The Computer Based Patient Record: A Strategic Issue in Process Innovation

Transcription

1 Journal of Medical Systems, Vol. 22, No. 6, 1998 The Computer Based Patient Record: A Strategic Issue in Process Innovation Claude Sicotte, 1,2 Jean Louis Denis, 1 and Pascale Lehoux 1 Reengineering of the workplace through Information Technology is an important strategic issue for today's hospitals. The computer-based patient record (CPR) is one technology that has the potential to profoundly modify the work routines of the care unit. This study investigates a CPR project aimed at allowing physicians and nurses to work in a completely electronic environment. The focus of our analysis was the patient nursing care process. The rationale behind the introduction of this technology was based on its alleged capability to both enhance quality of care and control costs. This is done by better managing the flow of information within the organization and by introducing mechanisms such as the timeless and spaceless organization of the work place, de-localization, and automation of work processes. The present case study analyzed the implementation of a large CPR project ($45 million U.S.) conducted in four hospitals in joint venture with two computer firms. The computerized system had to be withdrawn because of boycotts from both the medical and nursing personnel. User-resistance was not the problem. Despite its failure, this project was a good opportunity to understand better the intricate complexity of introducing technology in professional work where the usefulness of information is short lived and where it is difficult to predetermine the relevancy of information. Profound misconceptions in achieving a tighter fit (synchronization) between care processes and information processes were the main problems. KEY WORDS: computer-based patient record; reengineering care process; implementation study. INTRODUCTION Utilization of information technology (IT) is considered by more and more hospitals as an interesting solution to keep pace with financial constraints imposed upon them. Many hospitals are looking at the computerized patient records (CPR) as a good means to both enhance the quality of care and control costs. 1 Department of Health Administration, University of Montreal, P.O. Box 6128, Station Downtown, Montreal, Quebec, H3C 3J7, Canada. 2 To whom correspondence should be addressed /98/ $15.00/ Plenum Publishing Corporation

2 432 Sicotte, Denis, and Lehoux In this perspective, IT is seen as an important strategic issue offering process innovation opportunities that will facilitate the fundamental redesign of work.(1) The rationale behind the introduction of a comprehensive CPR is based on a reconfiguration of the care process, especially at the care unit level. This innovative process is aimed at better managing the flow of information within the care unit and between it and satellite departments (e.g., pharmacy, laboratories) upon which the care unit is dependent to fulfill its task, the care of patients. The IT is aimed at transforming processes that will enhance quality and productivity through the introduction of several mechanisms such as the automation of work processes and other care process innovations aimed at transcending physical and temporal boundaries specific to hospital work. The CPR project under study is of special interest in that its purpose was to make the paperless hospital a reality.(2) The main idea behind the system's design, development, and implementation was to completely eliminate the traditional paper patient record and to have physicians and nurses work in a completely electronic environment. The CPR is equivalent to a care process innovation of very large magnitude not often seen in hospital. It emphasizes faster care, greater flexibility, integration, and care tailored to the patient. While the implementation of the IT is attractive, it represents a radical reorganization of professional work. Numerous studies have underlined the difficulty obtaining the alleged benefits of this reorganization of the workplace.(2-5) In this paper, we will try to better understand the inherent complexities of the project in order to draw practical insights on the conditions that allow the successful application of the CPR, permitting process innovation able to further productivity and quality of care. The main focus in this paper is in-patient nursing care processes. BACKGROUND: THE PROJECT STORY Initiated by a Montreal University Hospital in the mid 80's, the computerized system was intended to completely replace the paper patient record and simultaneously sustain the planning and organization of delivery of care into in-patient units. Under the hospital's initiative, a consortium of four hospitals and two private computer firms received R&D funds from the Department of Industry to design, develop, and implement a CPR in the four participating hospitals. It was a commercial joint venture whereby each participant invested resources and participated in potential future financial benefits. At the time, the entire project was intended to take place over a 3-year period ( ). However, the project encountered serious delays in the development and implementation stages and a second grant was obtained to keep it going. The second phase took place between 1991 and The total cost for both phases of the project was over 45 million U.S. dollars. During the second phase, CPR implementation was attempted in several care units in each of the four participating hospitals. Both the nursing and medical personnel were then asked to use the new system. Finally after several months of various attempts, the computerized system had to be withdrawn because of boycotts from both the medical and nursing per-

