Heterogeneity in the Quality of Care for Patients With Multiple Chronic Conditions by Psychiatric Comorbidity
Domino, Marisa E. PhD*,†; Beadles, Christopher A. MD, PhD*,†; Lichstein, Jesse C. MSPH*; Farley, Joel F. PhD‡; Morrissey, Joseph P. PhD*,†,§; Ellis, Alan R. PhD, MSW†; Dubard, C. Annette MD, MPH∥
*Department of Health Policy and Management, Gillings School of Global Public Health
†Cecil G. Sheps Center for Health Services Research Center, University of North Carolina
‡Division of Pharmaceutical Outcomes and Policy, Eshelman School of Pharmacy
§Department of Health Policy and Management, Cecil G. Sheps Center for Health Services Research Center, University of North Carolina, Chapel Hill
∥Department of Informatics, Quality, and Evaluation, Community Care of North Carolina, Raleigh, NC
Present address: Christopher A. Beadles, Health Services Research and Development, Department of Veteran Affairs Medical Center, Durham, NC.
Supplemental Digital Content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Website, www.lww-medicalcare.com.
Supported by the Agency for Healthcare Research and Quality and North Carolina Community Care Networks in “Expansion of Research Capability to Study Comparative Effectiveness in Complex Patients,” PI: Dubard, Grant No. R24 HS019659-01. Additional funding for Dr Beadles was provided by Grant No. 5T32-HS000032 from the Agency for Healthcare Research and Quality and Grant No. TPP 21-023 from the Department of Veteran Affairs Office of Academic Affiliations. Additional funding for Ms. Lichstein was provided by 2T32NR008856 from the National Institute of Nursing Research at the National Institutes of Health.
J.F.F. has received consulting support from Daiichi Sankyo and Takeda Pharmaceutical Company for unrelated studies. M.E.D. receives funding from NCCCN and AccessCare for evaluation projects unrelated to the present manuscript. The remaining authors declare no conflict of interest.
Reprints: Marisa E. Domino, PhD, Department of Health Policy and Management, Gillings School of Global Public Health, University of North Carolina, 1104G McGavran-Greenberg Hall, 135 Dauer Dr., Chapel Hill, NC 27599-7411. E-mail: email@example.com.
Background: Little is known about the quality of care received by Medicaid enrollees with multiple chronic conditions (MCCs) and whether quality is different for those with mental illness.
Objectives: To examine cancer screening and single-disease quality of care measures in a Medicaid population with MCC and to compare quality measures among persons with MCC with varying medical comorbidities with and without depression or schizophrenia.
Research Design: Secondary data analysis using a unique data source combining Medicaid claims with other administrative datasets from North Carolina’s mental health system.
Subjects: Medicaid-enrolled adults aged 18 and older with ≥2 of 8 chronic conditions (asthma, chronic obstructive pulmonary disease, diabetes, hypertension, hyperlipidemia, seizure disorder, depression, or schizophrenia). Medicare/Medicaid dual enrollees were excluded due to incomplete data on their medical care utilization.
Measures: We examined a number of quality measures, including cancer screening, disease-specific metrics, such as receipt of hemoglobin A1C tests for persons with diabetes, and receipt of psychosocial therapies for persons with depression or schizophrenia, and medication adherence.
Results: Quality of care metrics was generally lower among those with depression or schizophrenia, and often higher among those with increasing levels of medical comorbidities. A number of exceptions to these trends were noted.
Conclusions: Cancer screening and single-disease quality measures may provide a benchmark for overall quality of care for persons with MCC; these measures were generally lower among persons with MCC and mental illness. Further research on quality measures that better reflect the complex care received by persons with MCC is essential.
With the passage of federal healthcare reform has come increased attention to bending the cost growth curve while maintaining or improving quality of care. Implicit in this emphasis is the identification of population groups in which both of these seemingly divergent goals may be achieved. One population group meriting greater attention is individuals with multiple chronic conditions (MCCs).
