This book describes a clinical information system designed for a comprehensive cancer center. The Oncology Clinical Information System (OCIS) was developed at The Johns Hopkins Hospital and is of general interest for several reasons. It has been used in the management of a high volume of critically ill patients for over 10 years; during the past five years it has operated seven days a week, 24 hours a day (an indication of how closely it has been integrated into patient management activities); and, finally, it supports a variety of integrated decision making tools that comprise one of the most extensive medical information systems in general use today. Although OCIS was implemented in a cancer center, it is important to note that the principles used in its development and the functions that it supports are applicable to most medical environments. It is believed, therefore, that a description of this system and a discussion of its implementation history will be helpful to both the developers and users of future clinical information systems.
Work on the OCIS began in 1975. At that time there were relatively few clinical information systems in operation, and the cost of interactive computing was high. Studies of these early systems suggested that their half-lives were approximately equal to the publication cycle; by the time a journal article appeared about a system, there was a fifty-fifty chance that it was no longer in use. In the mid 1970s the Technicon Medical Information System, which since has been accepted as the prototypical hospital information system, was still under evaluation. When Henley and Wiederhold surveyed automated ambulatory medical record systems in 1975, they could identify only 16 worthy of evaluation. Even in this highly selective study, by the time that they reported their findings, one of these systems was no longer in operation. Thus, when we began work on the OCIS, the experience base in medical informatics was narrow.
In the decade that followed, the cost of computers fell and their capabilities improved. The personal computer revolution of the early 1980s played a major role in increasing the public’s knowledge of computer technology. This computer literacy removed users’ fear and uncertainty, and it became easier to introduce new automated functions. A user community had been trained to accept automated tools based upon their value and contribution to the task at hand.
Since the time that work on OCIS began, we have gained considerable experience with computer technology; the modes of health care reimbursement have changed; the volume of data which must be assessed to make sound medical decisions has grown; and computer applications have become ubiquitous in medicine. While familiarity with personal computers has facilitated a greater acceptance of automation, it has not altered the role of the computer in the process of care delivery. Most of the commercially available systems of the early and mid 1980s continue to respond to the perceived needs of the 1970s. They support administrative functions, seek to reduce labor costs, and perform clinically oriented activities only fortuitously. In general, clinical systems today are structured in an environment that was not designed to support medical decision making. Conversely, the OCIS is designed to facilitate medical decision making and the necessary administrative functions are built around this structure.
We believe that information systems, augmented with knowledge processing applications, offer a solution to the present information management burden faced by most health care facilities. This burden distracts from the delivery of quality medical care, encourages reactive (as opposed to anticipatory) responses to medical problems, and results in the imperfect collection and utilization of medical knowledge. Furthermore, it is certain that the next generation of comprehensive systems cannot be produced unless the lessons of the past are built upon. With this in mind, we have prepared this book. The successes of OCIS are presented so that they can be adopted by other systems. The failures of OCIS also are described so that others can avoid repeating our mistakes.
This book is organized in three parts: (I) introductory and overview material, (II) functional descriptions of the OCIS components, and (III) a summary evaluation of OCIS, directions for the future, and a description of experience in porting the system to other cancer treatment facilities. A brief description of these parts and their associated chapters follows.
The purpose of Part I is to provide some background together with the environmental and structural information necessary to understand OCIS. The chapters offer the general audience an insight into the need for decision support systems in medicine and, in particular, for the treatment of cancer. They also provide an understanding of how OCIS was developed and what it looks like today. There are three chapters prepared by the editors.
Chapter 1 describes the background, environment, functions, and structural operation of OCIS. It also contains a brief summary of the chapters which follow.
Chapter 2 presents the philosophical foundation for decision support systems as applied to the clinical management of oncology patients. This chapter includes both the rationale for OCIS and an example of how the system is used in the routine management of patients.
Chapter 3 reviews the development history of OCIS. Because this is largely a software development activity, the chapter includes a review of software engineering and the software tools used with OCIS as well as a narrative of the system’s development.
Part II of the book contains a detailed description of OCIS. It presents the system from the perspective of its use in the Oncology Center; there are very few comparisons with other systems and no discussion of implementation details1. Each chapter contains a general introduction followed by a presentation of the OCIS tools in the context of their use. In most cases, the chapters have been written by the principal users of the applications. The material is organized as follows.
