Document View

Dramatic changes in health care and the growth of integrated delivery systems have intensified practitioners' efforts to access new information about more efficacious approaches that enhance discipline-specific and interdisciplinary contributions to patient outcomes. In the new healthcare environment, practitioners can no longer rely solely on clinical experience, pathophysiologic rationale, and opinion-based processes (Ellrodt et al., 1997; Feinstein & Horwitz, 1997). Practitioners also must learn to search the research literature, critically appraise research findings, and synthesize empirical and contextually relevant evidence. Practitioners need to question their current practices and find better alternatives (Barnsteiner, 1996). Critical thinking skills and evidence-based methods for making clinical decisions are essential for maximizing the quality and cost-effectiveness of care (Kessenich, Guyatt, & DiCenso, 1997; Sackett, Rosenberg, Gray, Haynes, & Richardson, 1996). The President's Advisory Commission on Consumer Protection and Quality in the Health Care Industry (1998, p. 169) reported that improving the quality of health care "requires a commitment to delivering health care based on sound scientific evidence and continuously innovating new, effective health care practices and preventive approaches." Evidence-based practice is the integration of "individual clinical expertise with the best available external clinical evidence from systematic research" (Sackett et al., 1996, p.71). The combined results from clinically relevant research, clinical expertise, and patient preferences produces the best evidence for ensuring effective, individualized patient care (Mulhall, 1998; Sackett & Rosenberg, 1995). Evidence-based practice is more likely to occur in practice settings that value the use of new knowledge and provide resources to access that knowledge.

Several national and international initiatives have been developed to facilitate evidence-based practice. In the past three decades there has been a tremendous increase in the number of clinical research studies, particularly studies using such methodologies as randomized clinical trials, meta-analysis, and study of patient outcomes. These research studies are the basis for the paradigm shift from the tradition and intuition-driven practice of physicians, nurses, and other health professionals, to the new paradigm of evidence-based practice. Researchers have sought to decrease gaps between the conduct of research and the use of research in practice settings. Many researchers are mentoring practitioners with the critique and synthesis of research and the development of guidelines for evidence-based practice. Although several models have been developed to guide practitioners in the research utilization process (Goode & Piedalue, 1999; Horsley, Crane, Crabtree, & Wood, 1983; Rosswurm, 1992; Stetler, 1994; Titler et al., 1994; White, Leske, & Pearcy, 1995), practitioners continue to have difficulty with synthesizing empirical and contextual evidence and with integrating evidencebased changes into practice (carniletti & Huffman, 1998; Mackey, 1998).

The evidence-based model described in this article is derived from theoretical and research literature related to evidence-based practice, research utilization, and change theory. The model guides practitioners through the entire process of changing to evidencebased practice, beginning with the assessment of the need for the change and ending with the integration of an evidence-based protocol (Figure 1). The authors developed and tested the usefulness of the model as they mentored nurses in defining and integrating evidence-based practice protocols at a regional medical center. The model might also be used in primary care or other settings in addition to acute inpatient units. A description of the model follows, along with an example of how nurses applied the model to implement an evidence-based protocol for hospitalized patients with acute confusion.

Overview of the Model

Step 1: Assess Need for Change in Practice

Practitioners' interest in a potential change in practice may be stimulated by awareness of patient preferences and dissatisfaction, quality improvement data, practitioner queries, evaluation data, or new research data. In Step 1, practitioners collect internal data and compare it with external data. When data indicate a problem with an aspect of practice, practitioners can assemble a team of stakeholders to participate in discussing and more clearly identifying the problem. Stakeholders may include discipline-specific or multidisciplinary practitioners, administrators, and patients who have a stake in the practice (Specht, Bergquist, & Frantz, 1995; Steelman, 1995). Group success can be enhanced by use of group-process techniques, such as structured brainstorming, flow-charts, and multivoting (Brassard & Ritter, 1994; McLaughlin & Kaluzny, 1999). Practitioners review the evidence using sources such as quality improvement Ql) and risk management (RM) data, utilization data, staff performance surveys, customer satisfaction surveys, agency resources, and strategic priorities.

