Age of Depressed Patient Does Not Affect Clinical Outcome in Collaborative Care Management

CLINICAL FOCUS: ADHD, DEPRESSION, PAIN, AND NEUROLOGICAL DISORDERS Age of Depressed Patient Does Not Affect Clinical Outcome in Collaborative Care Management DOI: 10.3810/pgm.2011.09.2467 Kurt B. Angstman, MS, MD 1 Kathy L. MacLaughlin, MD 1 Norman H. Rasmussen, EdD 1,2 Ramona S. DeJesus, MD 3 David J. Katzelnick, MD 2 1 Department of Family Medicine, Mayo Clinic, Rochester, MN; 2 Department of Psychiatry and Psychology, Mayo Clinic, Rochester, MN; 3 Division of Primary Care Internal Medicine, Mayo Clinic, Rochester, MN Abstract: Clinical response and remission for the treatment of depression has been shown to be improved utilizing collaborative care management (CCM). Prior studies have indicated that the presence of mental health comorbidities noted by self-rated screening tools at the intake for CCM are associated with worsening outcomes; few have examined directly the impact of age on clinical response and remission. The hypothesis was that when controlling for other mental health and demographic variables, the age of the patient at implementation of CCM does not significantly impact clinical outcome, and that CCM shows consistent efficacy across the adult age spectrum. We performed a retrospective chart analysis of a cohort of 574 patients with a clinical diagnosis of major depression (not dysthymia) treated in CCM who had 6 months of follow-up data. Using the age group as a categorical variable in logistic regression models demonstrated that while maintaining control of all other variables, age grouping remained a nonsignificant predictor of clinical response (P 0.1842) and remission (P 0.1919) after 6 months of treatment. In both models, a lower Generalized Anxiety Disorder-7 score and a negative Mood Disorder Questionnaire score were predictive of clinical response and remission. However, the initial Patient Health Questionnaire-9 score was a statistically significant predictor only for clinical remission (P = 0.0094), not for response (P = 0.0645), at 6 months. In a subset (n = 295) of the study cohort, clinical remission at 12 months was also not associated with age grouping (P 0.3355). The variables that were predictive of remission at 12 months were the presence of clinical remission at 6 months (odds ratio [OR], 7.4820; confidence interval [CI], 3.9301 14.0389; P 0.0001), clinical response (with persistent symptoms) (OR, 2.7722; CI, 1.1950 6.4313; P = 0.0176), and a lower initial Patient Health Questionnaire-9 score (OR, 0.9121; CI, 0.8475 0.9816; P = 0.0140). Our study suggests that using CCM for depression treatment may transcend age-related differences in depression and result in positive outcomes regardless of age. Keywords: depression; collaborative care management; primary care; mental health comorbidities Correspondence: Kurt B. Angstman, MS, MD, Department of Family Medicine, Mayo Clinic, 200 First St., Rochester, MN 55905. Tel: 507-284-2511 Fax: 507-538-8543 E-mail: angstman.kurt@mayo.edu Introduction Depression affects an estimated 121 million people 1 and was already the fourth leading contributor to the global burden of disease in 2000, not only as a stand-alone disorder, but also as a prominent risk factor or consequence of many other physical health diseases. Projections indicate that depression will become the second leading contributor to disability by 2020 (second only to heart disease), and is expected be the first by 2030. 2,3 Patient age affects the presentation and treatment of depression. With Sequenced Treatment Alternatives to Relieve Depression (STAR*D) data, evaluating the symptoms and comorbidities between 5 age groups, Husain et al 4 found that older patients (aged 51 65 and 66 75 years) reported more and 122 Postgraduate Medicine, Volume 123, Issue 5, September 2011, ISSN 0032-5481, e-issn 1941-9260 71508e

