Managed care systems rely on primary care providers as gatekeepers to make sensible decisions regarding the use of expensive health care resources. While this model has some intuitive appeal in terms of its potential for decreasing health care costs, it may not be applicable in VA medical centers, where patients are often medically complex and socioeconomically vulnerable. Thus, other strategies to integrate generalist and specialist care are required.
Our objective is to evaluate the effectiveness of an integrated model of primary care for veterans with alcohol dependence and/or depression in which mental health clinical nurse specialists (CNS) are placed within the General Medicine Clinic (GMC). Our primary outcomes are patients’ disease-specific mental health symptoms (Beck Depression Inventory: BDI) and satisfaction with care (RAND). Secondary outcomes include quality of care, health services utilization, and health care costs.
This is a randomized, controlled trial comparing patients in two GMC firms. After physicians in both firms were trained in the diagnosis and brief treatment of the two conditions, the two firms were randomized into one of two conditions. Patients in both firms are screened for depression. Patients who screen positive for depression and give written informed consent are enrolled. In the control firm, depression screening results are provided to the primary care physician. In the integrated primary care firm, results are shared with the primary care provider, along with having a mental health clinical nurse specialist in the GMC who is available to implement and support treatment decisions. Telephone interviews are conducted at three and 12 months after enrollment to collect outcome data by persons blinded to the study hypotheses. Data will also be collected using local VA databases. Data analysis will utilize generalized estimating equations to account for the repeated measures design, clustering of patients within physicians, and clustering of physicians within clinics.
We were unable to recruit adequate numbers of patients with alcohol problems. Thus, the alcohol component of the study was dropped. Enrollment was completed on August 1, 1999. Of the 271 patients with depression enrolled; we had three-month data for 247 (91%) patients. At three-month follow-up, change scores were not different for BDI (control = -0.06; intervention = -0.31, p=0.63). Change in the five RAND satisfaction sub-scales also showed no difference (p-values range from 0.14 to 0.66). Intervention patients averaged five more visits (p=.0059) (28% higher) and $451.50 higher overall costs (p=.0457) (25% higher), even though $79,000 for 1 FTEE in CNS costs were not included. Process analyses indicated that about one third of patients did not receive a depression diagnosis from the CNS, 28 percent were recommended for no intervention, and over 50 percent received no CNS follow-up after the enrollment visit.
Integrative care for depression fell well short of improving patient outcomes at three-months: being both ineffective and costly. Effective intervention may need to incorporate more direct, multi-leveled, and well-structured strategies to implement clinical guidelines for depression, provide systematic outcomes monitoring, and provide outcomes feedback to the clinician. At a minimum, these preliminary study findings suggest great caution should be used when deciding about implementing integrated psychiatric care for depression within VA primary care settings.
External Links for this Project
None at this time.