Depression in Corporate America: An Integrated Care Approach to Increase Productivity and Improve Outcomes

using integrated models that leverage primary care provider treatment with care management and mental health consultation. CONCLUSION: There is a need to reduce silos in the organization and financing of mental health care to prevent cost shifting that provides no benefit to patients, payers or providers. Poor mental health care will likely lead to a rise in absenteeism and presenteeism.

Supplements to the J o u rnal of Managed C a re Pharm a c y a re intended to support medical education and re s e a rch in areas of clinical practice, health care quality i m p rovement, or efficient administration and delivery of health benefits. The following standards are applied to all J M C P supplements to assure quality and assist readers in evaluating potential bias and d e t e rmining alternate explanations for findings and re s u l t s .
1 . Disclose the principal sources of funding in a manner that permits easy recognition by the re a d e r.
2 . Disclose the existence of all potential conflicts of interest among supplement contributors, including financial or personal bias.
3 . Describe all drugs by generic name unless the use of the brand name is n e c e s s a ry to reduce the opportunity for confusion among re a d e r s . 4 . Strive to re p o rt subjects of current i n t e rest to managed care pharmacists and other managed care pro f e s s i o n a l s .

.
Seek and publish content that does not duplicate content in the J o u rnal of Managed C a re Pharm a c y. 6 . Subject all supplements to expert peer re v i e w.

Table of Contents
Depression in Corporate America: An Integrated Care Approach to Increase Productivity and Improve Outcomes *A total of .20 CEUs (2.0 contact hours) will be awarded for successful completion of this continuing education program (ACPE Program No. 062-000-05-036-H04).
The articles published in this supplement represent the opinions of the authors and do not reflect the official policy or views of the Academy of Managed Care Pharm a c y, the authors' institutions, or Wyeth unless so specified.

Target Audience
Pharmacists working in all settings and other health care professionals

Learning Objectives
Upon completion of this program, participants will be better able to 1. recognize the importance and urgency of correctly diagnosing depression and related d i s o rders and treating patients early and aggressively until they achieve treatment to re m i s s i o n ; 2. understand the impact of depression on worker productivity, including absenteeism and presenteeism; 3. identify how employers can be proactive in decreasing the stigma of depression and associated diseases, thereby increasing awareness and diagnoses, to improve treatment outcomes, productivity, and job satisfaction; 4. describe how NCQA designs and monitors HEDIS measures for antidepressant medication management and gain an understanding of the implications of instituting best practices to improve HEDIS scores to employers, providers, and health plans; 5. discuss the direct and indirect value of depression pharmacotherapy and consider how the cost/benefit ratio can contribute to overall improvements in workforce health; and 6. understand the clinical and economic implications associated with treatment selection and integration of care across stakeholders.
O ne of the greatest challenges among employees and employers in today' s economic climate is the ability to pay for health care. Between 1991 and 2002, annual national expenditures on health care nearly doubled from $762 billion to $1.6 trillion. 1 The cost of health care has risen in all sectors, but there has been a dramatic increase in the perc e n t a g e of spending for prescription drugs compared with hospital and physician services over the decade. 2 T h e re is an urgent need to d e t e rmine the value being obtained for this increase in expenditure s while maintaining the quality of health care .

■ ■ Health Care and Employers
Employers are keenly aware of this rise in health care expenditure s because they are one of the major payers for health care. In the face of double-digit inflation, employers are transferring larger fractions of the cost, especially medication costs, to their employees. This t rend can unfortunately result in increased costs in other are a s , thus only shifting the cost.
The Institute of Medicine has re p o rted that, in the way we deliver health care, there are areas for impro v e m e n t . 3 For example, we do not offer preventive care for 50% of Americans, and the care that we do deliver is fragmented. 4 Employers must recognize the cost savings of providing preventive care versus the cost of untreated illnesses.
Medical errors, the rise in incidence and prevalence of cert a i n diseases, and new technologies for detection and treatment of illnesses for our aging population are also sources of rising cost to both the health care provider and patients alike. In addition to quality care, patients expect that health care providers work in a c o o rdinated fashion to optimize treatment. Unfort u n a t e l y, this does not occur in many instances.
Despite these difficulties and the rising cost of pharm a c e u t i c a l s , t h e re have been significant improvements in the treatment of many major disease states. Medications, in general, have been quite effective in improving the lives of patients with HIV, cancer, and heart disease. 5 For example, between 1980 and 2000, there has been a one-third decrease in the number of deaths per 100,000 population due to heart disease. The question remains as to whether we are able to achieve reductions in morbidity and acute care costs while increasing productivity and quality of life for patients with mental illnesses such as depression.

■■ Depression
One area in which there is potential room for improvement in cost savings to the employer and benefit to the employee is in the t reatment of depression. Depression is one of the most debilitating diseases that have significant effects on patients, family members, and society. Major depression is currently the leading cause of disability worldwide. 6 Individuals with depression also have i n c reased risk of morbidity and mortality resulting from psychosocial d i s t ress, comorbid diseases, loss of pro d u c t i v i t y / income, and suicide. 7 A c c o rding to the National Institute of Mental Health, depression affects 18.8 million Americans, or about 10% of the adult population. 8 The lifetime prevalence in the community sample is 16.2% according to the recent National Comorbidity Survey-Replication (NCS-R). 9