Depression and Collaborative Care Management longer-lasting episodes of major depression and medical comorbidities but had fewer psychiatric comorbidities. The Healthcare for Communities Household Telephone Survey of 9585 respondents noted that older adults (aged 65 years) with a mental illness diagnosis were less likely than younger adults to recognize a need for or to receive mental health care or counseling. 5 In a meta-analysis of 37 randomized studies, collaborative care management (CCM) for depression treatment was shown to be more effective than usual care. 6 Many of the studies referenced in the meta-analysis included a wide range of ages (18 90 years), but most had an average age of 40 to 50 years and many did not look at age as a predictor of response. 7 10 A notable exception regarding the age of study participants is the landmark Improving Mood-Promoting Access to Collaborative Treatment (IMPACT) trial, which focused on patients aged 60 years. 11 A prior CCM study involving 600 patients at 7 primary care clinics noted that age was not a predictor of treatment efficacy. 12 A recent publication pooled 3 randomized controlled trials on CCM for depression treatment and did not find any difference in effectiveness of the intervention comparing 18- to 59-year-olds with a cohort of patients aged 60 years. 13 In contrast, a 2009 IMPACT subset study that compared the young-old (60 74 years) with the old-old ( 75 years) found a similar 3-month response with CCM for depression treatment, but observed that the old-old were less likely to demonstrate response at 6, 12, 18, and 24 months, and were less likely to reach remission. 14 This study was undertaken to determine if there is a correlation between the efficacy of CCM and a patient s age at the time of depression treatment initiation. The hypothesis was that when controlling for other mental health and demographic variables, the age of the patient at implementation of CCM does not significantly impact clinical outcome and that CCM shows consistent efficacy across the adult age spectrum. Methods In 2007, the Institute for Clinical Systems Improvement (ICSI), along with 5 clinical sites across the state of Minnesota and 6 commercial health care plans, spearheaded the development of CCM for depression treatment. This was intended to improve depression management and augment the relationship between the patient, primary care provider, and psychiatrist. It was modeled from the IMPACT trial but was extended to all adults who met enrollment criteria. The CCM developed by ICSI allowed for monthly payment to the clinical sites for the CCM services. Since implementation in 2008, there have been a total of 85 clinics statewide that are now offering CCM for depression. Patients studied were from 1 of 3 primary care clinical sites (total patient population, 103 000) in Rochester, MN. The providers were from the Division of Primary Care Internal Medicine, Department of Family Medicine, or Division of Community Pediatrics and Adolescent Medicine at Mayo Clinic Rochester (Rochester, MN), and provided full-spectrum primary care, including obstetrics. The patient population was community based and was approximately 50% employees or dependents of Mayo Clinic Rochester. Adult patients (aged 18 years) who were diagnosed clinically with major depression or dysthymia and had a Patient Health Questionnaire-9 (PHQ-9 15 ) score of 10 were eligible to be enrolled in CCM. The CCM process involved weekly oversight by a psychiatrist, with medication or therapeutic changes managed by the primary care provider. Patient contact was dictated by the clinical scenario and PHQ-9 testing; some were contacted weekly and others monthly. Components of CCM included: a systematic screening and diagnostic approach, development of a patient registry, the introduction of a trained care manager, the utilization