Association Between Chronic Illness Complexity and Receipt of Evidence-based Depression Care
Jordan, Neil PhD*,†,‡; Sohn, Min-Woong PhD†,‡; Bartle, Brian MS‡; Valenstein, Marcia MD, MS§,∥; Lee, Yuri MA*; Lee, Todd A. PharmD, PhD‡,¶
*Department of Psychiatry & Behavioral Sciences
†Center for Healthcare Studies, Institute for Public Health and Medicine, Northwestern University Feinberg School of Medicine, Chicago
‡Center for Management of Complex Chronic Care, Hines VA Hospital, Hines, IL
§Department of Psychiatry, University of Michigan
∥Serious Mental Illness Treatment Research and Evaluation Center, Center for Clinical Management Research VA Ann Arbor Health System, Ann Arbor, MI
¶Department of Pharmacy Practice, College of Pharmacy, University of Illinois at Chicago, Chicago, IL
Supported through grant number R21HS017635 from the Agency for Healthcare Research and Quality. The content is solely the responsibility of the authors and does not necessarily represent the official views of the Agency for Healthcare Research and Quality.
The authors declare no conflict of interest.
Reprints: Neil Jordan, PhD, Mental Health Services & Policy Program, Northwestern University Feinberg School of Medicine, 710 N Lake Shore Dr 12th Flr, Chicago, IL 60611. E-mail: firstname.lastname@example.org.
Background: The rate of guideline concordance with antidepressant treatment for persons with depression is low. The problem may be even more pronounced for patients with depression and other multiple chronic conditions (MCC).
Objectives: To study, for persons with new depressive episodes, the association between MCC and the likelihood of receiving guideline-concordant depression treatment.
Research Design: Retrospective cohort study using Veterans Affairs administrative data.
Subjects: A total of 43,189 Veterans Affairs patients who had a new depressive episode during 2007 were included.
Measures: We assessed whether patients had an adequate supply of antidepressants during acute and continuation phases of depression treatment, which indicates guideline-concordant care. We determined the association between comorbid conditions and receipt of adequate antidepressant supplies after adjusting for potential confounders.
Results: Compared with patients with depression alone, those with comorbid cardiovascular/cerebrovascular disease, peptic ulcer/gastroesophageal reflux disease (GERD), or arthritis were 8%–13% more likely to receive adequate antidepressant supplies during the acute phase. Patients with depression and substance/alcohol abuse were 15% less likely to receive adequate supplies in the acute treatment phase. Those with cardiovascular/cerebrovascular disease or peptic ulcer/GERD were 9%–10% more likely to receive continuation phase guideline-concordant depression treatment. Patients with comorbid substance/alcohol abuse were 19% less likely to receive continuation phase guideline-concordant depression treatment. Relatively few of the most prevalent MCC clusters were significantly associated with receipt of guideline-concordant depression treatment.
Conclusions: There was no consistent association between specific clusters of chronic conditions and adequate antidepressant supplies. There continues to be need for practice-level and system-level interventions to increase quality of depression treatment, particularly among persons with certain comorbid conditions such as cardiovascular/cerebrovascular disease, peptic ulcer/GERD, and arthritis.
Depression and depression treatment are highly prevalent among Veterans.1–2 Guideline-concordant (GC) depression treatment rates among Veterans with new depressive episodes vary considerably, ranging from 35% to 61% during the acute phase (first 114 d after a major episode begins)3–5 and from 43% to 54% during the continuation phase (first 210 d).4,5
Some studies suggest that problems with guideline concordance may be more pronounced for depression patients with coexisting chronic physical conditions. Competing demands of comorbid physical conditions may reduce the odds of adequately adjusting depression care.6 Suboptimal depression care has been documented among patients with depression and comorbid heart failure,7 coronary heart disease,8 coronary artery disease,9 chronic obstructive pulmonary disease,10 and osteoarthritis.11 Conversely, patients with depression and diabetes have increased or equal likelihood of adequate antidepressant care.6,12,13
These studies focused on patients with depression and a single physical comorbidity. However, some research suggests that the effects of physical comorbidities on depression care quality may be compounded in patients with multiple chronic conditions (MCC). Studies on the effect of MCC on depression care indicate mixed findings.14–17 These inconsistent findings may be explained by variations in study samples and settings, data sources (self-report interview/survey vs. administrative), and range of chronic conditions studied. The objective of this study is to further elucidate the relationship between multimorbidity and depression care by examining the association between various clusters of MCC and receipt of guideline-concordant depression care among adults with new depressive episodes.