3 Computer Based Patient Record 433 sonnel. Despite the implementation's failure, this unsuccessful project allows us to draw practical lessons to better adapt the IT to professional work. METHOD Research Design The research design is multiple-case, which leads to a replication logic.(6) Each case is treated as an independent experiment that either confirms or annuls emerging conceptual insights. The research included all four hospitals taken as individual cases that attempted to implement the CPR. Data Collection Data were collected through individual interviews, focus group interviews, observations, and secondary documented sources. The primary source came from semi-structured interviews. At each hospital site, two types of respondents were interviewed: project team members (n = 10) who were responsible for the CPR development and implementation phases and nurse users (n = 24). Respondents were asked open-ended questions that enabled them to relate their stories of how CPR development and implementation were evolving in their hospital. Probing questions were asked so that further detail could be established. The 34 interviews were taped and typically lasted min. Data Analysis The research team analyzed the data by building individual case studies and then comparing each case. Using the interviews and secondary sources, the researchers wrote a case study for each site within an iterative process. A consensus was built up for each case through research team meetings involving two senior researchers and the two research assistants who conducted the interviews and wrote the cases. As a further check, a senior researcher read through the original interviews and formed an independent assessment of each case. Having no predetermined hypotheses, a cross-case analysis was used to develop conceptual insights. Several breaks were taken during the analysis process to ensure that an objective point of view was maintained. As the analysis evolved, the level of abstraction was elevated to develop these insights.(7) This iterative process led to the insights presented in this text. THE CPR's FUNDAMENTAL ROLE AND CHARACTERISTICS A hospital's information architecture is typically based on a series of information systems (IS) that are dedicated to meet the needs of specific departments (e.g.,

4 434 Sicotte, Denis, and Lehoux pharmacy, laboratories).(8) The Information System dedicated to the care unit is the patient medical record. It is the central database which the clinical team looks at to find the relevant information to ensure care planning, coordination, and delivery.(2,3,5) The patient record despite its traditional paper format remains a true information system that is well integrated in the work routines of nurses and physicians. The computerization of this paper IS can offer interesting benefits but necessitates an important reorganization of traditional work.(9) It is widely recognized that the diffusion and successful application of this kind of IT depends on significant changes in structure and administrative practice as well as a tight alignment with the computerization project.(10-13) The goal of the CPR project was to substitute an IS capable of managing the entire inpatient care episode in a completely electronic environment. The main features of the CPR, as conceived at the very beginning of the project, were: the complete elimination of the paper record to be replaced by an electronic document. the complete replacement by electronic communication links of clerical mechanisms ensuring communication within the care unit and with the specialized peripheral departments (e.g., pharmacy, laboratories, radiology). Each time the medical prescriptions or the nursing activities are modified, the nurse's kardex that contains all nursing interventions which must be done, is automatically updated and a new schedule is produced for the nursing team. the automatic transformation of raw data found in various organizational tools such as daily care schedules for patients, and individual case-load schedules for nurses. These continuously updated applications allow nurses to speed-up the care process and to continuously follow the care process in real time for each patient. To sum up, the project team initially emphasized the CPR s ability to decrease clerical workload and allow better organization and coordination of professional work, thereby liberating more time for patient care and ensuring better quality of care. RESULTS We first present the results by describing the project's evolution. Then, we analyze the several mechanisms needed to trigger the CPR to demonstrate the expected process innovations. A CPR is a complex information system to develop. It must assume several roles to meet the information needs of several parties (e.g., nurses, physicians, pharmacy). The development and implementation of the CPR was seen by the project team as an integrated system which needs to be carried out sequentially in several different applications corresponding to the different sections that one finds in a traditional paper patient record (e.g., laboratory results, nurse's observations). The project team decided to begin with what was seen as the first step in the care process: the collection of information about the patient's health state. A nurs-

