MEDICAL MANAGEMENT OF DEPRESSION

Disseminating Guidelines to Improve Diagnosis and Treatment

Depression is the number one mental health problem in the United States. Up to one in eight individuals may require treatment for depression in their lifetimes. [1] Despite the high prevalence of the disease, depression is under-diagnosed and undertreated by primary care providers, who are most likely to initially see these patients. Practitioners properly diagnose only one-third to one-half of those with a major depressive disorder. [2] The highest rates of depression are in the working-age group, age 25 to 44. [3] This results in huge indirect costs due to absenteeism and lowered productivity. Breakdowns of the costs of depression show that these costs to employers amount to $3,000 a year per depressed employee. [4]

Patients with cases of moderate to severe depression are generally good candidates for treatment with antidepressant medications. When pharmacological therapy is chosen, AHCPR guidelines define three phases of treatment: acute, continuation, and maintenance. [5] The acute phase lasts for 12 weeks to provide the opportunity to achieve a full remission, and allows the clinician to monitor drug response. Remission may be followed by relapse unless a four- to nine-month continuation phase of treatment is instituted. For certain patients, a maintenance phase of treatment must be adopted to prevent future recurrences of disease. [6]

In response to these challenges, this plan sought to increase the recognition of depression as a major illness as early as 1996. The plan found that 6.9 percent of its members received mental health services in 1995, a number below that of its major competitors (8 percent to 11 percent). At the same time, readmission rates for major affective disorders stood at 11.5 percent, far above a benchmark of 3.8 percent achieved by a competing plan. Further evidence of the importance of this disease was found in pharmacy data: the top two drugs prescribed by the plan were both antidepressants.

The plan designed a comprehensive program of practice guideline dissemination and educational activities to improve the quality of care for its members.

The plan initially used HEDIS 3.0 draft specifications to define its targeted population and performance measures. The plan used claims/encounter and pharmacy data. It targeted patients with a mental health claim showing either a primary or secondary diagnosis of mental health disorder (ICD-9-CM codes between 290.xx and 319.xx), and treated with an antidepressant medication.

It decided on three performance measures:

• The percentage of the targeted population, age 13 and older, who had at least three follow-up contacts with a primary care or mental health provider during the 12-week acute treatment phase. This was a measure of optimal practitioner contacts for medication management. As part of the management of the acute phase, it is an important variable in ensuring patient adherence to treatment
• The percentage of patients in the targeted population, age 18 and older, who remained on an antidepressant drug for a 12-week period. This measure looked at effective pharmacologic treatment for the entire acute phase of therapy
• The percentage of patients in the targeted population, age 18 and older, who remained on an antidepressant drug for at least six months. This was a measure of effective continuation phase treatment, based on guidelines calling for treatment of four to nine months after the complete resolution of symptoms

Baseline measurement, based on data from May 1996 to April 1997, showed that room for improvement existed. Thirty-seven percent met the criteria for the first measure, of optimal provider contacts. Forty-nine percent demonstrated effective acute phase treatment according to the criteria of the second measure. Only 26.2 percent had effective continuation phase treatment, as defined by the third measure.

A search for local benchmarks was fruitless; the plan decided on a performance goal of a 10 percent improvement in each of these measures.

The QI department analyzed the data to discover possible barriers to the appropriate management of depression. Barriers were divided into patient- and physician-based reasons.

Possible patient reasons for early discontinuation of treatment included:

• Side effects of antidepressant medications
• Lack of adequate information regarding treatment expectations
• Initial patient response to medication

Possible physician reasons included:

• Unfamiliarity with treatment guidelines
• Depression that is recognized but not formally diagnosed or coded
• Time constraints
• Reluctance to increase the numbers of visits under risk-sharing arrangements

The plan convened a focus group of primary care providers and psychiatrists to identify additional barriers. These included administrative barriers such as:

• Problems in accessing the plan's mental health network, making timely referrals by PCPs difficult
• A mental health provider directory that was often not up to date
• Closed panels of behavioral health providers

MULTI-PERSPECTIVE ANALYSIS Identified barriers often depend on who is doing the identifying. In this case, the health plan easily identified reasons attributable to patients and providers, while providers focused on barriers put in place by the health plan. (A survey of patients would undoubtedly point out the shortcomings of both the plan and its providers.) Although it is not clear how the structural problems attributed to the plan's system of behavioral health related to the poor performance of PCPs in managing depression, addressing practitioners areas of concern can go a long way towards achieving buy-in.