More than one in 1 American adults has ≥2 chronic conditions.1 Individuals with MCC experience lower quality of life,2,3 decreased functional status,4–6 poorer disease control,7,8 and greater mortality than individuals without MCC.1,9 These differences are observed despite greater healthcare utilization and expenditures among individuals with MCC.1,10 To preserve functional status, improve disease control, and potentially curb costs, health services researchers and policy makers are increasingly advocating improved access to coordinated, high-quality healthcare for individuals with MCC.
To date, discussion of improving healthcare for individuals with MCC has primarily focused on concurrent physical health conditions. Yet among Medicaid recipients with at least one of 5 prevalent medical conditions, two thirds also had a mental health condition, which was associated with higher Medicaid spending.11 Depression was the most prevalent of 13 chronic conditions in a Medicaid population, affecting almost 40% of Medicaid-enrolled nonelderly adults with ≥1 of the 13 conditions.11 In addition, persons with severe and persistent mental illness, such as schizophrenia, are at higher risk for developing physical health conditions such as hypertension, coronary heart disease, and diabetes,12–15 and experiencing a substantially shortened life expectancy.14,16 Despite this elevated risk and the fact that physical health conditions are associated with greater disability in the presence of concurrent mental health conditions,17 many patients with severe and persistent mental illness have undiagnosed and untreated medical comorbidities stemming from inadequate primary care access18,19 and lower rates of medication adherence.20 Even so, persons with comorbid physical and mental health conditions consume high levels of acute medical care services and incur high healthcare costs.17,21,22
The high costs and the high level of unmet need among persons with comorbid physical and mental health conditions suggest that this population deserves increased attention. In addition, little is known about the quality of care for persons with MCC, particularly those with severe and persistent mental illness. Our goal was to examine a number of evidence-based quality indicators for a diverse population of Medicaid enrollees with MCC including mental illness and to estimate differences by presence of mental illness and number of medical comorbidities.
To examine the quality of care received by patients with MCC, we conducted retrospective secondary data analysis using a unique data source, North Carolina Integrated Data for Researchers (NCIDR). NCIDR enhances North Carolina Medicaid claims and enrollment data for fiscal years 2008–2010 with information from 3 separate components of NC’s mental health system (https://www.communitycarenc.org/informatics-center/ncidr/): (1) data on stays in state psychiatric hospitals are available regardless of insurance coverage and fill in gaps among Medicaid claims, which do not contain psychiatric hospital stays for those age 21–64; (2) data on mental health services provided with state funds separate from Medicaid capture services provided to persons without insurance as well as to supplement Medicaid benefits; (3) administrative data from a regional behavioral health and developmental disabilities carve-out provides detailed claims for Medicaid-enrolled persons in participating counties. These additional data source provide greater opportunities to observe psychiatric diagnoses and services use than Medicaid claims alone.
From the merged data, we selected adults aged 18 and older with claims diagnoses for at least 2 of the following 8 target conditions: asthma, chronic obstructive pulmonary disease, diabetes, hypertension, hyperlipidemia, seizure disorder, depression, and schizophrenia. These conditions were selected based on prevalence, costliness, immediacy of treatment effects, and availability of claims-based quality measures. Seven of our 8 conditions were also identified among 13 conditions in nonelderly Medicaid recipients based on prevalence, costs, and ability to intervene; including depression and schizophrenia.11 Constraints on the overall project prevented inclusion of additional mental disorders. To improve specificity, we further restricted our sample to include only individuals who met stricter criteria for at least 1 target condition (herein referred to as the strict diagnosis definition): at least 1 inpatient diagnosis or at least 2 outpatient or emergency department diagnoses. Because many of the quality measures described below, such as medication adherence, could only be derived from Medicaid claims data, we included only persons with at least 1 month of Medicaid enrollment during the 36-month study period. We excluded person-months during which persons were enrolled in both Medicaid and Medicare, because of incomplete data on pharmacy use.