of treatment guidelines with standardized monitoring of improvement to ensure adequate therapy, and a consultative relationship with psychiatry and the primary care provider. Initial intake screening performed by a nurse care manager included self-rated screening tools for alcoholism (Alcohol Use Disorders Identification Test [AUDIT] 16 ), anxiety (Generalized Anxiety Disorder-7 [GAD-7] 17 ), bipolar disorder (Mood Disorder Questionnaire [MDQ] 18 ), and a review of the patient s psychiatric history and prior medications. Patients from the CCM database were analyzed retrospectively. Patients with a diagnosis of depression (not dysthymia) who had completed 6 months of CCM and had a 6-month follow-up PHQ-9 by October 2010 were eligible for the study. Of the original group of 638 patients, those with incomplete study variables were excluded. This left a cohort of 574 patients with complete data sets for analysis. Further subgroup analysis was performed on 295 patients who had PHQ-9 follow-up data available 1 year after initiation of CCM. The dependent variables in this study were if the patient had a clinical response at 6 months (defined as 6-month PHQ-9 score 50% of baseline) or was in remission of depression at 6 months (defined as PHQ-9 score 5). For the subgroup analysis, only remission was evaluated at 1 year. The independent variables were age group (categorized as age 30 years, 30 39 years, Postgraduate Medicine, Volume 123, Issue 5, September 2011, ISSN 0032-5481, e-issn 1941-9260 123

Angstman et al 40 49 years, 50 59 years, and 60 years), gender, marital status (married or not married), race (white or non-white), initial PHQ-9 score, AUDIT score, GAD-7 score, and MDQ score. The MDQ was considered negative if the patient had a score 6 and a negative response on questions 2 and 3; otherwise, it was noted as positive. Mann-Whitney testing was performed for numerical variables, as they were not in a normal distribution. Categorical variables were tested using Chi-square analysis. Means across multiple groups were tested with 1-way analysis of covariance. Multiple logistic regression analysis was used to identify independent predictors of clinical response and remission. Included in the regression models were age group, gender, marital status, race, initial PHQ-9 score, AUDIT score, GAD-7 score, and MDQ score. For the 295 patients with PHQ-9 data at 1 year, the variable of the 6-month PHQ-9 result (as clinical remission, those with response but not in remission, or those patients without clinical response and a 6-month PHQ-9 ratio to initial PHQ-9 of 0.5) was added as a categorical variable to the regression model. The study was approved by the institutional review board of the Mayo Clinic Rochester. Results The age groups were defined by patients who were aged 30 years (n = 148; 25.78%), 30 to 39 years (n = 156; 27.18%), 40 to 49 years (n = 109; 18.99%), 50 to 59 years (n = 94; 16.38%), and 60 years (n = 67; 11.67%). Table 1 shows the demographic and intake data and demonstrates that the group aged 30 to 39 years was more likely to be non-white (16.67%; P = 0.006). The youngest (age 30 years) and oldest (age 60 years) groups were less likely to be married (34.46% and 49.25%, respectively; P 0.0001). There was no statistical difference between the groups in gender (P = 0.106), but the group aged 40 to 49 years tended toward fewer women (65.14% men). The initial PHQ-9 scores were nearly identical in all age groups (P = 0.897), as was the percentage of negative MDQ scores (P = 0.067). There was a trend for a higher percentage of increased MDQ scores in the younger age group, which may be attributed to undiagnosed bipolar disorder (once diagnosed, it would remove them from CCM). The AUDIT score was more significantly elevated in the group aged 30 years (3.86) and decreased in the group aged 60 years (1.16; P 0.0001), as was GAD-7 