Predictors of metabolic monitoring among schizophrenia patients with a new episode of second-generation antipsychotic use in the Veterans Health Administration

Background: To examine the baseline metabolic monitoring (MetMon) for second generation antipsychotics (SGA) among patients with schizophrenia in the Veterans Integrated Service Network (VISN) 16 of the Veterans Health Administration (VHA). Methods: VISN16 electronic medical records for 10/2002-08/2005 were used to identify patients with schizophrenia who received a new episode of SGA treatment after 10/2003, in which the VISN 16 baseline MetMon program was implemented. Patients who underwent MetMon (MetMon+: either blood glucose or lipid testing records) were compared with patients who did not (MetMon), on patient characteristics and resource utilization in the year prior to index treatment episode. A parsimonious logistic regression was used to identify predictors for MetMon+ with adjusted odds ratios (OR) and 95% confidence intervals (CI). Results: Out of 4,709 patients, 3,568 (75.8%) underwent the baseline MetMon. Compared with the MetMon-group, the MetMon+ patients were found more likely to have baseline diagnoses or mediations for diabetes (OR [CI]: 2.336 [1.846-2.955]), dyslipidemia (2.439 [2.029-2.932]), and hypertension (1.497 [1.287-1.743]), substance use disorders (1.460 [1.257-1.696]), or to be recorded as obesity (2.052 [1.724-2.443]). Increased likelihood for monitoring were positively associated with number of antipsychotics during the previous year (FGA: 1.434 [1.129-1.821]; SGA: 1.503 [1.290-1.751]). Other significant predictors for monitoring were more augmentation episodes (1.580 [1.145-2.179]), more outpatient visits (1.007 [1.002-1.013])), hospitalization days (1.011 [1.007-1.015]), and longer duration of antipsychotic use (1.001 [1.001-1.001]). Among the MetMon+ group, approximately 38.9% patient had metabolic syndrome. Discussion: This wide time window of 180 days, although congruent with the VHA guidelines for the baseline MetMon process, needs to be re-evaluated and narrowed down, so that optimally the monitoring event occurs at the time of receiving a new episode of SGA treatment. Future research will examine whether or not patients prescribed an SGA are assessed for metabolic syndrome following the index episode of antipsychotic therapy, and whether or not such baseline and follow-up monitoring programs in routine care are cost-effective. Conclusion: The baseline MetMon has been performed for a majority of the VISN 16 patients with schizophrenia prior to index SGA over the study period. Compared with MetMon- group, MetMon+ patients were more likely to be obese and manifest a more severe illness profile.