In 1993, William A. Anthony, PhD, director of the Center for Psychiatric Rehabilitation at Boston University, declared that the 1990s should be called the “Decade of Recovery.”1 While many definitions of recovery were offered since then,2 the Center's work has been guided by a parsimonious definition of a person's recovery: “the development of new meaning and purpose in one's life as one grows beyond the catastrophe of a mental illness.”1
No matter what definition of recovery one prefers, the vision of recovery has entered the mental health field as a genuine hope for the millions of Americans living with mental illnesses. Fueled by the consumer self-help movement, more effective and less harmful medications, longitudinal research showing evidence for recovery outcomes, and new service delivery models, the recovery vision has manifested itself in people returning to work, living independently, and attending school and training programs.
Simultaneously, people are asking, “If recovery is real, can it be measured in real-world outcomes such as earnings, degrees, and reduction of benefits, as well as more intangible variables such as self-esteem, empowerment, and hope?” Furthermore, even if we agree on recovery outcome measures, how can we measure what factors are determining these outcomes? Are some programs better at helping people reach these outcomes and, if so, what intervention processes are most effective?
The Value of Measurement
In our quest for evidence-based mental health interventions, it is not enough to measure program structures and outcomes. We must know how the intervention's clinical process helps to bring about recovery outcomes. Monitoring and improving a program's outcomes are most relevant if we can link the clinical process to the outcomes achieved.
To deploy true recovery-oriented mental health services, providers need to examine the effects of their services and dynamically adjust those services to optimize recovery. Unfortunately, most programs have limited resources and are unable to perform these types of evaluation. Many mental healthcare programs, therefore, have not established effective, reliable measurements of their performance.
A Technologic Solution
Boston University, via the Center, has created ROMIS, a Recovery-Oriented Management Information System designed to assess intake, process, and outcome measurements in mental health services. Conceived during the 1990s, ROMIS was the result of two years of discussions among senior staff to describe and define various service processes (e.g., rehabilitation, case management, treatment) as accurately and completely as possible.
Karen Danley, a brilliant and innovative force in vocational rehabilitation who worked at the Center for 17 years until her death in 1997, said of those discussions, “We knew that if we could get down on paper a common, logical, and objective description of the service processes, we could create a management information system that would hold the field's feet to the fire in regards to facilitating recovery outcomes.”
Danley's words were prophetic, as the field in the past ten years has moved toward evidence-based practice and has begun to recognize the need to be able to describe and measure the processes producing recovery outcomes.3
Funded by consecutive grants from the Tower Foundation and the Fidelity Foundation, Center staff first developed, field-tested, and refined a paper-and-pencil version of ROMIS designed to measure recovery processes and outcomes. Next, we constructed a Microsoft Access version and repeated the field tests. The rapid shift from an office-based workforce to one that is much more mobile necessitated ROMIS's migration to a Web-based design (figure).
ROMIS is accessible by any Web browser and is being tested against Internet Explorer 6.0 for Windows, Firefox 1.5, and Safari 1.3. Our goal has been to make ROMIS quickly accessible, easily usable, and able to precisely track required intake, process, and outcome information. With this goal in mind, ROMIS is intuitively designed as a series of screens and drop-down menus that contain multiple variables regarding intake and process activity components. ROMIS can supplement information systems that already track demographic, administrative, and financial data, so that clinical processes and recovery outcomes can be incorporated within existing management systems. ROMIS will have a database dictionary that will tell how the data are stored so that anyone who can access the database should be able to retrieve data from ROMIS to use with any other application. The official launch of the new Web-based ROMIS (expected this fall) will be a great leap forward at the Center in terms of both flexibility and efficiency in our data collection.
ROMIS includes three modules.
Intake Module. The Intake Module includes a wide range of intake requirements, including role status, health status, residential status, financial status, clinical status, and demographics. Some of these items were chosen because they may change as a result of participation in recovery-oriented programs; these same items appear in the Outcomes Module.
Process Module. This is ROMIS's heart and soul. It is predicated on the assumption that the process between client and practitioner, regardless of the particular service process in which it is activated, consists of diagnosis, planning, and intervention activities that can be tracked. The essence of ROMIS is that by collecting data on the components of the interaction between the client and practitioner, measured by the adherence to a uniformly defined diagnostic, planning, and intervention process, information can be gained about the relationship between that process and recovery outcomes.