score (11.64 and 7.87, respectively; P 0.0001). Clinical response trended lower in the group aged 30 years (64.19%), but was not statistically different from the other age groups (range, 67.31% 75.23%; P = 0.262). Clinical remission also trended lower in the group aged 30 years (49.32%), but was not statistically different from the other age groups (range, 53.85% 62.69%; P = 0.152). Using the age group as a categorical variable in logistic regression models demonstrated that while maintaining control of all other variables, age grouping remained a nonsignificant predictor of clinical response (Table 2) and clinical remission (Table 3) at 6 months after diagnosis. In both models, a lower GAD-7 score and a negative MDQ score were predictive of clinical response and remission while controlling for all other variables. The initial PHQ-9 score was a statistically significant predictor for clinical remission (P = 0.0094), but not for response (P = 0.0645). Table 1. Demographic and Initial Clinical Data of Patients in Collaborative Care Management for Depression at 6 Months After Diagnosis N = 574 Age 30 Years n = 148 (25.78%) Age 30 39 Years n = 156 (27.18%) Age 40 49 Years n = 109 (18.99%) Age 50 59 Years n = 94 (16.38%) Age 60 Years n = 67 (11.67%) Age, years 24.56 34.21 44.45 54.10 69.45 Race, % white 92.57 83.33 95.41 91.49 94.03 0.006 Marital status, 34.46 71.79 75.23 74.47 49.25 0.0001 % married Gender, % female 79.05 76.92 65.14 74.47 70.15 0.106 Initial PHQ-9 score 15.30 (10 27) 15.18 (10 26) 15.09 (10 24) 15.64 (10 27) 15.30 (10 27) 0.897 (range) MDQ, % negative 88.51 94.87 90.83 95.74 97.01 0.067 AUDIT score (range) 3.86 (0 29) 2.69 (0 29) 2.47 (0 30) 2.78 (0 27) 1.16 (0 6) 0.001 GAD-7 score (range) 11.64 (0 21) 11.17 (0 21) 10.83 (0 21) 10.41 (0 21) 7.87 (0 18) 0.001 Clinical response 64.19 67.31 75.23 74.47 71.64 0.262 at 6 months, % Clinical remission at 6 months, % 49.32 53.85 61.47 61.70 62.69 0.152 Abbreviations: GAD-7, Generalized Anxiety Disorder 7-Item Scale; MDQ, Mood Disorder Questionnaire; PHQ-9, Patient Health Questionnaire-9. 124 Postgraduate Medicine, Volume 123, Issue 5, September 2011, ISSN 0032-5481, e-issn 1941-9260

Depression and Collaborative Care Management Table 2. Logistic Regression of Variables Associated with 6-Month Clinical Response for Depression Treatment Clinical Response N = 574 Of the original cohort, 295 patients had data at the time of the study for their 12-month follow-up PHQ-9 scores. Clinical remission was present in 174 (59.0%) patients. The percentage of patients who were married was significantly higher in the middle-age groups than in the youngest and oldest groups (70.69% 78.57% vs 20.67% and 48.78%, respectively; P 0.0001). The AUDIT score was significantly increased in the younger group (4.23) and was lowest Table 3. Logistic Regression of Variables Associated with 6-Month Clinical Remission for Depression Treatment Clinical Remission N = 574 Odds Ratio Odds Ratio Confidence Interval Confidence Interval Age, years 30 0.9332 0.5605 1.5540 0.7905 30 39 1.0000 1.0000 1.0000 40 49 1.4654 0.8337 2.5757 0.1842 50 59 1.3107 0.7333 2.3427 0.3612 60 1.0675 0.5522 2.0636 0.8460 Race, 0.8323 0.4532 1.5284 0.5538 white vs non-white Gender, 1.1295 0.7307 1.7459 0.5838 female vs male Marital status, 0.8928 0.5984 1.3321 0.5787 married vs not Initial PHQ-9 score 1.0485 0.9972 1.1025 0.0645 AUDIT score 1.0202 0.9720 1.0709 0.4180 MDQ score, % negative 0.4721 0.2381 0.9360 0.0316 GAD-7 score 0.9539 0.9187 0.9905 0.0140 Abbreviations: AUDIT, Alcohol Use Disorders Identifi cation Test; GAD-7, Generalized Anxiety Disorder 7-Item Scale; MDQ, Mood Disorder Questionnaire; PHQ-9, Patient Health Questionnaire-9. Age, years 30 0.9282 0.5696 1.5124 0.7647 30 39 1.0000 1.0000 1.0000 40 49 1.4129 0.8408 2.3743 0.1919 50 59 1.3934 0.8134 2.3871 0.2271 60 1.4197 0.7582 2.6582 0.2735 Race, 1.0942 0.6052 