Abstract
A collaborative care model (CCM) has been implemented for management of depression. This paper studies the impact that the CCM had on cost measures for the period of six months after initial diagnosis of depression compared to patients receiving usual care (UC). There was a significant increase in the CPT costs for the six months following diagnosis in the CCM group ($451.35 vs. $323.50, P < 0.001). The average CPT cost rank and CPT cost differential were also significantly increased in the CCM group. The adjusted means of the CPT costs were (when controlling for prior utilization) $452.11 for the CCM group and $322.09 for UC (P < 0.001).
In the CCM group; there were 161 patients (73.5%) that achieved a clinical response for their depression compared to the UC group, which had a 15.1% (18/119) response rate (P < 0.001). There also was a significant difference between the groups in those who were symptoms free of their depression (PHQ-9 score < 5), with the CCM having 59.4% of the patients symptom-free compared to 10.9% of the UC group (P < 0.001).
In this group of patients, CCM is associated with markedly improved clinical outcomes for depression, however with a modest short-term increase in CPT costs.
Many of the changes in the practice of medicine affect primary care. Mental health care management is an important part of overall primary care, as depression has been ranked ahead of coronary artery disease as the leading cause of disability and premature death worldwide. 1 However, current practices for the diagnosis and management of depression treat only 46% to 57% of patients, and only 18% to 25% receive adequate therapy. 2 The STAR*D trial, with aggressive management of depressed patients, suggests a theoretical possibility of 33% to 67% cumulative remission rates (defined as those free of symptoms of depression) from the original cohort.3-5
The challenges with chronic disease management for mental health are consistent with other diseases such as diabetes in adults and asthma in children. A systematic screening approach, 6 appropriate diagnostic testing, adequate therapy, development of a registry of patients, and a collaborative care model (CCM; involving the patient, primary care provider, psychiatry, and care managers) are all important for the management of this chronic disease. Several studies have shown that in depression, the use of a CCM has improved the therapeutic response to treatment and lowered long-term health care costs, while short-term costs have increased.7-10
With the need to improve quality of care and the importance of maintaining cost-effectiveness for chronic disease management, how is the impact of changes in primary care measured? We have looked at the relationship of innovations in primary care to return visits11,12 and cost measures.13,14 A prior evaluation of the CCM of care for depression demonstrated a significant increase in outpatient utilization in the first month after implementation of the model 15 however, those numbers were small and did not evaluate total health care costs or control for prior utilization.
The hypothesis for this article is that the CCM of care will have an increase in cost measures for the period of 6 months after initial diagnosis of depression, when controlling for prior utilization of clinical visits and costs for the preceding 6 months. As a part of the value of health care, the secondary hypothesis is that the quality of depression care will also be improved during this time frame compared with usual care.
Methods
In 2007, the Institute for Clinical Systems Improvement (ICSI), along with several clinical sites across the state of Minnesota and 6 payer organizations, spearheaded the development of a CCM for depression treatment. This was intended to augment the relationship between the patient, the primary care provider (PCP), and the psychiatrist. The CCM required the development of a depression registry, consistent testing of depressed patients using the Patient Health Questionnaire–9 (PHQ-9) screening tool, and the weekly oversight of a psychiatrist. Once a patient was enrolled within the CCM, he or she was also screened for comorbid psychiatric issues such as chemical dependency and other mood disorders (bipolar). The monitoring of depression was dictated by clinical need with the patient’s initial contacts weekly (more often if needed) or at a minimum of monthly. A key component of the CCM is the relapse prevention coordination between the care manager and the patient once he or she is in clinical remission. The CCM developed by ICSI allows for monthly payment to the clinical sites for their care management services. Four registered nurses were hired to assume the new role of care managers at these sites. Interactions with the PCPs were either by the electronic medical record or by personal contact. Weekly, on-site psychiatric review was provided by the Department of Psychiatry from Mayo Clinic Rochester.
