[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Table 1.  
Patient, Prescriber, and Practice Characteristics of New Users of Antipsychotic Medication
Patient, Prescriber, and Practice Characteristics of New Users of Antipsychotic Medication
Table 2.  
Antipsychotic Prescribing for Adults in Missouri Medicaid by Prescriber Specialty-Settinga
Antipsychotic Prescribing for Adults in Missouri Medicaid by Prescriber Specialty-Settinga
Table 3.  
Population-Based Metabolic Testing Rates
Population-Based Metabolic Testing Rates
Table 4.  
Odds of Failure to Test for Glucose Among New Users of Antipsychotic Medicationa
Odds of Failure to Test for Glucose Among New Users of Antipsychotic Medicationa
Table 5.  
Odds of Failure to Test for Lipids Among New Users of Antipsychotic Medicationa
Odds of Failure to Test for Lipids Among New Users of Antipsychotic Medicationa
1.
Newcomer  JW.  Second-generation (atypical) antipsychotics and metabolic effects: a comprehensive literature review.  CNS Drugs. 2005;19(suppl 1):1-93.PubMedArticle
2.
American Diabetes Association; American Psychiatric Association; American Association of Clinical Endocrinologists; North American Association for the Study of Obesity.  Consensus development conference on antipsychotic drugs and obesity and diabetes.  Diabetes Care. 2004;27(2):596-601.PubMedArticle
3.
Lean  ME, Pajonk  FG.  Patients on atypical antipsychotic drugs: another high-risk group for type 2 diabetes.  Diabetes Care. 2003;26(5):1597-1605.PubMedArticle
4.
Nasrallah  HA.  Metabolic findings from the CATIE trial and their relation to tolerability.  CNS Spectr. 2006;11(7)(suppl 7):32-39.PubMed
5.
Health Canada. Atypical antipsychotics: impaired glucose metabolism.  Can Adverse Drug React Newsl. 2011;11(4):3-6.
6.
Rosack  J. FDA to require diabetes warning on antipsychotics. http://psychnews.psychiatryonline.org/doi/10.1176/pn.38.20.0001a. Published October 17, 2003. Accessed April 11, 2016.
7.
MedWatch safety alerts for human medical products. Lilly. Zyprexa (olanzapine): Dear Healthcare Practitioner letter, March 1, 2004. http://www.fda.gov/medwatch/SAFETY/2004/Zyprexa_deardoc.pdf. Published March 1, 2004. Accessed November 28, 2005.
8.
Marder  SR, Essock  SM, Miller  AL,  et al.  Physical health monitoring of patients with schizophrenia.  Am J Psychiatry. 2004;161(8):1334-1349.PubMedArticle
9.
Woo  V, Harris  SB, Houlden  RL; Clinical & Scientific Section, Canadian Diabetes Association.  Canadian Diabetes Association position paper: antipsychotic medications and associated risks of weight gain and diabetes.  Can J Diabetes. 2005;29(2):111-112.
10.
American Diabetes Association; American Psychiatric Association; American Association of Clinical Endocrinologists; North American Association for the Study of Obesity.  Consensus development conference on antipsychotic drugs and obesity and diabetes.  J Clin Psychiatry. 2004;65(2):267-272.PubMedArticle
11.
Strategies to integrate physical health care into mental health (theme issue). J Clin Psychiatry. 2007;68(suppl 4). http://www.psychiatrist.com/JCP/TOC/Pages/t68s04.aspx. Accessed April 11, 2016.
12.
Phend  C. ADA: Metabolic monitoring guidelines for antipsychotics largely unheeded. MedPage Today.http://www.medpagetoday.com/MeetingCoverage/ADA/9746. June 10, 2008. Accessed April 11, 2016.
13.
Newcomer  JW. Antipsychotic medications and the risk of diabetes and cardiovascular disease. Paper presented at American Diabetes Association 65th Scientific Sessions; June 10-14, 2005.
14.
National Committee for Quality Assurance. Diabetes and cardiovascular disease screening and monitoring for people with schizophrenia or bipolar disorder. https://www.ncqa.org/report-cards/health-plans/state-of-health-care-quality/2015-table-of-contents/schizophrenia. Published 2014. Accessed April 11, 2016.
15.
Morrato  EH, Druss  B, Hartung  DM,  et al.  Metabolic testing rates in 3 state Medicaid programs after FDA warnings and ADA/APA recommendations for second-generation antipsychotic drugs.  Arch Gen Psychiatry. 2010;67(1):17-24.PubMedArticle
16.
Haupt  DW, Rosenblatt  LC, Kim  E, Baker  RA, Whitehead  R, Newcomer  JW.  Prevalence and predictors of lipid and glucose monitoring in commercially insured patients treated with second-generation antipsychotic agents.  Am J Psychiatry. 2009;166(3):345-353.PubMedArticle
