The use of research in clinical practice does not just refer to research utilization, nor is it synonymous with evidence-based practice. ⁹ Although use of research findings is an important component of both evidence-based practice and research utilization, both have limitations related to use in clinical practice. Research utilization was originally intended to be the application of a portion of the research in a way that was not related to the original research study. ²⁶ Research utilization models include a synthesis of research literature on a given topic to summarize research-based knowledge that can be used in clinical practice. ^{13. and 17.} The degree to which research utilization is integrated into practice can vary along a continuum. 4.^{9. and 13.} At one end of the continuum is the initial conceptualization of the literature (referred to as conceptual utilization). ² The next step of research utilization is thinking about research ideas and findings on a given topic (referred to as knowledge creep). The next level of research utilization is taking action to move to a decision about the implications of research synthesis (referred to as decision accretion), and finally, at the other end of the research utilization continuum is the implementation of research into clinical practice (referred to as instrumental utilization).

Evidence-based practice uses the best available evidence, which includes an individual’s expertise and other external clinical evidence from systematic research. ²² Research is a key component in delineating evidence-based nursing practices. The strongest source of evidence for evaluating interventions is from systematic reviews. ³⁴ These reviews include an appraisal of individual studies, and when these individual studies are combined, they provide a meta-analysis of the effectiveness of an intervention. However, the reality is that nursing, like other health care disciplines, lacks a comprehensive body of research to support all interventions performed in nursing practice. Therefore when there is insufficient or no research available, clinicians need to use other resources and sources of data to augment their problem solving. Other recommended sources of “evidence” include benchmarking data, clinical expertise, patient preferences, infection control data, standards (international, national, and local standards), quality improvement and risk data, retrospective or concurrent chart review, pathophysiology, and analysis of cost-effectiveness. ¹¹

Regardless of the source of data upon which to build evidence-based practices, it is essential to take into consideration the level of evidence or the strength of the scientific evidence. ³² The Agency for Healthcare Research and Quality (AHRQ) ¹ summarized the published reports on how to determine or grade the strength of evidence. The AHRQ report concluded that for any system grading the strength of the evidence, three elements should be included: quality, quantity, and consistency. Quality refers to the extent to which studies minimize bias and are valid studies. Quantity is the number of studies and total number of subjects in the study. Consistency refers to the extent to which findings are similar between different studies on the same topic. There is no one best method for determining the level of evidence; the process of evaluating evidence should employ the most relevant levels of an evidence grading system for the topic or type of procedure being assessed. ^{5. and 7.}

Quality improvement initiatives are a rich source of “evidence” reflecting clinical expertise, patient preferences, and local context. Nowadays quality improvement is an inherent part of “best care” for patients. ¹⁶ It is assumed that clinicians will use data from their own clinical practice to improve practice. ²² Furthermore, accrediting and certifying bodies expect that quality improvement initiatives are an intrinsic part of an organization’s patient care delivery. Quality improvement initiatives often arise from health care professionals’ experiences and insights. Analyzing aggregate patient population data is necessary to determine potential changes in processes and interventions needed for improved patient care. ¹⁶ In summary, the types of research sources and ways in which to integrate research findings into clinical practice will vary along a continuum.

The initial step of the research process is defining the research question or research problem. ¹⁵ The research question or problem reflects an identified problem related to patient care, nursing education, nursing administration, or any issue of nursing interest. Patient care or nursing practice problems generally address practice differences and what is ideal or desirable. Researchable questions often reflect clinical experiences, such as (1) How effective is triage in prioritizing patient acuity? (2) What type of pain management can be used for pediatric patients undergoing procedures in the emergency department (ED)? and (3) How effective are discharge instructions for ED patients? Researchers may also derive their research question based on focuses or problems identified in the nursing literature itself. ² Research studies often make recommendations for future studies when summarizing implications of the current study. Researchable questions or clinical problems yet to be addressed are often identified when reviewing several research articles on the current state of the science. In addition, several nursing and federal organizations have published recommendations for future research studies. The Emergency Nurses Association, American Association of Critical Care Nurses, American Nurses Association, Sigma Theta Tau International, and National Institutes of Health are examples of organizations that have identified research priorities.

The purpose of the literature review is to explore work conducted in a particular area of interest to further formulate or clarify the research problem. After critiquing previous research in a particular area, the researcher summarizes what has been previously studied and delineates how a proposed study will contribute to the state of the science. A good literature review critiques and summarizes other studies to see how they fit into the scope of the study being conducted. A thorough review reinforces the need for the study in light of what has already been done. A written literature review should include summaries of articles that differ from the proposed point of view. This indicates that the author conducted an exhaustive review of available knowledge. ²

Information sources for literature reviews include both primary and secondary resources. ¹⁷ A primary source of information is the description of an investigation written by the person who conducted it. A secondary source is a description of a study prepared by someone other than the original researcher. Literature reviews are very useful for examining the body of evidence for best clinical practices, as well as for identifying the existing gaps in a given area of content.