1.9784 0.7657 white vs non-white Gender, 1.0947 0.7282 1.6456 0.6636 female vs male Marital status, 0.8299 0.5690 1.2103 0.3327 married vs not Initial PHQ-9 score 0.9407 0.8983 0.9851 0.0094 AUDIT score 1.0321 0.9871 1.0791 0.1651 MDQ score, % negative 0.4419 0.2181 0.8954 0.0234 GAD-7 score 0.9634 0.9302 0.9978 0.0372 Abbreviations: AUDIT, Alcohol Use Disorders Identifi cation Test; GAD-7, Generalized Anxiety Disorder 7-Item Scale; PHQ-9, Patient Health Questionnaire-9. in those aged 60 years (1.46; P = 0.044). Although percentage of patients in clinical remission at 6 months was not statistically significant between the groups at 0.0777, there was a trend toward a lower percentage of patients in remission for the youngest age group (age 30 years). Clinical remission measured at 12 months in this subgroup was not significantly different between the age groups (P = 0.7316). Race, gender, initial PHQ-9 score, and MDQ score were not significant. There was a nonsignificant trend for a lower GAD-7 score in those aged 60 years (Table 4). Multiple logistic regression analysis of these 295 patients demonstrated that the age group was again not statistically predictive of clinical remission (P = 0.3355 0.9575) (Table 5). Intake screening scores of GAD-7, AUDIT, and MDQ and the demographic variables of age, gender, race, and marital status were not associated with the likelihood of clinical remission at 12 months. An increased initial PHQ-9 score was significantly associated with a lower odds ratio (OR) of clinical remission at 12 months (OR, 0.9121; confidence interval [CI], 0.8475 0.9816; P = 0.0140). The patient being in clinical remission with a 6-month PHQ-9 score 5 was significantly associated with increased likelihood of continued remission at 12 months (OR, 7.4280; CI, 3.9301 14.0389; P 0.0001). Patients who had clinical response at 6 months but still had symptoms with a PHQ-9 score 5 also had significantly improved odds of remission at 12 months compared with those who did not have a response by 6 months (OR, 2.7722; CI, 1.1950 6.4313; P = 0.0176). Discussion In our study, patient age group was not a significant predictor of the efficacy of CCM for depression treatment at 6 and 12 months. There was no significant difference in response and remission rates when age groups were evaluated, even when controlling for demographic and mental health comorbidities. The results supported our original hypothesis. Collaborative care management demonstrated similar effectiveness for depression treatment among adults, regardless of age group. Consistent with other CCM study results, such as the IMPACT study, this study also confirmed the sustained efficacy in depression treatment that can be achieved through CCM, as the remission rate was 50% even at 1 year. The single most predictive variable for determining clinical remission at 12 months after CCM implementation was the presence of remission at 6 months. This would suggest that aggressive early management of depression and other mental health comorbidities would most likely lead to improved Postgraduate Medicine, Volume 123, Issue 5, September 2011, ISSN 0032-5481, e-issn 1941-9260 125

Angstman et al Table 4. Demographic and Initial Clinical Data of Patients in Collaborative Care Management for Depression at 12 Months After Diagnosis N = 295 Age 30 Years n = 60 (20.33%) Age 30 39 Years n = 80 (27.12%) Age 40 49 Years n = 56 (18.98%) Age 50 59 Years n = 58 (19.66%) Age 60 Years n = 41 (13.90%) Age, years 24.17 34.20 44.68 54.14 69.12 Race, % white 90.00 83.75 94.64 91.38 95.12 0.1836 Marital status, 26.67 75.00 78.57 70.69 48.78 0.0001 % married Gender, % female 75.00 77.50 66.07 77.59 65.85 0.4057 Initial PHQ-9 score (range) 16.22 (10 27) 14.89 (10 26) 15.32 (10 24) 15.86 (10 27) 15.17 (10 24) 0.348 MDQ score, % negative 93.33 97.50 87.50 93.10 95.12 0.2256 AUDIT score (range) 4.23 (0 29) 