17.
Morrato  EH, Newcomer  JW, Kamat  S, Baser  O, Harnett  J, Cuffel  B.  Metabolic screening after the ADA’s consensus statement on antipsychotic drugs and diabetes.  Diabetes Care. 2009;32(6):1037-1042.PubMedArticle
18.
Morrato  EH, Nicol  GE, Maahs  D,  et al.  Metabolic screening in children receiving antipsychotic drug treatment.  Arch Pediatr Adolesc Med. 2010;164(4):344-351.PubMedArticle
19.
Morrato  EH, Druss  BG, Hartung  DM,  et al.  Small area variation and geographic and patient-specific determinants of metabolic testing in antipsychotic users.  Pharmacoepidemiol Drug Saf. 2011;20(1):66-75.PubMedArticle
20.
Nicol  GE, Campagna  EJ, Garfield  LD, Newcomer  JW, Parks  JJ, Morrato  EH.  The role of clinical setting and management approach in metabolic testing among youths and adults treated with antipsychotics.  Psychiatr Serv. 2016;67(1):128-132.PubMedArticle
21.
Moeller  KE, Rigler  SK, Mayorga  A, Nazir  N, Shireman  TI.  Quality of monitoring for metabolic effects associated with second generation antipsychotics in patients with schizophrenia on public insurance.  Schizophr Res. 2011;126(1-3):117-123.PubMedArticle
22.
Rubin  DM, Kreider  AR, Matone  M,  et al.  Risk for incident diabetes mellitus following initiation of second-generation antipsychotics among Medicaid-enrolled youths.  JAMA Pediatr. 2015;169(4):e150285.PubMedArticle
23.
Ray  WA.  Evaluating medication effects outside of clinical trials: new-user designs.  Am J Epidemiol. 2003;158(9):915-920.PubMedArticle
24.
 Clinical Classifications Software (CCS). Rockville, MD: Agency for Healthcare Research & Quality; 2014.
25.
ProviderPRO. http://www.healthcaredatasolutions.com/healthcare-databases/providerpro.html. Updated 2016. Accessed October 18, 2014.
26.
Morrato  EH, Brewer  S, Campagna  E,  et al.  Glucose testing for adults receiving Medicaid and antipsychotics: a population-based prescriber survey on behaviors, attitudes, and barriers [published online April 1, 2016].  Psychiatr Serv.PubMed
27.
American Diabetes Association.  Standards of medical care in diabetes—2011.  Diabetes Care. 2011;34(suppl 1):S11-S61.PubMedArticle
28.
Perneger  TV.  What’s wrong with Bonferroni adjustments.  BMJ. 1998;316(7139):1236-1238.PubMedArticle
29.
Gerrity  M.  Integrating Primary Care into Behavioral Health Settings: What Works for Individuals with Serious Mental Illness. New York, NY: Milbank Memorial Fund; 2014.
30.
Takach  M.  About half of the states are implementing patient-centered medical homes for their Medicaid populations.  Health Aff (Millwood). 2012;31(11):2432-2440.PubMedArticle
31.
Parks  J.  Gold Award: community-based program: a health care home for the “whole person” in Missouri’s community mental health centers: Missouri Community Mental Health Center Health Home Program, Jefferson City, Missouri.  Psychiatr Serv. 2015;66(10):e5-e8.PubMedArticle
32.
US Preventive Services Task Force. Final update summary: lipid disorders in adults (cholesterol, dyslipidemia): screening. http://www.uspreventiveservicestaskforce.org/Page/Document/UpdateSummaryFinal/lipid-disorders-in-adults-cholesterol-dyslipidemia-screening. Published 2015. Accessed September 1, 2015.
33.
US Preventive Services Task Force. Final update summary: diabetes mellitus (type 2) in adults: screening. http://www.uspreventiveservicestaskforce.org/Page/Document/UpdateSummaryFinal/diabetes-mellitus-type-2-in-adults-screening. Published 2015. Accessed September 1, 2015.
34.
Heinssen  RK, Goldstein  AB, Azrin  ST.  Evidence-Based Treatments for First Episode Psychosis: Components of Coordinated Specialty Care. Bethesda, MD: National Institute for Mental Health; 2014.
35.
 The cost-effectiveness of screening for type 2 diabetes. CDC Diabetes Cost-Effectiveness Study Group, Centers for Disease Control and Prevention.  JAMA. 1998;280(20):1757-1763.PubMedArticle
36.
Dillman  DA, Smyth  JD, Christian  LM.  Internet, Mail, and Mixed-Mode Surveys: The Tailored Design Method. 3rd ed. Hoboken, NJ: John Wiley & Sons, Inc; 2009.
37.
Nardone  M, Paradise  J.  Issue Brief: Medicaid Health Homes: A Profile of Newer Programs. Menlo Park, CA: Kaiser Commission on Medicaid and the Uninsured; 2014.
Original Investigation
July 2016