Treatment Options
The treatment of depression has changed significantly from the days of Freudian philosophy when mental illnesses were f requently attributed to nonbiological etiologies. 1 0 Older tre a t m e n t s such as insulin and electroconvulsive therapy were commonly used to treat many mental health disord e r s . 1 0 When the neuroc h e m i c a l basis for depression was discovered, drugs such as monoamine oxidase inhibitors and tricyclic antidepressants were c o n s i d e red the gold standard for the treatment of depre s s i o n . 1 1 U n f o rt u n a t e l y, these agents were associated with problems such as c a rdiac arrhythmias, significant sedation, anticholinergic eff e c t s , and orthostatic hypotension.
In the 1980s, the first drug in a class of selective sero t o n i n reuptake inhibitors (SSRI) was discovered and, again, dramatically changed the way we treated depression. These drugs not only t reated the depressive symptoms but also were significantly better tolerated by patients. Of course, SSRIs are not without their own side effects, and patients must still be monitored for symptoms such as sedation, agitation, headache, gastrointestinal pro b l e m s , and sexual dysfunction. 1 1 Challenges to Treatment Treatment of depression, however, is not solely dependent upon choosing the right medication for a patient. There are many challenges that exist in the treatment of mental illnesses, including d e p ression. Patients and providers alike must overcome the stigma of the disease and the antiquated views that suggest that depre s s i o n is simply a state of mind. Beyond the difficulty of identifying patients with depression, getting the patient to agree to re c e i v e t reatment is a significant hurdle in itself. In addition, it is diff i c u l t to convince patients to continue with their treatment when i m p rovement is not usually seen until after 4 to 6 weeks of therapy. Patients are often faced with short -t e rm side effects without much i m p rovement in their mood, thus resulting in discontinuation of the medication. F u rt h e r, recent re s e a rch indicates that the best method for t reating depression is a combination of psychotherapy and m e d i c a t i o n s . 1 2 , 1 3 P roviding patients with access to an adequate trial of psychotherapy is perhaps even more challenging than a c q u i r i n g medications and, again, impedes the process of pro v i d i n g o p t i m a l therapy for the disease. Finally, we have little outcome data pertaining to treatment options for depression. More re s e a rch is still needed to determine the impact of depression on absenteeism, p ro d u c t i v i t y, and the overall bottom line for payers. Employers are not willing to spend more dollars without reasonable expectation of an adequate re t u rn on their investment. Payers would benefit f rom data that demonstrate the value of spending more money on t reatment of depression and how to spend the money in ways that make both clinical and economic sense.

■ ■ Conclusion
Despite these gaps and barriers, we have opportunities, in terms of access, quality, and cost, to take advantage of newly emerg i n g m o d e l s of care to improve depre s s i o n t reatment. Emerging e v i d e n c e has shown that quality care can prevent relapse and integrated care models can improve outcomes for patients and health care systems. Minimizing costs while improving the health and quality of life for employees remains a significant challenge for most employers. We need to have more employers who are willing to look at the problem globally instead of simply shifting cost to their employees.

DISCLOSURES
The author received an honorarium for participation in the symposium upon which this article is based. He discloses that he has received grant/research support from numerous pharmaceutical companies, including Lilly, Pfizer, Merck, Sanofi, and AstraZeneca. I n order to understand how we can improve the treatment of d e p ression, it is useful to assess the problem from diff e re n t perspectives. We know much about the impact of depre s s i o n and its treatment from the patient and provider perspectives, but we rarely consider the ramifications of the disease on the purc h a s e r. P u rchasers have considerably more leverage than they curre n t l y e x e rt in the quality of care health systems deliver. In part i c u l a r, p u rchasers may be able to encourage health systems to pro v i d e better quality care by advocating for the adoption of integrated c a re models that are emerging for the treatment of depression and other chronic diseases. Before purchasers can advocate for these models, they must have evidence of the clinical and economic value these models pro d u c e .

■ ■ Depression in the Workplace
The private employer insures about 60% of Americans and is a major purchaser of health care for most working Americans. 1 Because 1 out of every 10 Americans suffer from depression annua l l y, the likelihood of a coworker being affected by depression in the workplace is high. 2 In fact, absenteeism from depression is estimated to be about 1.6 days of work lost per employee per month which is equivalent to about 1,500 lost days per month for a company with 1,000 employees. 3 Absenteeism results in i n c reased workload for other employees, reduced output, and lost income from hiring temporary workers.
In addition, reduced productivity at work, or "presenteeism," is a significant but underrecognized concern for employers. It has been re p o rted that employees who suffer from depression work at about 70% of their optimal pro d u c t i v i t y. 4 -7 Also, 62% of employees re p o rt decreased mental functioning that ultimately affects their output and time management.
In 2000, depression cost employers an estimated $51.5 billion. 8 A p p roximately 70% of that cost was due to absenteeism while the rest was attributed to productivity loss. This is a significant financial p roblem for employers, who are already paying $26.1 billion for d e p ression tre a t m e n t . 8 Employers are faced with a complex i s s u e -i n c reased numbers of employees being diagnosed with d e p ression, increased rates of treatment for depression, and i n c reases in pharmaceutical costs. Employers are seeking newer and better solutions to this problem, ones that can demonstrate s u fficient re t u rn on investment to warrant adoption.

