Examining quality of care for individuals treated for mental health using the HEDIS mental health quality measures.

Objective: This descriptive study examined the quality of care received by individuals with serious mental illness observed in clinical care using established Healthcare Effectiveness Data and Information Set (HEDIS) measures for individuals with serious mental illness.Methods: Administrative claims (Medicaid, Medicare, and commercial) from a national health and well-being company were used to identify adults with schizophrenia or bipolar disorder. Performance rates for five HEDIS mental health quality measures were computed. Sub-group analyses examined each HEDIS measure by those who were medication adherent vs non-adherent, and by typical vs atypical antipsychotics.Results: Eighty-nine percent of the Medicaid population received a diabetes screening (vs 79% for national benchmark Medicaid rates), 81% (vs 69%) received monitoring for diabetes, 88% (vs 79%) received monitoring for cardiovascular disease, 63% (vs 60%) were adherent with antipsychotic medication, and 34% (vs 61%) had a follow-up visit with a mental health practitioner within 30 days of a discharge. The rates for individuals with Medicare coverage were similar or marginally higher than those reported for those with Medicaid coverage, while rates for the commercially insured population were lower than the other groups.Conclusions: Most (>65%) individuals with serious mental illness received the recommended screening and monitoring for diabetes and cardiovascular disease. Barriers to and reasons for lack of follow-up should be investigated to guide future interventions to improve follow-up after hospitalization for individuals with serious mental illness.

Use of a Web-Based Shared Decision-Making Program: Impact on Ongoing Treatment Engagement and Antipsychotic Adherence.

OBJECTIVE: The authors examined the impact of a Web-based shared decision-making application, MyCHOIS-CommonGround, on ongoing outpatient mental health treatment engagement (all users) and antipsychotic medication adherence (users with schizophrenia). METHODS: An intervention study was conducted by comparing Medicaid-enrolled MyCHOIS-CommonGround users in 12 participating mental health clinics (N=472) with propensity score-matched adults receiving services in nonparticipating clinics (N=944). Medicaid claims were used to assess ongoing treatment engagement and antipsychotic adherence (among individuals with schizophrenia) one year prior to and after entry into the cohort. Multilevel linear models were conducted to estimate the effects of the MyCHOIS-CommonGround program over time. RESULTS: No differences during the baseline year were found between the MyCHOIS-CommonGround group and the matched control group on demographic, diagnostic, or service use characteristics. At one-year follow-up, engagement in outpatient mental health services was significantly higher for MyCHOIS-CommonGround users than for the control group (months with a service, 8.54±.22 versus 6.95±.15; ß=1.40, p<.001). Among individuals with schizophrenia, antipsychotic medication adherence was also higher during the follow-up year among MyCHOIS-CommonGround users compared with the control group (proportion of days covered by medication, .78±.04 versus .69±.03; ß=.06, p<.01). CONCLUSIONS: These findings provide new evidence that shared decision-making tools may promote ongoing mental health treatment engagement for individuals with serious mental illness and improved antipsychotic medication adherence for those with schizophrenia.

Hospitalization outcomes in patients with schizophrenia after switching to lurasidone or quetiapine: a US claims database analysis.

BACKGROUND: This study compared hospital admission rates among adult patients with schizophrenia who switched to antipsychotic monotherapy with lurasidone or quetiapine. METHODS: This retrospective cohort study used U.S.-based Truven Health MarketScan® Medicaid Multi-State Database (April 2010 through December 2012) and MarketScan® Commercial Claims and Encounters Database (April 2010 through October 2013). Continuous enrollment for 6-months before and after medication initiation was required. Treatment episodes ended after 6-months post lurasidone or quetiapine initiation, a 60-day treatment gap, or initiation of another antipsychotic. Length of treatment episodes (i.e., treatment duration) was compared using a t-test. All-cause, mental-health, and schizophrenia-related hospitalization rates, as well as costs, were compared between lurasidone- and quetiapine-treated patients using multivariable generalized linear models that adjusted for background characteristics. RESULTS: Quetiapine (n = 435) compared to lurasidone (n = 238) treatment was associated with increased all-cause (21% vs 13%, p < 0.05) and mental health-related hospitalizations (20% vs 12%, p < 0.05), but similar rates of schizophrenia-related hospitalizations (14% vs. 10%, p = 0.14). After adjusting for baseline covariates, quetiapine had 64% higher likelihood of all-cause hospitalizations (OR [odds ratio] = 1.64, 95% confidence interval [CI] 1.05-2.57, p = 0.03), 74% higher likelihood of mental health-related hospitalizations (OR = 1.74, 95% CI 1.11-2.75, p = 0.02), and a similar likelihood of schizophrenia-related hospitalization (OR = 1.35, 95% CI 0.82-2.22, p = 0.24). For those with hospital admissions, adjusted all-cause admission costs were higher for quetiapine when compared with lurasidone in both the Medicaid ($22,036 vs. $15,424, p = 0.17) and commercial populations ($23,490 vs. $20,049, p = 0.61). These differences were not significant. The length of treatment episodes was significantly shorter for quetiapine than lurasidone (115.4 vs 123.1 days, p < 0.05). CONCLUSIONS: In this retrospective claims database study, patients with schizophrenia who were switched to lurasidone had significantly fewer all-cause and mental health-related hospitalizations and similar rates of schizophrenia-related hospitalization compared with those switched to quetiapine. Patients switching to lurasidone had a significantly longer treatment duration rate than those switching to quetiapine. These results may reflect differences in efficacy or tolerability between lurasidone and quetiapine.