Chapter 4 (Clinical Data Management) presents the tools used to meet the clinical information needs described in Chapter 2. The primary author is the Manager of the OCIS and previous head of the clinical data coordinators, the group responsible for helping the physicians learn and use the system. Chapter 5 (Protocol-Directed Care) presents the tools used to provide advice on patient management based upon the rules formalized in treatment and research protocols. This chapter was prepared by one of the initial developers. Chapter 6 (Pharmacy System) describes a satellite oncology pharmacy run by the Department of Pharmacy. Although developed separately, the oncology pharmacy system is integrated with the other functions of OCIS. The authors are the past Director of the Oncology Pharmacy, a pharmacist, and the developer responsible for the implementation of this subsystem.
Chapter 7 (Hemapheresis System) describes the specialized tools required to manage the high volume of transfusions, product collection, and product-patient matching necessary in an oncology setting. The author is the Director of the Hemapheresis component of the Oncology Center.
Chapter 8 (General Administrative Functions) provides a description of the tools for patient scheduling, the management of a Tumor Registry, an OPD patient routing system, as well as a variety of other administrative functions. This chapter was prepared by the persons responsible for the various non-clinical activities.
Part III of the book contains two chapters.
Chapter 9 includes an evaluation summary and an overview of the future computing plans for the Oncology Center. The author is the Director of the Oncology Information Systems.
Chapter 10 recounts an experience in transporting OCIS functions to another cancer center. This chapter was prepared by an OCIS developer who installed the system in a second cancer center.
‘A review of the OCIS data structures is in B.I. Blum, R.E. Lenhard, Jr., and E.E. McColligan, An Integrated Data Model for Patient Care, IEEE Transaction on Biomedical Engineering, BME-32:277-288, 1985.
In summary, we note that OCIS represents a viable approach to meeting modern medical computing needs. It is one of the most extensive medical decision support systems in use; it has a rich developmental history; it supports a unique and useful array of decision-making tools; and it provides administrative functions that are a necessary adjunct to patient care in today’s medical environment. The success and long-term viability of OCIS are the result of both its initial orientation to patient management and its ability to adapt to a changing user demand.
We believe that our experience with OCIS will be useful for a variety of reasons. It illustrates how patient data can be managed to support medical decision making; it offers examples of information management tools that were developed inexpensively; it illustrates issues in both development and evaluation; and it demonstrates that the goal of a comprehensive clinical information system is realistic. Unfortunately, the development of OCIS (or any other large information system) takes time and is expensive. Consequently, a developer or user can have but limited hands-on experience. We hope, therefore, that the following chapters will assist the reader in expanding his conceptual view of clinical information systems. For the goal of this book is not to describe one system, rather it is to assist the medical community in better understanding how this type of system can aid in the primary mission of delivering and improving health care.
Naturally the development of a system as ambitious as the OCIS requires contributions from many people. In the material that follows, we must acknowledge that it was the interest, patience, and cooperation of the entire Oncology Center—faculty, clinical staff, administration, clerical personnel, the Information Systems’ staff, and the patients—that helped us mold this system into its present form. We would like to thank Mike Fox, Dena Fulton, Darleen Rose, and Debbie Hutson for their tremendous help in preparing the text and many of the figures for this book. Special thanks is also extended to Gloria Stuart, who not only is one of the authors of this book, but provided help in proofing and coordinating the graphics for a majority of the book’s chapters.
Most importantly, without the unwavering support of the Center’s Director, Albert H. Owens, Jr., M.D., the OCIS would not have been possible. Dr. Owens not only had the initiative to create the Johns Hopkins Oncology Center, but had the foresight to integrate a clinical information system into its initial design.
The implementation of OCIS was supported primarily out of patient care revenues. We did receive a gift from the Educational Foundation of America for which we are grateful. Development of some of the research-oriented OCIS functions was supported through a Cancer Center Support Grant (# 5P30CA06973) from the National Cancer Institute. In closing this preface, we the editors also would like to thank our wives, Karen, Peggy, and Harriet, for their tolerance and understanding that compulsive husbands sometimes demand.