■■ Integrated Care Models
T h e re are a number of potential solutions to improve the clinical and economic outcomes of depression treatment. Integrated care models that are currently available focus on 3 aspects of care that a re essential for the treatment of depression. The first step, using a screening tool, is to identify patients who may have depre s s i o n . As we know, the majority of patients who receive treatment for d e p ression do so from their primary care physician (PCP). 9 T h e I m p r oving Depression Treatment by Integrated Care

A B S T R AC T
OBJECTIVE: To identify the clinical and economic implications of depression in the workplace and review how integrated care models can improve overall patient outcomes.
SUMMARY: Depression is a significant financial burden to the employer due to lost days of work and decreased productivity. Employers are demanding return on the investment for their increasing health care expenditures. The cost of depression to employers may be contained by delivering care using integrated models that leverage primary care provider treatment with care management and mental health consultation.
CONCLUSION: There is a need to reduce silos in the organization and financing of mental health care to prevent cost shifting that provides no benefit to patients, payers or providers. Poor mental health care will likely lead to a rise in absenteeism and presenteeism. reason may be partly because managed care organizations are shifting the burden of depression and anxiety treatment away fro m mental health specialists to PCPs. 1 0 In addition, to reduce cost, many managed behavioral health care organizations limit the number of specialty care visits a depressed patient can make. 1 1 , 1 2 Even in short PCP visits, simple screening tools for depression can aid the PCP in recognizing patients who may be at risk for d e p ression or have clinically definable disease.
In our initial program designed 10 years ago, trained nurses identified patients with depression and provided care management to them over 2 years using a decision-tree care plan ( Figure 1). Nurses assessed disease severity, educated patients about tre a t m e n t options, and monitored their pro g ress over time. In today' s models, c a re managers are also supervised by either a psychiatrist or a p h a rmacist.
While pharmacists can provide patient education and m o n i t o r i n g , they also have unique expertise in providing a l t e rnatives for drugs that are intolerable or clinically ineffective. In the future, utilization of pharmacists may be the most costeffective solution for managing patients with complex medication regimens in the managed care setting. In addition, the shortage of psychiatrists may significantly increase the demand for advanced p h a rmacy practitioners to fill these care management needs.

Impact of Model
Integrated care models will in all likelihood increase the number of patients who are identified with depression and begin tre a tment. It has been estimated that 50% of all depressed patients in the primary care setting remain undiagnosed. 1 3 In addition, the education integrated care models provide will result in gre a t e r patient understanding about the importance of treatment completion, which will, in turn, curb early discontinuation rates. M o re frequent contacts with care managers along with counseling about medication side effects should result in increased compliance. Patients frequently list intolerable side effects as the p r i m a ry reason for discontinuing their treatment pre m a t u re l y.
A national survey found that, in patients with probable anxiety or depressive disord e r, only about 30% received some form of a p p ropriate tre a t m e n t . 1 4 Although identification of patients is an i m p o rtant first step, the field also needs to improve the tre a t m e n t patients receive once they are identified. The process of ensuring that patients actually fill a prescription and then take the medication a p p ropriately is complex, and we must find ways to bridge gaps in the process. Even after patients initiate therapy, there is still ro o m for potential complications. Care managers can facilitate appro p r i a t e medication switches and encourage patients to continue tre a t m e n t when the initial medication fails. They can also re c o m m e n d dosage increases, if needed, to optimize therapy. Additionally, they can simply encourage patients to continue with therapy during the time it can take for antidepressant medications to take full e ffect. Often, patients are tempted to discontinue their medications during this time because of intolerable side effects while experiencing little clinical benefit.

Outcomes Data
We evaluated the clinical and economic outcomes of an integrated c a re model to demonstrate its value to potential purchasers. Over 2 years, about 74% of patients who received our intervention were in probable remission compared with only 40% of patients who received usual care. In addition to improving emotional role functioning to close to population norms, the model had a s i g n i f i c a n t but smaller impact on physical functioning, possibly due to medical comorbidities commonly associated with depre s s i o n . The cost of the program is $130 per year with an additional $134 per year for incremental treatment. Accounting for inflation for year 2005, the total cost of the model was estimated to be $297 per year per treated depressed employee. We recognized that patients improved clinically, but what other benefits accrued and to whom? Over 2 years, the intervention decreased the number of hours of work lost in the previous month from ~20 hours to 4 hours, resulting in an average reduction of 12.3 days of absenteeism, with an annual value of $648 per participating employee ( Figure 2). Over 2 years, the intervention also improved productivity at work in the previous 2 weeks, resulting in an average increase of 8.2%, with an annual value of $1,982 per participating employee ( Figure 3).
It is possible for employers to calculate their re t u rn on investment for their depression care program based on a business case model that I have developed. The re t u rn on investment is dependent upon the ratio of annual savings with quality depre s s i o n c a re to the cost of the depression care program and the additional t reatment it stimulates. Return on investment is dependent upon various factors such as hourly wage, sick leave benefit, likelihood of increased revenue with increased pro d u c t i v i t y, likelihood of hiring temps to cover absent employees, the company's contribution to health plan premium, and the prevalence of depre s s i o n in the specific employee population. The re t u rn on investment calculation can be done for any company, health plan, or institution and is available at h t t p : / / w w w. d e p re s s i o n -p r i m a ry c a re . o rg / o rganizations/employers/ c a l c u l a t o r.

■ ■ Conclusion
Primary care providers are faced with a difficult task of a d d ressing a variety of health issues during every visit. Integrated care models such as the one presented here can be the part of the solution to optimize clinical and economic outcomes in the primary care treatment of depression.

ACKNOWLEDGMENTS
The author wishes to acknowledge Mark Schwartz, MBA, for his work on the model for this study. She also acknowledges the physicians, office staff, and patients of the following participating primary care practices:

DISCLOSURES
Funding for this study was contributed by the National Institute of Mental Health and the MacArthur Initiative on Depression and Primary Care and obtained by the author. The author received an honorarium for participating in the symposium upon which this article is based. She discloses that she has received grant/research support from NIMH, the MacArthur Foundation, and the Robert Wood Johnson Foundation.