2.50 (0 29) 2.64 (0 30) 2.86 (0 27) 1.46 (0 6) 0.044 GAD-7 score (range) 11.62 (0 21) 11.09 (0 21) 10.34 (0 21) 11.02 (0 21) 8.10 (0 15) 0.060 Clinical remission 48.33 56.25 55.36 67.24 73.17 0.0777 at 6 months, % Clinical remission at 12 months, % 56.67 56.25 60.71 56.90 68.29 0.7316 Abbreviations: GAD-7, Generalized Anxiety Disorder 7-Item Scale; MDQ, Mood Disorders Questionnaire; PHQ-9, Patient Health Questionnaire-9. long-term outcomes. Inadequate treatment of depression among patients seen in primary care was one challenge that has been addressed through the implementation of CCM. When using the regression model to control for mental health comorbidities, the age of the patient was not associated with changes in clinical response or remission at 6 or 12 months. However, our study showed a Table 5. Logistic Regression of Variables Associated with 12-Month Clinical Remission for Depression Treatment Clinical Remission N = 295 Odds Ratio Confidence Interval Age, years 30 1.4963 0.6589 3.3978 0.3355 30 39 1.0000 1.0000 1.0000 40 49 1.2720 0.5775 2.8017 0.5505 50 59 0.9787 0.4440 2.1575 0.9575 60 1.2929 0.5061 3.3030 0.5914 Race, 0.5502 0.2213 1.3681 0.1986 white vs non-white Gender, 1.2505 0.6727 2.3243 0.4798 female vs male Marital status, 0.8401 0.4652 1.5173 0.5635 married vs not Initial PHQ-9 score 0.9121 0.8475 0.9816 0.0140 AUDIT score 0.9780 0.9194 1.0404 0.4817 MDQ score, % negative 0.5873 0.1920 1.7959 0.3506 GAD-7 score 0.9690 0.9191 1.0216 0.2432 6-month PHQ-9 score Remission 7.4280 3.9301 14.0389 0.0001 Response without 2.7722 1.1950 6.4313 0.0176 remission No response 1.0000 1.0000 Abbreviations: AUDIT, Alcohol Use Disorders Identifi cation Test; GAD-7, Generalized Anxiety Disorder 7-Item Scale; MDQ, Mood Disorder Questionnaire; PHQ-9, Patient Health Questionnaire-9. nonsignificant trend toward lower clinical remission rate among patients aged 30 years. This same trend toward lower remission was observed during the IMPACT trial among patients with advanced age ( 75 years). It would be interesting to evaluate the factors that make these 2 groups of patients at opposite ends of the adult age spectrum less likely to achieve remission. The next steps in the progression of CCM include the need to enhance dissemination and sustainability of the model as Katon et al 19 stated in a recent review article. As this happens, it will be important to look further into whether the CCM needs to be adjusted based on age to have the greatest success. Future studies could also be directed at the impact of treating anxiety on improvement in depression outcomes. Prior studies by our group with smaller cohorts have demonstrated findings consistent with this study s, showing a correlation between clinical response and remission rates, and mental health comorbidities. With a group of 334 patients, we demonstrated an increased risk of the patient not responding to treatment at 6 months if there is an increased intake GAD-7 score or positive MDQ score. 20 Also, in studying those who initially responded to CCM treatment, there was an increased risk of repeat diagnosis of depression and readmission to CCM in those patients who had an increased GAD-7 score. 21 This study expands on that by demonstrating how there was a strong association between negative MDQ scores and lower GAD-7 scores with improved remission rates at 6 months, and that clinical remission at 6 months was most predictive of remission at 1 year. 126 Postgraduate Medicine, Volume 123, Issue 5, September 2011, ISSN 0032-5481, e-issn 1941-9260