Research Design and Study Cohort
Using a retrospective cohort design and methods similar to our previous work18,19 and the HEDIS 2008 guidelines for antidepressant medication management,20 we identified all Veterans Affairs (VA) patients who had a new (index) depressive episode during 2007. Patients were eligible if they were aged 18 years and above and met 1 of the 3 diagnostic criteria using ICD-9 depression codes (296.2x, 296.3x, 298.0x, 300.4x, 309.1x, 311.xx): had ≥1 principal diagnosis of depression in any setting, had ≥2 secondary diagnoses of depression on different days in any outpatient setting or in the emergency department, or had ≥1 secondary diagnosis of depression associated with any inpatient stay. Exclusion criteria are shown in Figure 1.
Identifying Complex Chronic Illness
Comorbid conditions were identified during the 12-month period before their index depressive episode using ICD-9 codes from AHRQ Clinical Classifications Software (http://www.hcup-us.ahrq.gov/toolssoftware/ccs/ccs.jsp). We identified chronic conditions managed through medication or psychosocial interventions: arthritis, cardiovascular and cerebrovascular conditions, diabetes, infectious disease, nonmelanoma cancer, peptic ulcer/gastroesophageal reflux disease (GERD), respiratory disease, and substance/alcohol abuse. A patient was deemed as having any of these conditions if they had 2 outpatient or 1 inpatient diagnosis. Patients were assigned to mutually exclusive clusters based on groupings of these 8 conditions.
Depression Guideline-concordant Antidepressant Management
We determined whether patients received an adequate days supply of antidepressants (Table 1) during the acute and continuation phases of their treatment. To identify acute phase concordance, we summed the days supply of antidepressants received, starting with the index antidepressant through all those dispensed within 114 days of the antidepressant index date. We applied previously developed methods to appropriately count days supply when patients seemed to have switched medications or received multidrug regimens.19 To be considered guideline-concordant during the acute and continuation phases, patients needed at least 84 days of antidepressant supply during the 114-day follow-up period and at least 180 days of antidepressant supply during the 210-day follow-up period, respectively.20,21
We used information from the 12-month period preceding the index depressive episode to characterize other patient characteristics such as race/ethnicity, age, sex, marital status, and VA medication copayment status (patient faced copayment for all, some, or no prescriptions received). We derived other measures that might be related to treatment adherence: number of unique medications received in the 45 days before their index date, number of prior year primary care and specialist visits (a marker for health services use intensity), and type of provider that made the index diagnosis (primary care physician, psychiatrist, other specialty care physician, other allied professional).
We calculated acute and continuation phase antidepressant guideline concordance and compared those rates across the comorbid condition groups. We then calculated adjusted odds ratios for each MCC cluster. We compared acute and continuation phase antidepressant guideline concordance across the 25 most prevalent clusters. Logistic regression models were estimated, using the depression-only group as the reference, to test whether each cluster’s guideline concordance rate was different from the reference, adjusting for the covariates listed above.
There were 43,189 Veterans with a depressive episode in the final cohort. This cohort was typical of the VA population more broadly in terms of demographic characteristics and comorbid conditions (Table 2).
Overall, 63.4% received guideline-concordant antidepressant treatment during the acute phase (Table 3). Compared with patients with depression alone, those with cardiovascular/cerebrovascular disease, peptic ulcer/GERD, or arthritis were significantly more likely to receive guideline-concordant antidepressant care. Conversely, patients with depression and substance/alcohol abuse were significantly less likely to receive guideline-concordant care than patients with depression alone.