Metabolic Testing for Adults in a State Medicaid Program Receiving AntipsychoticsRemaining Barriers to Achieving Population Health Prevention Goals

Author Affiliations
  • 1Department of Health Systems, Management, and Policy, Colorado School of Public Health, University of Colorado Anschutz Medical Campus, Aurora
  • 2Adult and Child Consortium for Health Outcomes Research and Delivery Science, Children's Hospital Colorado, University of Colorado Anschutz Medical Campus, Aurora
  • 3Department of Family Medicine, School of Medicine, University of Colorado Anschutz Medical Campus, Aurora
  • 4Department of Geography and Environmental Sciences, College of Liberal Arts and Sciences, University of Colorado, Denver
  • 5Department of Communication, College of Communication Arts and Sciences, Michigan State University, East Lansing
  • 6Department of Health Policy and Management, Rollins School of Public Health, Emory University, Atlanta, Georgia
JAMA Psychiatry. 2016;73(7):721-730. doi:10.1001/jamapsychiatry.2016.0538
Abstract

Importance  Medicaid quality indicators track diabetes mellitus and cardiovascular disease screening in adults receiving antipsychotics and/or those with serious mental illness.

Objective  To inform performance improvement interventions by evaluating the relative importance of patient, prescriber, and practice factors affecting metabolic testing.

Design, Setting, and Participants  A retrospective cohort study was conducted using Missouri Medicaid administrative claims data (January 1, 2010, to December 31, 2012) linked with prescriber market data. The analysis included 9316 adults (age, 18-64 years) who were starting antipsychotic medication. Secondary analysis included the subset of adults (n = 1813) for whom prescriber knowledge, attitudes, and behavior survey data were available. Generalized estimating equations were performed to identify factors associated with failure to receive annual testing during antipsychotic treatment (adjusted odds ratio [OR], <1 favor testing). Data analysis was performed from October 1, 2014, to February 18, 2016.

Exposure  Oral second-generation antipsychotics.

Main Outcomes and Measures  A medical claim for glucose or lipid testing occurring within 180 days before and after the antipsychotic prescription claim.

Results  The 9317 patients (mean [SD] age, 37.6 [12.0] years) initiated antipsychotic medication in a variety of prescriber specialty-settings: 24.3%, community mental health center (CMHC); 27.6%, non-CMHC behavioral health; 24.3%, primary care practitioners; and 23.8%, other/unknown. Annual testing rates were 79.6% for glucose and 41.2% for lipids. Failure to test glucose and lipids was most strongly associated with patient factors and health care utilization. To illustrate by using findings from glucose modeling (reported as adjusted OR [95% CI]), lower failure to receive testing was associated with older age (40-49 vs 18-29 years; 0.64 [0.55-0.74]), diagnosis of schizophrenia or bipolar disorder (0.55 [0.44-0.67]), cardiometabolic comorbidity (dyslipidemia, 0.28 [0.22-0.37]), hypertension (0.59 [0.50-0.69]), and greater outpatient utilization (>6 encounters vs none; 0.33 [0.28-0.39]). Analysis incorporating prescriber practice information found lower failure to receive glucose testing if the patient received care at a CMHC (0.74 [0.64-0.85]) or if the initiating prescriber was a primary care practitioner (0.81 [0.66-1.00]). However, the initiating prescriber specialty-setting was not associated with lipid testing.

Conclusions and Relevance  Compared with prior reports, progress has been made to improve diabetes screening, but lipid screening remains particularly underutilized. Medicaid performance improvement initiatives should target all prescriber settings and not just behavioral health.

Introduction

Antipsychotic medication contributes to increased metabolic risk.1 A decade ago, the American Psychiatric Association, American Diabetes Association, and other organizations issued recommendations2 for glucose and lipid monitoring with second-generation antipsychotics. Since then, metabolic risk information on antipsychotics has been broadly disseminated in the medical literature1,3,4 through warnings,57 recommendations,2,810 and medical education.1113 Diabetes screening became a Healthcare Effectiveness Data and Information Set (HEDIS) performance measure in 2014 for adults with serious mental illness served by Medicaid and receiving antipsychotics.14 HEDIS metrics, published by the National Committee for Quality Assurance, are used by more than 90% of US health plans to measure performance on important dimensions of health care.

Early studies1517 reported low rates of metabolic testing with little to no change following the warnings. Population-based studies1822 examining factors associated with receiving metabolic testing among Medicaid patients are limited. Research has examined patient characteristics associated with testing19,20 or has focused exclusively on patients with schizophrenia21 or on children.18,22 Prescriber and system factors affecting metabolic testing remain understudied. Knowledge of which patient, prescriber, practice, and/or system barriers are key bottlenecks in testing is needed to prioritize intervention strategies.

The first objective of the present study was to assess glucose and lipid testing rates among adults who received new prescriptions for antipsychotic medication in a state Medicaid program by practitioner specialty-setting to identify which prescriber segments to target for the greatest population health impact. The second objective was to determine which hypothesized patient, practice, and/or prescriber factors were associated with failure to test. The study was designed to inform national and state policymakers, clinical practitioners, patient advocacy groups, and other individuals responsible for ensuring that adults with serious mental illness receive appropriate clinical preventive services.

Box Section Ref ID

Key Points

  • Question What is the relative association of hypothesized patient, prescriber, and practice factors and annual glucose and lipid testing among adults receiving antipsychotics?

  • Findings In this retrospective cohort study of adults starting antipsychotic therapy in Missouri Medicaid, the strongest and most consistent factors associated with failure to use metabolic testing were patient characteristics (ie, younger age, lack of other risk factors, and infrequency of outpatient care). Gaps in metabolic testing were observed in all settings.

  • Meaning Medicaid education, performance improvement, and policy should target all care settings, and younger adults (<40 years) may be a particularly important group on which to focus prevention efforts because of low screening rates.

Methods
Study Population

This retrospective, observational cohort study was conducted among new users (age, 18-64 years) of oral second-generation antipsychotics in Missouri Medicaid using administrative claims data from January 1, 2010, to December 31, 2012 (eTable 1 in the Supplement). Oral second-generation antipsychotics (henceforth referred to as antipsychotics) included aripiprazole, asenapine maleate, clozapine, iloperidone, lurasidone hydrochloride, olanzapine, paliperidone, quetiapine fumarate, risperidone, and ziprasidone. The index antipsychotic was classified as high or low/medium metabolic risk using information from US Food and Drug Administration product labeling and the American Diabetes Association.2 On-label antipsychotic use was defined using Missouri Medicaid’s approved diagnosis list. The index date was the date of the first fill for an antipsychotic in 2011. A new user23 was defined as a patient with no claims for an antipsychotic prescription in the 180 days prior to the index date. Patients were excluded if they were Medicare dual eligible or had Medicaid eligibility for less than 180 days before and after the index date (the study period).