Socioeconomic Disparities and Metabolic Risk in Veterans with Serious Mental Illness.

Socioeconomic disparities were assessed in predicting metabolic risk among veterans with serious mental illness. Veterans with schizophrenia, schizoaffective, or bipolar disorders were identified in VISN 4 facilities from 10/1/2010 to 9/30/2012. Differences between patients with and without metabolic syndrome were compared using t-tests, Chi square tests and multivariate logistic regressions. Among 10,132 veterans with mental illness, 48.8% had metabolic syndrome. Multivariate logistic regression analysis confirmed that patients with metabolic syndrome were significantly more likely to be older, male, African-American, married, and receiving disability pensions but less likely to be homeless. They were more likely to receive antipsychotics, antidepressants, or anticonvulsants. Bivariate cross-sectional analysis revealed that patients with metabolic syndrome had higher rates of coronary artery disease, cerebrovascular disease, and mortality, and that metabolic syndrome was more often associated with emergency visits and psychiatric or medical hospitalizations. Demographics, socioeconomic status and medications are independent predictors of metabolic syndrome and should be considered in broader screening of risk factors in order to provide preventive interventions for metabolic syndrome.

Patient Characteristics Associated With Use of Lurasidone Versus Other Atypical Antipsychotics in Patients With Bipolar Disorder: Analysis From a Claims Database in the United States.

OBJECTIVE: To compare patient characteristics, medical comorbidities, health care utilization, and health care costs among patients with bipolar disorder who initiated lurasidone versus other atypical antipsychotics in usual clinical practice. METHODS: A retrospective analysis of administrative claims data was conducted using the US Optum Research Database (December 30, 2012, through February 27, 2014). Adult, commercially insured patients with bipolar disorder with an atypical antipsychotic prescription between June 28, 2013, and November 30, 2013, were included. The lurasidone cohort first included any patients with a lurasidone prescription; remaining patients were assigned to their first atypical antipsychotic (aripiprazole, olanzapine, quetiapine, risperidone, ziprasidone). Preindex patient characteristics comparisons to lurasidone were conducted with t tests (continuous variables) and χ² or Fisher exact tests (categorical variables). RESULTS: A total of 3,329 patients were included in this database analysis. A higher percentage of the lurasidone cohort (31.1%) had bipolar depression compared with the other cohorts (23.5%-28.0%). The lurasidone cohort had a statistically significantly higher percentage of patients with prior diabetes mellitus (13.3%) and lipid metabolism disorders (23.2%) than did the quetiapine cohort (8.4% and 16.3%, P < .01). In addition, the lurasidone cohort had significantly more prior antipsychotic polypharmacy (23.0% vs 6.7%-12.9%, P < .01) and atypical antipsychotic use (55.6% vs 11.8%-26.3%, P < .01) than other cohorts. The lurasidone cohort had a statistically significantly higher mean number of prior all-cause and mental health office visits (P < .001) and higher mean prior pharmacy costs than most cohorts (P < .01). CONCLUSIONS: Lurasidone-treated patients with bipolar disorder tended to have a more complex clinical profile, comorbidities, and prior treatment history compared to patients initiated with other atypical antipsychotics in this claims database study. This pattern of treatment may have reflected the overall clinical profile of lurasidone, the role perceived for lurasidone in the therapeutic armamentarium by practitioners, and the recent introduction of lurasidone into clinical practice during the study period.

Cardiometabolic comorbidities, readmission, and costs in schizophrenia and bipolar disorder: a real-world analysis.