Depression and Collaborative Care Management Physical symptoms often accompany depression and are predictive of both the presence of comorbid psychiatric disorder 22 (eg, somatoform and/or anxiety syndromes) and a higher probability of negative treatment outcome. 23 The negative association between physical symptoms and treatment outcome is often mediated by patient adherence to pharmacotherapy. In 2008, Rasmussen et al 22 detected significantly high ORs for physical symptoms, such as headache, chest pain, dizziness, sleep problems, shortness of breath, tired or low energy, and fainting spells, in rural primary care patients with psychiatric disorders. Although these individual symptoms were predictive of a psychiatric disorder, the most powerful correlate was the total number of physical complaints, especially robust when the family physician perceived the symptoms to be medically unexplained. 22 The use of the PHQ-9 in this investigation enabled assessment of current depression by screening for symptoms that occurred during the preceding 2 weeks. The importance of identifying current depression is punctuated by the fluctuating bidirectional transaction of depression and chronic disease, particularly more common among patients in primary care with chronic conditions, such as cardiovascular disease, diabetes, asthma, arthritis, and cancer. 24 Our assessment methodology is similar to that of the PHQ-8, used in the Behavioral Risk Factor Surveillance System (BRFSS) epidemiologic survey from 2006 to 2008, on the prevalence of depression in 45 US states, Puerto Rico, and the US Virgin Islands. 25 Although this study was designed as a retrospective review of CCM treatment, there were very few exclusion criteria and it may reflect current status of depression treatment utilizing CCM. The limitations of the current study include the lack of racial diversity, relatively small subgroups, and a single practice group. Also, there was not a structured diagnosis of depression or psychiatric comorbidities at intake. The study involved enrollment of adult patients in CCM and results cannot be generalized to those patients with depression who are aged 18 years. Further studies with larger group numbers, multiple sites, and larger patient populations would enhance the generalizability of the results of this study. Conclusion Collaborative care management is a proven effective intervention for treatment of depression. Our study suggested that using CCM for depression treatment may transcend agerelated differences in depression and result in positive outcomes regardless of age. These results were seen at 6 months in the study cohort and also at 12 months in a smaller cohort. The most significant predictor of clinical remission at 1 year after diagnosis was the presence of remission at 6 months after initiation of treatment. Acknowledgments James Rohrer, PhD provided critical review for this manuscript. Mr. Isaac Johnson assisted with abstraction and collection of the data. Conflict of Interest Statement Kurt B. Angstman, MS, MD, Kathy L. MacLaughlin, MD, Norman H. Rasmussen, EdD, Ramona S. DeJesus, MD, and David J. Katzelnick, MD disclose no conflicts of interest. References 1. World Federation for Mental Health. Mental health and chronic physical illness. The need for continued and integrated care. Woodbridge, VA: 2010. http://www.wfmh.org/2010docs/wmhday2010.pdf. Accessed July 6, 2011. 2. World Health Organization. Projections of mortality and burden of disease, 2004 2030. Geneva, Switzerland; 2008. 3. World Health Organization. The global burden of disease: 2004 update. Vol 2010. Geneva, Switzerland; 2008. 4. Husain MM, Rush AJ, Sackeim HA, et al. Age-related characteristics of depression: a preliminary STAR*D report. Am J Geriatr Psychiatry. 2005;13(10):852 860. 5. Klap R, Unroe KT, Unutzer J. Caring for mental illness in the United States: a focus on older adults. Am J Geriatr Psychiatry. 2003;11(5):517 524. 6. Gilbody S, Bower P, Fletcher J, Richards D, Sutton AJ. Collaborative care for depression: a cumulative meta-analysis and review of longerterm outcomes. Arch Intern Med. 2006;166(21):2314 2321. 