Quality Of Care And Health Care Utilization Among Children And Young Adults Using Antipyschotics Enrolled In Mississippi Medicaid

VA L U E I N H E A LT H 1 9 ( 2 0 1 6 ) A 1 - A 3 1 8

Library was also conducted to ensure comprehensiveness. After evaluation by independent reviewers, complete economic evaluations of oral antipsychotic medications were included in the final analysis.  Results: 24 studies were included in the final analysis. The trend observed in the pooled studies showed that risperidone, olanzapine and clozapine were the drugs most commonly considered cost-effective to treat schizophrenia. Aripiprazole and haloperidol were considered comparable to quetiapine or ziprasidone and less cost-effective than olanzapine and risperidone. After removal of comparisons that had sponsored drugs included, risperidone, olanzapine and clozapine were still considered the most cost-effective strategies to treat schizophrenia. The analysis of only cost-utility studies shows approximately the same results of the other analysis.  Conclusions: An analysis that consider first- vs. second-generation antipsychotics polled together might be biased by the different profiles of the specific drugs, not considering the heterogeneity of the group of second-generation antipsychotics. There seems to be a difference in the cost-effectiveness profiles between specific antipsychotic drugs. Risperidone, olanzapine and clozapine seem to be the most cost-effective drugs to treat schizophrenia in a pooled analysis. This result was robust to changes in funding. PMH73 Care-Providing And Prescribing Practices Of Physicians Treating Children And Adolescents With Attention-Deficit/Hyperactivity Disorder Patel A1, Medhekar R1, Chen H1, Aparasu RR1, Ochoa-Perez M2, Chan W3, Sherer J1, Alonzo J1 1University of Houston, Houston, TX, USA, 2Legacy Community Health Services, Houston, TX, USA, 3University of Texas Health Science Center, Houston, TX, USA

Objectives: To compare the care-providing and prescribing practices of physicians treating children and adolescents with attention-deficit/hyperactivity disorder (ADHD).  Methods: A retrospective cohort study was conducted using the GE-EMR data from 1995-2010. The cohort consisted of individuals ≤ 18 years of age, diagnosed with ADHD (ICD-9 CM: 314.xx) and received a prescription of stimulants or atomoxetine. The NCQA-HEDIS measure for ADHD was used to study the follow-up care. The prescribing physicians were compared on the basis of time-totreatment, follow-up care, and use of polypharmacy. Multivariate logistic regression was conducted to determine the association of physician specialty with these outcome measures.  Results: After applying the inclusion and exclusion criteria the cohort consisted of 66,719 children and adolescents diagnosed and treated for ADHD. 76% (N= 50,582) of these cases were identified by PCPs, 2.6% (N= 1,724) by child-psychiatrists and the rest by unknown specialty. 59% (N= 39,343) of the patients were prescribed ADHD medication on the day of diagnosis and the rest received delayed treatment, with a median time-to-treatment of 57 days (IQR: 22-190). 34% (N= 22,476) of the treated cases had follow-up visits in accordance with the HEDIS criteria. Patients identified by PCPs had a mean of 6.56 (SD= 3.67) follow-up visits while those identified by child-psychiatrists had a mean of 7.26 (SD= 4.10) visits during the 10-month period. Compared to PCPs, child-psychiatrists were 47% less likely to initiate the treatment on same day of diagnosis (OR: 0.53; 95% CI: 0.48-0.59) and identified more comorbid mental disorders (OR: 1.20; 95% CI: 1.02-1.41). Even after controlling for comorbid conditions, child-psychiatrists were 52% more likely to use psychotropic polypharmacy than PCPs (OR: 1.52; 95% CI: 1.35-1.71).  Conclusions: Disparities exist in the care-providing behavior of physicians treating children with ADHD. Additional studies are needed to understand the clinical consequences of these disparities and the implication of care coordination across different provider specialties. PMH74 Off-Label Drug Prescriptions In France Tuncel T Toulouse School of Economics, Toulouse, France

Objectives: We investigate physicians’ prescription decisions over on-label vs. off-label drugs in treatment of alcoholism, anxiety, bipolar, and depression and estimate welfare impact of off-label prescriptions.  Methods: We use French panel data on exhaustive prescriptions of a representative sample of 386 general practitioners to all of their patients with alcoholism, anxiety, bipolar, and depression from 2000 to 2008. Some drugs change label status for some indications during the sample period, they are off-label for an indication at the beginning of the sample period whereas they become approved for that indication afterwards. This allows us identify how a change in label status affects physicians’ prescription behaviour. The panel structure of the data allows us observe the differences in treatment outcomes across patients treated by off-label drugs and patients treated by approved drugs.  Results: The rate of off-label prescriptions is at least 18 % in each of these diseases. All of the physicians prescribe off-label drugs; however, there is some heterogeneity across physicians in their off-label prescriptions. There is also heterogeneity across patients treated by the same physician which provides evidence that off-label prescriptions are partially due to patient characteristics. The counterfactual simulations show that removing the off-label drugs from the choice set of physicians would lead to a substantial increase in the cost of prescription drugs in the treatment of these diseases, whereas it does not lead to an improvement in terms of treatment outcomes.  Conclusions: We find that off-label prescriptions lead to substantial decrease in cost of prescription drugs without leading to deterioration in treatment outcomes. PMH75 Quality Of Care And Health Care Utilization Among Children And Young Adults Using Antipyschotics Enrolled In Mississippi Medicaid Ramachandran S1, Banahan III B1, Hardwick S1, Noble S2 of Mississippi, University, MS, USA, 2Office of the Governor, Division of Medicaid, Jackson, MS, USA 1University