5 Computer Based Patient Record 435 ing patient history was thus the first application developed and implemented for the nurses. This choice was intended to better formalize and generalize an activity that the nursing management judged underused. The idea of first developing the nursing history was not welcomed by the floor nurses. They were more interested in other applications that would help them lessen their workload at the clerical level. They found, for instance, the CPR kardex more interesting. To keep a paper kardex up to date according to the latest medical and nursing orders is an intensive activity requiring hours of work. It is often dedicated to hand retranscription of already existing information. A computerized kardex was thus perceived by the floor nurses to deliver significant benefits such as the ability to create a list of each nurse's patients including the list of nursing actions scheduled automatically and in chronological order (daily planning scheduling). The prioritization of the CPR history showed a shift in the project's design. While the project team initially emphasized the CPR's ability to decrease clerical workload, the first application was about systematic and more thoroughly completed patient data collection. In the following text, we analyze in depth the interaction of the CPR project with work processes. To do this, we based our analysis on eight process innovation mechnanisms associated with IT identified by Davenport.(1) (cf. Figure 1) Automational Mechanism: Eliminating Human Labor From a Process The most commonly recognized benefit of IT is its ability to eliminate human labor and to produce a more structured process. In service environments, where processes are frequently defined by document flows, automational opportunities increasingly rely on removing paper from the process by substituting "work flow" software. In the perspective of eliminating all use of paper, the decision had been taken to develop a CPR that requires direct data entry and retrieval by nurses and physicians. The anticipated benefits were (a) elimination of any later retranscription of data that introduces time lag in the care process or possible errors linked to retranscription; and, (b) as data entry is done, the data can be electronically transformed and put into readily useable forms for various personnel, that is, the automated impact. The technical solution to facilitate the work of professionals not trained to use a keyboard was touch screen technology integrated in a visual interface. It 'logically' led to the subsequent decision of developing structured questionnaires to support data entry; a format compatible with the limited choice of a menu-driven input system. The nursing history became a closed format with a predetermined and preordered number of questions and with a predetermined choice of answers. The questionnaire was exhaustive covering all types of patients. The CPR was built so that the history needed to be completed in a continuous session and at an early stage in the patient's arrival. The idea of extensively documenting the patient's initial condition inevitably meant an increase in nursing time needed for each patient. This problem was exacerbated by the technical choice of the touch screen. The quantity of information

6 436 Sicotte, Denis, and Lehoux PROCESS INNOVATION MECHANISMS 1- Process automation (Direct data entry by nurses and electronic transformation and communication) 2- Analytical improvement INTENDED EFFECTS Decrease of clerical work load (Data retranscription) Elimination of time lag Decrease in costs linked to clerical work load Improvement of analytic abilities OBSERVED EFFECTS Increase in nursing clerical tasks Higher formalization of data collection Less flexibility in work organization High formalization and standardization of nursing cognitive process Nursing cognitive process Nurse de-skilling Information overload Less flexibility in work organization 3- Process sequence Acceleration of completion of the collection of nursing data Elimination of existing process parallelism (care delivery and patient assessment) Less flexibility in work organization 4- Tracking capability 5- De-localization (Elimination of geographical boundaries) 6- Integrative capability 7- Information capability 8- Intellectual capability Tracking in real time of information process status Consultation of patient files at a distance Horizontal case supervision and coordination accross various functions and departments Staff planning and allocation To build a knowledge data base (e.g. Protocols) Automation of people control Inoperative for nursing work Less flexibility in work organization because of localization constraints (Bed side terminals) Inoperative Increased control of staff Less flexibility in work organization Increase in nursing clinical task inoperative Fig. 1. A summary of the intended and observed effects associated with each of the process innovation mechanisms. contained on each screen was limited because of the screen surface that needs to be kept to ensure data entry with fingers. The large number of screens resulted in more navigational operations that take time and cognitive attention that resulted in increased user desorientation and greater difficulty in grasping the patient status during data retrieval.