Improving Depression Treatment by Integrated Care
Intervention Effects on Absenteeism Intervention Effects on Productivity  O rganizations such as the National Committee on Quality Assurance (NCQA) and Leapfrog are setting the s t a n d a rds on how to measure quality. For managed care o rganizations, NCQA sets the standards by which health care d e l i v e ry is evaluated through its accreditation programs. NCQA is a private, nonprofit organization that strives to improve the quality of health care through measurement, transpare n c y, and a c c o u n t a b i l i t y. The set of measurement standards, Health Plan Employer Data and Information Set (HEDIS), is developed with input from various groups including stakeholders and expert s . The Committee on Perf o rmance Measurement includes national e x p e rts who determine the measures that are appropriate for HEDIS. The Measurement Advisory Panels are composed of e x p e rts for each disease state, and they decide on the specific m e a s u res based on relevance, scientific evidence, and feasibility.

■ ■ HEDIS Measures for Depression
One of the HEDIS measures that evaluates the effectiveness of care is the antidepressant medication management measure (Table 1). 1 This measure is designed to evaluate the optimal number of practitioner contacts and the duration of acute phase and continuation phase treatment for persons with new episodes of depression treated with antidepressant medications. Accord i n g to the recent 2004 HEDIS perf o rmance re p o rt for the commerc i a l health plans, the average perf o rmance on acute phase tre a t m e n t m e a s u re is around 60%(

A B S T R AC T
OBJECTIVE: To review the role of the National Committee for Quality Assurance (NCQA) in ensuring the quality of care in the managed care setting and identify novel strategies to improve performance rates for Health Plan Employer Data and Information Set (HEDIS) measures, particularly in the area of depression. SUMMARY: NCQA, by regulating HEDIS measures, sets the standards by which managed care organizations evaluate their performance in providing care for their enrollees. The medication management measure for depression evaluates practitioner contacts and acute and continuation phase treatments for persons treated with an antidepressant. Despite increased detection and management of patients with depression, there is still room for improvement in HEDIS performance rates for this chronic disease.
CONCLUSION: NCQA hopes to improve collaboration among managed care organizations and managed behavioral health organizations In addition, NCQA regularly reevaluates the HEDIS measures using input from panels of experts. Incentive p r o grams for providers who deliver quality care may also help to improve HEDIS p e r f o rmance rates for depression. Research is under way to evaluate the feasibility and re t u rn on investment for pay-for-p e r f o rmance programs in depression.   A similar pattern is seen for follow-up rate after mental health hospitalization. If we compare these perf o rmance measures to those for non-mental health conditions, we can see that there is still room for improvement ( Figure 1). In 2003, the average p ro p o rtion of patients who received beta-blockers after a heart attack was high, with a rate of more than 94%. 2 P e rf o rmance rates for other chronic conditions such as asthma and diabetes management are also high compared with those for mental health m e a s u res. There has been little improvement over the past 5 years in any of the antidepressant medication management or follow-up m e a s u res. The question remains as to how we can improve our p e rf o rmance for mental health disorders as we have for non-mental health conditions. Although the measures for antidepressant medication management are not perfect, there are some advantages and good rationale for using them to measure the quality of care for patients with d e p ression. These measures ensure that we identify patients with a new episode of depression who should be treated for a specific length of time. 3 One of the aims for NCQA is to encoura g e collaboration among the managed care organizations and managed behavioral health organizations for the purposes of sharing data.
T h e re are several criticisms to using these measures, including the concern that the denominator, or the number of patients identified with new episodes of depression, is too low and does not reflect actual incidence rates in the population. 4 , 5 A re c e n t i m p rovement was the inclusion of telephone contacts into the optimal provider contacts measure. Of course, plans must be able to track telephone contacts in order to incorporate this number. Other problems that plans still face are the use of samples in p r i m a ry care practice and delays in diagnosis, mainly attributable to stigmatization of the illness in the community. NCQA perf o rms regular reviews of the depression measures in conjunction with its Behavioral Health Measurement Advisory Panel.

■■ Measuring and Encouraging Accountability and Quality of Care
A new direction of NCQA is to incorporate measurement and accountability into provider measurement and re w a rd pro g r a m s . P rograms such as Bridges to Excellence on the East coast and payf o r-p e rf o rmance programs in California aim to measure the quality of care given by providers and to re w a rd them (either the p roviders or medical groups) for good perf o rmance.
T h e re are 2 ways in which provider quality is being measure d in pay-for-p e rf o rmance programs. In the first approach, NCQA integrates administrative data received from the health plans. 6 T h e health plans receive a perf o rmance re p o rt, and the health plans re w a rd those groups based upon high perf o rmance rates. Another a p p roach is to use NCQA' s recognition programs where physi-   cians ask to be recognized for the quality of their care . 7 NCQA currently sponsors 3 recognition programs in diabetes, cardiovascular d i s e a s e / s t roke, and practice systems.

Comparing Trends in Behavioral and Nonbehavioral HEDIS Measures
In the content area of practice systems, physicians conduct medical re c o rd reviews or provide information about their practice setting. NCQA conducts surveys and random audits to determ i n e whether the physician or the practice group merits re c o g n i t i o n . Again, HEDIS measures are utilized, and the recognition pro g r a m s a re conducted in partnership with national organizations such as the American Heart Association and the American Diabetes Association. Of course, these programs are strictly voluntary and only those physicians or practice groups who meet the re c o g n i t i o n t h resholds are publicly re p o rt e d .
T h e re are a growing number of organizations that are using these recognition programs to recognize physicians. For example, an employer consortium, the Bridges to Excellence program, pays re w a rds to providers based on NCQA' s recognition pro g r a m s . 8 R e c e n t l y, NCQA received a grant from the Robert Wood Johnson Foundation to identify and test the feasibility of perf o rm a n c e m e a s u res for depression that would be suitable for pay for p e rf o rmance pro g r a m s . 9

■ ■ Conclusion
The identification and treatment of depression has improved with education, care management programs, and better utilization of a n t i d e p ressant medications. Future directions for NCQA with these projects include identifying potential indicators for depre s s i o n that assess the stru c t u re and process of depression care. We plan to incorporate outcome measures in order to determine whether patients improve clinically. We hope to meet the goals of NCQA in developing measures to improve quality of care and incre a s e accountability at all levels of health care .

