RESUMEN
OBJECTIVES: Italian regional health authorities annually negotiate the number of residency grants to be financed by the National government and the number and mix of supplementary grants to be funded by the regional budget. This study provides regional decision-makers with a requirement model to forecast the future demand of specialists at the regional level. METHODS: We have developed a system dynamics (SD) model that projects the evolution of the supply of medical specialists and three demand scenarios across the planning horizon (2030). Demand scenarios account for different drivers: demography, service utilization rates (ambulatory care and hospital discharges) and hospital beds. Based on the SD outputs (occupational and training gaps), a mixed integer programming (MIP) model computes potentially effective assignments of medical specialization grants for each year of the projection. RESULTS: To simulate the allocation of grants, we have compared how regional and national grants can be managed in order to reduce future gaps with respect to current training patterns. The allocation of 25 supplementary grants per year does not appear as effective in reducing expected occupational gaps as the re-modulation of all regional training vacancies.
Asunto(s)
Financiación Gubernamental , Necesidades y Demandas de Servicios de Salud , Internado y Residencia , Médicos/provisión & distribución , Regionalización , Especialización , Apoyo a la Formación Profesional , Predicción , Humanos , Internado y Residencia/economía , Italia , Modelos TeóricosRESUMEN
Objective: This study aims to develop and compare different models to predict the Length of Stay (LoS) and the Prolonged Length of Stay (PLoS) of inpatients admitted through the emergency department (ED) in general patient settings. This aim is not only to promote any specific model but rather to suggest a decision-supporting tool (i.e., a prediction framework). Methods: We analyzed a dataset of patients admitted through the ED to the "Sant"Orsola Malpighi University Hospital of Bologna, Italy, between January 1 and October 26, 2022. PLoS was defined as any hospitalization with LoS longer than 6 days. We deployed six classification algorithms for predicting PLoS: Random Forest (RF), Support Vector Machines (SVM), Gradient Boosting (GB), AdaBoost, K-Nearest Neighbors (KNN), and logistic regression (LoR). We evaluated the performance of these models with the Brier score, the area under the ROC curve (AUC), accuracy, sensitivity (recall), specificity, precision, and F1-score. We further developed eight regression models for LoS prediction: Linear Regression (LR), including the penalized linear models Least Absolute Shrinkage and Selection Operator (LASSO), Ridge and Elastic-net regression, Support vector regression, RF regression, KNN, and eXtreme Gradient Boosting (XGBoost) regression. The model performances were measured by their mean square error, mean absolute error, and mean relative error. The dataset was randomly split into a training set (70%) and a validation set (30%). Results: A total of 12,858 eligible patients were included in our study, of whom 60.88% had a PloS. The GB classifier best predicted PloS (accuracy 75%, AUC 75.4%, Brier score 0.181), followed by LoR classifier (accuracy 75%, AUC 75.2%, Brier score 0.182). These models also showed to be adequately calibrated. Ridge and XGBoost regressions best predicted LoS, with the smallest total prediction error. The overall prediction error is between 6 and 7 days, meaning there is a 6-7 day mean difference between actual and predicted LoS. Conclusion: Our results demonstrate the potential of machine learning-based methods to predict LoS and provide valuable insights into the risks behind prolonged hospitalizations. In addition to physicians' clinical expertise, the results of these models can be utilized as input to make informed decisions, such as predicting hospitalizations and enhancing the overall performance of a public healthcare system.
RESUMEN
This paper considers the organization and scheduling of a vaccination campaign during a pandemic emergency. We describe the decision process and introduce an optimization model, which showed a powerful multi-scenario tool for scheduling a campaign in detail within a dynamic and uncertain context. The solution of the model gave the decision maker the possibility to test different settings and have a configurable solution within few seconds, compared with the man-days of effort that would have required a manual schedule. Analysis of a real case study on COVID-19 vaccination campaign in northern Italy showed that the use of such optimized solution allowed to cover the target population within a much shorter time interval, compared to a manual approach.
