The federal government defines

Transcription

1 Rebecca P. Winsett Ann K. Cashion To order your copy, visit the ANNA Web site at This chapter on research in nursing was included in ANNA s Career Fulfillment in Nephrology Nursing: Your Guide to Professional Development (2 nd edition), which was published in September This chapter encourages nurses to participate in the research process by providing an overview of the nursing research process, including the steps to designing a research project, and answers to frequently asked questions. The objectives of the chapter are to: (a) define the nursing research process, including the systematic and universal knowledge concepts along with overviews of process improvement and evidence-based practice; (b) identify the steps to designing a research project, including identifying a clinical problem and formulating a research questions, conducting a review of the literature, and providing an overview of the research design; (c) provide an overview of the study methodology, including data analysis and protecting human subjects; (d) discuss available funding opportunities and the professional obligations that go with these opportunities; and (e) provide answers to frequently asked questions regarding nursing research to help demystify the process. The federal government defines research as the systematic collection of information in order to increase universal knowledge (Department of Health & Human Services, 1991). There are several complex concepts imbedded into that definition, but before we discuss the concepts we need to clarify the word research. The use of the noun research implies a systematic investigation to increase universal knowledge; one that is thoughtfully planned and executed. The noun research is not to be confused with the verb research, where one might state, I researched this and found. Our day is full of researching information and using that information to assist us in our work. But for the purposes of this chapter, we speak only to the noun research and the system of inquiry that surrounds our investigations to answer clinical questions. It is the noun research that often intimidates the clinical nurse. We hope to dispel some of the intimidating factors and light the flame of inquiry. To help us understand research, let s explore the Rebecca P. Winsett, PhD, is Associate Professor, College of Nursing, University of Tennessee Health Science Center, Memphis, TN. Ann K. Cashion, PhD, RN, FAAN, is Associate Professor and Chair, Department of Acute and Chronic Nursing, College of Nursing, University of Tennessee Health Science Center, Memphis, TN. She is a member of ANNA s Memphis Blues Chapter. process. Don t let it be daunting, as there are nurses out there who are professional researchers. Just like we go to medical experts when we have a medical concern, we can turn to our nurse research experts when we have a research concern. The first concept that we need to clarify is systematic. Our approach to data collection is methodical and organized in such a way as to answer the clinical question posed. There are sections in nursing research textbooks devoted to principles of systematic data collection and sampling strategies; key components of a research study. It is advisable for any nurse researcher (the authors included) to have on hand a general research textbook (LoBiondo-Wood & Haber, 2002; Polit & Beck, 2004). Referring to a textbook often helps clarify threats to your systematic collection of information. All researchers, clinical or basic, must determine the best way to collect study data so that the outcomes can be generalizable. If there is not enough attention paid to how the data is collected, the outcome is not reliable and thus not applicable to a larger population. The second concept to clarify is universal knowledge. What is universal knowledge? You can say that if your research adds new understanding for a given problem that helps broaden what we know for a population at large, then the new idea becomes part of the body of knowl- edge that is said to be universal. Counter this with an evaluation of how a program worked within your institution; for instance, your unit wants to assess whether instituting a self-medication program is effective on your renal floor. You found that the program or part of the program worked and now you want to write it up as research. Before the project can be categorized as research, several questions need to be answered. For example, has the project undergone the scrutiny of controlling the variables that may influence the outcome? Has the project followed the research process? Were the subjects aware that data was being collected on them? A key question to ask: was the data collected in a systematic manner? Did you get good outcomes because the sample only included those who successfully completed your program (thus introducing bias into the outcomes by the way that you chose your subjects)? Once these questions have undergone this type of scrutiny, can you then say that the outcomes were based on the intervention (self-medication)? These questions can then determine if your outcomes can be applied outside your individual organization and be a part of the body of universal knowledge. If so, then the project can be called a research study. Research studies are not usually determined after the data is collected; however retrospective reviews of NEPHROLOGY NURSING JOURNAL November-December 2007 Vol. 34, No