BACKGROUND: Serious mental illnesses are associated with increased risk of cardiometabolic comorbidities. The objective of this study was to evaluate the prevalence of cardiometabolic comorbidity and its association with hospitalization outcomes and costs among inpatients with schizophrenia or bipolar disorder. METHODS: This retrospective database analysis reviewed patients with an inpatient diagnosis of schizophrenia or bipolar disorder from the Premier Perspective® Database (4/1/2010-6/30/2012). Patients were categorized into 4 cohorts based on the number of ICD-9-CM cardiometabolic comorbidities (i.e., 0, 1, 2, or 3+). Outcomes included length of stay, mortality during the index hospitalization, healthcare costs, and 30-day all-cause readmission rates. RESULTS: Of 57,506 patients with schizophrenia, 66.1% had at least one cardiometabolic comorbidity; 39.3% had two or more comorbidities. Of 124,803 patients with bipolar disorder, 60.5% had at least one cardiometabolic comorbidity; 33.4% had two or more. Average length of stay was 8.5 (for patients with schizophrenia) and 5.2 (for patients with bipolar disorder) days. Each additional cardiometabolic comorbidity was associated with an increase in length of stay for patients with bipolar disorder (p < .001) but not for patients with schizophrenia. Mortality rates during the index hospitalization were 1.2% (schizophrenia) and .7% (bipolar disorder). Each additional cardiometabolic comorbidity was associated with a significant increase in mortality for patients with bipolar disorder (OR 1.218, p < .001), and a numerical increase in mortality for patients with schizophrenia (OR 1.014, p = .727). Patients with more cardiometabolic comorbidities were more likely to have a 30-day readmission (schizophrenia = 9-13%; bipolar disorder = 7-12%), and to incur higher costs (schizophrenia = $10,606-15,355; bipolar disorder = $7126-13,523) (all p < .01). CONCLUSIONS: Over 60% of inpatients with schizophrenia or bipolar disorder had cardiometabolic comorbidities. Greater cardiometabolic comorbidity burden was associated with an increased likelihood of readmission and higher costs among patients with schizophrenia or bipolar disorder, and an increase in length of stay and mortality for patients with bipolar disorder.

Major depressive disorder with subthreshold hypomanic (mixed) features: A real-world assessment of treatment patterns and economic burden.

BACKGROUND: To compare outcomes for individuals with major depressive disorder (MDD) with or without subthreshold hypomania (mixed features) in naturalistic settings. METHODS: Using the Optum Research Database (1/1/2009─10/31/2014), a retrospective analysis of individuals newly diagnosed with MDD was conducted. Continuous enrollment for 12-months before and after the initial MDD diagnosis was required. MDD with subthreshold hypomania (mixed features) (MDD-MF) was defined based on ≥1 hypomania diagnosis within 30 days after an MDD diagnosis during the one-year follow-up period, in the absence of bipolar I diagnoses. Psychiatric medication use, healthcare utilization, and costs during the one-year follow-up period were compared using multivariate logistic and gamma regressions, controlling for baseline differences. RESULTS: Of 130,626 MDD individuals, 652 (0.5%) met the operational definition of MDD-MF. Compared to the MDD-only group, the MDD-MF group had more suicidality (2.0% vs. 0.5%), anxiety disorders (46.8% vs. 34.0%), and substance use disorders (15.5% vs. 6.1%, all P<0.001). More individuals with MDD-MF were treated with antidepressants (83.6% vs. 71.6%), mood stabilizers (50.5% vs. 2.7%), atypical antipsychotics (39.0% vs. 5.5%), and polypharmacy with multiple drug classes (72.1% vs. 22.7%, all P<0.001). Individuals with MDD-MF had higher hospitalizations rates (24.2% vs. 10.5%) and total healthcare costs (mean: $15,660 vs. $10,744, all P<0.001). LIMITATIONS: The commercial claims data used were not collected for research purposes and may over- or under-represent certain populations. No specific claims-based diagnostic code for MDD with mixed features exists. CONCLUSIONS: Greater use of mood stabilizers, atypical antipsychotics, polypharmacy, and healthcare resources provides evidence of the complexity and severity of MDD-MF. Identifying optimal treatment regimens for this population represents a major unmet medical need.

The association between employee obesity and employer costs: evidence from a panel of U.S. employers.