Patients had a unique encrypted identifier that permitted linkage of their medical, pharmacy, and laboratory claim records. Prescribing health care professional identification was also available to link the practitioner and practice characteristics from market and survey data. The study received approval from the Colorado Multiple Institutional Review Board and adhered to Missouri’s Data Use Agreements. A waiver of documentation of consent for physicians completing the survey and a waiver of consent for analysis of the patient claims data were obtained from the Colorado Multiple Institutional Review Board.

Assessment of Metabolic Testing

Missouri Medicaid reimbursed for testing in all treatment settings. Annual testing was defined as a laboratory claim for glucose or lipid serum testing occurring in the 360-day study period.15 Metabolic testing rates were also assessed by emulating the HEDIS performance measures for adults, defined as “the percentage of members 18 to 64 years of age with schizophrenia or bipolar disorder who were dispensed an antipsychotic medication and had a diabetes screening test during the measurement year.”14

Assessment of Patient Characteristics

Age at index date, race, and sex were obtained from the administrative claims data. Comorbidity and health care utilization have been strongly associated19 with metabolic screening. Schizophrenia or bipolar disorder, diabetes mellitus, disorders of lipid metabolism, hypertension, and heart disease were identified using diagnosis codes ascertained from medical claims. Additional mental health conditions were identified from diagnosis codes using the Clinical Classifications Software (categories 650-662, with alcohol- and substance-related disorders merged) developed by the Agency for Healthcare Research and Quality.24 Concomitant psychotropic medication use was identified from pharmacy claims.

Assessment of Prescriber and Practice Characteristics

Pharmacy claims were used to characterize annual antipsychotic prescribing volume. Location of the index prescriber (in Missouri or a bordering state) and whether the patient received care at a community mental health center (CMHC) were flagged. ProviderPRO, a publically available health care professional database,25 was used to ascertain the practitioners’ specialty, age, and practice size. Practitioner specialty-setting was categorized as behavioral health in a CMHC, behavioral health (non-CMHC), primary care, and other/unknown.

A survey was fielded to enhance the available data and assess a range of health care professional, practice, and patient factors hypothesized to affect metabolic screening. It was mailed to all practitioners who initiated antipsychotic medication to Missouri Medicaid beneficiaries in 2011. Surveys were mailed in 2 waves: CMHC practitioners (late 2011-2012) and all practitioners (2013). The CMHCs were resurveyed in 2013. Attitudinal questions included screening intention, responsibility, knowledge, beliefs that screening will reduce risk (response efficacy), confidence in ordering and interpreting results (self-efficacy), and barriers. Methods and survey items have been described.26

Statistical Analysis

Descriptive statistics were performed to assess patient, prescriber, and practice characteristics among new users (primary cohort) and the subset for whom prescriber survey results were available (secondary cohort). Rates of glucose and lipid testing were determined. Annual rates of glucose testing were reported among adults without diabetes mellitus based on the number of ascertainable type 2 diabetes risk factors (nonwhite race, age ≥45 years, and presence of hypertension, dyslipidemia, and heart disease).27 The distribution of prescribers and patients initiating antipsychotic medication was displayed by specialty-setting.

Generalized estimating equations with a logit link were used to identify the association of patient and index prescriber characteristics with failure to have an annual glucose or lipid test (modeled separately). Clustering of patients by index prescriber was accounted for by using an exchangeable correlation matrix. All independent covariates were retained in the final models regardless of significance. Sensitivity analysis was performed among persistent users of antipsychotics, defined as having no gap in oral antipsychotic therapy lasting >30 days.15 All reported P values are from 2-sided hypothesis tests, and statistical significance was defined at α = .05. Adjustment for multiple comparisons was not performed.28 All statistical analyses used SAS, version 9.4 (SAS Institute Inc). The eMethods and eTables 2-5 in the Supplement provide details on study measures and sensitivity analyses. Data analysis was performed from October 1, 2014, to February 18, 2016.

Results

There were 9316 new users of antipsychotic medication, 1813 of whom had index prescribers who completed the survey. Table 1 summarizes patient, prescriber, and practice characteristics. The mean (SD) age of new users was 37.6 (12.0) years, 3335 (35.8%) were male, 1855 (19.9%) had a schizophrenia or bipolar disorder diagnosis, and 634 (6.8%) had therapy initiated in a state bordering Missouri. Antipsychotics with low or medium metabolic risk were prescribed for 8467 (90.9%) of the patients.

Table 2 and eTable 6 in the Supplement report the population distribution of antipsychotic prescribing and patient mental health profile by prescriber specialty setting within Missouri Medicaid. Antipsychotic prescribing by share of patients was fairly evenly distributed across specialty-setting: CHMC, 2263 (24.3%); non-CMHC behavioral health, 2574 (27.6%); primary care, 2261 (24.3%); and other/unknown, 2218 (23.8%). Specialty of index prescriber for the patients falling in the other/unknown category (n = 2218) was out-of-state (339 [15.3%]), emergency medicine (176 [7.9%]), neurology (83 [3.7%]), other (167 [7.5%]), and unknown (1453 [65.5%]). Mean new starts within Medicaid per prescriber per year was 16.8 for CMHC prescribers, 9.9 for non-CMHC behavioral health prescribers, 2.9 for primary care prescribers, and 4.2 for other/unknown.

