Article Title

Preventive Service Use Among People With Serious Mental Illnesses: Results From the PRIME Study

Publication Date



preventive service use, serious mental illness


Background/Aims: People with serious mental illnesses (SMI) experience excess morbidity and premature mortality resulting from preventable conditions. In response, concerns have been raised that people with SMI may not receive adequate preventive services. Understanding preventive service use patterns may help improve care delivery and reduce health disparities in this population.

Methods: Participants included all adults ≥ 18 years with at least one prior 12-month health care visit in a large integrated service system (Kaiser Permanente Northwest [KPNW], n = 366,194) or in participating federally qualified health centers and safety-net clinics in 18 states (OCHIN, n = 437,082). We used electronic medical record data to examine receipt of needed preventive care, measured by 12 services (e.g. flu shots, mammogram screening). We computed proportion of needed care by dividing the number of out-of-date services by the number of eligible services and multiplying by 100. Mean rates for four diagnostic groups –– schizophrenia spectrum disorders, bipolar disorders/affective psychoses, major depressive disorders, anxiety disorders –– were compared to patients without SMI diagnoses. All models were adjusted for patient characteristics (e.g. age, gender, Medicaid/Medicare status) and service use.

Results: At OCHIN, all SMI groups had lower rates of needed preventive services (schizophrenia: 34.12, bipolar disorders/affective psychoses: 35.68, major depressive disorders: 36.30, anxiety: 38.53) than non-SMI patients (39.97). Similarly, needed services did not differ or were lower among KPNW members with SMI, with bipolar disorders/affective psychoses (18.64) and major depressive disorders (18.63) showing significantly lower needed services than non-SMI members (20.64). Proportions of individuals with SMI completing cholesterol and diabetes screening were higher than individuals without SMI at both sites. At KPNW, individuals with SMI were more likely to have a recorded body mass index than those without SMI, but at OCHIN the reverse pattern was observed. Colorectal cancer screening did not differ by diagnosis but was more likely to occur at KPNW than at OCHIN. Blood pressure screening was common across diagnostic groups and sites. Flu vaccinations were uncommon, particularly at OCHIN.

Conclusion: Contrary to hypotheses, individuals with serious mental illnesses received more preventive services than individuals without mental illnesses. Adjusting for number of health care visits reduced, but did not eliminate, these differences.




June 28th, 2016


August 12th, 2016