We constructed a number of claims-based measures for the target conditions based on a review of the literature, feasibility of deriving measures from administrative data, and professional judgment of the study team. The study measures include (i) age-specific and sex-specific breast, colon, and cervical cancer screening, coded to be consistent with the American Cancer Society (ACS) guidelines (http://www.cancer.org/healthy/findcancerearly/cancerscreeningguidelines/american-cancer-society-guidelines-for-the-early-detection-of-cancer); (ii) disease-specific quality measures for diabetes, asthma, hyperlipidemia, hypertension, depression, and schizophrenia; and (iii) disease-specific medication adherence. The specific quality measures are further described in supplemental digital content (Table, Supplemental Digital Content 1, http://links.lww.com/MLR/A602). We measured adherence using the proportion of days covered measure,23,24 defined beginning with the first observed diagnosis or medication in each therapeutic area and adjusted for inpatient hospital days. Each measure indicates receipt of the test or procedure at any time during the 36-month study period or average medication adherence as diagnosis or first observed treatment. Models are adjusted for the actual number of months of Medicaid enrollment. Although the literature includes annual quality measures for singly diagnosed populations, we found generally low rates across single years in the MCC population; thus, we present results on use at the patient level within the 3-year study period.
We compared quality of care measures across persons with 1, 2, or ≥3 of the 6 medical (nonpsychiatric) comorbidities (asthma, chronic obstructive pulmonary disease, hypertension, hyperlipidemia, diabetes, or seizure disorder), using the strict definition described above, to those with psychiatric comorbidities (depression or schizophrenia). We included persons with both schizophrenia and depression only in the schizophrenia cohort, thus creating 3 distinct groups of persons with depression and no schizophrenia, schizophrenia with or without depression, and neither depression nor schizophrenia. We ran logit models of the receipt of each quality measure or ordinary least squares regressions of average adherence on age in quadratic form, race, ethnicity, sex, and indicators of the number of strictly defined medical comorbidities (1, 2, or 3 or more) interacted with indicators for depression or schizophrenia, controlling for the number of months enrolled in Medicaid during the study period. Only persons meeting the strict diagnostic criteria were included in disease-specific measures (eg, the receipt of A1C tests was only run on the sample of persons meeting strict criteria for diabetes). We then used predictions from the multivariate models to generate differences in each quality indicator by disease status and interactions, reported in Tables 1–5. Delta-method SEs for these estimates provides a measure of precision.
We first examined whether the quality of care measures varied by mental health status and by the number of physical health comorbidities. We hypothesized that the complexity of the overall constellation of conditions would reduce individual quality measures. We next examined disease-specific quality measures for physical health conditions, including medication adherence, and specific recommended tests or procedures, such as hemoglobin A1c tests (HbA1c) for persons with diabetes, minimally adequate treatment for depression, defined as ≥8 psychotherapy visits,25,26 and the use of assertive community treatment (ACT) for persons with schizophrenia. Each measure reflects use over the 3-year study period. We examined the average differences in quality of care measures by mental health and physical health indicators and their interactions. Finally, we examined quality of care measures for depression and schizophrenia as a function of the number of physical health comorbidities.
We conducted 2 sets of sensitivity analyses; first, we excluded months during which an individual spent >15 days in an inpatient facility or was in a residential facility such as a skilled nursing home. These individuals might be expected to have a lower probability of receiving screening or other quality measures, but possibly a higher probability of autofilled prescriptions, and thus higher medication adherence. Because these restrictions excluded only 4.9% of persons in the sample and produced results very similar to those reported for the full sample, they were not separately reported. The second sensitivity analysis included only the 43% of the sample enrolled on Medicaid for at least 80% of the study period, sometimes referred to as the continuously enrolled. The results are reported as supplemental digital content (Tables, Supplemental Digital Content 2–6, http://links.lww.com/MLR/A603; http://links.lww.com/MLR/A604; http://links.lww.com/MLR/A605; http://links.lww.com/MLR/A606; http://links.lww.com/MLR/A607), which repeat all analyses in continuously enrolled subsamples) and described in brief below. Because of the reduced generalizability of the continuously enrolled sample, we retain the results from the full sample as our main sampling approach.
This study was approved by the University of North Carolina Institutional Review Board.
The final sample included 188,531 unique persons with ≥2 of the 8 target chronic conditions (Table, Supplemental Digital Content 7, http://links.lww.com/MLR/A608, which contains demographic information). Average age was 43, 34% of the sample was male, 40% African American, and 2.6% Latino. The average length of Medicaid enrollment during the 3-year study period was 23 months. Although 77% of the sample was enrolled in Medicaid for at least 1 year, only 43% of the sample was continuously enrolled for at least 80% of the 36-month study period.