The proportion receiving guideline-concordant care during the continuation phase was lower (37.0%) and similarly varied by comorbidity. Compared with patients with depression alone, those with cardiovascular/cerebrovascular disease or peptic ulcer/GERD had significantly higher odds of receiving guideline-concordant treatment. Patients with comorbid substance/alcohol abuse were also significantly less likely to receive guideline-concordant care than patients with depression alone.
When patients were separated into mutually exclusive clusters based on how many of the 8 priority conditions they had (Table 4), the most common cluster was patients with depression plus a cardiovascular/cerebrovascular condition. Among the most prevalent clusters, there were 4 clusters for which the proportion of patients receiving acute phase guideline-concordant depression care exceeded 70%. The lowest rate of acute phase guideline concordance was for patients with depression plus diabetes but no other priority conditions (53.1%). Continuation phase guideline concordance ranged from 24.6% to 47.9%.
Relatively few of the most prevalent clusters of comorbid conditions were significantly associated with receipt of guideline-concordant depression treatment during the acute phase (Table 3). Patients with comorbid arthritis, cardiovascular/cerebrovascular disease, and peptic ulcer or GERD had 35% (99% CI, 1.04–1.74) higher odds of receiving guideline-concordant care than patients with depression alone. Patients with comorbid arthritis, cardiovascular/cerebrovascular disease, peptic ulcer or GERD, and diabetes had 40% (99% CI, 1.03–1.90) higher odds of receiving acute phase guideline-concordant care than patients with depression alone.
For guideline-concordant treatment during the continuation phase, 3 of the top 25 most prevalent comorbid condition clusters were associated with lower rates of guideline-concordant care, whereas none of the top 25 clusters were associated with higher receipt of guideline-concordant care. Patients with comorbid respiratory disease alone had 32% lower odds of receiving guideline-concordant care than patients with depression alone. Patients with comorbid respiratory disease and a cardiovascular/cerebrovascular condition had 28% lower odds of receiving guideline-concordant care than patients with depression alone. The odds of getting guideline-concordant treatment for persons with depression and comorbid substance or alcohol abuse alone was 33% lower than for patients with depression alone during the continuation phase.
Our goal was to examine whether coexisting chronic conditions influenced the receipt of guideline-concordant depression care in patients with a new episode of major depressive disorder. Most MCC clusters were not significantly associated with the likelihood of receiving guideline-concordant antidepressant care, although there were some clusters that significantly increased or decreased the likelihood of receiving guideline-concordant care relative to having depression alone. During the acute phase, only 2 of the 25 most prevalent clusters were significantly different than depression alone, and for both clusters, the presence of those comorbid conditions was associated with a higher likelihood of receiving adequate depression treatment. During the continuation phase, only 3 of the 25 most prevalent clusters were significantly different than depression alone, and all 3 clusters represented increased risk for not receiving adequate depression treatment.
Regardless of which comorbid conditions were present, overall rates of guideline-concordant care were suboptimal. Our observed rates of guideline concordance—63% during the acute phase and 37% during the continuation phase—are consistent with previous studies of depression guideline concordance among Veterans.3–5 Undertreatment of depression continues to be problematic in the VA, despite system-wide efforts during the past decade to improve the depression treatment quality. During the mid 2000s, the VA disseminated and implemented collaborative care for depression.22 More recently, the VA implemented the Primary Care-Mental Health Integration initiative.23 Although the impact of both efforts on adequate depression treatment is still being evaluated, there is evidence that collaborative care leads to more rapid improvement in depressive symptoms than standard care despite not significantly impacting the adequacy of antidepressant treatment.24 Unfortunately, neither initiative focuses on Veterans with depression and MCC. One reason we undertook this study was to help understand the extent to which comorbidity and, in particular, the presence of MCC impact depression treatment adequacy.