RESUMEN
Esophagogastroduodenoscopy (EGD) appropriateness in Open-Access System (OAS) is a relevant issue. The Gastropack Access System (GAS) is a new system to access gastroenterological services, based on the partnership between Gastroenterologists and GPs. This study aims to evaluate if GAS is superior to OAS in terms of EGDS appropriateness. Secondarily, we evaluated the diagnostic yield of EGDS according to ASGE guidelines. The GAS was developed in an area of Bologna where General Practitioners (GPs) could decide to directly prescribe EGDS through OAS or referring to GAS, where EGDS can be scheduled after contact between GPs and specialists sharing a patient's clinical information. Between 2016 and 2019, 2179 cases (M:F = 861:1318, median age 61, IQR 47.72) were referred to GAS and 1467 patients (65%) had a prescription for EGDS; conversely, 874 EGDS were prescribed through OAS (M:F = 383:491; median age 58 yrs, IQR 45.68). Indication was appropriate in 92% in GAS (1312/1424) versus 71% in OAS (618/874), p < 0.001. The rate of clinically significant endoscopic findings (CSEF) was significantly higher in GAS (49% vs. 34.8%, p < 0.001). Adherence to ASGE guidelines was not related to CSEF; however, surveillance for pre-malignant conditions was independently related to CSEF. All neoplasm were observed in appropriate EGD. GAS is an innovative method showing extremely high rates of appropriateness. ASGE guidelines confirmed their validity for cancer detection, but their performance for the detection of other conditions needs to be refined.
RESUMEN
This paper aimed to analyse the spatio-temporal patterns of the diffusion of SARS-CoV-2, the virus causing coronavirus 2019 (COVID-19, in the city of Bologna, the capital and largest city of the Emilia-Romagna Region in northern Italy. The study took place from February 1st, 2020 to November 20th, 2021 and accounted for space, sociodemographic characteristics and health conditions of the resident population. A second goal was to derive a model for the level of risk of being infected by SARS-CoV-2 and to identify and measure the place-specific factors associated with the disease and its determinants. Spatial heterogeneity was tested by comparing global Poisson regression (GPR) and local geographically weighted Poisson regression (GWPR) models. The key findings were that different city areas were impacted differently during the first three epidemic waves. The area-to-area influence was estimated to exert its effect over an area with 4.7 km radius. Spatio-temporal heterogeneity patterns were found to be independent of the sociodemographic and the clinical characteristics of the resident population. Significant single-individual risk factors for detected SARS-CoV-2 infection cases were old age, hypertension, diabetes and co-morbidities. More specifically, in the global model, the average SARS-CoV-2 infection rate decreased 0.93-fold in the 21-65 years age group compared to the >65 years age group, whereas hypertension, diabetes, and any other co-morbidities (present vs absent), increased 1.28-, 1.39- and 1.15-fold, respectively. The local GWPR model had a better fit better than GPR. Due to the global geographical distribution of the pandemic, local estimates are essential for mitigating or strengthening security measures.
Asunto(s)
COVID-19 , Hipertensión , Humanos , Anciano , SARS-CoV-2 , COVID-19/epidemiología , Pandemias , Italia/epidemiologíaRESUMEN
Kidney Exchange Programs (KEP) are valuable tools to increase the options of living donor kidney transplantation for patients with end-stage kidney disease with an immunologically incompatible live donor. Maximising the benefits of a KEP requires an information system to manage data and to optimise transplants. The data input specifications of the systems that relate to key information on blood group and Human Leukocyte Antigen (HLA) types and HLA antibodies are crucial in order to maximise the number of identified matched pairs while minimising the risk of match failures due to unanticipated positive crossmatches. Based on a survey of eight national and one transnational kidney exchange program, we discuss data requirements for running a KEP. We note large variations in the data recorded by different KEPs, reflecting varying medical practices. Furthermore, we describe how the information system supports decision making throughout these kidney exchange programs.