Almost 25% of persons aged 50 and older with MCC received Medicaid-funded colorectal cancer screening during the 3-year period (Table 1). Captured colorectal cancer screening methods range from annual fecal occult blood testing to colonoscopy on a 10-year schedule. Thus, if the ACS guidelines27 had been followed, we would expect to see 30%–100% of the sample screened during the 3-year period, depending on proportion of the sample using each method. Among women aged 40 and above, 41% had evidence of breast cancer screening during the 3-year period, despite ACS guideline recommendations of annual screening for women aged 40 and older. Because US Preventive Services Task Force guidelines recommend screening mammography every 2 years between ages 50 and 74,28 we additionally examined breast cancer screening rates within this age group, and found that 42% had evidence of mammography during the 3-year window. Cervical cancer screening, as evidenced by Medicaid-paid Pap smear claim, occurred for 30% of women aged 21–65 during the 3 years.
For both colorectal and breast cancer, screening rates among individuals with MCC were higher for those with depression and lower for those with schizophrenia, compared with persons without either psychiatric condition (Table 1). Rate of screening was positively associated with a greater number of medical comorbidities. These patterns were similar for breast cancer screening among women aged 50–74 (data not shown). Cervical cancer screening rates were slightly higher among women with depression compared with those without either psychiatric illness, but we saw no difference between women with and without schizophrenia.
The mental and physical health interactions (Table 2) indicate that depression had a positive association with both colorectal and breast cancer screening among people with one of the measured physical health conditions; the marginal difference from depression diminished as the number of medical conditions increased. Cervical cancer screening showed no difference by depression status across any of the measured levels of physical health conditions. In contrast, reductions in colorectal and breast cancer screening associated with schizophrenia were most pronounced among persons with 2 medical conditions. Women with schizophrenia and ≥2 medical conditions had higher rates of cervical cancer screening than women without either schizophrenia or depression.
Persons whose MCCs include depression had lower rates of adherence to medications in all classes examined (Table 3). When compared with persons with MCC without depression, we also found lower rates of HbA1C testing and nephropathy screening among those with comorbid diabetes, lower use of lipid profiles among those with depression and either diabetes or hyperlipidemia, and lower use of angiotensin-converting enzyme inhibitors and angiotensin receptor blockers among those with diabetes, hypertension, and depression. We found similar results for persons with schizophrenia and other target medical conditions, except for a higher rate of adherence for diabetes and hyperlipidemia among persons with comorbid schizophrenia. We found higher rates of liver function tests both for persons with depression and for those with schizophrenia, as compared with persons with MCC without either of these psychiatric impairments. Short-acting β-agonist overuse was lower among those with asthma and schizophrenia than among those with asthma and neither psychiatric impairment, but overuse increased with the number of medical comorbidities. Greater numbers of medical comorbidities were otherwise generally associated with higher scores on quality of care measures across target conditions.
We found variable differences by depression and schizophrenia status on disease-specific quality and adherence measures from models using mental and physical health interactions (Table 4). For some measures (eg, adherence to diabetes medications, short-acting β-agonist overuse) the differences between those depressed and those without either psychiatric condition increased with the number of medical comorbidities. For other measures (eg, lipid profiles), the association with depression decreased as the number of medical comorbidities increased. Differences between persons with schizophrenia and persons with MCC with other conditions were generally larger than the differences observed by depression status.
We estimate a proportion of days covered of 33% for persons with MCC with depression 46% among those with comorbid schizophrenia (Table 5). Adherence increased substantially in both classes as the number of medical conditions increased. Almost 40% of persons with depression received individual or group psychotherapy during the 3-year period, but this rate decreased with a greater number of physical comorbidities. The proportion receiving at least 8 psychotherapy visits among those with depression was low and constant across comorbidities, with a 1.6 percentage point decline among persons with ≥3 medical conditions. Twelve percent of persons with comorbid schizophrenia received ACT, and the receipt of ACT was unaffected by the number of medical comorbidities in this sample.
