Editors’ Note: The intent of this new column is to discuss topics at the intersection of technology and psychiatry – “Techiatry.” We’ve enlisted two leaders in this field to write for the column. Steven R. Daviss, MD, DFAPA (@HITshrink), is the chief medical informatics officer at M3 Information and chairs the American Psychiatric Association’s Committee on Mental Health Information Technology. James (Jay) H. Shore, MD, MPH, chairs the APA Committee on Telepsychiatry, is director of telemedicine at the Helen & Arthur E. Johnson Depression Center, and an associate professor of psychiatry at the University of Colorado at Denver, Aurora. Email them at cpnews@frontlinemedcom.com.
Medicine is late to the game when it comes to technology, specifically information technology. And psychiatry, even more so. Jay will talk in future columns about early use of telepsychiatry in the 1960s and since. But here in 2016, a surprisingly low percentage of us are using it to deliver care, despite the fact that half of the counties in the United States lack psychiatrists – and telemedicine has been shown to improve access to care.
Nonetheless, telemedicine and other uses of technology across all specialties is growing quickly, as usability, mobile technology, economics, and policy-making all converge. The integration of mental health care (including addiction treatment) with primary care is one of the driving forces in expanding access to the expertise that physicians trained in psychiatry possess. The collaborative care model of integrated care has the most evidence, making regular access to psychiatric consultants a weekly event.
Dr. Steven Roy Daviss
This exchange of information and knowledge between primary care and psychiatry is being formally incentivized by the Centers for Medicare & Medicaid Services (CMS) with proposed new codes to pay for this exchange, while the American Psychiatric Association has received a large grant from CMS to train 10% of its members in this care model.
Information technology is fundamental to this care model, because the efficient exchange of clinical information is important to optimize the capabilities and comprehensiveness of the clinical decision support provided by the psychiatrist to the primary care team.
As the team members learn what questions are asked and how the consultant arrives at her recommendations, they will become better at making these decisions on their own. They will learn how psychiatrists think and make decisions, weighing other medical, personal, social, family, and logistical aspects to guide the decision making process with the patient.
While this model of care is certainly helpful in expanding access to psychiatric expertise, there are other ways to achieve this access to expert knowledge. One of them is through electronic clinical decision support (CDS) tools. These are tools that apply clinical rules, algorithms, and other knowledge discovery processes to the information within the electronic health record (EHR) about a patient, with the goal of assessing and filling gaps in available patient information so that a set of possible recommendations can be delivered to the clinician.
Knowledge-based CDS tools apply clinical knowledge that comes from practice guidelines, textbooks, and the medical literature to what is known about the patient. The simplest CDS tool might be a rule that says, “IF patient is on lithium for bipolar disorder AND patient has current mood symptoms AND has not had a recent lithium level, THEN check a lithium level.” Applying and coding this rule into an EHR is fairly straightforward. A much more complex CDS tool could help the clinician think through the question, “What should I do next for this 32yo woman with hypertension and moderate depression who is symptomatic?”
Non–knowledge-based CDS tools use machine learning techniques, like neural networks, genetic algorithms, and natural language processing, to “learn” new clinical rules by going through a training process that inputs a large amount of clinical data and uses experts to “train” the system. Such a system was recently developed by IBM Watson and Memorial Sloan-Kettering Cancer Center to aid in developing recommendations for treating oncology patients.
The APA recently formed the CDS Product Workgroup (which I chair) to explore the feasibility of developing an electronic clinical decision support (CDS) tool that leverages the information and knowledge within the APA’s series of Practice Guidelines, DSM library, and other reference materials. This group will consider the necessary important clinical information sources – such as EHRs, personal health records, health information exchanges, claims and utilization data, patient generated data, mobile health apps, and clinical registries – from which to analyze patient-specific data and produce a set of ranked, evidence-based, annotated clinical suggestions.
The goal is to develop a CDS tool that is designed in a manner that ultimately benefits patients being treated by primary care practitioners, emergency practitioners, psychiatrists, and other medical specialists who treat patients with mental health and substance use disorders. Tools like this are being developed now in many specialties. Given the vast amount of psychiatric expertise within the APA, as well as the trove of content that exists within the publishing arm of the APA, the opportunity to make this more broadly available to medical practitioners is one that demands consideration.
Such an undertaking would require substantial time and commitment of resources, thus the task of the workgroup is to understand the pros and cons of developing this tool, and to explore its feasibility, including various business models to ensure that this CDS tool becomes a maintainable and sustainable product.
Bringing our expertise to the primary care settings, where most of our patients are treated, should greatly benefit the care of our patients, whether this is through collaborative care, clinical decision support, or telepsychiatry. In the same way that many people with diabetes do not require an endocrinologist, many with mental health conditions do not require a psychiatrist. Yet, primary care practitioners would certainly benefit from more help from us.
I will update readers of this column on our workgroup’s progress at the end of the year.