What factors determine success and failure in the development, adoption and use of technology in behavioral health care? What organizational and psychological dimensions must be addressed to align individual needs, group interests and organizational cultures? What is the role of the information technology professional in the development and adoption of technology in the traditionally low-tech, high-touch mental health disciplines? These questions will be explored in terms of a case study and a discussion of the IT professional's role as change agent.
Compass is an interactive database system that creates science-based, norm-referenced information regarding behavioral health status and treatment outcomes. Two companies own the technology and sell and license different versions of the technology in different markets. Compass creates information for the (pre- and post-) measurement and the ongoing monitoring of clients' response to treatment. Such information supports management of individual cases and caseloads as well as site, regional and national practice patterns for clinicians and health-care-provider companies, managed care companies and insurers. This data also informs employers and government agencies of the value of the health care services they are purchasing.
Context and Sources
The Compass technology emerged in a market that was based on the need to control health care costs. Gaining control of costs in the mental health arena was particularly problematic given the multiplicity of perspectives, the subjectivity and sensitivity of patients' needs and the apparently conflicting interests of the various parties. Simply put, patients wanted to feel better, clinicians wanted to cure disorders, employers wanted employees on the job and productive. Each group wanted what they wanted with the best financial consequences for themselves. Each group had its own ideas about how to bring about the consequences that they desired. Any behavioral health care company able to develop a means for aligning the interests of these groups, therefore, would have a powerful competitive advantage.
The research of Kenneth Howard at Northwestern University provided such a means. Howard, arguably the foremost expert on behavioral treatment outcomes, and his colleagues had accumulated a large body of research supporting two major findings: the Dose Response Theory and the Phase Theory. Dose Response Theory describes an observed logarithmic relationship between number of treatment sessions and client improvement -- similar to the relationship in numerous other areas showing increased application of resources yielding progressively fewer results. Phase Theory describes a causally related sequence of treatment targets, which accounted for the shape of the dose-response curve. Specifically, the three phases of treatment address (1) subjective well-being, (2) symptom relief and (3) improvement in functioning. This model allowed coordination of the conflicting concerns of patients, clinicians and employers. Compass thus provided a lingua franca based on common measures that allowed meaningful dialogue among the different stakeholders. If this conceptual system could be married to material and organizational systems to create a technology having the necessary functions, collaboration between these groups could replace conflict.
A spin-off company developed a second version of the technology. This version was designed for use by customer organizations with their own practice management systems. A number of changes were made to Compass in response to clinician provider groups' requests. The number of questions was reduced and a system developed for faxing the clients/clinicians' responses to the questionnaire and faxing back to the clinician a multidimensional report on the clients' status. Despite these accommodations of end-users' needs, Compass was vastly underutilized in several of these customer organizations. Extensive on-site interviewing of clinicians and managers revealed problems unique to each customer organization as well as problems that they had in common. These generic problems had to do with what some have termed the "hostile takeover" of mental health services by corporate America. At the sloganeering level, corporate managers viewed clinicians as profiting themselves by indulging their clients in "Woody Allen therapy." Clinicians viewed corporate managers as focused on the bottom line to the exclusion of patients' well-being. To say the least there was a clash of cultures. Successful commercialization of Compass thus required explicit and detailed understanding of these differences.
This connection allows the therapist to check client eligibility online. Case notes are stored in computer files, facilitating case manager and clinician discussion of treatment. New referrals and session authorizations can be processed quickly and efficiently. This facility eliminates some of clinicians' major complaints: numerous unanswered messages left on voice mail, repeated mailed requests for authorizations that get lost in the MCO offices or approved after the clinician has seen the client for several sessions. This latter situation can generate a need to obtain retroactive authorization -- another burdensome process. Plans are underway to include a hand-held POV device that docks at a workstation and immediately generates a treatment progress report, as well as entering information directly into the client's computerized file. The companies involved in designing and implementing this further adaptation of the technology expect to have 10,000 clinicians involved in the next two to three years.
What is a change agent? According to Everett Rogers, a change agent is someone who links together the change agency and the client system. In this case Compass serves to bridge or link the MCO, the clinician, the patient, the insurer and the employer. IT professionals and IT tools, such as Compass, serve a critical communication function that enables providers, MCOs, and other stakeholders to "read from the same page." The task of promoting innovation and collaboration may be framed by adaptation of Rogers' Sequence of Change Agent Roles as follows.