In 2008, the Depression Improvement Across Minnesota Offering a New Direction (DIAMOND) project, coordinated by ICSI, was implemented at 2 clinical sites at Mayo Family Clinics in Rochester, Minnesota. These clinics care for approximately 41 000 patients and have physicians and mid-level providers in family medicine, primary care internal medicine, and community pediatrics and adolescent medicine. The patients are a community-based population and are approximately 50% Mayo Clinic employees or their dependents.
Medical records of 338 patients were abstracted for 2 cohorts of patients. All patients were adult patients who had received a diagnosis of depression since March 2008, had a screening score of 10 or greater on the PHQ-9, 16 and had given permission to have their medical records reviewed. The CCM treatment group (n = 219, 64.8%) includes those who completed at least 6 months of treatment. The usual care (UC) group (n = 119, 35.2%) were patients from the clinical sites where CCM was implemented (thus having the same medical providers during the same time period), but they did not receive collaborative care management.
The first dependent variable in this study was the summation of Current Procedural Terminology (CPT) code costs for the 6 months after the index date. Since direct comparison of billing data could be biased, the standard national fees as listed by the Centers for Medicare & Medicaid Services (CMS) were used. CPT codes that were not listed in the CMS database were excluded. For example, the codes 99213 and 99214 are listed in Minnesota with fees of $43.73 and $67.61, respectively. Since the total costs measured had unequal variances between the groups studied, rank cost measures were also compared. The lowest cost ranked as a 1. The differential between the pre– and post–index date CPT costs was calculated and also studied with rank cost measures.
The independent variables include age, gender, prior history of depression (within 2 years), initial PHQ-9 score, the number of outpatient medical office visits in the 6 months prior to the index visit, the summation of CPT code costs for the 6 months prior to the index visit, and the type of care given (CCM vs UC). Medical and psychiatric comorbidity was not specifically reviewed in this study; however, prior utilization (clinic visits and CPT costs) was evaluated.
Cost comparisons were controlled by analysis of covariance for the patient’s prior history of utilization, as determined by the number of outpatient visits and the total costs in the 6 months prior to diagnosis.
Statistical analysis used χ2 testing for the categorical variables. Since the data were not in a normal distribution, Mann-Whitney testing was performed for numerical variables. Two trained medical secretaries were used as abstractors. All information was obtained from the patient’s electronic medical record. The study was approved by the institutional review board.
Results
There was no difference noted between the CCM and UC groups in gender, marital status, PHQ-9 level at initial diagnosis, average CPT costs, cost ranking, outpatient clinical visits for the preceding 6 months, or history of depression diagnosis in the 2 years prior to diagnosis. A statistically significant difference in age was noted between the CCM and the UC groups (40.6 vs 37.7 years, P = .03; Table 1).
Characteristics of Collaborative Care Manager (CCM) Patients Versus Usual Care (UC)
There was a significant increase in the CPT cost basis for the 6 months following diagnosis in the CCM group ($451.35 vs $323.50, P < .001). The average CPT cost rank for the CCM group was 185.67 and for the UC group was 135.40 (P < .001). The cost differential was also significantly increased in the CCM versus the UC groups, $246.10 versus $105.62 (P < .001), as well as the differential ranking of 185.26 and 139.07 (P < .001; Table 2).
Results of Collaborative Care Manager (CCM) Patients Versus Usual Care (UC) for Current Procedural Terminology (CPT) Costs 6 Months After Diagnosis
By using analysis of covariance, in controlling for age (since this was significantly different between the groups), the number of clinical outpatient visits, and the summation of CPT costs for the 6 months prior to index date, the adjusted means of CPT costs in the 6 months following diagnosis were $452.11 (confidence interval [CI], $406.96 to $497.26) for the CCM group and $322.09 (CI, $260.74 to $383.44) for UC (P < .001). Cost ranking was also increased when controlled for the above variables in CCM versus UC: 185.16 (CI, 173.40 to 196.92) versus 136.33 (CI, 120.35 to 152.31; P < .001). The CPT cost differential also demonstrated an elevated adjusted means for the CCM group ($242.41; CI, $197.27 to $287.56) versus UC ($112.40; CI, $51.05 to $173.75; P < .001; Figure 1).
