RESEARCH LETTER
Adults in the United States with severe mental illness (SMI), such as schizophrenia and bipolar disorder (totaling approximately 7 million), are estimated to die, on average, 25 years earlier than the general population, largely of premature cardiovascular disease.1 The Institute of Medicine2 has called for improvements in health care for this population. Severe mental illness is associated with elevated risk for type 2 diabetes mellitus.3 Treatment with antipsychotic medications contributes to risk, with most evidence focused on second-generation antipsychotic medications, but similar increases in risk are reported with older and newer medications.4 The American Diabetes Association5 recommends annual diabetes screening for patients treated with antipsychotic medications, and public health administrators have targeted this population for improved health screening.6 To our knowledge, no studies have examined screening rates in this highest-risk population of adults with SMI because of limitations in public health medical records. We examined diabetes screening among publicly insured adults with SMI taking antipsychotic medications using matched administrative data for physical and mental health care services in a large health care system. We measured diabetes screening prevalence among patients with SMI treated with antipsychotic medications and assessed characteristics predictive of screening.
METHODS
This retrospective cohort study analyzed data from the California Medicaid (Medi-Cal) and Client and Service Information systems using the 2 study periods January 1, 2009, to December 31, 2009 (period 1), and October 1, 2010, to September 30, 2011 (period 2). Data from period 2 were used to characterize diabetes screening in the subgroup without diabetes mellitus in period 1. Following approval by the UCSF (University of California, San Francisco) Committee of Human Research, the State of California Committee for the Protection of Human Subjects, and the California Department of Health Care Services’ Data and Research Committee, the latter department combined these databases, deidentified data, and created our analytic data set. The following criteria characterized the cohort: (1) age 18 years or older, (2) diagnosis of SMI by a psychiatrist, (3) prescription of an antipsychotic medication at least once during period 1 and period 2, (4) mental health care use during both study periods, (5) Medi-Cal enrollee, and (6) non–dual eligibility for Medicare (because of unavailable Medicare laboratory billing data). The primary outcome measure was evidence of diabetes screening via glucose-specific fasting serum test (Current Procedural Terminology[CPT] code 82947, 82948, 82950, or 82951) or glycated hemoglobin test (CPT code 83036). A secondary outcome was nonspecific screening (eg, nonfasting metabolic panel) (CPT code 80048, 80050, or 80053). Poisson regression was used to estimate the relative prevalence of diabetes screening for each predictor. We estimated associations for receiving diabetes-specific screening vs nonspecific or no screening and then contrasted any screening vs none. Using statistical software (Stata, version 13.2; StataCorp LP), robust standard errors were used to account for clustering of outcomes by county and to accommodate the use of a Poisson model for a binary outcome.
RESULTS
Of 50 915 study participants, 15 315 (30.1%) received diabetes-specific screening (Table 1). Almost one-third, 15 832 (31.1%), received no form of glucose screening in a yearlong period. The strongest correlate of diabetes-specific screening was having at least 1 outpatient primary care visit during the period examined (adjusted prevalence ratio, 1.80; 95% CI, 1.62–2.00; P <.001) (Table 2).
TABLE 1.
Demographic and clinical characteristics of adult Medi-Cal recipients with severe mental illness who received antipsychotics, by diabetes screening status*
| Characteristic | Total (N=50915) |
Diabetes specific screening† (N=15315) |
Non-specific diabetes screening†† (N=19768) |
No diabetes screening (N=15832) |
|---|---|---|---|---|
| Sex | ||||
| Female | 27813 (54.6) | 9018 (32.4) | 10922 (39.3) | 7873 (28.3) |
| Male | 23102 (45.4) | 6297 (27.3) | 8846 (38.3) | 7959 (34.4) |
| Race/ethnicity | ||||
| Asian | 6420 (12.6) | 1983 (30.9) | 2421 (37.7) | 2016 (31.4) |
| Black | 9773 (19.2) | 2865 (29.3) | 3468 (35.5) | 3440 (35.2) |
| Hispanic | 10022 (19.7) | 3142 (31.4) | 3775 (37.7) | 3105 (31.0) |