The results of the sensitivity analysis on the continuously enrolled subsample varied somewhat from those reported here (Tables, Supplemental Digital Content 2–6, http://links.lww.com/MLR/A603; http://links.lww.com/MLR/A604; http://links.lww.com/MLR/A605; http://links.lww.com/MLR/A606; http://links.lww.com/MLR/A607); which report results of all analyses from the continuously Medicaid-enrolled subsample). Depression generally had a similar or larger difference in cancer screening, quality, and adherence measures to those reported here, and schizophrenia generally had smaller effect, although both terms lost significance in several models. The effects of medical comorbidities on cancer screening were generally smaller and those on disease-specific measures were generally larger.
This work serves as a starting point in comparing cancer screening, single-disease quality of care measures, and medication adherence among persons with varying combinations of physical and psychiatric conditions. It contributes to the growing literature on the role of mental illness on the receipt of guideline-concordant care among persons with MCC.29,30 In summary, quality of care metrics were generally, but not always lower among those with depression or schizophrenia, and often higher among those with increasing levels of medical comorbidities. A number of exceptions to these trends were noted. From the complex interactions and patterns of care we observed, 4 key points emerge.
First, greater burden of disease is not always associated with lower levels of quality of care measures used in our study. This result is consistent with other works31 finding a positive association between the number of comorbidities and a composite quality measure. Although persons with MCC may have more complicated and sometimes conflicting medical regimes, they may also experience greater economies of scope or opportunities for care. We often found better medication adherence among those with more, not fewer, medical comorbidities (Tables 3 and 5). This pattern may reflect the existence of a subgroup of patients with MCC who are more experienced interacting with the healthcare system. For example, these patients may be better at establishing formal and informal caregiver networks or using aids such as weekly pillboxes to manage their care. These skills may make such patients more likely to receive care consistent with established guidelines.
Second, among persons with MCC, we generally, but not always, found detrimental quality associated with poorer mental health. It is by now well known that major depression is associated with decreased medication adherence,20 and these results also seem to apply among persons with MCC. Adherence rates to antidepressant and antipsychotic medications reported here were somewhat lower than adherence rates in other published studies.24,32,33 Somewhat paradoxically, in almost half of the quality and adherence measures, differences by depression status were greatest among persons with a single chronic condition and decreased as the number of conditions increased. We also found that persons with depression had generally greater rates of cancer screening, possibly reflecting greater opportunities for screening with a greater number of healthcare contacts. These results are somewhat at odds with prior studies of cancer screening by depression status,34 which found slightly lower rates of breast cancer screening and no difference in colorectal cancer screening. Their study used a survey measure of depression symptoms rather than administrative diagnoses, however.
Third, comparing people with comorbid schizophrenia to those with MCC without schizophrenia or depression also yielded several surprising findings. While persons with schizophrenia had generally poorer cancer screening rates, they often had better adherence to medications, such as diabetic agents and medications for hyperlipidemia. This finding may reflect a system that has converged around the most acutely ill to improve care and subsequent health outcomes. Alternatively, it may reflect the greater awareness of metabolic-related side effects associated with the use of many of the second-generation antipsychotic medications, although these rates are still noted to be suboptimal.35,36
Finally, we saw that number of medical conditions had mixed effects on evidenced-based services among those with a psychiatric impairment. Persons with a greater burden of chronic physical health conditions generally had higher medication adherence, but also had lower use of psychotherapy for those with depression. This likely reflects greater use of the medical system and reduced contact with the specialty mental health system among those with more physical health demands.37 In contrast, persons with schizophrenia are more likely to use the specialty mental health system than to use primary care,38 regardless of their burden of physical health conditions, and thus we did not see a similar reduction in the use of psychosocial therapies typically provided by specialty mental healthcare.
