Gestión de camas hospitalarias durantela pandemia en Navarra con el apoyo demétodos matemáticos de predicción / Management of hospital beds during the pandemic in Navarra with the support of mathematical prediction methods

Durante la pandemia por coronavirus, en Navarra se utilizaron modelos matemáticos depredicción para estimar las camas necesarias, convencionales y de críticos, para atender alos pacientes COVID-19. Las seis ondas pandémicas presentaron distinta incidencia en la población, ocasionandovariabilidad en los ingresos hospitalarios y en la ocupación hospitalaria. La respuesta a laenfermedad de los pacientes no fue constante en cada onda, por lo que, para la predicción decada una, se utilizaron los datos correspondientes de esa onda.El método de predicción constó de dos partes: una describió la entrada de pacientes alhospital y la otra su estancia dentro del mismo. El modelo requirió de la alimentación a tiempo real de los datos actualizados. Los resultados delos modelos de predicción fueron posteriormente volcados al sistema de información corporativotipo Business Intelligence. Esta información fue utilizada para planificar el recurso cama y lasnecesidades de profesionales asociadas a la atención de estos pacientes en el ámbito hospitalario.En la cuarta onda se realizó un análisis para cuantificar el grado de acierto de los modelospredictivos. Los modelos predijeron adecuadamente el pico, la meseta y el cambio detendencia, pero sobreestimaron los recursos necesarios para la atención de los pacientes enla parte descendente de la curva. El principal punto fuerte de la sistemática utilizada para la construcción de modelospredictivos fue proporcionar modelos en tiempo real con datos recogidos con precisión porlos sistemas de información que consiguieron un grado de acierto aceptable permitiendo unautilización inmediata.(AU)

Hospital preparedness during epidemics using simulation: the case of COVID-19.

This paper presents a discrete event simulation model to support decision-making for the short-term planning of hospital resource needs, especially Intensive Care Unit (ICU) beds, to cope with outbreaks, such as the COVID-19 pandemic. Given its purpose as a short-term forecasting tool, the simulation model requires an accurate representation of the current system state and high fidelity in mimicking the system dynamics from that state. The two main components of the simulation model are the stochastic modeling of patient admission and patient flow processes. The patient arrival process is modelled using a Gompertz growth model, which enables the representation of the exponential growth caused by the initial spread of the virus, followed by a period of maximum arrival rate and then a decreasing phase until the wave subsides. We conducted an empirical study concluding that the Gompertz model provides a better fit to pandemic-related data (positive cases and hospitalization numbers) and has superior prediction capacity than other sigmoid models based on Richards, Logistic, and Stannard functions. Patient flow modelling considers different pathways and dynamic length of stay estimation in several healthcare stages using patient-level data. We report on the application of the simulation model in two Autonomous Regions of Spain (Navarre and La Rioja) during the two COVID-19 waves experienced in 2020. The simulation model was employed on a daily basis to inform the regional logistic health care planning team, who programmed the ward and ICU beds based on the resulting predictions.

A mathematical model for simulating daily bed occupancy in an intensive care unit.

OBJECTIVES: To develop a mathematical model for simulating the daily bed occupancy in an intensive care unit. DESIGN: Data collection and retrospective analysis to develop a mathematical model for simulating daily bed occupancy in an intensive care unit. SETTING: We studied all admissions to the intensive care unit at the Hospital of Navarra over a 9-yr period. PATIENTS: Six-thousand three-hundred adult patients consecutively admitted to intensive care units at a tertiary care hospital. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: The large set of data collected comprises an arrivals file, a patient file, and a bed occupancy file. The arrival file records the number of patients admitted to the intensive care unit each day, by admission type, and by day of the week. The patient file contains records for all patients admitted to the intensive care unit during the study period: Admission type, admission and discharge dates, age, sex, Acute Physiology and Chronic Health Evaluation II score within the first 24 hrs, infections during hospitalization, and mortality. We used these two files to fit appropriate statistical models of arrival rates and length of stay by patient type. Based on this statistical analysis and the representation of the intensive care unit as a queuing problem, we built a simulation model. The bed occupancy file records the number of occupied beds at 4:00 PM each day. We used this file to validate the simulation model by testing the similarity of the real and simulated output data. The simulation model also includes bed management decisions related to patient discharge. RESULTS: We obtained a valid simulation model that reproduced on a computer the patient flow through the intensive care unit at the Hospital of Navarra. This computerized simulation model can be used to study the intensive care unit bed occupancy profile and can be used as a reliable sizing and capacity analysis tool. As an example, we present the problem of estimating the number of beds needed to meet an increase in patient arrivals at the intensive care unit because of different causes. CONCLUSIONS: It is possible to develop simulation models that can be used to predict future intensive care unit resource needs.