Theories and conceptual frameworks provide a structure or blueprint to guide the study of clinical problems. A theoretical framework defines the concepts and proposes relationships between those concepts to provide a systematic view of a phenomenon. ¹⁵ It further enables the researcher to link the findings to a body of knowledge. This framework consists of the definition of concepts and propositions about the relationships of those concepts, a way to organize rules or beliefs about what is observed, and a systematic method to organize information about a particular aspect of interest in a research study. ¹⁵

Two components of a theory are concepts and propositions. Concepts, the building blocks of a theory, are abstract characteristics, categories, or labels of things, persons, or events. Examples of nursing concepts are health, stress, adaptation, caring, and pain. Propositions are statements that define the relationships among concepts. A set of propositions may state that one concept is associated with another or is contingent upon another. Examples of theories used in nursing are psychoanalytic theory, the theory of relativity, the theory of evolution, the theory of gravity, learning theory, systems theory, and the theory of homeostasis. The power of theories lies in their ability to explain the relationship of variables. ² Theories can stimulate research by giving direction. Questions and ideas formulated about what will occur in specific situations are called hypotheses. In research, hypotheses are tested to determine whether the information fits the theory. ¹⁵

Conceptual frameworks represent a less formal, less well developed system for organizing phenomena. They contain concepts that represent a common theme but lack the deductive system of propositions that identify the relationship among concepts. Conceptual frameworks are more or less a map for the proposed study. ¹⁵ The groundwork for more formal theories often evolves from conceptual frameworks.

Before a problem is researched, it must be narrowed, refined, and made feasible for study. The research interest can be stated as a research question or a hypothesis. The research question in a study should identify key independent and dependent variables. An independent variable is what is assumed to cause or thought to be associated with the dependent variable. Changes in the dependent variable are presumed to depend on the effects of the independent variable. The dependent variable is what a researcher wants to explain or understand. Research questions should be specific and not attempt to measure too much, because data analysis may be complex and be confusing to interpret. ² For example, a research question might be, What effect does the presence of the parent in the child’s room have on the child’s experience of pain during fracture reduction? The dependent variable is the child’s pain experience and the independent variable is the presence of the parent in the room. The dependent variable is explained through its relationship with the independent variable. It is known that many factors affect a child’s perception of pain, but only one independent variable (parent’s presence) is intended to be measured in the proposed research question.

Often the dependent variable can have multiple causes. A study may be designed to examine several factors and their influence on a phenomenon. For example, a researcher may want to know whether experience with triage or an educational program concerning triage influences ability to accurately perform triage. Both independent variables (education and experience) can influence triage performance ability (dependent variable). ²

Several dependent variables can be designated as measures of treatment effectiveness. An example of multiple dependent variables identified in a research question is, Does a comprehensive triage system have an influence on length of stay in the ED, patient satisfaction, and patient outcome? Length of stay, patient satisfaction, and patient outcome are all dependent variables by which triage effectiveness is measured. ²

A hypothesis expands upon a research question, because it is a prediction of the relationship or differences between two or more variables. ^{15. and 17.} This prediction of expected outcomes is the basis of the research process. Hypotheses, which often stem from theories, are possible solutions or answers to research problems. The hypothesis is a prediction of the nature of the relationship between several variables that is intended to be identified before the initiation of the research study. For example, one hypothesis might be that pediatric patients who are promised a reward at the end of a suturing procedure will cooperate and be more compliant than pediatric patients who are not promised a reward. In this example, the researcher is not only delineating whether a relationship between rewards and behavior exists but is also predicting outcomes from this relationship. The null hypothesis indicates that the two populations (samples) have the same mean. The null hypothesis plays a major role in testing the significance of differences between the treatment and control groups. The assumption at the outset of the experiment is that no difference exists between the two groups (for the variable being compared). Therefore the null hypothesis, stated as “there will be no relationship between” or ”there will be no difference in …,” is often generated for statistical purposes, data analysis, and discussion. ^{2.15. and 17.}

The methods section of a research study reflects how the researcher plans to or did implement the research study to answer the research questions or hypotheses. The components of the methodology section include research design, subjects, measures used to collect data, and study procedures. There are two major categories of research designs: quantitative and qualitative (Tables 6-2 and 6-3). In the context of this chapter, the focus will be quantitative studies. ²