Identification of psychiatric patients with high mortality and low medical utilization: a population-based propensity score-matched analysis

• Main Authors: Jong-Yi Wang (Author), Cheng-Chen Chang (Author), Meng-Chen Lee (Author), Yi-Jhen Li (Author)
• Format: Book
• Published: BMC, 2020-03-01T00:00:00Z.
• Subjects: article
• Online Access: Connect to this object online.
• Notes: 10.1186/s12913-020-05089-6; 1472-6963
• Keywords: Category of mental disorders; Mortality; Medical utilization; Potential years of life loss; Expenditure survival ratio; Public aspects of medicine; RA1-1270
• Source: BMC Health Services Research, Vol 20, Iss 1, Pp 1-12 (2020)

Abstract Background The decreased life expectancy and care costs of mental disorders could be enormous. However, research that compares mortality and utilization concurrently across the major category of mental disorders is absent. This study investigated all-cause mortality and medical utilization among patients with and without mental disorders, with an emphasis on identifying the psychiatric category of high mortality and low medical utilization. Methods A total of 570,250 individuals identified from the 2002-2013 Taiwan National Health Insurance Reearch Database consistuted 285,125 psychiatric patients and 285,125 non-psychiatric peers through 1:1 dual propensity score matching (PSM). The expenditure survival ratio (ESR) was proposed to indicate potential utilization shortage. The category of mental disorders and 13 covariates were analyzed using the Cox proportional hazard model and general linear model (GLM) through SAS 9.4. Results PSM analyses indicated that mortality and total medical expenditures per capita were both significantly higher in psychiatric patients than those in non-psychiatric patients (all P <.0.0001). Patients with substance use disorders were reported having the youngest ages at diagnosis and at death, with the highest 25.64 of potential years of life loss (YPLL) and relevant 2904.89 of ESR. Adjusted Cox model and GLM results indicated that, compared with anxiety disorders, affective disorders and substance use disorders were significantly associated with higher mortality (HR = 1.246 and 1.064, respectively; all P < 0.05); schizophrenia was significantly associated with higher total medical expenditures per capita (P < 0.0001). Thirteen additional factors were significantly associated with mortality or utilization (all P < 0.05). Conclusion Substance use disorders are the category of highest YPLL but notably in insufficient utilization. Health care utilization in patients with substance use disorders should be augmented timely after the diagnosis, especially toward home and community care. The factors related to mortality and utilization identified by this study merit clinical attention.