DISCLOSURES
The author received an honorarium for participation in the symposium upon which this article is based. She discloses no potential bias or conflict of intere s t relating to this article. D e p ression is one of the leading causes of disability in the United States, resulting in estimated medical costs of $26.1 billion annually. 1 I n d i rect costs are even higher, with more than $50 billion attributed to lost productivity and absenteeism among depressed employees. 2 The ultimate payer of these costs is the employer n o t the health plan since the majority of these costs are not incurred within the health care systems but in the workplace. Health care expenditures re p resent the fastest g rowing costs among employee benefits, but lost labor costs are much gre a t e r. In the long run, there f o re, the primary objective of managing depression should be to create a new value pro p o s i t i o n by examining the impact of health status-and cost-upon work p ro d u c t i v i t y.
The relationship of health status to employee productivity is vital to understand. In the coming years, the workforce will actually be shrinking so employers will need to keep their employees healthy if productivity is to be optimized. If depre s s e d patients continue to wind up on disability, the predicted decline in skilled, middle-aged employees will result in a serious deficit in the workforce, making it very difficult to sustain a competitive e n t e r p r i s e .

■ ■ Integration of Health Management
In order to better appreciate the impact of health upon p ro d u c t i v i t y, we will need to stop thinking of people (or employees) as diagnoses or disease entities. A total health measure must be c reated and quantified, and the focus has now shifted to measuring f u n c t i o n a l i t y. This can only be achieved by integrating several key activities (Table 1). For example, a thorough health risk assessment is essential in order to identify potential (or existing, but as yet undetected) cases of disease and ultimately avoid some or all of the direct and indirect costs that will eventually occur if no i n t e rvention is implemented. A health risk assessment also enables employers and health plans to target chronic illnesses for wellness and promotion activities to reduce the total disease b u rden. These programs may not be effective for all conditions so thought must be given to prioritizing disease states that may actually be significantly influenced by decreased risk (e.g., obesity, diabetes, and depression). After prevention, the focus shifts to disease management since we will never be able to eradicate all illnesses. The prevalence and severity of some chronic conditions will almost certainly rise in the coming years in spite of all eff o rts, as the workforce continues to age. The next link in this integrated chain then becomes disability management. While most companies continue to re g a rd disease management and disability management as 2 diff e re n t entities with distinct data sets and processes, they should really be analyzed and handled together. Fort u n a t e l y, this situation is i m p roving in the corporate world as companies such as United Technology begin to integrate these activities.

Integrated Health Management Requires Linking Several Key Activities
The final piece of integrated health care management can also be viewed as the sum total of the process. It is called "health and p roductivity management" (HPM), and it involves the integration of data and services specific to all other activities. HPM re q u i re s the deconstruction of these other "silos" or "stovepipes" so that corporations can analyze outcomes from a broader perspective than is currently possible. It is only through this level of integration that one can rapidly and accurately quantify medical cost off s e t s , d e c reased hospitalizations, and, most import a n t l y, improved work p e rf o rm a n c e .
The integration and reconfiguration of employee health and p roductivity has also led to the identification of the importance of o rganizational health and culture. For many employers, this is a f o reign term but it can have a major influence upon overall health. The term "organizational health and culture" re p resents an enhanced understanding of how the work environment can impact health status and functioning. Organizational health and c u l t u re is embodied in the apparent values of the workplace, the way employees interact, the means by which information is t r a n s m i t t e d , and how work is ultimately accomplished.
The success of health promotion and disease management p rograms will depend essentially on the culture in which these activities take place. For example, one needs to ascertain if upper management genuinely supports HPM activities and whether or not they are truly willing to invest in employees' health. For the f u t u re, it has become clear that employers must believe that employees are the corporation' s greatest asset and invest in HPM a c c o rdingly if they are to remain a viable enterprise.

■■ Specific Steps
T h e re are many steps involved in the integration of these activities. Busting the silos is a requisite if data is to be shared and used e ff e c t i v e l y. In the current situation, departmental incentives are typically to shift costs from one silo to the next (e.g., move employees f rom disability to workers compensation, turning a medical claim into a disability claim), which is of no economic benefit to the company overall. Once data is shared, it is then viewed within the context of the company' s own demographics. Cost stru c t u res, risk profiles, and other demographics will vary g reatly among corporations and in diff e rent regions of the c o u n t ry . A company' s specific demographics will often dictate w h e re in the HPM system the health dollars should be invested (e.g., health promotion versus disease management).
A diff e rent philosophical approach may also be necessary to integrate disease prevention with disease management. For instance, some health policy experts believe that the pre v i o u s focus upon heavy utilizers in a health care system is misguided. While it is true that 20% of a beneficiary pool will be re s p o n s i b l e for generating roughly 80% of the medical costs, the real challenge is to identify individuals in the remaining 80% of the population who are at increased risk for becoming heavy utilizers. 3 The goal is to keep these employees healthy well into the future. If the e n t i re eff o rt is directed toward controlling medical costs, the system, the employer, and the company are doomed to failure . The current demographics are not in the employers' favor. Technology and associated costs are working against them as well.
A final component to consider is how the re t u rn on these investments will be measured. Conservative estimates suggest that nonmedical costs are at least twice as high as direct medical e x p e n d i t u re s . 4 H i s t o r i c a l l y, employers have only been concern e d with the latter. If and when mechanisms are implemented to m e a s u re the success of HPM activities, it will become evident that the lost productivity costs should be the biggest target for employer i n t e rventions. Only then will employers genuinely begin to make employee health an explicit part of their business.

