Characterization of prolonged hospital stays on a Medical Oncology Department in Portugal

Prolonged hospital stays represent a significant challenge in oncological care, consuming resources and potentially reflecting systemic inefficiencies in discharge planning and palliative care access. This study aims to characterize prolonged hospitalizations and identify independent predictors of prolonged stay through comparison with non-prolonged hospitalizations during the same period. This comparative cohort study includes all hospitalization episodes with discharge dates between January 1st and June 30th, 2023. Prolonged hospitalization was defined as hospital length of stay (LOS) ≥ 16 days (the definition of “prolonged hospitalization” was tailored to the specificities of the department enrolled - mean + 1 standard deviation). The non-prolonged group (LOS < 16 days) served as the control group. Univariate comparisons used Mann-Whitney U, chi-square, and Fisher’s exact tests. Independent predictors were identified by binary logistic regression. Of 434 eligible episodes, 138 (31.8%) were prolonged (mean LOS 34.5 ± 21.6 days) and 296 (68.2%) were non-prolonged (mean LOS 7.6 ± 3.5 days). Independent predictors of prolonged hospitalization included intra-hospital department transfer (OR 7.88, 95% CI 1.97–31.6, p = 0.004), Palliative Care Unit (PCU) referral (OR 5.33, 95% CI 1.66–17.1, p = 0.005), social services involvement (OR 4.71, 95% CI 2.47–9.0, p < 0.001), and family meetings (OR 3.00, 95% CI 1.41–6.4, p = 0.004). In this cohort, prolonged hospitalization was most strongly associated with systemic and organizational factors—including social support deficits and palliative care access—rather than with markers of irreducible clinical complexity. These associations highlight potential targets for coordinated intervention in discharge planning and palliative care resource allocation, pending prospective validation.