PURPOSE: To estimate the employer costs associated with employee obesity. DESIGN: The study used cross-sectional analysis of employee health-risk assessment, disability, workers' compensation, and medical claims data from 2006 to 2008. SETTING: The study took place in the United States from 2006 to 2008. SUBJECTS: A panel database with 29,699 employees drawn from a panel of employers and observed for 3 years each (N = 89,097) was used. MEASURES: Workdays lost owing to illness and disability; the cost of medical, short-term disability, and workers' compensation claims; and employees' adjusted body mass indices (BMI) were measured. ANALYSIS: We model the number and probability of workdays lost from illness, short-term disability, and workers' compensation events; short-term disability and workers' compensation payments; and health care spending as a function of BMI. We estimate spline regression models and fit results using a third-degree fractional polynomial. RESULTS: Probability of disability, workers' compensation claims, and number of days missed owing to any cause increase with BMI above 25, as do total employer costs. The probability of a short-term disability claim increases faster for employees with hypertension, hyperlipidemia, or diabetes. Normal weight employees cost on average $3830 per year in covered medical, sick day, short-term disability, and workers' compensation claims combined; morbidly obese employees cost more than twice that amount, or $8067, in 2011 dollars. CONCLUSION: Obesity is associated with large employer costs from direct health care and insurance claims and indirect costs from lost productivity owing to workdays lost because of illness and disability.

Assessment of treatment-seeking behavior and healthcare utilization in an international cohort of subjects with overactive bladder.

UNLABELLED: Abstract Objective: To assess the association between incontinence severity, treatment-seeking behavior, and healthcare resource utilization (HRU) among participants with overactive bladder (OAB) in eight countries. RESEARCH DESIGN AND METHODS: A cross-sectional online survey of subjects ≥18 years old in Australia, Europe, and North America, who had a past OAB diagnosis and/or experienced ≥1 urinary incontinence (UI) episode in the preceding 12 months, were eligible to participate. Subjects contacted for the survey were primarily from a voluntary medication monitoring registry, MediGuard. Predominantly stress incontinence subjects were excluded. Incontinence severity was assessed by the number of UI episodes over 3 days and grouped as 0 ('dry'), 1-2, 3-4, and ≥5 UI episodes/day. Subject demographics, employment status, comorbidities, treatment-seeking behavior (past OAB diagnosis; spoken to healthcare provider [HCP]), and HRU (diagnostic tests; HCP visits in 6 months before screening) were analyzed by incontinence severity. RESULTS: Overall, 1341 subjects with OAB (mean age 54.5 years; 70.7% female) were surveyed; 20.2%, 47.7%, 18.8%, and 13.3% of subjects reported 0, 1-2, 3-4, and ≥5 UI episodes/day, respectively. Employment status and comorbidities were significantly (p < 0.05) associated with incontinence severity. The two measures of treatment-seeking behavior were significantly (p < 0.05) associated with incontinence severity groups; the proportion of subjects with a past diagnosis of OAB were 35.8%, 44.8%, 52.4%, and 64.0% in the 0, 1-2, 3-4, and ≥5 UI episodes/day groups, respectively; and 59.0%, 63.6%, 65.9%, and 78.1% of subjects in the respective UI severity groups talked to a HCP about their OAB symptoms. Multivariate linear regression analyses showed a positive and consistent association between incontinence severity and HRU; subjects reported a mean of 2.7, 4.1, 4.4, and 7.7 diagnostic tests overall (p < 0.001), and a mean of 1.4, 2.2, 2.7, and 4.0 HCP visits in the 0, 1-2, 3-4, and ≥5 UI episodes/day groups, respectively (p < 0.001). A potential limitation of the study is the cross-sectional survey methodology which limits the ability to draw causal inferences from the results. Additionally, since this is a web-based survey it is possible respondents who have access to/are familiar with technology were more likely to be enrolled. CONCLUSIONS: Incontinence severity was positively associated with both treatment-seeking behavior and HRU among subjects with OAB.

Identifying patients at risk for high medical costs and good candidates for obesity intervention.

PURPOSE: To develop a risk-scoring tool to identify in a base year patients likely to have high medical spending in the subsequent year and to understand the role obesity and obesity reduction may play in mitigating this risk. DESIGN: Cross-sectional analysis, using commercial claims and health risk assessment data. SETTING: United States, 2004-2009. SUBJECTS: Panel of 192,750 person-year observations from 116,868 unique working-age employees of large companies. MEASURES: Probability of high medical expenses (80th percentile or above) in the following year; adjusted body mass index (BMI). ANALYSIS: Generate risk scores by modeling the likelihood of high next-year expenses as a function of base-year age, sex, medical utilization, comorbidities, and BMI. Estimate the effect of simulated bariatric intervention on patient risk scores. RESULTS: Individuals with higher BMI were more likely to be categorized in the very high risk group, in which the average annual medical expense was $8621. A weight-loss intervention transitioning a patient to the next lower obesity class was predicted to reduce this risk by 1.5% to 27.4%-comparable to hypothetically curing a patient of depression or type 2 diabetes. CONCLUSION: A logistic model was used to capture the effect of BMI on the risk of high future medical spending. Weight-loss interventions for obese patients may generate significant savings by reducing this risk.