■■ Employers' Perspective
In an effort to characterize the attitudes and beliefs of administrators, the Institute for Health and Productivity Management conducted a cross-sectional survey in 2002 of corporate medical d i rectors, benefits directors, human re s o u rces directors, and associated wellness personnel. Their survey was designed to quantify the perceived medical reasons underlying employee absenteeism and lost productivity ( Table 2). The results appeare d

Promoting Health and Productivity for Depressed Patients in the Workplace
to reflect a wide variety of corporate backgrounds and experiences. The corporate survey revealed, for instance, that musculoskeletal conditions were believed to be the leading cause of absenteeism. This may be due to antiquated perceptions about the amount of physical labor currently perf o rmed in the workplace, but it may also reflect an increase in the prevalence of osteoarthritis, as one may come to expect from an aging population. Mental health was the #2 reason listed for absenteeism and pregnancy was third , though that is generally considered to be a short -t e rm disability as opposed to a chronic illness. Other conditions that were also identified with high absenteeism rates were re s p i r a t o ry conditions (e.g., allergies, chronic obstructive pulmonary disease, and asthma), g a s t rointestinal (GI) problems, and cardiovascular illness. The other question on the survey asked participants to rank the leading causes for lost productivity or "presenteeism," as it is now commonly known. The response of administrators provided additional testimony to the enormous impact that depression has on the workplace (Table 3). Presenteeism is believed to be a much bigger economic factor than absenteeism, in general, and in the s u rv e y, mental health was the leading reason listed for decreased p e rf o rmance. Depression, specifically, was commonly cited, suggesting that depressed employees will often show up for work but they're "not really there," or at least are not fully pro d u c t i v e . Musculoskeletal problems were perceived to be a major cause of absenteeism as well, followed once more by re s p i r a t o ry and GI p roblems. It is interesting to note that migraine headaches appear on this list as well. The medical literature has shown that a relatively small expenditure for migraine treatment can have a tremendous re t u rn on investment for the employer. 5 The relationship of indirect costs to direct costs can vary depending upon the chronic condition in question. In the early 1990s, re s e a rchers began comparing these relative costs for disease states and this pre l i m i n a ry data suggests indirect costs (i.e., lost productivity) are the primary cost factor for migraine headaches, arthritis, and depression ( Figure 1). 6 In recent times, the methods for measuring presenteeism have improved s u b s t a nt i a l l y, and the medical community is beginning to appre c i a t e j u s t how large the impact of depression is on this aspect of work p e rf o rmance. One can anticipate that this economic factor will only gain importance as additional re s e a rch is devoted to this t o p i c . Yet another area that re s e a rchers have only begun to unravel is the impact of pharmaceuticals upon pro d u c t i v i t y. As pre v i o u s l y mentioned, depression is widely recognized as a major cause of d i s a b i l i t y. The good news is that medications can have a very favorable-and measurable-effect on worker pro d u c t i v i t y. 7 I m p rovements seen with antidepressants actually appear to be much greater than with medical treatments used to manage a n x i e t y, migraine headaches, and hypertension ( Figure 2). 7 O n c e m o re, it is hoped that contemporary re s e a rch methods will be employed in the near future to improve our understanding of this i s s u e .

■■ Conclusion
D e p ression is a big cost factor for employers, but it also re p re s e n t s a big cost-saving opport u n i t y. In the developed world, depre s s i o n is viewed as a leading cause of disability, and its influence upon p resenteeism is particularly profound. With a combination of p h a rmacotherapy and psychotherapy, depression is eminently t reatable but oftentimes goes undiagnosed, untreated, or subopti-  Promoting Health and Productivity for Depressed Patients in the Workplace mally treated, and the associated costs become unnecessarily high.

Direct and Indirect Costs of Major Illness in the Workplace
Integrated care models can result in substantial cost savings and a much more productive workforc e .

DISCLOSURES
The author received an honorarium for participation in the symposium upon which this article is based. He discloses no potential bias or conflict of intere s t relating to this art i c l e . D e p ression is widely acknowledged in the corporate world as a topic worthy of much attention. 1 , 2 Most of us are a w a re that improvements are desperately needed in the way we approach depressed employees, but few of us sincere l y believe we can do anything about it. I strongly believe that it is possible to change this situation, and I would like to share our experience at PPG Industries with you. It may not be an entire l y scientific approach, but it is a practical one and should be thought of as more of a cultural anthropology endeavor than a bioscientific e x p e r i m e n t .

■ ■ Embracing Change
The first challenge in improving depression management in the workplace is to try to change the relationships between all stakeholders-patients, providers, pharmacists, psychiatrists, medical practices, plan purchasers, and data managers. Many aspects of c a re are carved out these days, but we need to all get together and center on the patient' s well-being, focusing on an improved continuum of care. Historically, we all have looked at the situation f rom our own perspectives and, from an employer' s vantage point, cutting costs or services has seemed the most logical appro a c h . H o w e v e r, the quality of care often suffers when we approach the situation in this manner. And bad care and bad perf o rmance is really the most expensive and wasteful solution.
So the first step is to bring all the diff e rent players to the table to reconstitute the flow of information. We need to have the right i n f o rmation at the right time in the right place when it is needed. Sharing and reintegrating the data among all the players is pivotal if we are ultimately to effect change in this pro c e s s .