A number of limitations should be noted. Estimates were based on information in administrative data, and therefore do not capture all services that may have been received during times of Medicaid ineligibility or through other programs, such as free breast and cervical cancer screening programs. In addition, guideline-concordant screening rates may be underestimated due to clinical factors that would exclude individuals from screening recommendations but were unobservable in available data; such as prior hysterectomy or double mastectomy. For example, DuBard et al39 found that as much as 50% of a sample of Medicaid eligible women aged 50–64 had prior history of a hysterectomy from chart review, rendering them ineligible for cervical cancer screening. If unobservable differences are correlated with our key variables, this could partially explain these results (eg, if women with schizophrenia had disproportionately high rates of hysterectomy). Our comorbidity indicators may reflect residual differences between comorbidity groups that are not accounted for by included covariates. Consistent with this fact, our models are not intended to be causal; rather, they are intended to examine the relationships between comorbidity levels and service use and quality received by persons with varying combinations of the 8 target disorders. Finally, and perhaps most importantly, these data do not contain information on health outcomes and quality of life which may be most important to persons with MCC.
Although cancer screening and single-disease guideline measures are important indicators of quality of care received and laudable benchmarks for improving care for persons with MCCs, there are a number of complexities inherent in treating and measuring the quality of care for this complex population. Treating MCC represents both a challenge and an opportunity to achieve greater quality, possibly at lower cost. In the context of the MCC conversation, mental illness complicates the receipt of high-quality care in a way that requires increased attention. Further effort should be devoted to identifying the specific obstacles to high-quality care and simultaneously advancing the science of quality measurement in this growing and costly population.
1. Anderson G.Chronic Care: Making the Case for Ongoing Care.2010.Princeton, NJ:Robert Wood Johnson Foundation and Johns Hopkins University.
2. Fortin M, Bravo G, Hudon C, et al..Relationship between multimorbidity and health-related quality of life of patients in primary care.Qual Life Res.2006;15:83–91.
3. Fortin M, Bravo G, Hudon C, et al..Psychological distress and multimorbidity in primary care.Ann Fam Med.2006;4:417–422.
4. Bayliss E, Bayliss M, Ware J, et al..Predicting declines in physical function in persons with multiple chronic medical conditions: what we can learn from the medical problem list.Health Qual Life Outcomes.2004;2:47.
5. Kadam U, Croft P.Clinical multimorbidity and physical function in older adults: a record and health status linkage study in general practice.Fam Pract.2007;24:412–419.
6. Marengoni A, Von Strauss E, Rizzuto D, et al..The impact of chronic multimorbidity and disability on functional decline and survival in elderly persons. A community-based, longitudinal study.J Intern Med.2009;265:288–295.
7. Wong ND, Lopez V, Tang S, et al..Prevalence, treatment, and control of combined hypertension and hypercholesterolemia in the United States.Am J Cardiol.2006;98:204–208.
8. Kerr E, Heisler M, Krein S, et al..Beyond comorbidity counts: how do comorbidity type and severity influence diabetes patients’ treatment priorities and self-management?J Gen Intern Med.2007;22:1635–1640.
9. Menotti A, Mulder I, Nissinen A, et al..Prevalence of morbidity and multimorbidity in elderly male populations and their impact on 10-year all-cause mortality: the FINE study (Finland, Italy, Netherlands, Elderly).J Clin Epidemiol.2001;54:680–686.
10. Wolff J, Starfield B, Anderson G.Prevalence, expenditures, and complications of multiple chronic conditions in the elderly.Arch Intern Med.2002;162:2269–2276.
11. Boyd C, Leff B, Weiss C, et al..Clarifying multimorbidity patterns to improve targeting and delivery of clinical services for Medicaid populations.Center for Health Care Strategies Inc.2010Faces of Medicaid.
12. Stang PE, Brandenburg NA, Lane MC, et al..Mental and physical comorbid conditions and days in role among persons with arthritis.Psychosom Med.2006;68:152–158.
13. Astles J.Extended nurse prescribing: improving care for older people.Br J Nurs.2006;15:150–154.
14. Kelly D, Boggs D, Conley R.Reaching for wellness in schizophrenia.Psychiatr Clin North Am.2007;30:453–479.
15. Lett H, Blumenthal J, Babyak M, et al..Depression as a risk factor for coronary artery disease: evidence, mechanisms, and treatment.Psychosom Med.2004;66:305–315.
16. Druss BG, Bornemann TH.Improving health and health care for persons with serious mental illness: the window for US Federal Policy Change.JAMA.2010;303:1972–1973.