2 large databases have been done and do have scientific merit. It goes back to the importance of the research question. Some questions are adequately answered by a look back at data, but most will require a prospective inquiry. The research question and number or amount of data needed to answer the question will dictate the design. Research is a process in which the investigator actively determines the question to answer, creates a data collection and analysis plan as well as a plan to control for events (known as threats) that would influence the outcome (Burns & Grove, 2005). Evaluating a system or a program on the unit is still an excellent way to focus on clinical outcomes. Here you have to distinguish process improvement from research. To a novice this may be difficult because these two methodologies overlap in some ways, but by and large, process improvement serves the institution and is not considered to increase generalizable knowledge. The information gleaned from evaluating your practice is no less important; however the label of research sets the standards for a different guiding body. Once we move into the realm of research, oversight by an institutional review board and human subject protection are required. Research questions and the study methodology do not have to be grand; we just stretch what we know so that it is available to others to know. Once we clarify the definitions of process improvement and then evidence-based practice methodologies, understanding the research process will be much clearer. Process Improvement Process improvement is a method to recognize and prioritize problems on your clinical unit or within the institution (Davidoff & Batalden, 2005). The goal of process improvement is to fix problems that can improve patient service or outcomes. There are methodologies developed just for process improvement and you may be familiar with a few of these; for example: FOCUS PDCA, rapid cycle PDCA, Six Sigma, or Lean (Medical Risk Management Associates, 2006; Pyzdek, 2003; Zidel, 2006). These methodologies provide guidelines to help improve the systematic collection of data for process improvement projects, and can, in some instances, be considered research, but by-in-large, do not meet the scrutiny to be considered research. Process improvement is very important and is not to be considered less valuable than nursing research. Even though the processes are similar, it is important to remember that some questions need a process improvement methodology and other questions require a nursing research methodology. Successful process improvement strategies may warrant publication, but these are not generally publishable in nursing research journals. Evidence-Based Practice In nursing, we also search for best practices by using evidence to guide our clinical practices. Evidence-based practice differs from research in that we are examining the published literature about a given topic in order to help make decisions about patient care (Sackett, Rosenberg, Gray, Haynes, & Richardson, 1996). The methodology of determining and evaluating the evidence starts with a question and proceeds to a thorough search of the published literature. The literature or evidence is ranked and critiqued, and determinations of whether practice changes are needed are based on the conclusions of the evidence reviewed. Using evidencedbased practice is a sophisticated use of the published research. Having a solid evidence-based practice textbook helps to clarify the processes that are needed to acquire and critique the evidence (Malloch & Porter- O Grady, 2006; Melnyk & Fineout- Overholt, 2005). Both process improvement and critiquing evidence to modify practice are important for nurses, and each has some overlap with the nursing research process. One is not more important than the other whatever question that you ask requires that you use the methodology that best answers the question. This chapter focuses on the nursing research process, and if you are a skilled user of process improvement or evidencebased practice, you will recognize many of the steps of the nursing research process. Understanding the research process may help you to identify and clarify a researchable question and provide the guidelines for you to institute a research study in your unit. Not all nurses are required to conduct nursing research, but part of our professional obligation is to understand and critique the research published in our field and participate in determining if findings are applicable to our settings. The American Nurses Association (ANA) asserts that participating in research is critical to provide the knowledge needed for practice, and that the educational level of the practicing nurse determines how the nurse can best contribute to nursing research (Council of Nurse Researchers & Council of Nursing Practice, 1993). The research educational standards differ for each nursing degree level. For example, the associate degree nurse may not have a formal research course but is taught the importance of constructing and following policy and procedures within each setting. The baccalaureate nurse has preparation to read and critique the published literature to develop the culture of inquiry. The educational standards for the master s level nurse has changed from designing and implementing a small research project to developing a sophisticated approach to synthesizing the research literature to modify nursing practice. No matter the level of education, your role in following the protocols of a nursing research project being done on your unit is just as important as those who participated in the design of the study. Not every nurse needs to lead a research team; however, if the desire is there, use your resources around you to help you be successful. 636 NEPHROLOGY NURSING JOURNAL November-December 2007 Vol. 34, No. 6