As reported in Table 1, the demographic profile of the secondary cohort was similar to that of the primary cohort with the exception that individuals in the secondary cohort were more likely to be white (1536 [84.7%] vs 7323 [78.6%]), receive care at a CMHC (1021 [56.3%] vs 4589 [49.3%]), and receive their index prescription from a CMHC practitioner. Patients in the secondary cohort were more likely to have a practitioner who prescribed more antipsychotics (552 [30.4%] with ≥1500 Medicaid prescription claims in 2011 vs 2229 [23.9%] for all prescribers). Based on the survey results, 1296 patients (71.5%) were initiated on antipsychotic therapy in practices that used an electronic health record system; however, only 18.0% had prescribers who operated within a shared mental health and medical care facility. Intention to screen was high, with most patients having an index prescriber stating they would definitely order glucose (1371 [75.6%]) and lipid profile (1347 [74.3%]) tests at annual follow-up. Patients forgetting to complete laboratory work scored highest on perceived barriers (524 patients [28.9%] had an index prescriber who strongly agreed). Many (1330 [73.4%]) patients had an index prescriber who strongly disagreed that “metabolic screening is not a priority for me or my organization.” Nearly 80% of patients had an index prescriber who was very confident in performing clinical tasks related to screening (eg, interpreting blood glucose laboratory values and diagnosing diabetes).

Table 3 presents population-based metabolic testing rates. Annual glucose testing was higher than lipid testing among new users of antipsychotics (79.6% vs 41.2%, respectively). Among adults without diabetes mellitus, annual glucose testing ranged from 68.1% (no ascertainable type 2 diabetes risk factors) to 92.8% (≥3 risk factors). Using the simulated HEDIS measure, 92.4% of adults with schizophrenia or bipolar disorder receiving antipsychotic medication were screened for diabetes. Annual glucose monitoring for people with schizophrenia who already had diabetes was lower (69.7%). The proportion of patients receiving no annual glucose testing was similar across prescriber specialty-settings (Table 2), as well as the proportion receiving no lipid testing: 60.6% for CMHC prescribers, 54.0% for non-CMHC behavioral health prescribers, 59.1% for primary care prescribers, and 61.2% for other/unknown prescribers.

Table 4 and Table 5 present the adjusted odds of failure to test for glucose and lipids during the year. Odds ratios less than 1.0 indicate factors favoring testing. Examining the adjusted analysis in the primary cohort, increasing age, schizophrenia or bipolar disorder, alcohol or substance abuse disorder, cardiovascular-related clinical condition, care provided at a CMHC, use of mood stabilizers, and greater frequency of health care utilization were the strongest factors favoring glucose testing. Results for failure to perform lipid testing differed from glucose testing in that (1) schizophrenia or bipolar disorder was not associated with lipid testing, (2) an alcohol or substance abuse disorder or an index prescriber who wrote fewer antipsychotic prescriptions was associated with a failure to test for lipids, and (3) greater frequency of emergency department encounters and hospitalizations was associated with a lack of lipid testing in the outpatient setting. Sensitivity analyses performed among persistent users of antipsychotics (32.1% of patients) found similar findings regarding factors associated with a failure to receive glucose or lipid testing (eMethods, eTable 4, eTable 5, and eTable 9 in the Supplement).

Analysis of the secondary cohort of patients whose index prescriber responded to the survey permitted the investigation of an expanded set of prescriber and practice factors hypothesized to be associated with metabolic testing. See eTables 7 and 8 in the Supplement for results of additional prescriber (age, sex, screening intention, perceived barriers, and attitudes toward metabolic screening) and practice (size, use of electronic health records, and shared mental and medical health facilities) characteristics included in the adjusted models. Having an index prescriber who was a primary care practitioner was associated with a higher likelihood of glucose testing; however, prescriber specialty had no association on the odds of lipid testing. None of the other practice characteristics or attitudinal statements assessed were associated with the odds of metabolic testing with the exception that having an index prescriber who strongly disagreed that screening added complexity to the clinical workload was associated with a failure to receive glucose testing.

Discussion

To our knowledge, this is the first systematic investigation of patient, prescriber, and practice-level barriers affecting metabolic screening in Medicaid from a state system perspective. Higher rates of annual glucose and lipid testing were found compared with prior reports indicating progress in mitigating cardiometabolic risk among adults receiving antipsychotics. Approximately 80% of adults underwent glucose testing (an absolute increase in testing rates of approximately 30% compared with 2005-2006 claims data from Missouri Medicaid15); using the HEDIS metric, this rate exceeded 90% for adults with schizophrenia or bipolar disorder. Improvement in lipid testing was more limited; approximately 41% of patients received an annual lipid test (an absolute increase of approximately 10% compared with the 2005-2006 claims data15). An antipsychotic with low or moderate metabolic risk was initiated in 90.9% of the patients, which is another possible indicator of intentions to mitigate risk.