Our results do not support the competing needs hypothesis that predicts lower guideline-concordant care for persons with multiple comorbidities or the opposite hypothesis that comorbidities do not affect GC depression care. We additionally found that the number of comorbid conditions in each cluster were not associated with either increased or decreased GC care (data not shown). Our results rather suggest that individual conditions are important for GC care. For example, many of the clusters that included substance/alcohol abuse were consistently associated with lower odds of GC care. Although this finding is consistent with other literature which notes that the use of antidepressants for treating persons with comorbid depression and alcohol or other substance dependence is controversial,25,26 our analysis is the first to demonstrate greater risk for inadequate depression care in a population with depression, substance/alcohol abuse, and other comorbid conditions.
There are several limitations that should be noted. Lack of statistical power due to small numbers of patients in several clusters limited our ability to detect statistical significance in many comparisons where there was a substantial difference in the receipt of GC care (eg, 53% vs. 63% in diabetes-alone vs. depression-alone groups). Our study population was Veterans who primarily used the VA to obtain their healthcare, so results may not generalize to Veterans who primarily received Medicare, Medicaid, or private insurance reimbursed health services or to non-Veteran populations. It is also possible that some Veterans who receive depression treatment primarily outside the VA are receiving guideline-concordant depression care. Although with administrative data we were unable to control for depression severity, which may impact the likelihood of receiving guideline-concordant depression care, we have no reason to believe that depression severity is substantially different across MCC clusters, as there were no meaningful differences in the distribution of depression ICD-9 diagnosis codes across MCC clusters (data not shown). Similarly, we were unable to capture differences in severity of comorbid conditions, which hinders our ability to know how much the treatment of a particular comorbid condition competed with that patient’s depression treatment. Collapsing all cardiovascular and cerebrovascular conditions into 1 condition category limits our ability to acknowledge the impact of specific conditions in that category (eg, congestive heart failure), but doing so allowed us to have large enough clusters to detect effects for these related conditions. Another limitation is that there may be other reasons why particular MCC groups were more or less likely to receive GC depression care, such as health behaviors or the types of providers involved in their treatment, which should be explored in future research.
In conclusion, there seems to be only a relative few clusters of comorbid conditions that signal significantly higher or lower risk for receiving adequate depression care. Given the relatively low rates of adequate depression care, particularly in the continuation phase, there continues to be need for practice-level and system-level interventions to increase quality of depression treatment. To improve the rate of depression patients receiving guideline-concordant depression care, it may be beneficial for existing collaborative care and care management efforts to target subgroups that are least likely to receive such care so that providers can focus on the issues preventing them from receiving adequate depression treatment.
1. Hankin CS, Spiro A, Miller DR, et al..Mental disorders and mental health treatment among US Department of Veterans Affairs outpatients: The Veterans Health Study.Am J Psych.1999;156:1924–1930.
2. Blow FC, Owen RE.Specialty Care for Veterans With Depression in the VHA 2002 National Depression Registry Report.2003.Ann Arbor, MI:VHA Health Services Research & Development.
3. Chermack ST, Zivin K, Valenstein M, et al..The prevalence and predictors of mental health treatment services in a national sample of depressed Veterans.Med Care.2008;46:813–820.
4. Busch SH, Leslie D, Rosenheck R.Comparing the quality of antidepressant pharmacotherapy in the Department of Veteran Affairs and the private sector.Psychiatr Serv.2004;55:1386–1391.
5. Charbonneau A, Rosen AK, Ash AS, et al..Measuring the quality of depression care in a large integrated health system.Med Care.2004;41:669–680.
6. Rost K, Nutting P, Smith J, et al..The role of competing demands in the treatment provided primary care patients with major depression.Arch Fam Med.2000;9:150–154.
7. Bogner HR, Ford DE, Gallo JJ.The role of cardiovascular disease in the identification and management of depression by primary care physicians.Am J Geriatr Psychiatry.2006;14:71–78.
8. Gill JM, Chen YX, Lieberman MI.Management of depression in ambulatory care for patients with medical comorbidities: a study from a national Electronic Health Record (EHR) network.Int J Psychiatry Med.2008;38:203–215.
