Dr. Kenneth Gersing uses Dukes electronic behavioral healthcare management system to integrate scheduling, eligibility and registration management, clinical care at all levels, billing and claims management, regulatory management, and quality improvement and outcomes management. The system includes the most comprehensive set of behavioral health assessments and outcome measurements, 130 codified clinical knowledge tables, a decision-support rules engine to help guide clinical practices, and a clinical outcomes data warehouse (global database) which anonymizes and brings in data from all of the vendors customers for retrospective decision support and benchmarking.
Duke University Behavioral Health
The decision-support rules engine can prompt and alert clinicians to perform specific activities based on the preferred workflowas a task at the start of their documentation process, as a prompt during their charting, or as a reminder prior to signing their note. In simple terms, the decision-support engine can set rules.
The global database includes over 150,000 patients and 1.2 million visits; the vendor projects this will grow to 500,000 patients and 6 million visits in two years.
The combination of the behavioral health functionality, codified knowledge tables, decision-support rules engine to ensure collection of critical data and the global database provides all the tools to collect and evaluate data to improve patient care.
Gersing used codified treatment and outcomes data from the electronic medical-record system for 522 patients with a primary diagnosis of major depressive disorder single episode to determine treatment patterns used in their care, the results/outcomes of their care, and the factors that influenced their outcomes. All of the patients chosen to follow were first episode major depression whose treatment began as monotherapy on one of eight front-line antidepressant medications. Three treatment strategies: monotherapy, augmentation and switching, were noted and compared with respect to their therapeutic efficacy.
Results: 42% of the patients experienced improvement with monotherapy within three months; 75% improved within the entire period examined. The median number of days from the initiation of treatment to first improvement or change in therapy was 56 days. Over the course of their therapy, approximately one-half stayed on monotherapy, two-fifths augmented and nearly 10% switched medicines. Those with more severe diagnoses were more likely to respond, while those with comorbid relationship problems or anxiety were less likely. Those who used higher doses of their medication had lower response rates.
The reduced probabilities of response for those who augment or switch from their initial therapy highlights the importance of examining factors that may account for differential rates of response. The significance of race highlights the need for sensitivity to factors which may affect the course of therapy. Other factors, such as comorbid diagnoses and adverse events, which may be related to the likelihood of successful therapy, also merit attention.
Gersing will use this information for additional decision-support practice guidelines within the EMR. He also employed a biostatistician to develop advanced statistical models to control independent variables to determine comparative effectiveness of antidepressant medications.
To date, the American Recovery and Reinvestment Act of 2009, or stimulus law, funds for behavioral healthcare EMRs appears minimal. Medicare inpatient EMR incentives totally exclude behavioral healthcare providers. Although the details are not yet fully defined, the stimulus law outpatient incentives favor community rural health and medical providers seemingly to the exclusion of behavioral health caregivers. However, a number of grants and contracts through various federal health and informational technology agencies appear to be in the works, including numerous comparative-effectiveness opportunities. Gersing plans to pursue these funding paths to develop technologies to bring the global database information back to the clinicians desktop in real-time. Once this tool is developed and the global database reaches a critical data mass (Gersing believes this is 500,000 patients), the clinician will be able to see statistically valid probability of outcome data by selected intervention based on the patients characteristics and treatment history.
A. Deo Garlock
Director, Department of Psychiatry and Behavorial Sciences
Duke University Behavioral Health
Send us a letter
Have an opinion about this story? Click here to submit a Letter to the Editor, and we may publish it in print.