■ ■ Surveying the Landscape
The next step in improving the management of depression is to use the data we've shared to assess the landscape pert a i n i n g specifically to our depressed employees. In the case of PPG, we used published data available in the medical literature but also relied upon Medstat (a health information company) to assemble data specific to our employee pool. Medstat integrated all the relevant and available costs (e.g., medical, prescription dru g , absenteeism, and disability expenditures) and ranked the physical and mental conditions based on costs to the employer.
The impact of depression and other affective disorders on our expenses at PPG was predictably large, and we wanted to figure out the risk factors predisposing employees to these conditions and ascertain whether or not these were reversible. Historically, s t ress has been closely associated with depression, but we also wanted to look at the relative influence of other risk factors and try to identify causes that we could modify. 3 , 4 Our re s u l t s demonstrate that stress is, in fact, most closely associated with the p revalence and severity of depression but that few people were re a d y, willing, and able to change. In general, if employees demonstrate a willingness to change certain risk factors, then the D e p ression Management in the Wo r k p l a c e : A Case Study

A B S T R AC T
OBJECTIVE: To review a case of a depression management program at PPG Industries with the potential to improve outcomes on functioning and productivity in the workplace.
SUMMARY: The need for major improvements in depression management has been well established. While most in the corporate world are aware of these deficiencies, the perception is that this fragmented system is difficult to change. PPG Industries, a medium-sized, Fortune 500 company manufacturing paints, stains, and sealants, has launched a successful and comprehensive program in the workplace that has improved outcomes for depressed employees.
The PPG approach is practical, addressing each step in the process one at a time. This process began by establishing a close working relationship between all entities responsible for employee health, many of which are currently carved out. This enabled the company to have a comprehensive view of their depressed population and to examine their functional outcomes as well. In an effort to help identify depressed employees, the company educated their physicians and care managers and launched a confidential Web site to reduce mental health stigma and aid in employee screening and education. A multidisciplinary team was then assembled to facilitate the treatment of these patients. All of these steps were measurable, and the preliminary results of this process are summarized.
CONCLUSION: A disease management program that incorporates a multifaceted, collaborative approach to treat depression is feasible and may improve care and decrease cost to employers.
KEYWORDS: Depression, PPG, Disease management, Antidepressant utilization, Primary care workplace is mature for the deployment of a program. If there is a high prevalence but little readiness to change (as was demonstrated with stress and depression at PPG), then we need to focus on creating awareness and providing education.
We also looked at the mental health landscape through dru g utilization patterns. Working closely with Caremark, our p h a rmacy benefits management company, we analyzed antid e p ressant utilization (as well as utilization of other medication classes) and quickly realized that our depressed population was s u ffering from many other medical comorbidities as well. They w e re very heavy utilizers in general. We also analyzed the demographics of who was receiving prescriptions, as well as who was writing them (i.e., prescribers' medical specialty). We learn e d , for instance, that men and women were equally likely to re c e i v e an antidepressant prescription and that the likelihood of doing so i n c reased with age, peaking in the geriatric population ( Figure 1). Analysis of the provider data revealed that most antidepressants w e re being prescribed by providers in the fields of family medicine and internal medicine, so we were now aware that the p r i m a ry care setting was the front line ( Figure 2).
Another source of vital information came from a regional study known as the Southwestern Pennsylvania Depression Report (SPDR). The SPDR demonstrated that the highest hospitalization rate for depression was in the group aged 30 to 39 years, a demographic that is at the peak of their productive years. Using the classic depression quality indicators from the Health Plan Employer Data and Information Set, the SPDR revealed that only 10% to 38% of patients received follow-up for their antidepre s s a n t t reatment after discharge, and 13% of these patients were readmitted within 30 days of discharge. This study pro v i d e d s t rong evidence of the poor quality of care that depressed patients received. Follow-up was inadequate and coordinated care was l a c k i n g .
A comprehensive analysis of our particular landscape, theref o re, was fairly consistent with what has been re p o rted nationally. 5 , 6 The health care delivery system for mental health, including d e p ression, is very fragmented, with little collaboration between p r i m a ry care, specialty care, employee assistance, and pharm a c y. M a n y, if not all, components have been isolated or carved out, and t h e re are no identifiable incentives for integrating diagnosis, medication management, therapy, and follow-up. What we have ultimately created is a shattered mirror where it is very difficult to accurately recompose the original image.