After examining internal data, practitioners assess the need for a change in practice by comparing internal data with external data in benchmarking databases. Benchmarking entails collecting comparable performance data and "sharing of performance information to identify operational and clinical practices that lead to the best outcomes" (Czarnecki, 1996, p. 2). Healthcare organizations have recently begun mandatory external benchmarking using indicators from one of over 200 JCAHO-approved performance measurement systems (Joint Commission on Accreditation of Healthcare Organizations, 1999). Comparison of internal and external data may substantiate current practice or support the need for a change in practice. If data are inadequate, collecting additional internal data to identify the problem may be necessary (Titter et aL, 1994). The Table highlights Step I processes of the model that nurses used in the example of a change in practice for patients with acute confusion. Stakeholders included nurse managers and staff nurses concerned about the care of confused patients. They collected and analyzed internal and external data related to confusion. Identifying that a problem in caring for confused patients existed in their hospital, they committed to developing an evidence-based change.

Step 2: Link Problem with Interventions and Outcomes

Practitioners need to define the problem using the language of standardized classifications and then link the problem with classification of interventions and outcomes. Classification systems help to define the concepts of a science and organize the knowledge (McCloskey, 1995). They also facilitate communications among practitioners, provide standards for determining the effectiveness and cost of care, and identify needed resources (Maas & Johnson, 1998). National databases have primarily consisted of medical classifications, such as the International Classification of Diseases (ICD), the Diagnostic and Statistical Manual of Mental Disorders (DSM), and the Current Procedural Terminology (CPT) (McCloskey, 1995). Patient outcomes often are linked to episodic physician interventions, although multiple providers deliver health care across a continuum of care. To verify specific accountabilities for cost effectiveness and quality of care, longitudinal measurements of multiple disciplines are needed (Maas & Johnson, 1998).

Enlarge 200% Enlarge 400%

Standardized nursing languages are being refined to provide nursing data elements for automated patient records (Johnson & Maas, 1997; McCloskey & Bulecheck, 1999; North American Nursing Diagnosis Association, 1999; Saba & McCormick, 1996). Nursing Outcomes Classification (NOC) and Nursing Interventions Classifications (NIC) facilitate the linkages of outcomes and interventions appropriate for identified nursing diagnoses (Johnson & Maas, 1997).

In Step 2 of the practice change for patients with confusion (Table), the nurses referred to NIC and NOC. They linked acute confusion with the intervention of delirium management and tentatively selected several nursing activities listed under delirium management. The nursing activities served as process indicators during the quality monitoring process in Step 5. The nurses selected outcomes of cognitive orientation and safety. These outcomes were measured by a confusion rating scale, fall rates, and use of restraints. The selection of potential interventions and patient outcomes are based primarily on clinical judgment (Johnson & Maas, 1998) and system priorities and resources.

Step 3: Synthesize Best Evidence

In Step 3 of the model, selected interventions and outcomes are refined. The best research evidence is synthesized and combined with clinical judgment and contextual data. The problem, potential interventions, and desired outcomes become the major variables for reviewing the research literature. Steps taken before conducting the literature search include clarifying the specific topic and identifing criteria for including a reference in the review (Slavin, 1995). In the critical appraisal of the literature, practitioners evaluate the strengths and weaknesses of studies and identify gaps and conflicts in the available knowledge. The use of a structured critique worksheet facilitates the recording of the major components of each critique and organizes information for the synthesis of the evidence. An example of such a worksheet appears in Figure 2. The last section of the worksheet contains a rating scale for the quality of evidence, as well as questions about the benefits and feasibility of using the research findings. This evidence rating scale was adapted from the rating scale used in AHCPR research reviews (United States Preventive Services Task Force, 1989).

Numerous electronic databases are available for literature searches. Medline is the largest biomedical, electronic database, referencing over 4,000 journals. It can be accessed without additional cost through the Intemet Grateful Med website (U.S. National Library of Medicine, 1999). Systematic literature reviews are available on the Internet. The Cochrane Library is one example of a proprietary resource for systematic reviews of research about health care interventions. It is accessible on-line (The Cochrane Collaboration, 1998). Other systematic reviews completed by independent researchers are in other Internet resources such as MEDLINE, CINAHL, the On-line Journal of Knowledge Synthesis in Nursing, and Best Evidence. The AHCPR evidence-based practice guidelines are also available on-line to practitioners (Agency for Health Care Policy and Research, 1999). Their rigorous, evidence-based approach has made the AHCPR guidelines the "gold standard."