Higher rates of metabolic testing are good news; however, screening is a first step toward reducing cardiometabolic risk in this population. Over the past decade, there has been significant investment to prevent and reduce chronic disease among persons with mental illness through better integration of medical and behavioral health care. Federal funding initiatives have accelerated efforts to integrate care; these include Primary and Behavioral Health Care Integration service grants from the Substance Abuse and Mental Health Services Administration, the Health Home Initiative (section 2703) under the Affordable Care Act, and the Centers for Medicare & Medicaid Services Comprehensive Primary Care Initiative.29 Missouri was an early adopter of the Health Home model targeting beneficiaries with a behavioral health condition.30 In 2015, the Missouri CMHC Health Home Program received the American Psychiatric Association Gold Award for “leadership in establishing health homes to provide integrated care to its CMHC clients.”31(pe5) These models may hold promise both for increasing screening and for providing treatment to those who screen positive. However, the direct causal effects of federal and state policies should be interpreted with caution until formal trend analysis testing is performed. Furthermore, findings should be extrapolated with caution to other states and settings with fewer resources than those found in Missouri.

A notable finding of the present study is understanding who is prescribing antipsychotics. Approximately 75% of patients initiated therapy with a prescriber not practicing in a CMHC (the designated health home for coordinated care), and approximately half of the patients initiated therapy with a nonbehavioral health care professional. Federal and state investment to prevent and reduce cardiovascular disease among persons with mental illness has focused on psychiatrists practicing in community mental health settings. However, the percentage of untested patients was not concentrated in one care setting more than another. Therefore, education and new performance improvement initiatives within Medicaid will need to target all care settings, including primary care, to improve rates of screening among patients treated with antipsychotic medications.

Failure to receive metabolic testing was most strongly associated with patient characteristics and factors affecting frequency of health care utilization. Lack of testing was highest among younger adults (<40 years) with fewer chronic conditions. One possible reason for low rates of testing may be the result of conflicting messaging on whether these adults should be targeted for screening. Some guidelines call for use of screening tests in all patients receiving antipsychotics regardless of mental health diagnosis (ie, American Diabetes Association, American Psychiatric Association Consensus Conference,2 and US Food and Drug Administration drug labeling); some guidelines prioritize screening only in adults with schizophrenia or bipolar disorder (ie, HEDIS14), and other guidelines have triaged screening based on age, ethnicity, and risk factors without mental illness cited as a risk factor (eg, US Preventive Services Task Force32,33 and American Diabetes Association,27 although antipsychotic use is referenced as a risk factor for overweight adults). Resolving guidance discrepancies will be necessary to establish target screening rates for the population of adults receiving antipsychotics, especially given the diversity of prescriber specialties. Discrepancies in screening recommendations may also explain, in part, the disparities observed between annual glucose and lipid testing rates since differences in annual rates may relate to differences in the recommended frequency of testing (eg, annually vs every 5 years).

Because state policymakers consider interventions, it is equally important to learn that several hypothesized prescriber and practice barriers (eg, lack of agreement on risk perception, added complexity to the clinical workload, confidence in interpreting laboratory values and diagnosing metabolic disorders, and prescriber perceptions that patients forget to be tested or do not view testing as important) were not strongly associated with failure to test once patient demographic and clinical characteristics were considered. Findings from this study indicate that interventions to achieve population-based screening goals for adults receiving antipsychotics should target younger (<40 years), relatively healthier adults with fewer chronic conditions who have the lowest rates of screening. Previous studies34 have recommended that emphasis should be given to metabolic risk factors, such as prediabetes, at first-episode psychosis to promote appropriate prevention. Diabetes screening is more cost-effective at younger ages.35

The results of this research are subject to limitations. A potential limitation of the analysis of prescriber and practice factors using survey data is nonresponse selection bias.36 To address this concern, we conducted a formal analysis (eTables 10 and 11 in the Supplement). Responders had more Medicaid antipsychotic prescription claims than did nonresponders and were more likely to be a CMHC or primary care practitioner. Another limitation is that measurement issues—specifically, the selection and availability of patient, physician, and practice factors as well as social desirability bias in the survey responses—may have contributed to a weakening of the strength of correlations. Antipsychotic prescribers within Missouri Medicaid and the state health care system in which they practice may not be nationally representative. Although Missouri’s model of care is similar to that of other states with health home initiatives,37 having strong statewide leaders in Missouri emphasizing the importance of health care for people with serious mental illness is unique. Medicare dual-eligible patients were excluded owing to incomplete data, so findings may not be generalizable to all new users of antipsychotics. Adherence to fasting recommendations for glucose and cholesterol testing is unknown. Finally, this study focused on the upstream issue of screening because this issue is where there has been a barrier to care. Further research is needed to investigate metabolic treatment barriers following screening.

Conclusions

Adults with serious mental illness should receive appropriate clinical preventive services, including regular and early glucose and lipid testing. Metabolic risk information on antipsychotics has been broadly disseminated over the past decade. Still, as is often the case in the diffusion of a new practice, knowledge does not necessarily equal behavioral change. Progress has been made to improve diabetes testing, but lipid testing remains particularly underutilized in adults receiving antipsychotic medication. Given the large proportion of antipsychotics prescribed by non–mental health care professionals, education and performance improvement initiatives should also be targeted for these settings. Consistency and redundancy in messaging about the advisability of annual screening is important, and professional associations, federal agencies, and health care organizations should reach consensus on testing recommendations so that they reinforce the same message. Further research is needed to ensure that patients with positive results of screening tests receive good follow-up care.