■ ■ Changing the Workplace Environment
Although the stigma surrounding mental illness may have diminished somewhat in recent years, it still exists, and we need to deal with it eff e c t i v e l y. To reduce stigma in our workplace, we emphasized the benefits of early recognition through We b -b a s e d s c reening and the availability of functional rehabilitation. As studi e s have shown, however, it is not enough to simply screen for d e p ression-a program of coordinated care must also be in p l a c e . 7 , 8 Integration must occur, there f o re, between primary care and behavioral health specialists, and effective tools must be placed in the hands of providers to enhance treatment and patient education. Most of all, we strived to provide depressed patients with an experience of continuous pro g ressive care to p revent them from falling through all the traditional cracks in the s y s t e m .
At the point of service, we elected to train our nurses to fill the role of care managers. The goal was to enable them to serve as facilitators to help patients or employees navigate through the system, accessing support services such as the PPG Employee Assistance Program (EAP). We didn't want the nurses to become p h a rmacists, providing expert medication management. We didn't want them to become psychiatrists, providing formal diagnosis a n d c o m p rehensive treatment plans. And we didn't want them to be psychologists either, delivering behavioral therapy. We trained our nurses to be facilitators and educators, helping to elevate depre s s i o n to the level of other conditions such as high cholesterol and high This emphasis upon depression, wellness, and the role of nurses in our system was also conveyed effectively to our employees. We let them know about the accessible Web-based screening tools for depression and that standardized guidelines for treatment were readily available. We also gave them a good idea of what to expect when they went to a physician seeking help for their depre s s i o n .
The integration of health care specialists re q u i res that care f u l thought also be devoted to identifying where the expert support should come from. From my perspective, expert support can be p rovided by psychiatrists, psychologists, or pharmacists, depending on your specific institutional re s o u rces. I believe that p h a rmacists should seriously consider developing a role in managing depression that is analogous to certified diabetes e d u c a t o r s . 9 P h a rmacists could be formally trained to arrive at worksites or come to employers with the expressed purpose of educating them about depression and its appropriate treatment. I would encourage pharmacists to consider this role and the additional value it would bring to patients and to the pro f e s s i o n .

■ ■ The PPG Experience
One year ago, we embarked on this ambitious project, bringing all the stakeholders together for the first time. This first step was critical but we had to get all of them together, including the b e h a v i o r a l health carve-out, if we were to make it work (Figure 3).
The method we chose for training the trainers (i.e., the nurses) was to put together a series of seminars, which included a day-long conference as well as 9 hours of Web-site instru c t i o n . The Web casts, in part i c u l a r, re p resented a more cost-eff e c t i v e a p p roach to us, and we used the opportunity to emphasize such topics as appropriate use of the screening tool (PHQ9), mechanics of effective collaborative care eff o rts, and methods to impro v e access to employer assistance programs. None of this was re v o l ut i o n a ry -it was just a very practical approach to laying down the foundation for our interv e n t i o n .
S i m u l t a n e o u s l y, we also had to reach out to the primary care physicians to improve their ability to identify and successfully manage depressed patients. The logistics of reaching out to a larg e pool of physicians is often daunting, so we wanted to make our a p p roach more tangible by identifying exactly how many practices and providers we were talking about. We wanted to make everyone more comfortable by demonstrating that it was a finite number of physician practices that we were reaching out to: 62, in our case, would cover 50% of our employees. Our health plan (Highmark) then started working with the larger primary c a re practices first, sending out liaison personnel to train the physicians in the use of screening tools and circulating the AMA toolkit for depression. Highmark also created pre s c r i p t i o n pads to facilitate the necessary re f e rrals (e.g., to educational and support s e rv i c e s ) .
Another feature of our primary care intervention was to open up direct access for patients to behavioral health specialists. A 24-hour telephone support service was also created for patients with more urgent needs.
Once patients were initiated on medication, we wanted to make sure that they gave the antidepressant a full therapeutic trial and, if successful, that they continued to take it for a minimum of 6 months. The only good medicine is the medicine people actually take, and studies have shown that most patients stop their antid e p ressant before 6 months. 6 C o n s e q u e n t l y, we emphasized to p roviders and care managers that they needed to convince patients that they are active players in treating their own chronic conditions. In surg e ry, the surgeon knows what to do, and a general assembly is not usually re q u i red to decide which surg e ry is appropriate. In chronic care, successful treatment re q u i res

Measuring Outcomes
As mentioned pre v i o u s l y, we have established a variety of objectives for the various stages in the disease management p rocess. Aw a reness, early recognition, treatment, follow-up, and monitoring adherence have all been cited. For the sake of evaluating our relative success, we had to identify or create methods to measure our impact upon every step in this pro c e s s .  (Table 1). There was also a significant increase in the number of employees using the Web-based screening tool (from 5 to 42) and in the number contacting EAP for assistance (from 42 to 157). Several data points, there f o re, suggested to us that the a w a reness intervention was translating into increased access and i n c reased utilization of frontline serv i c e s . This increase in awareness and assessment also appears to be equating to an increase in the number of verified depression cases. During the past year, we saw a 30% increase in the number of cases opened by EAP (from 63 to 82). As other experts have mentioned, lost productivity re p resents the greatest cost to employers, and we are beginning to witness significant impro v ements in this re g a rd as well. 1 0 F rom a survey administered to the EAP cases, we learned that employees experienced a substantial Return to Function: Impact on Health and Productivity i n c rease in functionality at home and in the workplace (Table 2). S e l f -re p o rts of absenteeism suggest that lost work days have also diminished among our depressed employees. Although our findings in the productivity spectrum are favorable, we hope that m o re sophisticated tools can be developed in the future that will give us a more comprehensive look at the impact of depression on functionality in the workplace. We hope that a measure of physician p e rf o rmance can be integrated some day as well.

■ ■ Conclusion
I hope that my description of the PPG experience has convinced you from a practical perspective that it is possible to effect change in the way depression is managed by employers. The first and most crucial step is to get all the stakeholders together to share their information. Once this data is analyzed, we need to pursue a collaborative approach to improving detection, treatment, and follow-up. The employee or patient must be motivated to become an active member of this treatment team. If such collaboration can be accomplished, improved care and reduced costs are very much achievable objectives.

DISCLOSURES
Funding for this study was provided by the 2003 Innovation Award from the Institute for Occupational and Environmental Health, obtained by the author. The author received an honorarium for participation in the symposium upon which this article is based. He discloses no potential bias or conflict of interest relating to this article.