| Other | 5197 (10.2) | 1602 (30.8) | 2016 (38.8) | 1579 (30.4) |
| White | 19503 (38.3) | 5723 (29.3) | 8088 (41.5) | 5692 (29.2) |
| Age, years§ | ||||
| 18–27 | 8635 (17.0) | 2026 (23.5) | 3293 (38.1) | 3316 (38.4) |
| 28–47 | 21959 (43.1) | 6424 (29.3) | 8525 (38.8) | 7010 (31.9) |
| 48–67 | 20106 (39.5) | 6816 (33.9) | 7872 (39.2) | 5418 (27.0) |
| 68+ | 215 (0.4) | 49 (22.8) | 78 (36.3) | 88 (40.9) |
| County Type | ||||
| Urban | 49474 (97.2) | 14943 (30.2) | 19091 (38.6) | 15440 (31.2) |
| Rural | 1441 (2.8) | 372 (25.8) | 677 (47.0) | 392 (27.2) |
| Psychiatric Diagnosis | ||||
| Anxiety Disorder | 1966 (3.9) | 573 (29.2) | 808 (41.1) | 585 (29.8) |
| Bipolar Disorder | 7369 (14.5) | 2306 (31.3) | 2838 (38.5) | 2225 (30.2) |
| Major Depressive Disorder | 11645 (22.9) | 3520 (30.2) | 4632 (39.8) | 3493 (30.0) |
| Other | 3626 (7.1) | 906 (25.0) | 1518 (41.9) | 1202 (33.1) |
| Schizophrenia Spectrum | 26309 (51.7) | 8010 (30.5) | 9972 (37.9) | 8327 (31.7) |
| Co-morbid substance abuse | ||||
| Yes | 9238 (18) | 2687 (29.1) | 3812 (41.3) | 2739 (29.7) |
| No | 41677 (82) | 12628 (30.3) | 15956 (38.3) | 13093 (31.4) |
|
Co-morbid dyslipidemia or hypertention¶ |
||||
| Yes | 733 (1.4) | 271 (37.0) | 322 (43.9) | 140 (19.1) |
| No | 50194 (98.6) | 15044 (30.0) | 19446 (38.7) | 15692 (31.3) |
| Antipsychotic Medication¥ | ||||
| Second-generation antipsychotic (SGA) |
49273 (96.8) | 14886 (30.2) | 19178 (38.9) | 15209 (30.9) |
| First-generation antipsychotic (FGA) only |
1642 (3.2) | 429 (26.1) | 590 (35.9) | 623 (37.9) |
|
Evidence of primary care outpatient healthcare utilization |
||||
| Yes | 32896 (64.6) | 11786 (35.8) | 13823 (42.0) | 7287 (22.2) |
| No | 18019 (35.4) | 3529 (19.6) | 5945 (33.0) | 8545 (47.4) |
In this table we present column percentages in the total column, and row percentages in the screening columns. The percentages across rows or columns may not add exactly to 100% because of rounding.
Defined by evidence of HgA1c or fasting glucose test in the past year, excluding people with pre-existing evidence of diabetes.
Defined by evidence of a non-fasting glucose test in the past year, excluding people with pre-existing evidence of diabetes.
These age categories were those provided by the California Department of Health Care Services to the study investigators
The American Diabetes Association and American Psychiatric Association recommend annual diabetes screening for anyone taking second-generation antipsychotic (SGA) medications.
Defined by ICD-9 diagnosis of these co-morbid conditions or prescription of medications to treat the diagnosis.
TABLE 2.
Potential factors associated with diabetes screening status†
| Adjusted Prevalence Ratio of Diabetes Specific vs. No Diabetes-Specific Screeningi |
Adjusted Prevalence Ratio of Any Diabetes Screening vs. Noneii |
|||||
|---|---|---|---|---|---|---|
| Characteristic | PR (95% CI) | P-pwise | P | PR (95% CI) | P-pwise | P |
| Sex(a) | . | <0.00005 | . | <0.00005 | ||
| Female | 1.15 (1.10–1.20) | <.00005 | . | 1.08 (1.06, 1.09) | <.00005 | . |
| Male | Ref | . | . | Ref | . | . |
| Race/ethnicity(a) | . | <0.0001 | . | <0.00005 | ||
| Asian | 1.02 (0.94, 1.11) | 0.56 | . | 0.96 (0.91, 1.02) | 0.20 | . |
| Black | 1.00 (0.92, 1.09) | 0.99 | . | 0.92 (0.90, 0.94) | <.00005 | . |
| Hispanic | 1.12 (1.01, 1.24) | 0.03 | . | 1.00 (0.97, 1.04) | 0.96 | . |
| Other | 1.05 (1.00, 1.09) | 0.04 | . | 0.98 (0.95, 1.02) | 0.32 | . |
| White | Ref | . | . | Ref | . | . |
| Age, years§(a) | . | <0.00005 | . | <0.00005 | ||
| 18–27 | Ref | . | . | Ref | . | . |
| 28–47 | 1.23 (1.17, 1.30) | <.00005 | . | 1.09 (1.07, 1.12) | <.00005 | . |
| 48–67 | 1.43 (1.31, 1.55) | <.00005 | . | 1.17 (1.14, 1.21) | <.00005 | . |
| 68+ | 0.93 (0.72, 1.21) | 0.62 | . | 0.93 (0.82, 1.06) | 0.27 | . |
|
Evidence of primary care outpatient healthcare utilization(b) |
. | <0.00005 | . | <0.00005 | ||
| Yes | 1.80 (1.62, 2.00) | <.00005 | . | 1.48 (1.36–1.61) | <.00005 | . |
| No | Ref | . | . | Ref | . | . |
Each adjusted model depends upon the specific variable and their position along with directed acyclic graph (DAG or causal graph). We created a DAG to identify confounders and mediators of the predictors of interest (available upon request).
These age categories were those provided by the California Department of Health Care Services to the study investigators
Diabetes-specific screening (N=15,315) vs. no diabetes-specific screening [includes both non-specific or no screening] (N=35,600).