Back to top
Article Information

Corresponding Author: Elaine H. Morrato, DrPH, MPH, Department of Health Systems, Management, and Policy, Colorado School of Public Health, University of Colorado Anschutz Medical Campus, Mail Stop B119, 13001 E 19th Pl, Aurora, CO 80045 (elaine.morrato@ucdenver.edu).

Submitted for Publication: October 13, 2015; final revision received February 19, 2016; accepted February 21, 2016.

Published Online: May 11, 2016. doi:10.1001/jamapsychiatry.2016.0538.

Author Contributions: Dr Morrato and Ms Campagna had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Morrato, Campagna, Dickinson, Thomas, Miller, Druss.

Acquisition, analysis, or interpretation of data: Morrato, Campagna, Brewer, Dickinson, Thomas, Miller, Dearing, Lindrooth.

Drafting of the manuscript: Morrato, Campagna, Miller.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Morrato, Campagna, Dickinson, Lindrooth.

Obtained funding: Morrato Lindrooth.

Administrative, technical, or material support: Campagna, Brewer, Miller.

Study supervision: Morrato, Dickinson.

Conflict of Interest Disclosures: Dr Morrato has received consulting fees and travel funds from Merck and Janssen Pharmaceutical. Dr Morrato and Ms Campagna have received research funding from Janssen Pharmaceuticals.

Funding/Support: Research funding was obtained through grants R21 MH 097045 and R44 AG038316 from the National Institutes of Health (NIH) and grant K12 HS019464 from the Association for Healthcare Research and Quality (AHRQ).

Role of the Funder/Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The contents of this article are the authors’ sole responsibility and do not necessarily represent official NIH or AHRQ views.

Previous Presentation: The results of the study were presented in part at the 31st International Conference on Pharmacoepidemiology and Therapeutic Risk Management; August 26, 2015; Boston, Massachusetts.

Additional Contributions: Joseph Parks, MD (State of Missouri Department of Social Services), John W. Newcomer, MD (Florida Atlantic University), and Rhonda Driver, RPh (State of Missouri Department of Social Services), contributed substantially to study design and data interpretation. There was no financial compensation.