The plan decided to focus on an educational program emphasizing the many aspects of depression in primary care, offering solutions to increase patient compliance with treatment.

The plan embarked on an internal guideline development effort. It convened a multidisciplinary advisory panel to make recommendations on the use of national guidelines. The draft guideline was submitted to PCPs and employer representatives for comments before it was finalized and distributed to the network.

The plan held a series of educational seminars designed to educate PCPs on the diagnosis and treatment of depression. It made attendance at these seminars a component of the PCPs' quality bonus.

UNEXPECTED BENEFITS Although pediatric patients were not a focus of this activity or its ultimate measurements, a large number of pediatricians attended the educational seminars nonetheless. They expressed a desire to receive more information on how to treat their younger patients. In response, the plan scheduled additional seminars, and cooperated with the local chapter of the pediatric association to address childhood depression as a topic at their annual meeting.

The provider newsletter ran a detailed article on how the plan compensated providers for treating patients with depression. The article encouraged providers to see their patients more often, and noted how they could get reimbursed by billing with accurate codes.

The plan shared data from the focus group with its mental health vendor. Together, they explored possible solutions relating to the production of an up to date provider directory that accurately reflected the status of open panels.

The first remeasurement, based on data from May 1997 to April 1998, showed mixed results. The first measure, optimal provider contacts, was essentially unchanged at 36.8 percent. However, the other two measures showed statistically significant improvement (2-tailed Fisher Exact Test; p<.00l). The second measure, of effective acute phase treatment, increased to 60 percent. The third measure, of effective continuation phase treatment, increased to 42.2 percent.

CHANGING HEDIS SPECIFICATIONS Several changes were made to the final HEDIS specifications used in the remeasurement period, compared to the draft specifications used during the baseline measurement. These included changing the age of the targeted members in the first measure from 13 and older to 18 and older; requiring one follow-up contact to be with the prescribing provider; and changing the gaps allowable in continuous enrollment. The plan explicitly addressed these changes in its QIA, and noted that they did not substantively affect the results.

The plan noted that the improvement in the appropriateness of pharmacological treatment was encouraging, but that much room for improvement remained in addressing the need for appropriate provider contacts.

Additional explanations for PCPs' performance were identified by the plan. It found that PCPs often use new pharmacological agents differently - in terms of duration and dosage - compared with practices in academic medical centers. Also many practitioners performed follow-up over the phone, rather than at the office.

The mental health directory has been updated on a quarterly basis, and was made available on a "fax on demand" system. New providers and open panel status were reflected in the new directories.

Mailings of guidelines to practitioners have continued, as have educational seminars on depression. In these offerings, the plan focused on the importance of practitioner contact for patient compliance.

Despite these efforts, 1999 measures did not show improvement. Optimal practitioner contacts decreased slightly to 34 percent; effective acute phase treatment stayed at 60 percent, and effective continuation phase treatment dropped slightly to 41 percent.

The plan notes the continued importance of making sure that PCPs know treatment guidelines, and that they will be paid for follow-up mental health visits. They have decided to convene a number of practitioner focus groups in an effort to better understand barriers and educational needs.

[1] - Department of Health and Human Services, Centers for Disease Control and Prevention, Depression in Primary Care: Volume I Detection and Diagnosis Practice Guideline Number 5, (Rockville, MD: Department of Health and Services, April 1993), AHCPR Publication no. 93-0550.

[2] - Ibid.

[3] - Williams, RA and Strasser PB, "Depression in the Workplace-Impact on Employees," Journal of the American Association of Occupational Health Nurses, 47 (November 1999): 526-537.

[4] - Ibid.

[5] - National Committee for Quality Assurance, HEDIS 2000 vol 1 (Washington, D.C.: National Committee for Quality Assurance, 2000), 45.

[6] - Ibid.