Camila Larrazabal Melgar is a PhD student at the University of the Balearic Islands and a research assistant at the Health Research Institute of the Balearic Islands (IdISBa) in Palma de Mallorca. Her research focuses on the performance of health care providers and systems.
Analysis of hospital length of stay (LOS) is a key metric for assessing the efficiency and performance of healthcare systems. While clinical factors are well-known drivers of LOS, emerging evidence highlights the significant influence of non-clinical, organisational factors in shaping hospital resource use - particularly in the context of acute conditions such as sepsis. In an era of ageing populations and rising health and hospital expenditure - Germany is one of the EU countries with the highest rates of inpatient care - addressing avoidable variability in LOS is not only a matter of operational improvement, but also of system sustainability.
As part of her doctoral research, a three-month academic stay at the Institute of Health Economics at Leibniz University in Germany will contribute to a broader understanding of how health system structures interact with clinical practice to shape patient outcomes and resource use. The analysis will draw on observational microdata from 2005 to 2015 covering all acute care providers in 16 German Länder.
The current research is inspired by previous research conducted in Spain, which examines the variability of LOS for acute sepsis in the Spanish National Health System (SNHS) using a comprehensive dataset of acute hospital discharges (2003-2015) across 17 autonomous communities. Using survival analysis and Cox proportional hazards regression, the study disentangles the effect of organisational proxies - including day of week, proximity to holidays, seasonality and hospital complexity - on the probability of discharge. The results show that LOS is not solely determined by patient characteristics such as age, gender or severity of comorbidity, but is significantly influenced by the institutional context in which care is provided. The implications are twofold. First, LOS variation may serve as a proxy for identifying inefficiencies related to organisational structures rather than patient needs. Second, these findings support policy interventions targeting hospital workflow and administrative design to improve care efficiency and optimise public resource allocation.
