Abstract
Background Prolonged hospitalization in patients with hematologic and collagen diseases presents significant challenges for patient recovery and healthcare resource utilization. Perioperative oral management (POM) has been suggested to reduce systemic complications; however, its impact on length of hospital stay remains unclear. Methods This retrospective case-control study analyzed 278 patients admitted to the Department of Hematology and Rheumatology at Kagoshima University Hospital between November 2009 and January 2025. Patients who received POM during hospitalization were compared with those who did not. Prolonged hospitalization was defined as a length of stay in the top 25% of the overall distribution. Multivariable logistic regression analyses were conducted to evaluate the association between POM and prolonged hospitalization, adjusting for age, sex, BMI, smoking status, diabetes mellitus, transplantation type, and disease categories. To minimize selection bias and confounding, inverse probability of treatment weighting (IPTW) based on propensity scores was applied. Results Of the 278 patients, 184 received POM. The absence of POM was significantly associated with prolonged hospitalization in both unadjusted and multivariable-adjusted analyses (odds ratio [OR]: 3.70, p < 0.001). After applying IPTW, covariate balance between the POM and non-POM groups was adequately achieved (standardized differences < 0.1 for all covariates). Logistic regression analysis after IPTW confirmed that the absence of POM remained significantly associated with prolonged hospitalization (OR: 1.22, p < 0.001), demonstrating the robustness of this association. Conclusion Perioperative oral management during hospitalization was independently associated with a reduced risk of prolonged hospitalization in patients with hematologic diseases. These findings suggest that incorporating POM into inpatient care may contribute to improved clinical outcomes and more efficient utilization of hospital resources. Further prospective studies are warranted to validate these findings.