7. Akerblad AC, Bengtsson F, Ekselius L, von Knorring L. Effects of an educational compliance enhancement programme and therapeutic drug monitoring on treatment adherence in depressed patients managed by general practitioners. Int Clin Psychopharmacol. 2003;18(6):347 354. 8. Hunkeler EM, Meresman JF, Hargreaves WA, et al. Efficacy of nurse telehealth care and peer support in augmenting treatment of depression in primary care. Arch Fam Med. 2000;9(8):700 708. 9. Katon WJ, Von Korff M, Lin EH, et al. The Pathways Study: a randomized trial of collaborative care in patients with diabetes and depression. Arch Gen Psychiatry. 2004;61(10):1042 1049. 10. Rost K, Nutting P, Smith J, Werner J, Duan N. Improving depression outcomes in community primary care practice: a randomized trial of the quest intervention. Quality Enhancement by Strategic Teaming. J Gen Intern Med. 2001;16(3):143 149. 11. Unutzer J, Katon W, Callahan CM, et al; IMPACT Investigators. Improving Mood-Promoting Access to Collaborative Treatment. Collaborative care management of late-life depression in the primary care setting: a randomized controlled trial. JAMA. 2002;288(22):2836 2845. 12. Simon GE, Ludman EJ, Tutty S, Operskalski B, Von Korff M. Telephone psychotherapy and telephone care management for primary care patients starting antidepressant treatment: a randomized controlled trial. JAMA. 2004;292:935 942. 13. Ell K, Aranda MP, Xie B, Lee PJ, Chou CP. Collaborative depression treatment in older and younger adults with physical illness: pooled comparative analysis of three randomized clinical trials. Am J Geriatr Psychiatry.18(6):520 530. Postgraduate Medicine, Volume 123, Issue 5, September 2011, ISSN 0032-5481, e-issn 1941-9260 127

Angstman et al 14. Van Leeuwen Williams E, Unutzer J, Lee S, Noel PH. Collaborative depression care for the old-old: findings from the IMPACT trial. Am J Geriatr Psychiatry. 2009;17(12):1040 1049. 15. Spitzer RL, Kroenke K, Williams JB. Validation and utility of a selfreport version of PRIME-MD: the PHQ primary care study. Primary Care Evaluation of Mental Disorders. Patient Health Questionnaire. JAMA. 1999;282(18):1737 1744. 16. Reinert DF, Allen JP. The alcohol use disorders identification test: an update of research findings. Alcohol Clin Exp Res. 2007;31(2):185 199. 17. Spitzer RL, Kroenke K, Williams JB, Löwe B. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med. 2006;166(10):1092 1097. 18. Hirschfeld RM, Calabrese JR, Weissman MM, et al. Screening for bipolar disorder in the community. J Clin Psychiatry. 2003;64(1):53 59. 19. Katon W, Unutzer J, Wells K, Jones L. Collaborative depression care: history, evolution and ways to enhance dissemination and sustainability. Gen Hosp Psychiatry. 2010;32(5):456 464. 20. Angstman K, DeJesus R, Rohrer J. Correlation between mental health comorbidity screening scores and clinical response in collaborative care treatment for depression. Mental Health in Family Medicine. In press. 21. Angstman K, MacLaughlin K, Williams M, Rasmussen N, DeJesus R. Anxiety and length in treatment impact early re-admission to collaborative care treatment for depression. J Prim Care and Comm Health. In press. 22. Rasmussen NH, Bernard ME, Harmsen WS. Physical symptoms that predict psychiatric disorders in rural primary care adults. J Eval Clin Pract. 2008;14:399 406. 23. Huijbregts KM, van der Feltz-Cornelis CM, van Marwijk HW, et al. Negative association of concomitant physical symptoms with the course of major depressive disorder: a systematic review. J Psychosom Res. 68(6):511 519. 24. Johnson S. Medically Unexplained Illness: Gender and Biopsychosocial Implications. Washington, DC: American Psychological Association; 2007. 25. Centers for Disease Control and Prevention (CDC). Current depression among adults United States, 2006 and 2008. MMWR Morb Mortal Wkly Rep. 2010;59(38):1229 1235. 128 Postgraduate Medicine, Volume 123, Issue 5, September 2011, ISSN 0032-5481, e-issn 1941-9260