Drăgan Cristian Ovidiu, Mihai Laurențiu Stelian, Popescu Ana-Maria Camelia, Buligiu Ion, Mirescu Lucian, Militaru Daniel
Faculty of Economics and Business Administration, University of Craiova, 200585 Craiova, Dolj, Romania.
Healthcare (Basel). 2025 Jan 7;13(2):102. doi: 10.3390/healthcare13020102.
BACKGROUND/OBJECTIVES: Globally, healthcare systems face challenges in optimizing performance, particularly in the wake of the COVID-19 pandemic. This study focuses on the analysis and forecasting of key performance indicators (KPIs) for the County Emergency Clinical Hospital in Craiova, Romania. The study evaluates indicators such as average length of stay (ALoS), bed occupancy rate (BOR), number of cases (NC), case mix index (CMI), and average cost per hospitalization (ACH), providing insight into their dynamics and future trends.
We performed statistical analyses on quarterly data from 2010 to 2023, employing descriptive statistics and stationarity tests (e.g., Dickey-Fuller), using ARIMA models to forecast each KPI, ensuring model validation through tests for autocorrelation, heteroscedasticity, and stationarity. The model selection prioritized Akaike and Schwarz criteria for robustness.
The findings reveal that ALoS and BOR demonstrate seasonality and are influenced by colder months, and it is expected that the ALoS will stabilize to around five days by 2025. Moreover, we predict that the BOR will range between 46 and 52%, reflecting these seasonal variations. The NC forecasts indicate a post-pandemic recovery but to below pre-pandemic levels, and we project the CMI to stabilize at around 1.54, suggesting a return to consistent case complexity. The ACH showed significant growth, particularly in the fourth quarter, driven by inflation and seasonal costs, and it is projected to reach more than RON 3000 by 2025.
This study highlights the utility of ARIMA models in forecasting healthcare KPIs, enabling proactive resource planning and decision-making. The findings underscore the impact of seasonality and economic factors on hospital operations, offering valuable insights for improving efficiency and adapting to post-pandemic challenges.
背景/目标：在全球范围内，医疗保健系统在优化绩效方面面临挑战，尤其是在新冠疫情之后。本研究聚焦于罗马尼亚克拉约瓦县急诊临床医院关键绩效指标（KPI）的分析与预测。该研究评估了诸如平均住院时间（ALoS）、床位占用率（BOR）、病例数（NC）、病例组合指数（CMI）以及每次住院平均费用（ACH）等指标，深入了解它们的动态变化和未来趋势。
我们对2010年至2023年的季度数据进行了统计分析，采用描述性统计和平稳性检验（如迪基 - 富勒检验），使用自回归积分移动平均（ARIMA）模型预测每个KPI，并通过自相关、异方差和平稳性检验确保模型验证。模型选择优先考虑赤池信息准则和施瓦茨准则以确保稳健性。
研究结果表明，ALoS和BOR呈现季节性，且受较冷月份影响，预计到2025年ALoS将稳定在约5天左右。此外，我们预测BOR将在46%至52%之间波动，反映出这些季节性变化。NC预测显示疫情后有所恢复，但低于疫情前水平，我们预计CMI将稳定在约1.54左右，表明病例复杂性将恢复到一致水平。ACH呈现显著增长，特别是在第四季度，受通货膨胀和季节性成本推动，预计到2025年将超过3000罗马尼亚列伊。
本研究凸显了ARIMA模型在预测医疗保健KPI方面的实用性，有助于进行前瞻性资源规划和决策。研究结果强调了季节性和经济因素对医院运营的影响，为提高效率和应对疫情后挑战提供了有价值的见解。
