Do health care professionals’ perceptions help to measure the degree of overcrowding in the emergency department? A pilot study in an Italian University hospital

The NEDOCS score is the total of seven variables recorded at a single point in time and entered into a formula to generate the score [25]. These included, as fixed values, 25 ED beds and 528 hospital beds for adults and children, reflecting the installed capacity. The other values used for the score were total patients in the ED, total number of requested hospital admissions from the ED, number of respirators in the ED, longest admission time and waiting room time for the last patient called. The NEDOCS score was calculated every hour at real-time points during the study period and graded as follows: 0–20 not busy, 21–60 busy, 61–100 extremely busy but not overcrowded, 101–140 overcrowded, 141–180 severely overcrowded and 181–200 dangerously overcrowded [23]. The NEDOCS score was compared to the subjective evaluation of ED overcrowding (perceived ED crowding) by the ED staff (physicians, nurses and nursing aides/nursing assistants) Questionnaires were administered by a student/trainee and included the following variables: time at which the questionnaire was answered, role (physician, nurse, nursing assistant), and an index of perception of ED overcrowding rated on a 0–10 cm Visual Analogue Scale (VAS) [26]. The latter reflected the respondent’s level of “feeling rushed” or “pressured” at that particular point in time. The questionnaire was administered twice per shift for three daily shifts (8:00–14:00, 14:00–20:00, 20:00–8:00) during the period from February 19 to March 7, 2018. For comparison purposes with the results of the NEDOCS scale, each individual VAS score was multiplied by 20, as suggested by Weiss et al. [24]. The perception of overcrowding was graded as follows: 0–40 = not busy, 40–80 = busy, 80–120 = extremely busy, not overcrowded, 120–160 = overcrowded, 160–200 = severely overcrowded, and ≥ 200 dangerously overcrowded. The NEDOCS and perceived overcrowding scales were administered during the same time period. The staff participating in the study had over two years of experience working in the ED. Healthcare professionals absent for health reasons (illness, pregnancy, etc.) or on holiday during the study period were also excluded.

The results were expressed as the means and standard deviations (SD) for normally distributed data and as medians and interquartile ranges [1Q 3Q] for skewed data. Categorical variables were summarized using counts and percentages. Quantitative variables were compared among the three healthcare professional groups by an analysis of variance followed by the application of a multiple comparison test or non-parametric Kruskal–Wallis test. Bonferroni correction was used for multiple comparisons. Chi-squared or, when at least one expected frequency in a fourfold table was less than five, the Fisher's Exact test were performed to compare categorical variables (VAS and NEDOCS score graded into six classes, including “not busy”, “busy”, “extremely busy, not overcrowded”, “overcrowded”, “severely overcrowded”, “dangerously overcrowded”) from among groups (physicians, nurses, nursing assistant). Non-parametric Spearman coefficients (r_s) were calculated to measure the correlation between the two overcrowding measures. The concordance between the overcrowding classification derived from NEDOCS score and healthcare professionals’ crowding perceptions using the six-point VAS scale was evaluated in each ED using unweighted and weighted Cohen’s kappa coefficient and the corresponding 95% confidence interval (95% CI) [27, 28]. The weights assigned were calculated according to Cohen’s method using linear weights (κ_w) [29]. The κ-coefficient interpretation was performed on the basis of the study by Landis and Koch [30]. The following levels of agreement were considered appropriate for judging the extent of the agreement: κ less than 0.0, poor; 0.0 ≤ κ ≤0.2, slight; 0.21 ≤ κ ≤ 0.4, fair; 0.41 ≤ κ ≤ 0.6, moderate; 0.61 ≤ κ ≤ 0.80, substantial; and κ more than 0.80, almost perfect. Statistical analysis was performed using the statistical package for the social sciences (IBM Corp., IBM SPSS Statistics for Windows, Version 23.0 Armonk, New York USA). We also used the VassarStats website program for Statistical Computation to calculate concordance measurements (http://​vassarstats.​net/​kappa.​html). A two-sided value of p ≤ 0.05 was considered statistically significant.

The ED organizational model has been re-examined, and specific monitoring of emergency department functioning has been established, paying particular attention to access, flow and treatment times. Tools and the strategies used to manage ED crowding are described in the Hyperflow management plan (Table 4). The fluidity of the intake pathways for the different priority codes, green, yellow and red will be closely examined. This will be under the supervision of a Flow Manager (FM), an experienced nurse responsible for managing patients pre triage, periodically re-examining their clinical conditions in synergy with the clinical–healthcare team and, when necessary, redefining priority codes. Post-triage will also be a concern as the FM would constantly review and redefine priorities in accessing medical care. Furthermore, the FM can improve outflow by acting as the interface between the ED and the various specialized outpatient clinics, the hospital Bed Managers (BM) and maintaining a good working relationship with the spoke hospitals of the metropolitan area. BM facilitate correct and swift allocation of patients in the appropriate care settings by checking and controlling activation of the availability of suitable beds. Case Managers (CM) are involved in the planned discharge of patients from the ED, assessing the potential risks and difficulties that this process might incur [34].

The BM, the CM and the CDCA are in constant communication as they share the same objectives of simplifying and facilitating incoming patient flows (patient – in), ensuring continuity of care during hospital admission (patient – stay) and easing outflow, and maintaining and guaranteeing suitable continuity of care and assistance (patient – out). The communicative interfaces of the BM are the ED doctors, the CM, the CDCA and the Medical Department doctors.

The CM is a nurse possessing extensive knowledge of the services and organizational structures of the HUB and primary healthcare. The CM ensures appropriate and swift management of hospitalized patients with a view to their ongoing healthcare and facilitated discharge in a suitable primary care setting with proper information. The CM, through updated and precise awareness of the situation of patients held for recovery in the Medical Department, facilitates the use of tools and resources for the planning of discharges. The CM draws on the collaboration of clinical nurses suitably trained to become aware of patients with discharge difficulties and rapidly activates the social/healthcare services of primary care. In such cases, the professionals advise the CM swiftly (24–48 h from hospitalization or as soon as possible) whenever the following occur:

The Case Manager refers to the BRASS Index Scale1 for early assessment of a difficult discharge case. The Case Manager interfaces with the professionals in charge of the patient’s case to achieve a planned discharge and simultaneously observe caregivers’ needs for therapeutic education in relation to family members’ conditions and the need to activate aids or devices at home. In the case of necessity, the Case Manager supports activation through the formal structured channels of the CDCA of suitable multidimensional assessments. The Case Manager maintains an ongoing and structured interface with the Bed Manager, informing the latter of the state of discharges from the hospital departments.