis, linking criteria in a GP's assessment of patients presenting with influenza like illness, to immediate management decisions and patient outcomes. Objective: Assessment, refinement and validation of triage tools to guide GP referral of patients with influenza like illness during a pandemic in readiness for use should widespread illness exceed health care capacity (surge). Method: GPs participating in the Clinical Practice Research Datalink (CPRD) will record their assessment and management of patients with influenza like illness. We will explore linking CPRD to the HPA supra-regional microbiology network database to provide linked-anonymised results on microbiological investigations for respiratory pathogens. Analysis: Weekly cumulative analyses are planned using CPRD data. Three monthly analyses are planned using linked HES data as a gold standard. Univariate and multivariate analyses using unconditional logistic regression will be used to investigate the association between triage criteria threshold values and primary outcomes (hospital admission, and death) and secondary outcomes (length of stay and augmented levels of care (high dependency / intensive care)). The threshold values of triage criteria will be refined by comparing the receiver operator characteristics at various thresholds of abnormality (e.g. respiratory rate >30, >35, >40). The discriminatory value of existing and refined triage tools will be compared using logistic regression. Triage tools will be compared for their ability to predict outcomes using Area Under the Receiver Operating Characteristic Curve (AUROC) comparisons. The sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) will be calculated for triage tools using different score thresholds. Analysis will be reset between pandemic waves.
Study URL:Back to listing