The purpose of the synthesis of the research studies is to determine whether the strength of the evidence supports a change in practice. The results of studies can be pooled only if the studies are similar in design. In the absence of strong evidence, practitioners need to weigh benefit to risk factors. They also need to consider the feasibility of implementing the findings in their own practice setting. The synthesis only brings together the existing evidence. It cannot create new evidence or knowledge. Thus, if most of the evidence is weak, additional research may be needed before making decisions to change practice or policies. If the research synthesis indicates sufficient research evidence to support a change in practice with desirable benefits and minimal risks, practitioners can proceed in designing the change.

Enlarge 200% Enlarge 400%

In the example of the protocol for acute confusion (Table), the nurses, with guidance from nurse researchers employed by the hospital, completed a thorough literature search of quantitative and qualitative studies focused on delirium and patient safety. They critiqued the research and synthesized evidence of quantitative studies. This evidence was combined with qualitative findings, clinical judgment, and contextual data. Based on this evidence, they decided to develop and pilot test a protocol that was feasible for nursing staff to implement and offered maximum benefit with minimal risk to patients.

Step 4: Design a Change in Practice

After synthesizing the best evidence, practitioners describe the process variables or detailed sequence of care activities in the change in practice, usually in the format of a protocol, procedure, or standard (Specht et al., 1995; Steelman, 1995). The practice environment, its resources, and feedback from stakeholders are essential considerations when designing a change. Decreasing complexity of the protocol increases the likelihood of its acceptance. Only activities addressed in the evidence base are included in the protocol (Horsley et al., 1983, p. 42). Likewise, the protocol is designed to guide care only for populations similar to those in the evidence base. The evidence base is used to guide practitioners in identifying anticipated discipline-- sensitive and interdisciplinary patient outcomes of the practice change. Those outcomes are clearly delineated as desired outcomes or reduction in undesired outcomes. The more relevant the outcomes are to the organization, the more likely the practice will be accepted.

Enlarge 200% Enlarge 400%

If the change in practice affects a standard of care in a large hospital, a pilot demonstration of the change on one or two units is advised. The pilot test allows practitioners to influence adaptation of the change to fit their practice needs (The President's Advisory Commission on Consumer Protection and Quality in the Health Care Industry, 1998), giving practitioners a sense of ownership of the change process and contributing to smoother integration of the change. The plan for the pilot delineates necessary structural components for implementation, that is, equipment, documentation forms, personnel, other resources, and associated costs. Successful implementation partially depends on suitability of the new practice for the particular health care site (Cipperley, Butcher, & Hayes, 1995; Cook, Greengold, Ellrodt, & Weingarten, 1997) and on obtaining administrative support when costs will increase (Horsley et aL, 1983).

The plan for the pilot test also includes the timing and delegation of specific activities for obtaining agency approvals, preparing the test sites, and evaluating the results. The evaluation plan for the pilot test includes a study of quality improvement Ql) with process and outcome indicators and surveys of patient satisfaction and staff responses to the change in practice. Some indicators will be identical to those in existing QI data to provide for pre- and postimplementation comparisons. The indicators pertain to activities specified in the protocol, resources needed to follow the protocol, anticipated patient outcomes, and associated costs. Practitioners prepare the QI data-collection instrument and surveys, specifying data sources and acceptable samples. Data collectors are trained in use of the QI data-collection instrument, and interrater reliability is established. Clinically significant indicators are selected and are planned for data analysis procedures.

In the example of confusion in hospitalized elderly patients (Table), nurses from the two units participated in designing the protocol and planning for its implementation and evaluation. The project team identified cognitive orientation and safety as target patient outcomes. They selected nursing activities for delirium management that were supported by the evidence base. Information about the pilot test was communicated to collaborative practice groups and administrators for their review and approval. Nursing staff on the two test units had inservice training to implement the protocol. A 2-month implementation period followed.

