A mixed-methods evaluation of urgent care delivered through telephone based digital triage

Problem

England’s urgent care is delivered via a two-step triage model, where initial (primary) triage is conducted by a non-clinician, this contrasts with other countries which do not widely use non-clinician led triage. In both models, software based digital triage is widely used by call takers to support the provision of referral and/or self-care advice, based on the patient’s symptoms. Despite wide adoption of digital triage, there is limited evaluation of patterns of use, triage outcomes, and patient experience, and patients’ subsequent use of healthcare following triage. Understanding these outcomes in the context of two-step triage may help to identify where the delivery can be improved which is particularly important, given the pressures faced by urgent and emergency care.

Approach

A mixed methods study including an analysis of routine data from four urgent care providers in England to evaluate patterns of triage outcomes, including clincians’ overriding of: 1) primary triage outcomes and 2) digitally recommended triage outcomes. A qualitative study using semi-structured interviews and thematic analysis was used to explore patients’ experiences. A follow-on study of patients ED attendance and hospitalisation following two-step triage using Hospital Episodes Statistics (HES) data is planned.

Findings

Non-clinician triage was found to be risk averse, however, in calls about certain symptoms clinical risk appeared to be underestimated. There was substantial variation between clinicians in how likely they were to override urgency levels from the urgency assigned by the non-clinician, as well from the digital recommendation. Complexity in two-step triage, and variation in call takers conduct of triage was evident in patients’ experiences, additionally the patient’s confidence and knowledge were seen to influence the triage outcome urgency. Additionally, it is hoped that preliminary findings relating to patients subsequent use of emergency departments and hospitalisation following two-step triage will be presented.

Consequences

This research has identified potential clinical risk and the highlighted the importance of clinician triage within the two-step model. The very high variation in how clincians use digital triage suggests inconsistency in care provision; further research is required to better understand why this occurs to improve the safety and consistency of care. Service providers should focus monitoring, auditing, and training on key areas of risk identified. Exploring patients service use following triage is additionally expected to better understand how patients go on to use the health system after using telephone based digital triage services in urgent care.