Adjusted means for collaborative care manager (CCM) versus usual care (UC) when controlled for prior utilization (age, clinical outpatient visits, and Current Procedural Terminology costs).
The definition of response in this study is the decrease of the initial PHQ-9 score by 50% or more. Of the 219 members of the CCM group, there were 161 patients (73.5%) who achieved a response for their depression monitoring. This is in contrast to the UC group, which had only a 15.1% (18/119) response rate (P < .001; Figure 2).
Percentage of clinical response and symptom-free patients over 6 months, comparing collaborative care management (CCM) versus usual care (UC).
Since clinical remission is the absence of depressive symptoms for a period of 2 months, we were unable to completely document remission in this study because we evaluated only the 6-month follow-up PHQ-9 score. In comparing the data available, there is a marked significant difference between the groups in patients who were free from symptoms of their depression. In the CCM patients, 130 (59.4%) had a follow-up PHQ-9 score of less than 5. This compares with 13 (10.9%) in the UC group (P < .001; Figure 2). Compliance with obtaining 6-month PHQ-9 data was 200 (91.3%) versus 36 (30.3%) between the CCM and UC groups (P < .001).
Discussion
After the diagnosis of depression and enrollment into a CCM, one explanation of an increased utilization of health care resources is the “activation” of the patient. There is close monitoring of the patient by the care manager, encouraging follow-up visits as needed with the PCP and/or specialty visits. The difficulty with UC for depression is that patients get lost to follow-up and do not have an opportunity to improve clinically. Since the quality of depression care can be measured and followed with the PHQ-9, monitoring by the care manager has a significant impact in the proactive treatment of this illness.
With the significant improvement in response and symptom-free measurements at the 6-month time frame, CCM has a significant quality advantage over UC. Although the adjusted mean CPT cost summation was $130 more in the CCM, in our CMS cost structure, this is approximately 2 additional 99214 visits over a 6-month time frame.
Significant changes in the processes of health care continue to be an important component of health care reform. Many of the changes are effective at the primary care level. Having a tool to evaluate the quality and effectiveness of process changes, along with the costs as demonstrated in this article is imperative in determining appropriate and effective change. We are anticipating using the method of CPT cost analysis to evaluate other care processes, such as diabetes care.
This study may not be able to be generalized to other clinical sites. The patient population was not a significantly diverse population in ethnicity, medical or psychiatric comorbidity, or lack of health care insurance. The population was composed entirely of adults and cannot be generalized to a pediatric population. This study included practitioners in one community, and the results may not be able to be generalized to other sites or practitioners. Since the data were collected using billing information, charges from outside the institution were not captured. We are currently looking into factors that may predict which patients will have an increased utilization after diagnosis of depression and will compare this clinical practice site to other sites within our health system.
Conclusions
Utilization of a CCM in the treatment of primary care patients with the diagnosis of depression has a significant impact on the outcome of care as measured by clinical response and remission. There appears to be a modest increase in cost utilization as measured by actual costs, cost rankings, and cost differentials, even when controlling for prior utilization. This may be explained by “activation” of the patient in management of their depression, but further analysis of the reasons for the increased utilization needs to be conducted. CPT cost utilization appears to be a reasonable assessment tool for evaluation of practice model changes in primary care.
Footnotes
Acknowledgements
Kelly Amunrud and Julie Maxon abstracted the data analyzed in this study from electronic medical records.
The authors declared no potential conflicts of interests with respect to the authorship and/or publication of this article.
The authors received no financial support for the research and/or authorship of this article.