Factors associated with persistent high use of emergency departments: a nationwide retrospective cohort study.

OBJECTIVES

To identify patient and attendance characteristics that predict persistent high use of emergency departments (EDs) in England and to quantify the relationship between these factors and the duration an individual remains a high user.

DESIGN

Nationwide retrospective cohort study using a Cox proportional hazards model with random effects.

SETTING

EDs across England, using linked national healthcare datasets including hospital admissions, General Practice (GP) prescribing data and mortality records from January 2016 to December 2019.

PARTICIPANTS

The total national dataset comprised approximately 150 million ED attendances by 30 million adult patients (aged>18). From this, a random sample of 5 million individuals was analysed. In the 2017 calendar year, 1 766 208 individuals attended the ED, of which 91 034 (5.2%) met the criteria for high use (defined as five or more attendances within a 12-month period).

PRIMARY AND SECONDARY OUTCOME MEASURES: The primary outcome was the length of time an individual remained in the 'high user' status. Predictors were measured using HRs to determine the likelihood of transitioning out of high-user status (where an HR<1.00 indicates a longer duration of high use).

RESULTS

Several factors significantly predicted prolonged high-user status. Strong predictors (expressed as HRs) included: use of hypnotics and anxiolytics (HR=0.64), antipsychotics (HR=0.67), opioid analgesics (HR=0.73) and non-opioid analgesics (HR=0.74), being a 'burst' attender (HR=0.55), social deprivation (HR=0.81), leaving the department after refusing treatment (HR=0.77), leaving before being treated (HR=0.86) and having 'nothing abnormal detected' during the visit (HR=0.88).

CONCLUSIONS

Persistent high ED use is strongly associated with prescriptions for analgesics and mental health medications, as well as specific attendance behaviours like 'burst' patterns and leaving without treatment. These findings suggest that healthcare providers should prioritise proactive care plans, enhanced mental health support and specialised pain management strategies to intervene before patients become long-term frequent users. Further research into the efficacy of these specific interventions for 'burst' attenders is warranted.