Having the best evidence at the point of care for all care providers is important because readily available evidence improves clinicians' performance. Studies indicate that the quality of nursing care depends on the information available to the nurses at the point of care.1-4 To facilitate evidence-based practice, hospitals have implemented various health information systems (HISs) such as electronic medical records, computerized provider order entry (CPOE), decision support systems, and bar-code medication administration (BCMA) systems, despite the high costs of implementation and maintenance after the design and build.5-7
Many HISs have been developed to ensure quality care and to assist in the continuity of patient care, but often developers design and develop the systems without considering users' workflow, tasks, and preferences.8-10 When the HIS is implemented in healthcare institutions without considering usability, it decreases users' satisfaction, acceptance, job performance, efficiency, and quality of work.10,11 Eventually, these unwanted outcomes foster new errors, workarounds, and unintended consequences.12-14 Campbell and colleagues15 reported in a study of CPOE systems that unintended adverse consequences such as additional work for clinicians and unfavorable workflow issues were generated by the failure of the system design and implementation to address complicated clinical workflow. Koppel and colleagues14 identified the creation of workarounds to BCMA because it did not include nurses' tasks and preferences in the design and build. Possible consequences of workarounds were wrong medication administration, wrong dosing and timing, and wrong formulations that threatened patient safety. Thus, a local adaptation process is necessary to reduce the unwanted outcomes after the implementation of the HIS. Despite the importance of the implementation and the adaptation processes of HIS, few studies have been conducted to develop a framework.
Quality nursing care is closely associated with high-quality patient information that nurses share when they are planning future care.16 Good nursing documentation encourages nurses to evaluate their patients' progress on an ongoing basis.17 A study by Griffiths and Hutchings18 implied that poor documentation might lead to bad nursing outcomes. It is of paramount importance to have a nursing documentation system that can store quality patient information and ensure patient safety. To have such a system, it is essential that nurses' workflow, tasks, and preferences be incorporated into the system during implementation as well as design and planning stages.
To demonstrate the benefits of applying a formally defined framework for system implementation and local adaptation approach, we have provided a case example of the step-by-step application of the workflow-oriented framework (WOF) to a nursing documentation system implementation project in Spaulding Rehabilitation Hospital (SRH).
BACKGROUND
Theoretical Underpinnings
TECHNOLOGY ACCEPTANCE MODEL
The Technology Acceptance Model (TAM) is an inductive model19 adapted by Davis from Ajzen and Fishbein's20 Theory of Reasoned Action Model to specifically study the use of computer software. This theory states that there is a correlation between actual behavior and intention, because a behavior is determined by a person's intention to perform the behavior.
The TAM model explains that perceived usefulness and perceived ease of use determine an individual's behavioral intention to use a system. To increase the end-users' performance and acceptance, it is important to design a system that is not only useful but also easy to use.
HEALTH INFORMATION SYSTEMS AND LOCAL ADAPTATION
Many theories and research on HIS implementations emphasize end-users' early involvement in designing and developing the system to improve job efficiency, performance, and quality of the function that the system is designed to provide.9,21-23 However, clinicians are not often involved in designing and developing an HIS because of their workload and the vendors' conventional system development processes that reduce user involvement. As a result, vendors develop HIS software without the clinicians' input.
To accommodate end-users' preferences, the vendors usually provide some degree of flexibility to meet the needs of a local hospital setting by allowing users to incorporate specific needs into the system. These specific needs are identified through a workflow analysis.8
A WORKFLOW-ORIENTED FRAMEWORK
A WOF was developed based on a framework for effective implementation and local adaptation of decision support systems developed in a previous study.8 The WOF has five distinctive phases: workflow analysis, system configuration, adaptation, system implementation, and adjustment. The adaptation and adjustment phases are repeated as often as needed until users are satisfied with a system, as measured before and after an actual system implementation (Figure 1).