References
1.
Newcomer  JW.  Second-generation (atypical) antipsychotics and metabolic effects: a comprehensive literature review.  CNS Drugs. 2005;19(suppl 1):1-93.PubMedArticle
2.
American Diabetes Association; American Psychiatric Association; American Association of Clinical Endocrinologists; North American Association for the Study of Obesity.  Consensus development conference on antipsychotic drugs and obesity and diabetes.  Diabetes Care. 2004;27(2):596-601.PubMedArticle
3.
Lean  ME, Pajonk  FG.  Patients on atypical antipsychotic drugs: another high-risk group for type 2 diabetes.  Diabetes Care. 2003;26(5):1597-1605.PubMedArticle
4.
Nasrallah  HA.  Metabolic findings from the CATIE trial and their relation to tolerability.  CNS Spectr. 2006;11(7)(suppl 7):32-39.PubMed
5.
Health Canada. Atypical antipsychotics: impaired glucose metabolism.  Can Adverse Drug React Newsl. 2011;11(4):3-6.
6.
Rosack  J. FDA to require diabetes warning on antipsychotics. http://psychnews.psychiatryonline.org/doi/10.1176/pn.38.20.0001a. Published October 17, 2003. Accessed April 11, 2016.
7.
MedWatch safety alerts for human medical products. Lilly. Zyprexa (olanzapine): Dear Healthcare Practitioner letter, March 1, 2004. http://www.fda.gov/medwatch/SAFETY/2004/Zyprexa_deardoc.pdf. Published March 1, 2004. Accessed November 28, 2005.
8.
Marder  SR, Essock  SM, Miller  AL,  et al.  Physical health monitoring of patients with schizophrenia.  Am J Psychiatry. 2004;161(8):1334-1349.PubMedArticle
9.
Woo  V, Harris  SB, Houlden  RL; Clinical & Scientific Section, Canadian Diabetes Association.  Canadian Diabetes Association position paper: antipsychotic medications and associated risks of weight gain and diabetes.  Can J Diabetes. 2005;29(2):111-112.
10.
American Diabetes Association; American Psychiatric Association; American Association of Clinical Endocrinologists; North American Association for the Study of Obesity.  Consensus development conference on antipsychotic drugs and obesity and diabetes.  J Clin Psychiatry. 2004;65(2):267-272.PubMedArticle
11.
Strategies to integrate physical health care into mental health (theme issue). J Clin Psychiatry. 2007;68(suppl 4). http://www.psychiatrist.com/JCP/TOC/Pages/t68s04.aspx. Accessed April 11, 2016.
12.
Phend  C. ADA: Metabolic monitoring guidelines for antipsychotics largely unheeded. MedPage Today.http://www.medpagetoday.com/MeetingCoverage/ADA/9746. June 10, 2008. Accessed April 11, 2016.
13.
Newcomer  JW. Antipsychotic medications and the risk of diabetes and cardiovascular disease. Paper presented at American Diabetes Association 65th Scientific Sessions; June 10-14, 2005.
14.
National Committee for Quality Assurance. Diabetes and cardiovascular disease screening and monitoring for people with schizophrenia or bipolar disorder. https://www.ncqa.org/report-cards/health-plans/state-of-health-care-quality/2015-table-of-contents/schizophrenia. Published 2014. Accessed April 11, 2016.
15.
Morrato  EH, Druss  B, Hartung  DM,  et al.  Metabolic testing rates in 3 state Medicaid programs after FDA warnings and ADA/APA recommendations for second-generation antipsychotic drugs.  Arch Gen Psychiatry. 2010;67(1):17-24.PubMedArticle
16.
Haupt  DW, Rosenblatt  LC, Kim  E, Baker  RA, Whitehead  R, Newcomer  JW.  Prevalence and predictors of lipid and glucose monitoring in commercially insured patients treated with second-generation antipsychotic agents.  Am J Psychiatry. 2009;166(3):345-353.PubMedArticle
17.
Morrato  EH, Newcomer  JW, Kamat  S, Baser  O, Harnett  J, Cuffel  B.  Metabolic screening after the ADA’s consensus statement on antipsychotic drugs and diabetes.  Diabetes Care. 2009;32(6):1037-1042.PubMedArticle
18.
Morrato  EH, Nicol  GE, Maahs  D,  et al.  Metabolic screening in children receiving antipsychotic drug treatment.  Arch Pediatr Adolesc Med. 2010;164(4):344-351.PubMedArticle
19.
Morrato  EH, Druss  BG, Hartung  DM,  et al.  Small area variation and geographic and patient-specific determinants of metabolic testing in antipsychotic users.  Pharmacoepidemiol Drug Saf. 2011;20(1):66-75.PubMedArticle
20.
Nicol  GE, Campagna  EJ, Garfield  LD, Newcomer  JW, Parks  JJ, Morrato  EH.  The role of clinical setting and management approach in metabolic testing among youths and adults treated with antipsychotics.  Psychiatr Serv. 2016;67(1):128-132.PubMedArticle
21.
Moeller  KE, Rigler  SK, Mayorga  A, Nazir  N, Shireman  TI.  Quality of monitoring for metabolic effects associated with second generation antipsychotics in patients with schizophrenia on public insurance.  Schizophr Res. 2011;126(1-3):117-123.PubMedArticle
22.
Rubin  DM, Kreider  AR, Matone  M,  et al.  Risk for incident diabetes mellitus following initiation of second-generation antipsychotics among Medicaid-enrolled youths.  JAMA Pediatr. 2015;169(4):e150285.PubMedArticle
23.
Ray  WA.  Evaluating medication effects outside of clinical trials: new-user designs.  Am J Epidemiol. 2003;158(9):915-920.PubMedArticle
24.
 Clinical Classifications Software (CCS). Rockville, MD: Agency for Healthcare Research & Quality; 2014.
25.
ProviderPRO. http://www.healthcaredatasolutions.com/healthcare-databases/providerpro.html. Updated 2016. Accessed October 18, 2014.
26.
Morrato  EH, Brewer  S, Campagna  E,  et al.  Glucose testing for adults receiving Medicaid and antipsychotics: a population-based prescriber survey on behaviors, attitudes, and barriers [published online April 1, 2016].  Psychiatr Serv.PubMed
27.
American Diabetes Association.  Standards of medical care in diabetes—2011.  Diabetes Care. 2011;34(suppl 1):S11-S61.PubMedArticle
28.
Perneger  TV.  What’s wrong with Bonferroni adjustments.  BMJ. 1998;316(7139):1236-1238.PubMedArticle
29.
Gerrity  M.  Integrating Primary Care into Behavioral Health Settings: What Works for Individuals with Serious Mental Illness. New York, NY: Milbank Memorial Fund; 2014.
30.
Takach  M.  About half of the states are implementing patient-centered medical homes for their Medicaid populations.  Health Aff (Millwood). 2012;31(11):2432-2440.PubMedArticle
31.
Parks  J.  Gold Award: community-based program: a health care home for the “whole person” in Missouri’s community mental health centers: Missouri Community Mental Health Center Health Home Program, Jefferson City, Missouri.  Psychiatr Serv. 2015;66(10):e5-e8.PubMedArticle
32.
US Preventive Services Task Force. Final update summary: lipid disorders in adults (cholesterol, dyslipidemia): screening. http://www.uspreventiveservicestaskforce.org/Page/Document/UpdateSummaryFinal/lipid-disorders-in-adults-cholesterol-dyslipidemia-screening. Published 2015. Accessed September 1, 2015.
33.
US Preventive Services Task Force. Final update summary: diabetes mellitus (type 2) in adults: screening. http://www.uspreventiveservicestaskforce.org/Page/Document/UpdateSummaryFinal/diabetes-mellitus-type-2-in-adults-screening. Published 2015. Accessed September 1, 2015.
34.
Heinssen  RK, Goldstein  AB, Azrin  ST.  Evidence-Based Treatments for First Episode Psychosis: Components of Coordinated Specialty Care. Bethesda, MD: National Institute for Mental Health; 2014.
35.
 The cost-effectiveness of screening for type 2 diabetes. CDC Diabetes Cost-Effectiveness Study Group, Centers for Disease Control and Prevention.  JAMA. 1998;280(20):1757-1763.PubMedArticle
36.
Dillman  DA, Smyth  JD, Christian  LM.  Internet, Mail, and Mixed-Mode Surveys: The Tailored Design Method. 3rd ed. Hoboken, NJ: John Wiley & Sons, Inc; 2009.
37.
Nardone  M, Paradise  J.  Issue Brief: Medicaid Health Homes: A Profile of Newer Programs. Menlo Park, CA: Kaiser Commission on Medicaid and the Uninsured; 2014.
×