Frequent User in Pronto Soccorso e Bed Blocker nei reparti di degenza: analisi del profilo e strategie preventive sul territorio attraverso le nuove piattaforme digitali e le COT distrettuali

Background In recent decades, the progressive aging of the population and the higher levels of urbanisation and pollution are increasing the number of patients suffering from chronic diseases, which has a significant impact on the use of health care services, in particular Emergency Departments (EDs). In Italy, despite significant improvements in the emergency care system, the cuts in numbers of hospital beds over the last 30 years and the increase in numbers of fragile patients (adults and children) have led to a rise in ED attendances. Published evidence from many countries shows that frequent ED use increases the risk of adverse effects such as hospitalisation, functional decline, and complications related to treatment and procedures. A significant proportion of hospital attendances are inappropriate and the respective ailments could be managed by non-acute health care services outside the hospital setting; such inappropriate attendances, therefore, drive up costs and increase inefficiency. The causes of frequent ED use are multifactorial: although many patients have chronic medical problems, these are often combined with marked psychosocial morbidity. Against this backdrop, the Covid-19 pandemic caused a significant fall in overall ED utilisation, with reduced volumes of up to 50% in some countries. The largest proportional reduction in ED users was in preventable ED attendance, including accident and traumatic injuries, probably as a result of a reduction in motor vehicle travel and fewer work activities, but also to time-dependent illnesses such as stroke or cardiac complaints among the oldest age groups, possibly due to concerns about Covid-19 acquisition in hospital. In Italy and in other countries, the Covid-19 pandemic also influenced the usage of Emergency Medical Services (EMSs). Despite a copious international literature on the frequent use of EDs, there is no single definition of a Frequent User (FU). The choice of threshold values is often subjective, generally based on previous literature or the distribution of ED attendances in a given period. More often a FU is considered a user with ≥4 or 5 attendances at the hospital, with both physical and mental issues, but various alternatives, from 2 to 12 attendances per year or six-month period, have also been chosen. Similar studies have also been conducted in Italy and the Netherlands showing that FUs represent a small percentage of hospital attendances, but nevertheless form a high proportion of the total ED costs. Among children, available studies conducted in Italy report that non-urgent attendances account for 27.6% of ED users and 58.2% of total paediatric attendance episodes. Non-urgent attendances have been negatively associated with crowding and costs, causing longer waiting times and greater dissatisfaction among both parents and health workers. The Italian National Health Service (NHS) is structured on three levels: the first includes the Central Government and the Ministry of Health, the second comprises the twenty Regional Governments and the third consists of the Local Health Authorities (LHAs) together with independent hospitals. The NHS is primarily funded through public taxation and is guided by the principles of universal coverage, solidarity, and human dignity. Each LHA includes at least one non-independent hospital and one or more Local Health Districts (LHDs), which provide primary care services (vaccination and screening, specialist consultations, family planning counselling, home care) and coordinates General Practitioners (GPs) and Primary Care Paediatricians (PCPs). Primary care physicians may work individually or in operational and multidisciplinary associations to ensure full ac-cess to care, 24 hours a day, 7 days a week. The availability of beds in hospital for acute patients and outside hospital for post-acute assistance is a major issue for the Italian NHS, especially in metropolitan areas like Rome, the most populous municipality in Italy, where three large LHAs administer health care services. Various Italian studies have explored the problem by analysing some of the characteristics of the adults and children FU but there is still insufficient consideration in the literature of the following points: number of attendances, usage of EMS, level of urgency, and appropriateness of attendance, geographical distribution and social-clinical characteristics. Covid-19 diagnosis was also investigated for a possible influence on ED attendances. The first study of this project investigates ED attendances in Rome; it describes the characteristic of the FU population and defines a FU profile (for adults and children), which highlights the differences between FUs and non-FUs, and identifies factors linked to FU status and appropriate ED attendance. The second analysis was aimed to identify the clinical and social characteristics of FUs and to quantify and compare the variation in the probability of being FU attributable to GPs and LHDs. The third part of the study described an example of strategic interventions of FU case management in a pilot phase carried out between 2023 and 2024. Materials and methods Study design and data collection A retrospective cohort study was carried out during 2022 of the