Any diabetes screening [includes both diabetes-specific screening and non-specific diabetes screening] (N=35,083) vs. no screening (N=15,832)
Controlling for 3 main demographic variables (sex, race/ethnicity, age) and county type (rural/urban), unless it is the predictor variable of interest.
Controlling for main demographic variables, county type, psychiatric diagnosis, comorbid substance abuse, and comorbid metabolic disorders.
DISCUSSION
In this large cohort study of adults with SMI taking antipsychotic medications in the California public mental health care system, almost 70% were not screened for diabetes mellitus using validated screening measures. Individuals with SMI who had at least 1 primary care visit in addition to mental health services were more than twice as likely to be screened than those who did not. This observation supports the value of burgeoning efforts to integrate behavioral health and primary care. Growing evidence supports the value of screening for diabetes mellitus in higher-risk populations, such as those receiving treatment with antipsychotic medications, including first-generation and second-generation agents that commonly result in co-occurring obesity. Future studies should explore barriers to screening in this vulnerable population.
Acknowledgments
This study was initiated during a State Quality Improvement project to integrate primary care and mental health care called the California Mental Health Care Management Program (CalMEND), which was a collaboration between the Department of Mental Health and the Pharmacy Benefits Division of Health Care Services. Martha Shumway, PhD (Department of Psychiatry, University of California, San Francisco) performed initial data cleaning and coding, for which she received compensation. Amy J. Markowitz, JD (Clinical and Translational Research Career Development Program, University of California, San Francisco) provided expert editing, for which she received compensation. We acknowledge CalMEND staff for their assistance, in particular John Igwe for combining administrative databases without compensation.
Dr. Mangurian is supported by Career Development Award K23MH093689 from the National Institutes of Health (NIH), by the UCSF (University of California, San Francisco) Hellman Fellows Award for Early-Career Faculty, and by UCSF–Clinical and Translational Science Institute (UCSF-CTSI) grant UL1 TR000004 from the National Center for Advancing Translational Sciences (NCATS), NIH. Ms. Creasman and Dr. Fuentes-Afflick are supported by UCSF-CTSI grant UL1 TR000004 from the NCATS, NIH. Dr. Fuentes-Afflick is supported by center grant P60MD006902 from the NIH. Dr. Schillinger is supported by center grants P30DK092924-01 and P60MD006902 from the NIH. Dr Newcomer reported serving on the data safety monitoring boards for Amgen, Bristol-Myers Squibb, and Merck and reported receiving honoraria from VIVUS, Cleveland Clinic, American Physician Institute, CME Outfitters, CMEology, American Psychiatric Association, and American Society for Clinical Psychopharmacology (all outside of the present work). Dr Knapp reported serving as medical director of the California Department of Mental Health during the study period. The funding sources 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. The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of the National Institutes of Health.
Footnotes
Disclosures
No other disclosures were reported.
Contributor Information
Dr. Christina Mangurian, University of California, San Francisco (UCSF) Department of Psychiatry; UCSF Center for Vulnerable Populations at San Francisco General Hospital and Trauma Center.
Dr. John W. Newcomer, Charles E. Schmidt College of Medicine at Florida Atlantic University.
Dr. Eric Vittinghoff, UCSF Department of Epidemiology and Biostatistics.
Dr. Jennifer M. Creasman, UCSF Department of Obstetrics and Gynecology.
Dr. Penelope Knapp, University of California, Davis, Department of Psychiatry and Behavioral Sciences.
Dr. Elena Fuentes-Afflick, UCSF Department of Epidemiology and Biostatistics; UCSF Department of Pediatrics.
Dr. Dean Schillinger, UCSF Center for Vulnerable Populations at San Francisco General Hospital and Trauma Center; UCSF Department of Medicine.
REFERENCES
- 1.Colton CW, Manderscheid RW. Congruencies in increased mortality rates, years of potential life lost, and causes of death among public mental health clients in eight states. Preventing chronic disease. 2006;3(2):A42. [PMC free article] [PubMed] [Google Scholar]
- 2.Institute of Medicine. Committee on Crossing the Quality Chasm: Improving the Quality of Health Care for Mental and Substance-Use Conditions. Washington DC: National Academies Press; 2006. [Google Scholar]
- 3.Osborn DP, Wright CA, Levy G, King MB, Deo R, Nazareth I. Relative risk of diabetes, dyslipidaemia, hypertension and the metabolic syndrome in people with severe mental illnesses: systematic review and metaanalysis. BMC Psychiatry. 2008;8:84. doi: 10.1186/1471-244X-8-84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Newcomer JW. Second-generation (atypical) antipsychotics and metabolic effects: a comprehensive literature review. CNS Drugs. 2005;19(Suppl 1):1–93. doi: 10.2165/00023210-200519001-00001. [DOI] [PubMed] [Google Scholar]
- 5.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. Vol 27: Diabetes Care. 2004:596–601. doi: 10.2337/diacare.27.2.596. [DOI] [PubMed] [Google Scholar]
- 6.Parks J, Radke AQ, Mazade NA. Measurement of health status for people with serious mental illness. Alexandria, VA: National Association of State Mental Health Program Directors; 2008. [Google Scholar]