17. Merikangas KR, Ames M, Cui L, et al..The impact of comorbidity of mental and physical conditions on role disability in the US adult household population.Arch Gen Psychiatry.2007;64:1180–1188.
18. Nasrallah HA, Meyer JM, Goff DC, et al..Low rates of treatment for hypertension, dyslipidemia and diabetes in schizophrenia: data from the CATIE schizophrenia trial sample at baseline.Schizophr Res.2006;86:15–22.
19. Salsberry PJ, Chipps E, Kennedy C.Use of general medical services among medicaid patients with severe and persistent mental illness.Psychiatr Serv.2005;56:458–462.
20. Grenard J, Munjas BA, Adams JL, et al..Depression and medication adherence in the treatment of chronic diseases in the United States: a meta-analysis.J Gen Intern Med.2011;26:1175–1182.
21. Kathol R, Saravay SM, Lobo A, et al..Epidemiologic trends and costs of fragmentation.Med Clin North Am.2006;90:549–572.
22. Kessler RC, Ormel J, Demler O, et al..Comorbid mental disorders account for the role impairment of commonly occurring chronic physical disorders: results from the National Comorbidity Survey.J Occup Environ Med.2003;45:1257–1266.
23. Benner JS, Glynn RJ, Mogun H, et al..Long-term persistence in use of statin therapy in elderly patients.JAMA.2002;288:455–456.
24. Martin BC, Wiley-Exley EK, Richards S, et al..Contrasting measures of adherence with simple drug use, medication switching, and therapeutic duplication.Ann Pharmacother.2009;43:36–44.
25. Kessler RC, Berglund P, Demler O, et al..The epidemiology of major depressive disorder: results from the National Comorbidity Survey Replication (NCS-R).JAMA.2003;289:3095–3105.
26. Fortney JC, Harman JS, Xu S, et al..The association between rural residence and the use, type, and quality of depression care.J Rural Health.2010;26:205–213.
29. Ciechanowski PS, Katon WJ, Russo JE.Depression and diabetes impact of depressive symptoms on adherence, function, and costs.Arch Intern Med.2000;160:3278–3285.
30. Kilbourne AM, Welsh D, McCarthy JF, et al..Quality of care for cardiovascular disease-related conditions in patients with and without mental disorders.J Gen Intern Med.2008;23:1628–1633.
31. Higashi T, Wenger NS, Adams JL, et al..Relationship between number of medical conditions and quality of care.N Engl J Med.2007;356:2496–2504.
32. Wu C-H, Farley JF, Gaynes BN.Evaluating the association of initial benzodiazepine use and antidepressant adherence among adults with anxiety disorders.J Exp Clin Med.2012;4:338–344.
33. Domino ME, Martin BC, Elizabeth W-E, et al..Increasing Time Costs and Copayments for Prescription Drugs: An Analysis of Policy Changes in a Complex Environment.2009.Chapel Hill, NC:UNC.
34. Aggarwal A, Freund K, Sato A, et al..Are depressive symptoms associated with cancer screening and cancer stage at diagnosis among postmenopausal women? The Women’s Health Initiative Observational Cohort.J Women Health.2008;17:1353–1361.
35. Mitchell AJ, Delaffon V, Vancampfort D, et al..Guideline concordant monitoring of metabolic risk in people treated with antipsychotic medication: systematic review and meta-analysis of screening ractices.Psychol Med.2012;42:125–147.
36. Morrato E, Druss B, Hartung D, 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;67:17–24.
37. Ettner SL, Azocar F, Branstrom RB, et al..Association of general medical and psychiatric comorbidities with receipt of guideline-concordant care for depression.Psychiatr Serv.2010;61:1255–1259.
38. Frank RG, Glied SA.Better But Not Well: Mental Health Policy in the United States since 1950.2006.Baltimore, MD:The Johns Hopkins University Press.
39. DuBard CA, Schmid D, Yow A, et al..Recommendation for and receipt of cancer screenings among Medicaid recipients 50 years and older.Arch Intern Med.2008;168:2014–2021.
quality of care; multiple chronic conditions; mental illness; Medicaid
Supplemental Digital Content
© 2014 by Lippincott Williams & Wilkins.
Highlight selected keywords in the article text.