ED attendances in 2021 for adults and the ED attendances in 2022 for children. The LHA Roma 1 geographical area in Rome was chosen for the analysis as it is one of the most populous areas in Italy, containing 13 EDs (of 22 in the Rome metropolitan area), including 8 paediatric EDs, with an aging index (number of population aged >64 years per 100 individuals aged <14 years) of 192 (the Italian mean is 183.3). The study population consisted of all patients with residency in the LHA Roma 1 geographical area who were admitted to any of its 13 local EDs. Records of the attendances at two other EDs close to the LHA Roma 1 area were added to the total to include potential ED attendances of LHA Roma 1 residents outside the main metropolitan area. A digital platform was used to extract the ED data from the Lazio Region official data flows for emergency attendances. The data are pseudo-anonymised: although the ID code of each patient is represented by an encrypted string, it is still possible to connect health events attributable to the same individual. Using this pseudo-anonymised ID, subsequent attendances of the same individual in 2021 were counted and classified according to the number of attendances made. In the first analysis, records included information on the following variables for each patient: - Number of attendances: a FU is defined as having ≥4 attendances per year, according to the literature; - Demographic characteristics: age, gender; - Arrival mode: by EMS or not by EMS; - Triage code: in 2001, a 24-h nurse-led triage system was introduced by the Italian Ministry of Health to evaluate a patient’s level of urgency, with assessment resulting in the assignment of a priority code. Since 2019, a transition from colour codes to numerical codes (1, 2, 3, 4, 5) has gradually been introduced; - Appropriateness of attendance: according to a visiting physician evaluation, all patients who were admitted to a hospital ward, had refused admission to a hospital ward or died in ED, were considered as appropriate; - Diagnosis of Covid-19: defined by any positive swab during ED attendance. All variables are mandatory in each patient’s attendance record, so there were no missing data. Single-specialism EDs (ophthalmology and obstetrics) were excluded from the study as they could affect the appropriateness of the results. The analysis was performed separately between individuals <18 years old and adults. Among children demographic characteristics the gender was classified as binary (male or female) according to the LHA registry; age was classified according to the age stages of the National Institute of Child Health and Human Development: 'infancy' (birth to 12 months); “toddler” (13 to 24 months); “early childhood” (25 months to 5 years), “middle childhood” (6 to 12 years), and “adolescent” (12 to 15 years). In the second analysis for each patient in the cohort the following potential risk factors were assessed: gender, age, socioeconomic status (high, middle-high, medium, middle-low, low) and the presence of chronic or multiple-chronic conditions. The socioeconomic level was calculated at the census tract level, based on the methodology developed by Nicola Caranci et al. This index integrates multiple socio-economic indicators de-rived from national census data, including educational attainment, employment status, home ownership versus rental, household overcrowding, and family structure. It provides a composite measure of socio-economic disadvantage within small geographic areas. Among patients with multiple chronic conditions, high clinical complexity was defined as a five-year mortality risk higher than 10%, based on the number and type of chronic dis-eases. The cohort was derived from the Healthcare Emergency Information System, which collects all attendances to emergency services and patient data. The cohort was linked to the automated databases of Lazio Region residents who receive NHS assistance, thus allowing researchers to obtain information related to chronic or multiple chronic diseases, GP and LHD of each patient, and socioeconomic status based on the residence address. A deterministic record linkage procedure with anonymous identification codes was used to merge the data from different information systems. To preserve privacy, each individual identification code was subsequently and automatically deidentified, and the conversion table was deleted, leaving only fully anonymized data available to researchers. Regarding the geographical analysis, the administrative-territorial division of the LHA Roma 1 was used to examine the association between FU prevalence and urban settings, as previously described. Each of the six LHDs of LHA Roma 1 is divided into Geographical Units (GUs, in Italy called “Zone Urbanistiche”), as defined by the Municipality of Rome. This represent the smallest territorial unit for which population data are available in Italy and many other countries. The third analysis is a prospective cohort study conducted between January 2024 and January 2025. The FU selection criteria are the same of previous analyses. Initial patient contact was mediated by GPs who acted as facilitators. In this first phase, the number of patients recruited was opportunistic, based on the availability of participating GPs (participation in the project was voluntary). Information regarding FU health problems was also collected, allowing the attendances to be divided by main