Workflow analysis is the process of identifying data and information that end-users collect and the tasks end-users perform in the real world. Workflow-oriented framework includes requirements of input and output data format, clinical tasks, categories of information to be displayed, information sharing, and others.24,25 This process identifies the goals of the end-users and their methods of communicating with the system. Identified tasks and information should be broken down into detailed parts to reduce the cognitive load of the end-users when they interact with the system.10 The system configuration process is defined as hardware and system software configuration for implementation. During the adaptation phase, data, information, and tasks collected in the workflow analysis are tested and validated by a representative group of end-users. The continuous adjustment phase is followed by system implementation. During this phase, end-users identify and request modification of content in the system functions to fit their needs.
The workflow analysis conducted to develop this framework was guided by the object-oriented analysis (OOA) method. Object-oriented analysis is one of the most recent techniques for analyzing problems and specifying requirements for an information system.26,27 An important goal of requirements modeling is an understanding of the system and its behavioral requirements.28 These requirement specifications can be represented by the Unified Modeling Language such as use-case and activity diagrams.29 The workflow analysis is critical in system development because the direction provided by this analysis determines whether the system will be appropriate and specific to user requirements. In addition, misunderstanding user requirements is one of the major causes of project failure, budget overruns, and user dissatisfaction.30,31
METHODOLOGY
Applying the Workflow Oriented Framework
SETTING
Spaulding Rehabilitation Hospital is an academic acute rehabilitation hospital with 190 certified beds and a 160-average daily census (2011 statistics). Spaulding Rehabilitation Hospital offers six different rehabilitation programs: five for adult patients and one for pediatric patients. Adult programs encompass various types of injuries including cardiac, musculoskeletal, stroke, brain injury, and spinal cord injuries.
Nurses at SRH have had experience using various computerized systems in daily practice: the hospital information system, patient classification system, scheduling system, medication-dispensing system, and others. The implementation of MediServe Software Company's nursing documentation system (MediServe, Chandler, AZ) started in March 2008 and went live in November 2008 on all adult units.
APPLYING THE WOF
(1) Step I: Workflow analysis is defined as identifying data, information that end-users collect, and tasks end-users perform in daily practice.
To analyze the nursing documentation process at SRH, a nursing documentation committee was formed. It consisted of clinical nurse managers, charge nurses from each shift, and supervisors who met twice a month from March until May and once a week between June and August.
During the meetings, informatics-trained nurses facilitated discussions adopting a think-aloud technique and recording a log of the committees' comments, suggestions, and likes or dislikes of the new electronic documentation system under review. An example of an item reviewed was the need for nurses in the spinal cord injury unit to document "respiratory care," with the committee's suggestion to create a document section under the general nursing care category. Because this documentation section was not needed in the workflow of other specialty units (this respiratory care related to ventilator use), the respiratory care section was created but placed under a special category.
In May, after six committee meetings, the functional specifications were identified. A use-case diagram was created for the system behavior, and an activity diagram was created for the action flow of the system. In addition, a data representation format was determined during this phase. For example, the committee selected a flowsheet format of documentation rather than a form-based charting template for nursing documentation (Figure 2) because the flowchart format (a) facilitated documentation, (b) facilitated viewing of historical data on a screen, and (c) meshed well with the nursing workflow.
(2) Step II: System configuration is defined as identifying and preparing required hardware and system software for implementation.
The nursing documentation system was configured based on the functional specifications and the use cases identified during the committee meetings. Based on the requirement specification, client-server architecture was considered appropriate to support various configuration needs of the new system. Client-server architecture denotes a network architecture in which the client is a workstation upon which end-users run applications, and the server is a powerful computer that processes data and information requested by a client. A prototype of the new documentation system was installed in Window XP platforms and Microsoft SQL server in a testing environment (Microsoft, Redmond, WA).
(3) Step III: The adaptation process is defined as testing and validating data, information, and tasks that were identified during workflow analysis.
Clinically possible scenarios were written by a researcher and validated by the clinical nurse managers, charge nurses, and supervisors. A patient who has multiple complicated wounds was a typical example of a developed scenario. This scenario tested whether the nursing documentation system contained enough content to document multiple wounds. The prototype system was adjusted iteratively by clinical scenarios.
