Inflammatory ratios as predictors of length of hospitalization in psychiatric patients: A multicenter study

Burgeoning evidence underscores the role of inflammation in psychiatric disorders where inflammation is linked to some clinical parameters such as disease severity and treatment resistance. However, the relationship between inflammation and the length of hospitalization remains poorly understood.

This study aimed to investigate whether elevated inflammation at the time of admission, as indicated by blood cell ratios which are easily accessible markers of peripheral inflammation, predicts the length of stay (LOS) in acutely ill psychiatric patients.

This multi-centre study was conducted at two state hospitals using retrospective data. A total of 497 inpatients were screened, and clinical records at admission were carefully reviewed to exclude any intervening infections or inflammatory conditions. Patients were excluded if they had significantly elevated acute-phase reactants, symptoms of infection, received treatment for suspected infections, used anti-inflammatory drugs, or had physical examination findings suggestive of infection. Additionally, patients with alcohol/substance use disorders and those hospitalized after a suicide attempt were not included. Ultimately, the final sample comprised 167 individuals. CRP levels and NLR, MLR and PLR ratios derived from routine laboratory tests.

The length of hospitalization ranged from 10 to 85 days, with a mean of 30.43 ± 14.45 days. A significant positive correlation was found between LOS and NLR (r = 0.453, p <.001), PLR (r = 0.351, p <.001), and MLR (r = 0.292, p <.001), even after controlling for age, gender, diagnosis, comorbid medical conditions, and study site. However, CRP levels did not correlate with LOS (r = 0.025, p =.762). Hierarchical regression analysis revealed that adding the set of immune ratios significantly improved the model's predictive value (p <.001), with immune ratios explaining an additional 10.8% of the variance in LOS, even after controlling for other factors. No significant differences in NLR, PLR, or MLR were observed across diagnostic groups (p =.47, p =.52, and p =.15, respectively). No significant difference was observed between diagnostic groups in terms of inflammatory ratios.

Our findings indicate that an elevated inflammatory response, as reflected by blood cell ratios, is associated with prolonged hospitalization in all patients, regardless of their diagnosis. These results emphasize the importance of addressing inflammatory processes in psychiatric care. Targeting inflammation as a modifiable risk factor may offer a new therapeutic avenue to reduce LOS. Future studies should focus on prospective designs and randomized controlled trials to determine whether targeting inflammation can directly reduce LOS and improve recovery in psychiatric populations.