7 Computer Based Patient Record 437 To sum up, few automational effects were possible. The only one was an automatic transfer of an ensemble of patient health status information into the kardex useful to contextualize nursing actions (e.g., patient's allergies). In fact, the project team was more interested in automating the process of patient history than the automatic data transfer supporting team communication. The problem seen by nursing management was that the collection of data was not complete enough. The CPR was thus used in order to formalize the nursing process but it also led to less flexibility in individual tasks and team work organization. Each nurse had to reserve a continuous long period in order to do this clerical task. In most of the care units, the organization of nursing work had to be reorganized to absorb the surplus of clerical work. We present in Fig. 1 the summary of the results for the automational mechanism. Analytical Mechanism: Improving Analysis of Information and Decision Making In a process that involves the analysis of information and decision making, IT can lead to an array of sophisticated analytical resources that permit more data to be incorporated in and analyzed during the decision-making process. The CPR nursing history could have been a helpful tool that diminishes the burden of clerical tasks and improves the analytical abilities of the users. However, the realized application had negative impacts. Because of the highly structured fashion within which the CPR was built, clinical judgment was completely determined by the system. It was not the nurse who guided the choice and order of questions, it was the system. The CPR became a tool structuring data entry rather than a flexible tool enabling nurses to improve their clinical investigation capabilities. Furthermore, the format and rigidity of the questionnaire had an impact on the meaning of clinical information. Usually the nurses' data collection process is aimed at identifying information useful to guide the nurses' actions during the care process. Conversely, the CPR patient history was designed so that pre-formated questions always needed an answer. Often the information collected appeared "normal" or "without problems." A new layer of clinical information (normal) was thus introduced at the clerical level within the CPR. Usually the nurses filter the information so only the essential information (pathological) is kept. Information overload was a direct consequence of this formalization that modified the meaning of clinical information entered in the patient record. Furthermore, the CPR, in uniformly modeling the manner in which the patient history was captured in the patient record, was an attempt at automating the cognitive process that nurses use to collect data and to systematize expert thinking. The solution offered by the CPR can be compared to a form of work computerization leading to the de-skilling of the user. The consequence of this formalization of the cognitive process was the introduction of rigidity in work routines and less flexibility in individual and group selforganization of time and care activities.

8 438 Sicotte, Denis, and Lehoux Sequential Mechanism: Changing Process Sequence, or Enabling Parallelism IT can enable changes in the sequence of processes or transform a process from sequential to parallel in order to achieve process cycle-time reductions. It is frequently possible to design in parallel components that once had to be designed sequentially. In this respect, the CPR nursing history led to less flexibility in nursing work. The nursing management's demand was to complete the CPR history at the very beginning of the patient stay. The traditional routine was different. The nursing history was usually built in two steps: at the patient arrival to the unit, collection of information important for immediate care was collected (e.g., allergies); and later, a cumulative process of data collection took place during each patient-nurse contact which gradually increased knowledge about each patient. Thus a parallel process already existed. While carrying on nursing care activities, a concurrent patient documentation process was taking place. Thus, the CPR's new constraint to block a long straight continuous period of time for nursing history meant less flexibility in the care process. Tracking Mechanism: Closely Monitoring Process Status IT allows one to effectively execute some process that requires a high degree of information monitoring and tracking. In hospitals, knowing when information will be produced can be crucial to decision making. Often, when a physician is taking care of a patient and is missing information about laboratory or radiology examination results he will ask a nurse to check when the results will be available in order to better plan the next intervention. In this matter, it is the CPR's applications which allow real time communication with peripheral departments that can be useful. The potential of a CPR patient history was thus not helpful because it gathers useful information at the beginning of the patient's stay but its use is shortlived. The important information is to know the patient's on-going state of health. In fact, the CPR tracking potential was used more often by nursing management as a means to track individuals rather than to gather information. We will more fully analyze this aspect with the seventh mechanism: informational. Geographical Mechanism: Coordinating Processes Across Distances A key benefit of IT has been the ability to overcome distances. The possibility to consult the patient file at a distance and to update the care program was among the alleged benefits of the CPR. In fact, these benefits were more suitable for the physicians than the nurses. A physician is personally in charge of his patients but does not remain in one location within the hospital. Thus, the ability to consult patient records from a distance can be a real benefit for a physician. Conversely, nursing is carried out by a team covering a location for a 24-hr period. While on

