Background: The prediction of violence among psychiatric inpatients using biophysiological indicators is warranted for re-examinations longitudinally. This study aims to explore factors associated with the occurrence of violence and subsequent medical impacts in psychiatric inpatients.
Methods: Inpatients diagnosed with either schizoaffective disorder or bipolar mania were admitted to acute wards in a professional psychiatric care setting. A longitudinal analysis was applied to construct predictive models with blood biochemistry tests upon admission. Medical records and an administrative database were used for analyses.
Results: Triglycerides were found to be a significant predictor of violence inception, which demonstrated a multivariate-adjusted odds ratio of 0.988 per mg/dL increment. Psychiatric inpatients with a higher level of triglycerides were less likely to have violent behaviors while more serious medical impacts were found once violence occurred. The elevated medical expenses derived from violence were negatively correlated with the level of cholesterol upon admission. A U-shape relationship was found between medical impacts and the combination of serum triglycerides and cholesterol.
Conclusion: The study provides useful predictors for early pre-screening of potential violence cases among acute psychiatric inpatients and therefore offers various angles for future strategic management of care plans in psychiatric medical settings.