The 6-month duration of the adaptation process was divided into two periods. The first 3 months (June to August) were spent building a useful and facile nursing documentation system with committee members. Training all nursing staff at SRH using scenarios and test patients to familiarize nursing staffs with the system was done in September and October.
Continuous modifications and adjustments of the system functionality were made based on comments, suggestions, and actual practices provided by the committee members.
(4) Step IV: System implementation is defined as an application that is being implemented and used in practice.
The system went live on November 11, 2008, and interfaced with an existing HIS at SRH. Admission information, including demographics of patients, was fed automatically by the existing electronic Admission Discharge Transfer system.
(5) Step V: The adjustment process is defined as the process whereby end-users modify in detail the contents in the system functions to fit their needs.
After the system went live, nursing staff who had actively used the documentation system on a daily basis encountered various problems that were expected. Some of the problems were simple, such as requiring changes of words and expressions in the labels of various parts of the system. The failure to populate a data item in multiple relevant places in the system was an example of a more complicated and crucial problem.
The informatics-trained nurses collected user feedback and categorized it into technical problems or functionality problems with the collected feedback and problems prioritized by significance and solution feasibility. Examples of technical problems were (1) inability to print a nursing shift note, (2) length of time the system took to generate one patient shift report (60 seconds), and (3) system slowness occurring when a nurse tried to navigate from documentation to documentation. The functionality problems were (1) inability to log in, (2) character limitation that prevented full documentation of patient information for a nursing shift note, and (3) the need for additional flowsheets to document a patient who had more than four wound sites. The informatics-trained nurses and the hospital information system teams collaborated to resolve the issues in a timely manner.
This process is critical and necessary to refine the system further even after implementation. Listening and adjusting the implemented system based on nursing staff' input encouraged nurses to adapt quickly to the system.
DISCUSSION
This case study illustrates how the formal process of system implementation described in the workflow-oriented framework is applied to real-world HIS implementation tasks and the resulting benefits from such an exercise. End-user' preferences, which were missing in the initial designing phase of the system on the vendor side, were incorporated into the HIS and included the data presentation and documentation format before the final implementation. The vendor system was fashioned into a user-centered and user-preferred documentation system through the step-by-step processes of system adaptation and adjustment.
Four distinctive benefits resulted from this case study. The first step of WOF, workflow analysis, helped nurses make critical decisions based on their actual practice32 (ie, flowsheet format vs a charting template). For example, a flowsheet format of documentation displays clinical data in time and date order, but a charting template does not. This flowsheet format was aligned with nursing practice because of their need to view clinical data in chronological order for planning care. Figure 3 shows an example of nurses' preferences. This type of nursing shift category allows nurses to access their daily documentations quickly and easily. Second, use-case and activity diagrams created while utilizing the OOA technique during workflow analysis helped nurses to capture their workflow, tasks, and preferences precisely.26,27 Third, the iterative adaptation process, step III of WOF, allowed the identified needs to be integrated into the system. By incorporating nurses' needs and preferences, their perceptions of the ease of use and usefulness of the system increased, which are two key determinants in TAM that eventually result in system acceptance. Fourth, the adjustment period, step V of WOF, helped to integrate end-users' needs in detail into the system to fit real-world practices.
LIMITATIONS
This case study was performed in the same healthcare setting where the WOF was developed. It is not generalizable until thorough evaluation of the WOF is performed in prospective studies that include the HIS implemented based on this framework.
FUTURE STUDY
The evaluation of end-user satisfaction with the system and measurement of unwanted outcomes before and after implementation following steps of WOF in variety healthcare settings should be initiated.
CONCLUSION
This article demonstrates benefits of a nursing documentation system implementation guided by the WOF. Benefits of following WOF for a HIS implementation are capturing nurses' actual practice, identifying precise tasks and preferences of nurses, and incorporating nurses' detailed needs into the HIS. Utilizing the WOF is useful in two ways. First, the HIS will be a better fit for nurses' real-world practices when nurses are able to influence the modification and adjustment of a HIS during implementation. This is true even if the HIS was initially designed without incorporating their needs. Second, the system's accommodation of nurses' actual practices will increase acceptance of the HIS, ensuring successful implementation.
REFERENCES