9 Computer Based Patient Record 439 duty, each nurse stays on the unit. The ability to access the CPR from a distance is thus far less interesting for nurses than it is for the physicians. Another aspect linked to the geographical dimension was the decision to introduce bedside terminals. Nursing management wanted to synchronize care and data entry by eliminating the wide-spread nursing habit of using a piece of paper as a memorization tool. This tool enabled the nurse (1) to concentrate on care delivery without going continously to the patient record for guidance, (2) to note meaningful information about the patient status, and (3) to note what action has been done and when. At a suitable time, the pertinent information (2 & 3) is transcribed into the patient record. This habit is well adapted to the nature of the nursing task: a repetitive task (e.g., to give medication), fragmented (with several patients) and often interrupted (to take care of another patient in urgent need, help a colleague or see a physician). The necessity to do data entry at the patient's bedside limited the flexibility of nursing work. It resulted in loss of time (repetition of logging procedures into the CPR), loss of control of the nurse's own work, higher task fragmentation, and localization constraints. The bedside terminals became a handicap to personal mobility in comparison with the paper environment. It is interesting to underline that several nurses claim that the time they take to transcribe their notes into the patient paper record improves quality of care. They can look back to ensure that everything has been done correctly. The incessant logging into the CPR during patient contacts may tamper with the careful reassessment of the care process. Integrative Mechanism: Coordination Between Tasks and Processes More and more, organizations are finding it difficult to radically improve process performance for highly segmented tasks split across many jobs and are moving to a case management approach. In this type of process, an individual or a team completes or supervises all aspects of a service delivery process. In these applications, information on various aspects of the process stored in data bases spread throughout the organization is consolidated in desktop workstations. The concept of case management is already popular within hospital nursing. It was not explicitly expressed by the project team. However, a complete CPR conceived as a unique data base constantly kept up to date, should allow for a case management approach. The CPR nursing history, however, is a remote intervention with regard to the on-going care process and was not the right kind of application to obtain the benefits of IT Informational Mechanism: Capturing Process Information for Purposes of Understanding Information Technology can be used not just to eliminate human labor from a process but also to complement it.(14) IT can be used within a process to capture information about process performance which can then be analyzed by humans.

10 440 Sicotte, Denis, and Lehoux In this matter, one potential use of a CPR is its ability to monitor the entire workload of a care unit. This information can be used to better manage the available personnel so that the workload can be allocated uniformly during a shift. The CPR can be used to eliminate workload peaks in which case the staff available would be insufficient. In this matter, the potential of a CPR history is limited. However, because nursing management was preoccupied with the objective of increasing the volume of nursing histories, the IT was used to control the nurses' activities, namely, to know the identity of nurses doing a history, the completion level of each history, and the time of task completion. The CPR's automatic registration of user identity and login time were used by nursing management as a device for distant controlling of the floor nurses. The use of this IT capability is important in terms of work flexibility. Upon completion, several nursing interventions and their time of administration must be documented in the patient record (e.g., medication). In the paper environment, the nurses have some flexibility. They usually note the time of administration on a piece of paper and then later, at a suitable time, transcribe this information into the patient record. The internal clock of the CPR eliminated this kind of flexibility because it automatically registered the time the nurse inputted that the care had been given. A tight synchronization between care delivery and data entry was thus enforced by the technology. This practice introduced more rigidity in workflow and increased clerical time because for each patient the nurse had to log into the system. Usually, the clerical task of transcribing information into the record is delayed to a quieter time during the shift when the nurse can update all her files at the same time. Intellectual Mechanism: Capturing and Distributing Intellectual Assets Knowledge-intensive activities are often not treated as processes. Nevertheless, a number of organizations are beginning to try to capture and distribute knowledge more broadly and consistently. The idea of building a database of care protocols for nurses and physicians was presented as an innovation of the CPR. The idea was to have specialized personnel construct state-of-the-art protocols that would be available on the system for consultation. The goal was to make expert knowledge available across a hospital and even a network of several hospitals. The prioritization of the CPR patient history was not the kind of application that would allow the attainment of these benefits. DISCUSSION The CPR did not offer the advantages that were touted at the beginning of the project. A project which was initially sold as facilitating nursing work, improving coordination between nursing and medical activities, improving quality of care, and di-

