BACKGROUND/RATIONALE:
Second-generation antipsychotic (SGA) medications are widely used to treat psychotic disorders but are associated with metabolic side effects such as weight gain, glucose dysregulation, and hyperlipidemia that may contribute to the high rates of cardiovascular disease observed in individuals with serious mental illness (SMI). Adherence to guidelines for regular monitoring for these metabolic side effects has been inadequate. Patient-centered care, characterized by an effective partnership between clinicians and patients that promotes active participation by patients in their own care, improves health outcomes and satisfaction in the general population. In order to increase rates of monitoring for metabolic side effects of SGAs, we designed a patient-centered computerized tool, COMP-SGA, which provided veterans with SMI with personalized health information on how well their care adhered to expert recommendations on metabolic monitoring. COMP-SGA used principles shown to enhance usability in persons with cognitive impairments.

OBJECTIVE(S):
The major goals of this randomized controlled trial were to determine the effect of exposure to COMP-SGA compared to enhanced treatment as usual (e-TAU) on: (Aim 1) rates of monitoring for and identification of metabolic abnormalities associated with the metabolic side effects of SGAs; (Aim 2) patterns of patient-centered communication around monitoring; and (Aim 3) veterans' self-efficacy in communicating with their prescribers about monitoring, their preferences for obtaining health information and participating in decision-making about monitoring, and perceptions of their prescribers' participatory decision-making styles around monitoring.

METHODS:
Veterans with psychiatric disorders prescribed SGAs and in regular contact with their prescribing psychiatrists or nurse practitioners in two VA outpatient mental health clinics were recruited for this study. Half of veteran-participants were randomly assigned to the intervention condition, COMP-SGA, in which they viewed a brief computer program that provided personalized health information on how well their care adhered to guidelines for monitoring of metabolic side effects of SGAs that was designed to facilitate discussion with prescribers about appropriate monitoring. To activate veterans to initiate this conversation, they were given the option of printing out a summary report of their monitoring status and the metabolic test results that were reviewed with them that could be shared with their prescriber. The other half of participants received enhanced treatment as usual (e-TAU) consisting of printed information on the metabolic side effects of SGAs and general recommendations for monitoring. Veteran-participants were exposed to COMP-SGA or e-TAU up to 3 times, at least 4 months apart, immediately prior to regularly scheduled visits with their prescribers over the one-year study period. A single prescriber visit for each participant was audiotaped and coded with the Roter Interaction Analysis System (RIAS) to characterize patterns of patient-prescriber communication around monitoring for metabolic side effects.

FINDINGS/RESULTS:
We met recruitment goals by enrolling all 17 prescribers invited to participate and 239 veterans, of whom 119 were randomized to COMP-SGA and 120 were randomized to e-TAU. At baseline, overall the mean age of veterans was 54 (+/-8) years, 89% were male, 53% were non-white, and 68% were ever married. In terms of primary psychiatric diagnosis, 30% had a psychotic disorder, 32% had bipolar disorder, 26% had major depression, and 12% had PTSD. The most commonly prescribed SGAs were quetiapine and risperidone. The study groups did not differ significantly on any demographic or clinical characteristics at baseline. Over the one-year study period, 96% of veterans randomized to COMP-SGA and 95% randomized to e-TAU were exposed to intervention at one research visit; 86% and 83%, respectively, completed two research visits, and 62% and 56%, respectively, completed three research visits. For those randomized to COMP-SGA, the average time to view the program before an appointment with their prescriber was 11.3 (+/-5.6) minutes. Among those randomized to COMP-SGA who viewed the program before only one visit with their prescriber, 58% printed out the optional summary report to share with their prescriber; 44% who viewed COMP-SGA before two visits printed the report; and 36% who viewed COMP-SGA before three visits printed the report. Throughout the study, only 11 (9%) veterans randomized to COMP-SGA and 13 (11%) randomized to e-TAU were withdrawn from the study, primarily because the SGA was discontinued by the prescriber.

For Aim 1, there were no differences between the groups in the percentage of days during the study period in which monitoring for each of the metabolic side effects adhered to guideline recommendations (p > 0.05); adherent days were high in both groups, ranging from 80-100%. There were also no differences between the groups in identification of metabolic abnormalities associated with SGAs; however, prevalence of metabolic abnormalities including elevated body mass index (90% in COMP-SGA, 92% in e-TAU), elevated blood pressure (69% and 66%, respectively), elevated fasting blood glucose or glycosylated hemoglobin (31%, 36%), elevated LDL cholesterol (39%, 25%), reduced HDL cholesterol (61%, 59%), and elevated triglycerides (50%, 54%) was notably high in both groups.

For Aim 2, we obtained audio tapes of one visit between veterans and prescribers for 77% of those randomized to COMP-SGA and 76% assigned to e-TAU. Whereas there were no differences by treatment condition with regard to mentions of weight (p=0.106), blood pressure (p=0.995), or blood sugar (p=0.083) during the visit, there were more mentions of LDL cholesterol (p<0.001), HDL cholesterol (p<0.001), and triglycerides (p=0.005) in those exposed to COMP-SGA compared to e-TAU. In addition, there were significantly more mentions of a report, list, or pamphlet in those exposed to COMP-SGA (p<0.001).

For Aim 3, there were no differences between the groups in terms of veterans' self-efficacy in communicating with their prescribers about metabolic monitoring (p=0.627), their preferences for obtaining health information (p=0.985) and participating in decision-making about monitoring (p=0.482 and 0.996), and perceptions of their prescribers' participatory decision-making styles around monitoring (p >0.05).

IMPACT:
The computerized intervention evaluated in this study enhanced patient-provider communication around screening for metabolic side effects of SGAs and can thus serve as a template for engaging veterans with SMI in the self-management of their physical as well as psychiatric care to facilitate improvements in overall health outcomes.

---

External Links for this Project

---

PUBLICATIONS:

VA Cyberseminars

1. Kreyenbuhl JA. A Patient-Centered Approach to Improve Screening for the Metabolic Side Effects of Second-Generation Antipsychotic Medication. [Cyberseminar]. 2012 Oct 1. [view]

Conference Presentations

1. Kreyenbuhl JA, Dixon LB, Medoff DR, Rotondi A. A Patient-Centered Approach to Improve Screening for Side Effects of Second-Generation Antipsychotics. Paper presented at: VA Mental Health Annual Conference; 2011 Aug 1; Baltimore, MD. [view]
2. Kreyenbuhl JA, Dixon LB, Brown C, Medoff DR, Fang L. A Patient-Centered Approach to Improve Screening for Side Effects of Second-Generation Antipsychotics. Paper presented at: AcademyHealth Annual Research Meeting; 2013 Jun 1; Baltimore, MD. [view]
3. Kreyenbuhl JA. A Patient-Centered Health Technology Intervention to Improve Screening for the Metabolic Side Effects of Second-Generation Antipsychotic Medications. Paper presented at: American Psychiatric Association Annual Meeting; 2011 May 1; Honolulu, HI. [view]

DRA: Mental, Cognitive and Behavioral Disorders, Health Systems
DRE: Prevention, Prognosis, Diagnosis, Treatment - Observational
Keywords: Mental health care service, Pharmaceuticals, Severe mental illness
MeSH Terms: none