Objectives: In March 2015, the Office of the Inspector General (OIG) issued a report concerning the use of antipsychotics. The OIG highlighted several quality

A195

measurement areas including indications for use, appropriate dosage and duration of use, monitoring, polypharmacy, side-effects and age appropriate use. The purpose of this project was identifying quality measures concerning these seven areas and testing measure performance in the eligible Mississippi Medicaid population.  Methods: A retrospective analysis during the 2014 calendar year was conducted using Mississippi Medicaid administrative claims and beneficiary eligibility data. Individuals below the age of 21 years with at least one month of eligibility in Medicaid during the study period and antipsychotic prescription claims were included. Quality measures calculated included metabolic screening for new and existing antipsychotic prescriptions, availability of a supporting diagnosis, and antipsychotic polypharmacy. Measure specifications proposed by the National Collaborative for Innovation in Quality Measurement and developed by HEDIS were used to address OIG identified areas identified.  Results: A total of 19,009 beneficiaries (51% female, 49% male, 54% African Americans, 44% Caucasians) were identified as taking antipsychotics during the study period. Overall, 67% of beneficiaries did not have a primary indication for antipsychotic use. Only 14.1% had one or more follow-up visits, 3% were using two or more antipsychotics concurrently, 14.1% had baseline metabolic screening performed prior to filling antipsychotic prescriptions, only 14.9% had both glucose and lipid levels tests performed during the year in which antipsychotics were prescribed. Only 0.11% of children <  age 5 years had antipsychotic medications prescribed.  Conclusions: Study results indicate considerable opportunity for improvement in quality of care identified in the OIG report. Several challenges exist in this area include health care access, continuity of care, and awareness of quality measures amongst providers. PMH76 The Power Of The Measure Incubator: Leveraging “Big Data” To Assess Care Quality Measures For Persons With Dementia Kravetz A1, Pickering MK2, Perfetto EM2 of Maryland, School of Pharmacy, and Visiting Fellow at OptumLabs, Baltimore, MD, USA, 2University of Maryland, School of Pharmacy, Baltimore, MD, USA 1Univeristy

Objectives: Better dementia-care quality measurement is needed. Unfortunately, measure development is time consuming and expensive. Using the NQF Measure Incubator at OptumLabsTM, the feasibility of evaluating medication-use measures for persons with dementia was examined. It was hypothesized that trends in measure rates over time would correspond to program implementation and rates vary based upon how explicitly specifications define populations.  Methods: For two existing measures (antipsychotic and anticholinergic/TCA use), annual rates were calculated, 2001-2015, using the OptumLabs Data Warehouse, which includes over 150 million nationally representative commercial and Medicare Advantage (MA) enrollees. Dementia cases were identified using specifications based on NQF #2111. Subgroup analyses were performed based on age, race, gender, geographic region, and insurance type. Sensitivity analyses were conducted on eligibility criteria based on number of header diagnoses and psychoses exclusions specified.  Results: Antipsychotic use in persons with dementia was 13-18% in 2001-2006. In 2007, the rate dropped to 16% and continued to decline, remaining somewhat constant at 13-15% up to 2015. Stratifying by insurance, MA enrollees had higher rates than commercial. Rates were higher amongsth those >/=65 years of age. Cases <65 years of age had increasing rates from 7% in 2001 to up to 20% in 2015. Differences were not seen by race or geographic region. Similar results were noted for anticholinergic/TCA use. Case identification was not sensitive to the number of header diagnoses used. Rates varied by 1-2 difference based on header diagnosis number. For NQF #2111, results did not differ based on denominator psychoses exclusions.  Conclusions: The Measure Incubator provides efficiency in quality measure development and assessment. Leveraging this resource allowed for examination and modification of specifications for two dementia-specific care quality measures within a relatively short timeframe at relatively low cost. Measures can be tested and adapted efficiently with this resource. Additional work is needed to understand clinical implications. PMH77 Quality Of Care And Health Care Utilization Among Foster Children In Mississippi Medicaid Ramachandran S1, Banahan III B1, Shah R1, Nunna S1, Hardwick S1, Noble S2 1University of Mississippi, University, MS, USA, 2Office of the Governor, Division of Medicaid, Jackson, MS, USA

Objectives: Children placed in foster care are among our nation’s most vulnerable populations. They tend to have more serious mental and medical health conditions when compared to other children. The objective of this project was to compare the health care utilization of foster and non-foster children enrolled in Mississippi Medicaid and compare them to other children in Medicaid.  Methods: A retrospective analysis was conducted using Mississippi Medicaid administrative claims for calendar years of 2013 and 2014. Beneficiaries were included if they were less than 21 years of age and eligible for services for at least one month during the observation year. Foster children were identified using Medicaid eligibility codes (003, 005, 026). All study variables were weighted based on age and race in order to provide the best comparison of health care utilization.  Results: Nearly 450,000 children were included in the study in both 2013 and 2014. The study population comprised of 49% males in both the foster and non-foster groups. At least 48% of the foster children had at least one mental-illness, compared to 18% of the other children. Prevalence rates for mental health illnesses such as psychosis, bipolardisorder, depression, ADHD were all higher in the foster population. 20% of foster children received at least one mental health medication as compared to 7% of non-foster children. A significantly higher proportion of foster children used each class of mental-health drugs such as barbiturates, antidepressants, antipsychotics, and anticonvulsants.  Conclusions: The findings are consistent with other studies. It is important to provide effective and appropriate mental health screenings, assessment and treatment for all children, especially those in foster care. The results indicate that although use of mental health drugs is higher in foster