Step 5: Implementing and Evaluating Change in Practice

Implementation of the pilot study will be more successful if the coordinator of the pilot study closely monitors the process and is available to staff on the pilot study units to answer questions. Followup reinforcement of the practice change by the coordinator is essential. After the protocol has been in use for the designated time, patient and staff surveys and QI study are conducted. Then, data are analyzed and displayed in charts or bar graphs to facilitate data interpretation. Following analysis, practitioners interpret the results by deciding whether there were differences in the indicators before and after the pilot study. Were the necessary structural variables provided? Do the data indicate that the new protocol was implemented as intended (McCollam, 1995; Specht et al., 1995; Steelman, 1995)? If yes, what effect did the new protocol have on patient outcomes? An inappropriately implemented protocol, because of misunderstanding or lack of endorsement, can do more harm than good (Cook et al., 1997; Ellrodt et al., 1997). When considering the results, practitioners must remember that outcomes can be affected by numerous factors other than the intervention, such as characteristics of patients, staff, interpersonal aspect of care, and the setting (Sidani & Braden, 1998). QI studies are not designed to control for all these influences but can give an indication of the protocol's effectiveness in a particular setting. In addition to QI data, practitioners evaluate the results of staff opinion surveys at participating sites (McCollam, 1995). Endorsement by respected peers is essential for successful implementation of the change in practice (Cook et al., 1997; Specht et al., 1995).

The decision to adapt, adopt, or reject the change is based on feedback from staff on the pilot units, managers, and pilot coordinators, QI and survey data, cost data, and recommendations from stakeholders. Feasibility, benefits, and risks are considered when evaluating the data. Personnel opinions of the implemented change provide information about acceptability or the need for modifications (McCollam, 1995). QI and cost data indicate whether the care and outcomes improved at a reasonable cost to the system. Based on all the evaluation data, practitioners make recommendations to adapt, adopt, or reject the change in practice (McCollam, 1995; Specht et al., 1995; Steelman, 1995).

In the example of confusion in hospitalized elderly patients (Table), two nurses from each unit served as pilot-study coordinators. They monitored the 2-month implementation process and obtained informal staff nurse feedback. The unit coordinators participated in collecting and analyzing the QI data. All project activities progressed as planned.

Step 6. Integrate and Maintain Change in Practice

If the results of the pilot study support integration of the new practice into standards of care, change strategies are initiated. Even the smallest change has a domino effect and people affected by the change often perceive it as disruptive. Practitioners who are change agents need to consider the cultural climate of the organization as they attempt to integrate practice innovations. For example, impediments to change are greatest in large bureaucratic organizations where changes are made by the top-down method (Bennis, 1993). Change is more likely to be accepted when people participate in making the change (Rogers, 1995). The evidence-based model described in this article encourages participation of stakeholders throughout the various steps of the model. The results of the pilot study enhance stakeholder confidence in the effectiveness of the change and the feasibility of making the change in their environment. Administrators and practitioners prefer concise summaries of recommendations, supporting evidence, and anticipated benefits (Ellrodt et al., 1997). Ongoing communication with stakeholders is vital to the acceptance of change. Adherence to the details of the organization's operations and approval processes ensures a smooth and speedy integration of the change into the organization's standards of care. Informal leaders need to participate in the diffusion process. Continuing education and staff-inservice education facilitate changes in practitioners' behavior and reinforce implementation of the new evidence-based practice.

Maintaining the change is ensured by providing practitioners with the necessary resources to implement the change, by monitoring the process and outcomes, and by rewarding quality performance with incentives (Greco & Eisenberg, 1993). In the integration and maintenance of the protocol for confused patients (Table), specific actions were implemented to enhance communication, education, and monitoring of the change. The initial meetings were with the staff nurses on the pilot units to obtain their feedback about feasibility, benefits, and needed revisions. The revised protocol was then presented to the standards of practice council for approval. Concise written and oral presentations were prepared for administrators and collaborative practice groups. Inservice sessions about the new protocol were presented to nursing staff on all hospital units. Nurses decided to conduct periodic QI monitoring of the implementation and outcomes.

Conclusions

The momentum is escalating in support of evidence-based practice that will improve the quality of patient care and enhance clinical judgment. Practitioners must know how to obtain, interpret, and integrate the best available research evidence with patient data and clinical observations. The evidence-based model described in this article was derived from theoretical and research literature. The model was successfully applied by nurses who were implementing change to evidence-based nursing practice. It may serve as a useful framework for other practitioners seeking to change to evidence-based practice in a variety of settings.

Patient outcomes must reflect discipline-specific and interdisciplinary accountabilities. Nurses' contributions to patient outcomes will be measured when nurses consistently use standardized language in defining patient problems, interventions, and outcomes. Practitioners need time and support to access databases and synthesize the best evidence for making changes in practice. Administrators must provide the infrastructure for evidence-based practice to develop and diffuse throughout an organization. Collaboration between researchers and practitioners within and among disciplines should enhance the diffusion of practice innovations.