11 Computer Based Patient Record 441 minishing costs, resulted in information overload and standardization, clerical task load increase, work organization rigidity, and expert autonomy negation. Two fixed ideas produced the inability of the project team to profit from the beneficial potential of process innovation associated with the IT These ideas which drove the strategic decisions leading to the CPR design and implementation errors were the paperless hospital and the professionalization of nurses. The idea of the paperless hospital took on so much importance that it led to a unilateral view of the CPR as an electronic document aimed at completely eliminating the paper record. Crucial decisons such as direct data entry by professionals, the touch screen and bedside terminals resulted from this vision. As a consequence, the main drawback was a partial vision of the IT. Whereas it was the communication capability that was more suitable to operationalize the majority of process innovation mechanisms to enhance flexibility in nursing work leading to better quality of care and higher productivity, this communication device was overlooked and completely put aside. Practical insight #1: The Dual Information Role of a CPR: Documentation and Communication The ability to take advantage of process innovation capabilities offered by computerized clinical information is linked to the necessity of distinguishing the presence of different types of information. A CPR contains contextual information that allows nurses to grasp a patient's health status so that they can adapt their interventions and keep their care activities up to date. Furthermore, a CPR needs to ensure the coordination of care. The communication role of the CPR is then crucial at the group level, connecting nursing, physicians, and external peripheral departments. Having two roles means different realities in terms of the meaning, the richness and the manner with which the information needs to be transferred into software language. The information useful for coordination can be very limited. Key words can be easily used, especially with specialized and well-trained workgroups like nurses who quickly understand what needs to be done with a limited quantity of highly coded language; coordination information is more easily codified within a computerized system than contextual information. The information needed to ensure work coordination is based on well-known organizational and professional routines easily codified within a computerized CPR. Furthermore, this is the communication information which is essential to sustain the care processes. A CPR designed to improve the process of delivering care should therefore focus its efforts on this type of information. The other main factor that explains the project's negative outcome was the nature of the reorganization of work that was required with the deployment of IT The CPR project decisions were made in order to implement a new model of nursing. This model was inspired from a vision promoted by the nursing elite: the professionalization of nurses. This new rationale which led to the reorganization of nursing through the CPR meant an important break from the initial concept of the CPR. The prioritization of the CPR history was a crucial moment for the project's success. It is, according to us, the main error which led to the complete withdrawal