health issue. Only data regarding ED attendances made between 2023 and 2024 were collected; single-specialty attendances were not considered, and patients <16 years of age were excluded to avoid confounding the results. A descriptive analysis of the overall sample and a sub-analysis of the care group were performed, analyzing variables such as ED attendances and reasons of attendances, both as absolute numbers and as percentages. Due to the large sample size, it was not deemed useful to perform a univariate or multivariate analysis. Statistical analysis Microsoft® Excel® v.2016 MSO and STATA v. 17.0 were used for data analysis. The cumulative number of ED attendances was computed for each patient ID, and patients with ≥4 attendances were classified as FUs. Descriptive analysis for adults and children was performed on all variables recorded. Descriptive statistics, such as mean, SD, frequency and percentage, were used to describe the demographics and ED attendance characteristics of the sample. For inferential analysis, both for adults and children, given the large sample size, statistical significance was determined at a level of p=0.001. Welch’s t-test was used to test differences in mean age among FUs and non-FUs. Pearson’s χ2-test was used to investigate differences in categorical variables among FUs and non-FUs. Univariate analysis was performed for any potentially associated factor. Multivariable logistic regression was performed for all factors identified with the significance level set at p<0.001. The second analysis included a multilevel logistic model (patient < GP < District) to quantify the variability in FU behaviour attributable to LHDs and primary care physicians and to identify the role of social and clinical determinants (gender, age, socioeconomic level and chronic conditions). Age was considered as a continuous variable. The effect of individual variables was expressed as Odds Ratios (OR); variance components estimated by multilevel models were expressed as Median Odds Ratios (MORs). The MOR quantifies between-cluster variation by comparing two patients from two randomly chosen, different clusters. Consider two persons with the same covariates, chosen randomly from two different clusters. The MOR represents the median odds ratio between the person of higher propensity and the person of lower propensity. MOR values are always ≥1.00; if MOR = 1.00, there is no variation between clusters, while larger values indicate greater variation. MORs were estimated for both the “empty” model, which includes a random intercept only, and the full model, which includes all patient risk factors. Statistical analyses were conducted using SAS software (SAS Institute Inc., North Carolina). In the analysis for children, a four-level model was performed, with observations clustered as follows: attendance15 (MOR < 1 year old: 3.05 and MOR > 15 years old: 3.32), EMS usage (MOR EMS: 5.65). Pilot study Regarding the case management pilot study, 745 FUs were identified in the population and 62 patients agreed to participate in the project and be contacted by our staff. The FUs have an average of 5.6 attendances in 2023, while in 2024 it decreased to 2.6. The patients were divided in different groups based on the main clinical diagnosis. In 38.7%, the attendances were caused by cardiovascular problem, followed by respiratory and kidney/urinary tract problems, both at 16.1%. Psychiatric problems were the main reason for 7 patients (11.3%), while the remaining 17.7% of patients had an issue belonging to a different macro-category. Among them, 20 had 4 attendances in 2023, 21 patients had 5 attendances, 9 had >5 attendances in a year, 12 patients had more than 7 attendances. The highest number of attendances was 14; overall the average number of attendances was 5.7. However, in 2024 51 patients (82.2% of the sample), had <4 attendances, while the remaining 11 reported from a minimum of 4 to a maximum of 11 attendances, and the average was 2.3. Moving on to the univariate analysis, no significant differences were found in the reduction of attendances between 2023 and 2024, although the FU group decreased from the average 3.4 to 3. Discussion Adults FUs represented 2.7% of the overall population but were accounted for 11.3% of attendances in 2021, in line with the literature, most frequently in the 50-59 age group. The risk of being FU increased with the patient's clinical complexity and with low and medium-low socioeconomic status. Generally, patients with higher socioeconomic status were less likely to attend the ED. This is consistent with other studies, where the odds of avoidable hospital admissions were higher among patients with lower socioeconomic status compared with other groups. Prior studies suggest that patients with low socioeconomic level perceive ED assistance to be cheaper and more accessible than ambulatory care and are often more likely to use EDs for non-urgent conditions. Non-urgent triage codes were more frequent among FUs (from 8.1% in the FU≥4 group to 14.4% in the FU≥10 group), and mental disorders were common in a substantial proportion of FUs (to 15.9% in the FU≥10 group). Psychomotor agitation and social issues were important diagnoses associated with FUs, but the results of “symptoms, signs and ill-defined conditions” and “external causes of injury and supplemental classification” diagnosis groups, as well as the main issues on admission, such as fever, chest