Telephone based digital triage in urgent care provision: A routine data analysis of patients’ service use and patterns of triage advice

Problem

Urgent care services that provide out of hours care often use digital triage. This involves a health care service staff member using a ‘digital triage tool’ to help refer the patient to an appropriate service to receive health care, based on the patient’s symptoms. In part digital triage helps to manage high demand within the healthcare system whilst improving patients’ access to care.

Despite wide adoption, there has been limited research into patterns of use and factors that influence the recommended urgency of advice that is generated within digital triage tools, particularly in relation to gender, ethnicity and level of deprivation.

Using large routine datasets from NHS care providers in England, this research aims to understand how patients use digital triage services. It will investigate characteristics of users, their presenting symptoms, and factors that influence the priority and urgency of advice. Datasets spanning before and after the emergence of Covid-19 will be analysed in order to explore changes in these outcomes since the start of the pandemic.

Telephone based care has been of increasing importance during the Covid-19 pandemic; digital triage is central to telephone based care and if optimised has potential to improve care for patients whilst reserving the need for face to face contact for when it is necessary.

Approach

This is an observational cross-sectional study that utilises routine data for approximately 200,000 patients from four English services, spanning April 2019 – October 2020. The dataset comprises anonymised call records from urgent care providers for patients who have been triaged. The data includes the following variables, which will be descriptively analysed: patient demographics (age, sex, level of deprivation, ethnicity), presenting symptoms and triage advice generated.

Subgroup analysis will be conducted to better understand factors that influence the priority and urgency of advice, including factors such as: age, sex, symptom type (for example chest pain or abdominal pain).

Changes over time, prior to and during the Covid-19 pandemic will be analysed using time series regression, building in appropriate ‘step terms’ to account for changes such as: pandemic control measures, local changes within services, and seasonal changes that may affect service use.

Findings

Early findings are expected in July 2021 and will relate to two key areas: 1) Service users’ characteristics and presenting symptoms before and after the start of the Covid-19 pandemic 2) Patterns of triage advice, including factors that influence the priority and urgency of advice given to patients.

Consequences

This study will make recommendations for services and policymakers based on findings relating to 1) patterns of service use, for example under-use amongst certain sub-groups and 2) findings relating to in-built bias within digital triage tools, which could highlight areas for improvement within digital triage tools.