The study was conducted at the Emergency Department (ED) of Townsville University Hospital (TUH), Queensland, Australia. TUH is the tertiary referral centre for North Queensland servicing a catchment of ~ 700,000 people20. TUH had an annual ED census of 91,997 in 2020–2021. 21
Patient selection and data abstraction
This is a retrospective, single-centre study of patients who frequently presented to TUH ED between October 1st, 2016 - March 31st, 2017. Two investigators (VG1 and KT) independently determined patient eligibility for inclusion into FSS, Pos-FSS, or non-FSS groups. (Fig. 1), collated qualitative descriptions of stressors from patient charts and analysed data.
A consensus definition of frequent ED presenters is lacking although there have been attempts to operationalise the concept 22. For this study, we defined a frequent ED presenter as someone who presents 3 or more times in a calendar month to ED. This definition was used for operational convenience since the organisation where this study took place produces an administrative report each month identifying these individuals. The report is called the ED recidivist report (EDRR) and contains details relating to patient’s name, number of presentations, arrival date and time for each presentation, and discharge diagnosis. All the patients on the EDRR were stratified into one of the three groups (FSS, Pos FSS and non FSS, see Fig. 1) based on Stephenson and Prices’1 description of FSS. Frequent presenters with FSS included presentations ranging across non-cardiac chest pain, benign palpitations, nonspecific abdominal pain, non‐ulcer dyspepsia, physical symptoms of anxiety and depression, nonspecific symptoms (e.g. “funny turns”), symptoms with undiagnosed organic pathophysiology, no obvious pathology and non-specific presenting issues. Patients with an equivocal clinical diagnosis, such as cases with one or more presentations with FSS but an established underlying pathology that could result in similar symptoms (e.g. non-cardiac chest pain in patients with established ischaemic heart disease), were classified as possible FSS (pos-FSS) group. The presentation for these patients may have been due to FSS, and these patients were included in the study to enrich our understanding. If none of the presentations for a patient met the criteria for FFS, the patient was categorised as non-FFS group. Patients were also classified as non-FFS if they had been diagnosed as having severe mental disorder (e.g., Bipolar disorder, major depression, psychotic episode or schizophrenia), suicidal ideation or known drug seeking behaviour. The classification of cases as FSS, Pos FSS or Non-FSS was carried out by two clinical psychologists (VG1 and KT).
Since there was no published Australian literature regarding operationalizing FSS amongst ED presentations at the time of study, previous EDRRs were used to familiarize the two psychologists regarding FSS using re-presentation reports and patient charts. This helped create shared understanding of FSS as described in the patient selection algorithm (Fig. 1). The psychologists undertook chart reviews of 144 frequent presenters identified through one of the earlier EDRRs to assess for inter-rater reliability. Presentations were independently rated as being FSS or Non-FSS and inter-rater reliability was calculated using Pearson's co-efficient of correlation (r = 0.9).
Data Analyses
The data were analysed using Statistical Packages for Social Science (SPSS) version 26 and Microsoft Excel 10. Descriptive statistics were used to describe frequent ED presentations among those characterised as FSS, pos-FSS and non-FSS. Means with standard deviations were used for continuous variables and counts with proportions were used for categorical variables. Differences between frequent presenters categorised as FSS, Pos-FSS and non-FSS were assessed using independent chi-square tests or ANOVA, as appropriate.
Person—time at risk of ED presentation during the study period was calculated for each frequent presenter, and then separately for those categorised as FSS, Pos-FSS and non-FSS. Subsequently, the incidence rate of ED presentation was calculated for the three groups (number of ED presentations/person- time at risk). Rates are presented per 100-person days. The relative risk of ED presentation (with 95%CI) was calculated for those categorised as FSS compared to non-FSS. Relative risks were also calculated for pos-FSS compared with non-FSS.
Qualitative data regarding psychological distress indicators for those included in the study as FSS were descriptively analysed. Psychological distress indicators were defined as indicators for psychological distress noted by the treating ED clinician in the patient’s chart. For example, “ the patient appeared highly anxious” or “patient reports a recent break up in relationship”. These were extracted directly from patient charts as recorded by the treating ED clinician (usually an ED nurse or a doctor). Researchers (VG1 and KT) then thematically grouped into the following categories: Anxious (Anxious mood and high anxiety); depressed, financial stress, living /residential stressor; psychosocial family/relationship issues; PTSD and psychosocial distress (including distressed and tearful). The last category was a category of exclusion when the distress indicator could not be classified in any other category. Categories were not mutually exclusive and patients could have more than one psychological distress.