Step 1. Develop Need for Change
External pressures created much of the need for change in the mental health field. These pressures can be viewed as "sticks and carrots." "Sticks" are clinicians' loss of income and loss of autonomy in the shift to corporate health care. Some clinicians believe that to survive they must cooperate with managed care companies. Some clinicians also believe that to thrive they must adapt to the changes. The most successful adoption of new behavior, however, occurs when the individual wants to change. That means there must be "carrots" such as improved patient care, more referrals, lowered operating costs and opportunities to effectively communicate about clinical concerns.
Step 2. Establish Information Exchange Relationship
The IT professional must demonstrate sincere interest in the environment or social context of the provider. IT has to be perceived as an honest broker -- as trustworthy. Recent articles in a wide range of journals discuss the need for IT professionals to reassess their roles and responsibilities. IT change agents need to develop working relationships with users. Specifically they must
Comfort is also critical. How much does a therapist know about using a computer? How much training in basic computer skills do the providers need? Demonstrating efficiency and even increased quality of care is not enough. People have to be comfortable with a technology or they will not use it, will not use it effectively or will use it improperly.
Confidentiality, the third factor, is a basic ethical value imbued in clinicians. Attempting to reassure clinicians' worried about lack of confidentiality by talking about "128 bit government certified G-2 encryption technology" means nothing to non-technical people. These concerns have to be acknowledged and meaningfully addressed.
Two more factors relevant to the mental health provider population are age and associational patterns. Therapists in their 30s are more comfortable with ideas like common data sets and with using computers than are therapists in their 50s. Individual therapists are less likely to see the benefits of adopting new technology. Clinicians of any age form groups in order to get referrals from MCOs. When groups reach about 10 providers, they find that they need computer-based technology. Adoption of new technology usually happens out of necessity, not desire.
Step 3. Diagnose Problems
This step may require the "detective work" of unobtrusively observing the IT tools in use. To understand implementation and design problems, IT people may spend time in providers' offices or convening focus groups. Problems can include providers not being able to afford computers with expensive monthly costs. It was a revelation to executives from one of the companies selling Compass when they calculated how little money a Ph.D. psychologist makes, given average reimbursement rates and average caseloads. This directly affects the investment providers are likely to make in computer technology.
Other problems can be more subtle, falling within the traditional purview of HCI. For example, are views logical for the way clinicians think about organizing information in their knowledge domain? Once a technology is implemented, it is important to go back and ask the user, "Is there anything you can suggest to make this tool better? Can you suggest improvements?"
It is important to expect that some groups who are interested in adopting new tools will have problems and will need support in making the transition. Expecting groups to solve problems by themselves is counterproductive. Another example of operationalizing intent occurred when 10 groups of providers were piloting Compass. Two groups were very successful, two groups failed miserably and the rest fell in between. IT compared the most successful and the least successful groups to find out reasons for the differences. As follow-ups of crucial importance they provided training and support for the unsuccessful groups and published the best practices and made them available to all pilot participants.
Step 5. Stabilize Adoption and Prevent Discontinuance
IT people may need to iterate the previous steps in order to have the adoption become a permanent part of the culture instead of a grafted addition. It should be clear that what is being described is an ongoing process, not a one-time project. At this point in the process IT can also pursue new uses for the tools. One strategy might be providing continuing education online or setting up discussion groups around clinical or practice issues. Opportunities like these would be extensions of the original uses of the tools. Such extensions also help to imbed the new technology in the culture.
The last point involves communication. It is important to stay in touch with providers after implementation. Also, publish the results of implementation to the wider community despite the extra time involved in doing so.
Step 6. Achieve Ongoing Collaboration
One significant change of Rogers' model is the last step. Rogers proposed this to be "termination of change agent role," meaning that the clients become self-reliant and able to serve as their own change agents. In the projects described above, the sixth step was achieving a partnership and ongoing collaboration -- for further implementation and development of the technology.
Although some of the problems and solutions discussed in this article apply specifically to the field of behavioral health care, the lessons learned by these companies in their efforts to implement Compass seem relevant to information technology applications in other fields. Successful development and implementation requires of the IT professional frequent and ongoing communication with clients, serious investigation of the context (culture) of end users and investing significant time and other resources in non-technical issues.
Bulletin of the American Society for Information Science