pain or dyspnea, may have been influenced by the COVID-19 pandemic. This may also have affected time-dependent conditions such as stroke or cardiac complaints, possibly due to concerns about acquiring COVID-19 in hospital. However, all patients included in the analysis were equally exposed to the pandemic waves, and no systematic differences were expected between subgroups that could have introduced bias. Furthermore, other studies in the Lazio Region comparing ED attendances during the COVID-19 waves found a sharp reduction in ED attendances, except for pneumonia. The increase of physical and mental morbidities was associated with higher ED attendance rates, supporting previous evidence that individuals with chronic diseases and psychiatric disorders are more likely to attend EDs. This also applies to individuals with socioeconomic deprivation. Some authors have investigated the importance of social support for older adults in the ED, although in a systematic review there was no significant association be-tween ED attendance and social support. Regarding the geographical analysis, this study characterized variation using more granular geographic unit of analysis. This may be considered a first step toward developing specific public health strategies aimed at improving appropriate healthcare utilization for specific populations and communities, as implemented in Paris by the Île-de-France Regional Health Agency. Further researches are needed to analyze the association among FUs, population density, income and medical services or GP offices in the same area. Children In our study, pediatric FUs are more likely to be two years younger than non-FUs (7.7 and 5.3, p=0.001), while there are no other differences in demographic data analysis regarding access condition. These findings are in line with other Italian studies in which patients who visited the ED for non-urgent attendances are more likely to be younger and male. Our study shows that up to 62% of pediatric ED attendances were inappropriate for FUs and non-FUs. This result is consistent with other studies results, which reported up to 80% inappropriate attendances for FUs and non-FUs. These rates of inappropriate access to EDs are even much higher than in adults and indicate a great opportunity to improve the appropriateness of care and patient safety. Finally, in our study, a higher rate of appropriateness for access was observed in FUs compared to non-FUs in EDs, this is likely due to the higher prevalence of serious urgent codes (code 1) compared to non-FU patients. The nested logistic regression and variance analysis show both PCP/GP and District factors seem to play a small role in determining a greater number of ED attendances. Pilot study As recommended in the literature, the care process for FUs is based on a multi-professional approach, allowing different strategies and professional skills to work synergistically. For example, throughout their journey, patients can interact with various professionals who contribute in various capacities to their care, such as doctors, nurses, social workers, physiotherapists, etc. This organization is also supported by a case management strategy through the active call of patients by nursing staff (IFeC-Institutional Health and Care Professional) specifically trained to manage community and home care. While the results cannot be extended to the entire sample, the subpopulation characterized by neurological and psychiatric conditions may benefit more from a humanized care and case management process due to the IFeC support. To our knowledge, this pilot study is the first to have systematically described a FU managing strategy in the Italian context through the new IFeC skills, and in synergy between hospitals and non-hospital community facilities. These studies have some limitations. The retrospective design allowed investigation of the predictors of ED attendance only at a single point in time and exclusively among LHA residents, thereby excluding homeless individuals without residency, foreigners, and people formally resident in LHA Roma 1 but whose healthcare services were provided by other LHAs in Rome. Only the main diagnosis was considered, and inaccuracies in the clinical dataset may have led to underreporting of some morbidities. The potential influence of proximity to EDs and primary care services on ED attendance was not evaluated. The sample evaluated in this pilot phase is small and not generalizable. Conclusion Frequent ED use represents a major challenge for healthcare system management. Analysis of ED attendances and the socioeconomic and geographical characteristics of FUs highlights the need for new approaches to address key issues such as socioeconomic inequalities, improvement of housing and employment conditions, and structural factors including the strategic placement of primary care services and improved transportation. This project identified the potential risk factors predictive of disproportionate ED use in order to support policymakers in anticipating the needs of specific patient groups or categories. The FU phenomenon requires a systematic approach that utilizes all available professional skills. While it is not yet possible to draw any generalizable conclusions from the pilot study, the preliminary results appear encouraging, and lays the foundation for expanding the project to collect more data in order to have a more representative sample.