12 442 Sicotte, Denis, and Lehoux of the CPR in the four hospitals and the project's failure. This decision had a series of repercussions that the project team had not anticipated. It resulted in several effects that weakened mechanisms that were potentially able to improve care processes by giving work more flexibility and effectiveness. These mechanisms were the ones by which the hospital's initial objectives, namely cost reduction and quality of care enhancement, should have been met. The utilitarian role of the CPR, promoted in the early phase of the project focused on the decrease in clerical tasks and the facilitation of work. By promoting the professional role of nursing, the project discarded the constraints of daily work. The economic profitability initially considered was replaced by an idealized vision of nursing promoting a model of work organized according to other values. The underlying logic explaining the choice of developing the patient history conformed to the medical model, the exemplary model of professionalism. The collection of data is necessary to establish the list of patient problems (nursing diagnosis) that precede the elaboration of the care plan (medical prescription). While this model is slowly adopted by physicians during their long medical training, nursing management attempted to use a more authoritarian strategy in which the professional model is imposed by IT in a working context. The attempt to model a field of knowledge within an information system (IS) rather than stay close to the organization of work made the project far more difficult in terms of design and development. It resulted in a distinction between the characteristics of the task and the IS. The concept of the IS was built with the vision of the nursing elite. The design of the CPR did not benefit from the possible input of the users, the floor nurses; development was done only with nursing management. The CPR's design was achieved on the basis of logical ideas emerging from the professional model and was not considered with respect to the constraints of daily nursing. The idealized vision had to prevail despite the reality of the constraints of daily nursing practice in the care units. These constraints were rapidly discarded when brought as inputs to influence the CPR design. It is interesting to underline that the justification of imposing the nursing history was not based on an organizational legitimacy specific to each hospital. Rather than adhering to the hospital authorities, the project team imposed the nursing history on the floor nurses and used external professional nursing corporation norms to legitimize their decisions. Practical insight #2: Rationale Behind Work Reengineering The professional project that was the underlying model for redesigning nursing was too far removed from the daily constraints of nursing practice. The project team attempted to define the nature of information from an idealized point of view rather than work closely with the delivery process. In this manner, the CPR information architecture was inspired from the perspective of how nursing is taught and promoted in academic institutions and professional corporations rather than from the work site where nursing is truly practiced. A more comprehensive and integrated approach is needed to better understand the potential and limits of the IT, the constraints of nursing work and how closely related these two aspects must be. It

13 Computer Based Patient Record 443 is an important lesson to be remembered: the way the CPR was designed was highly normative as it tried to impose a new reality, producing uniformity and predictability in the thought and behavior pattern of nurses. It was one of the main reasons for the nurses' great reluctance to use the system. Another lesson drawn from the project is the necessity of a balanced involvement of the various groups of users. A more balanced vision of daily nursing practice and its professional transformation is essential to develop a workable CPR. This result reaffirms the principle of coherence that must exist between an IT and the work process within which it needs to be integrated. ACKNOWLEDGEMENT The research reported in this text was supported by a research grant ( ) from the Social Sciences Humanities Research Council of Canada. REFERENCES 1. Davenport, T.H. Process Innovation. Reengineering Work Through Information Technology, Harvard Business School Press, Boston, IOM, The Computer-Based Patient Record, Institute of Medicine, National Academy Press, Washington, Ball, M.J., and Collen, M.F. Aspects of the Computer-Eased Patient Record, Springer-Verlag, New York, Kaplan, B., Development and acceptance of medical information systems: An historical overview. J. Health Hum. Res. Admin. Summer:9-29, Drazen, E.L., Metzger, J.B., Ritter, J.L., and Schneider, M.K. Patient Care Information Systems, Springer-Verlag, New York, Yin, R.K., Case Study Research (revised edition), Sage Publications, Beverly Hills, Einsenhardt, K.M., Building theories from case-study research. Acad. Manag. Rev. 14: , Bourke, M.D., Strategy and Architecture of Health Care Information Systems, Springer-Verlag, New York, Preuss, G.A., Labor, skills, and information in service delivery: An examination of hospital care. Acad. Manag. Proc. 97: , Barley, S.R. Technology as an occasion for structuring: Evidence from observations of CT scanners and the social order of radiology departments. Admin. Sci. Quart. 31:78-108, DeSanctis, G., and Scott Poole, M., Capturing the complexity in advanced technology use: Adaptative Structuration Theory. Organ. Sci. 5(2): , Ettlie, J.D., and Reza. M., Organizational integration and process innovation. Acad. Manag. J. 35(4): , Orlikowski, W.J., The duality of technology: rethinking the concept of technology in organizations. Organ. Sci. 3(3): , Zuboff, S., In the Age of the Smart Machine, Basic Books, New York, 1988.