Measuring migraine-related quality of care across 10 health plans.

OBJECTIVES: To refine a previously published standardized quality and utilization measurement set for migraine care and to establish performance benchmarks. STUDY DESIGN: Retrospective application of the migraine measurement set to health plan data in order to assess patterns of health service utilization. METHODS: Measurement specifications were applied to data from 10 health plans for measurement year 2009. RESULTS: Of the 2.9 million continuously enrolled members of the health plans, 138,004 (4.7%) met inclusion criteria for the migraine population. Of these, 26% did not have a migraine diagnosis, but were utilizing migraine drugs; 12% had a computed tomography scan within the year (range 8%-25% across plans); and 8% had magnetic resonance imaging (range 6%-11%). Nearly 18% of the migraineurs had 1 or more visits to an emergency department/urgent care center for migraine; few (6%) were followed up with primary care visits. Approximately one-fourth of the migraineurs were not being routinely monitored by a physician. Medication utilization also was examined for members of the migraine population with pharmacy benefits. A significant proportion (42%) were given a migraine preventive, 38% had at least 1 prescription for a triptan, and 2% of those on triptans were potentially overutilizing the medication. Among patients aged 18 to 49 years who were given triptans, 3% had a cardiac contraindication; this percentage rose to 7% for patients aged 50 to 64 years. CONCLUSIONS: This study demonstrates the value of standardized measures in identifying potential quality issues for migraine care, including underdiagnosis, overutilization of imaging, and underutilization of preventive drugs.

EQ-5D™-derived utility values for different levels of migraine severity from a UK sample of migraineurs.

BACKGROUND: To estimate utility values for different levels of migraine pain severity from a United Kingdom (UK) sample of migraineurs. METHODS: One hundred and six migraineurs completed the EQ-5D to evaluate their health status for mild, moderate and severe levels of migraine pain severity for a recent migraine attack, and for current health defined as health status within seven days post-migraine attack. Statistical tests were used to evaluate differences in mean utility scores by migraine severity. RESULTS: Utility scores for each health state were significantly different from 1.0 (no problems on any EQ-5D dimension) (p < 0.0001) and one another (p < 0.0001). The lowest mean utility, - 0.20 (95% confidence interval [CI]: -0.27 - -0.13), was for severe migraine pain. The smallest difference in mean utility was between mild and moderate migraine pain (0.13) and the largest difference in mean utility was between current health (without migraine) and severe migraine pain (1.07). CONCLUSIONS: Results indicate that all levels of migraine pain are associated with significantly reduced utility values. As severity worsened, utility decreased and severe migraine pain was considered a health state worse than death. Results can be used in cost-utility models examining the relative economic value of therapeutic strategies for migraine in the UK.

Costs associated with outpatient, emergency room and inpatient care for migraine in the USA.

BACKGROUND: Data on the average US costs of an outpatient visit, emergency room (ER) visit or hospitalization for migraine are scant, with the most recent available values based on healthcare charges reported from 1994 data. METHODS: We estimated healthcare costs associated with outpatient and ER visits and inpatient hospitalizations related to migraine retrospectively obtained from the 2007 Medstat MarketScan Commercial Claims & Encounters database. Tabulated costs reflected payments from insurers, patients and other sources. All costs were adjusted to 2010 US dollars. RESULTS: The estimated mean cost (95% CI) for migraine-related care per outpatient visit (N = 680,946) was $139.88 ($139.35-140.41); per ER visit (N = 88,128) was $775.09 ($768.10-782.09); and per inpatient hospitalization (N = 5516) was $7317.07 ($7134.96-7499.17). The most frequently coded procedures at outpatient and ER visits were subcutaneous or intra-muscular injection, and for hospitalizations was computed tomography. Estimated annual US healthcare costs in 2010 for migraine associated with: outpatient visits were $3.2 billion, ER visits were $700 million, and inpatient hospitalizations were $375 million. CONCLUSIONS: Direct healthcare costs associated with patient visits and hospitalizations for migraine headaches have increased since previously published estimates. Further research is needed to understand the current overall healthcare cost burden per patient and within the US population.