9. Sambamoorthi U, Shen C, Findley P, et al..Depression treatment patterns among women Veterans with cardiovascular conditions or diabetes.World Psychiatry.2010;9:177–182.
10. Pirraglia PA, Charbonneau A, Kader B, et al..Adequate initial antidepressant treatment among patients with chronic obstructive pulmonary disease in a cohort of depressed Veterans.Prim Care Companion J Clin Psychiatry.2006;8:71–76.
11. Gleicher Y, Croxford R, Hochman J, et al..A prospective study of mental health care for comorbid depressed mood in older adults with painful osteoarthritis.BMC Psychiatry.2011;11:147–156.
12. Jones LE, Turvey C, Torner JC, et al..Nonadherence to depression treatment guidelines among Veterans with diabetes mellitus.Am J Manag Care.2006;12:701–710.
13. Harman JS, Edlund MJ, Fortney JC, et al..The influence of comorbid chronic medical conditions on the adequacy of depression care for older Americans.J Am Geriatr Soc.2005;53:2178–2183.
14. Gill JM, Klinkman MS, Chen YX.Antidepressant medication use for primary care patients with and without medical comorbidities: a national electronic health record (EHR) network study.J Am Board Fam Med.2010;23:499–508.
15. Ani C, Bazargan M, Hindman D, et al..Comorbid chronic illness and the diagnosis and treatment of depression in safety net primary care settings.J Am Board Fam Med.2009;22:123–135.
16. Vyas A, Sambamoorthi U.Multimorbidity and depression treatment.Gen Hosp Psychiatry.2011;33:238–245.
17. Loeb DF, Ghushchyan V, Huebschmann AG, et al..Association of treatment modality for depression and burden of comorbid chronic illness in a nationally representative sample in the United States.Gen Hosp Psychiatry.2012;34:588–597.
18. Jordan N, Lee TA, Valenstein M, et al..Effect of depression care on outcomes in COPD patients with depression.Chest.2009;135:626–632.
19. Jordan N, Lee TA, Valenstein M, et al..Effect of care setting on evidence-based depression treatment for Veterans with COPD and comorbid depression.J Gen Intern Med.2007;22:1447–1452.
20. .HEDIS 2008 Technical Specifications Volume Two.2007.Washington, DC:National Committee for Quality Assurance.
21. .VHA/DoD Clinical Practice Guideline for Management of Major Depressive Disorder (MDD).2008.Washington, DC:Office of Quality and Performance, Department of Veteran Affairs; Quality Management Directorate, United States Army MEDCOM.
22. Smith J, Williams JW, Owen RR, et al..Developing a national dissemination plan for collaborative care for depression: QUERI Series.Implement Sci.2008;3:59.
23. Zivin K, Pfeiffer PN, Szymanski BR, et al..Initiation of primary care-mental health integration programs in the VA Health System: associations with psychiatric diagnoses in primary care.Med Care.2010;48:843–851.
24. Hedrick SC, Chaney EF, Felker B, et al..Effectiveness of collaborative care depression treatment in Veterans’ Affairs primary care.J Gen Intern Med.2003;18:9–16.
25. Pettinati HM.Antidepressant treatment of co-occurring depression and alcohol dependence.Biol Psychiatry.2004;56:785–792.
26. Davis L, Uezato A, Newell JM, et al..Major depression and comorbid substance use disorders.Curr Opin Psychiatry.2008;21:14–18.
antidepressants; chronic disease; evidence-based medicine; adherence; depression
© 2014 by Lippincott Williams & Wilkins.
What does "Remember me" mean?
By checking this box, you'll stay logged in until you logout. You'll get easier access to your articles, collections,
media, and all your other content, even if you close your browser or shut down your
To protect your most sensitive data and activities (like changing your password),
we'll ask you to re-enter your password when you access these services.
What if I'm on a computer that I share with others?
If you're using a public computer or you share this computer with others, we recommend
that you uncheck the "Remember me" box.
Highlight selected keywords in the article text.
Data is temporarily unavailable. Please try again soon.