3 Novice researchers may gain expertise by participating as study coordinators or research assistants. Teaming up with a well-seasoned nurse investigator is another way of gaining experience. Most principal investigators wet their research appetites as study team members prior to taking the lead position. Research Process - The Steps to Design a Research Project Identifying a Clinical Problem and Formulating a Research Question Identifying the practice problem and crafting the research questions are the most important steps of the research process. The steps we discuss are listed in Table 1. No one is better at identifying a practice problem than a practicing nurse (Council of Nurse Researchers & Council of Nursing Practice, 1993). This process begins with the nurse asking the question what would happen if? or does this always happen when? One of the best ways to identify a problem of interest is to think about what bothers you at work; oftentimes, this creates passion for investigation. Realize that as a novice investigator, it will take perseverance and attention to detail to complete the research process. Narrow your focus; there is no better advice to give to a novice researcher than those 3 words. Once you identify a problem area, write down several questions of interest and seek advice on what is feasible to pursue. As we discussed earlier, relying on research experts to help guide you through the research process is suggested. Discussing the importance of your question helps identify the significance of the problem and purpose of your study. The results of the study need to have the potential to contribute to a larger body of knowledge and specifically to patient care. One significant roadblock for novice researchers is identifying a problem that is too large and too unfocused. For example, a large problem would be Determine why hemodialysis patients become septic. Renal nurses can think of lots of reasons why patients become septic and want to explore each and every one of them and then see that the question is far beyond their ability to answer. This global problem appears monumental so that it is easy to consider dropping the investigation all together. If you take a more narrow approach and state the problem as identifying the clinical characteristics associated with septicemia in hemodialysis patients, this question seems feasible. The authors highly recommend that the problem and research questions err on the side of a narrow focus of interest. One does not have to cure world hunger to add to universal knowledge. Formulating the research question that is narrow in scope, with clearly identified variables to study creates the guide for the rest of the process. Spend your energy in creating a researchable question. The attention that you pay to the research question will set in motion how you will accomplish the research project. The question drives the so what and thus your review of the literature. The research question drives the study design, thus determining the choice of your methodology and data collection. If your research question is unclear, then you will feel lost in the rest of the process. You will also need to focus on the practical issues related to resources Table 1 The Steps in the Research Process Problem formulation-research question Review of the Literature Methodology Choose the appropriate design Prepare the procedural steps Assess validity threats Data collection and data processing Data analysis Interpreting the data Writing the results needed to accomplish your project. Resources such as money, time, ability to collect data, and statistical support will need to be explored with your administrators as well as your medical, nursing and other colleagues who may be asked to help recruit patients or administer the interventions. There may be other issues that you will have to address as you move further in the planning stages, but most feasibility issues can be overcome if identified and addressed prior to starting the study. Review of the Literature You may have started with reading the scientific literature that led to your clinical question or you may have asked the clinical question and then went to the literature to see what has already been published on your topic. Either way, using published research can help you as you finalize your research question. You can focus on particular topics once you have: obtained a clearer picture of what you would like to study, evaluated the literature to substantiate that the study needs to be done, identified a unifying conceptual framework, and gathered ideas for study design. Replicating a study that has already been published can be a good way to determine if findings are applicable in a different setting or with a different population, fostering valida- NEPHROLOGY NURSING JOURNAL November-December 2007 Vol. 34, No

4 tion and strengthening the overall body of evidence. There are several methods that can be used to find published articles. Commonly, researchers conduct an internet search using online access to medical databases such as MEDLINE or the Cumulative Index to Nursing and Allied Health Literature (CINAHL). Search engines like Pub- MED or OVID are user-friendly and nurses can easily obtain printed abstracts and publications on any medical topic of interest. Other sources for information include websites that collect, analyze, or distribute information about diseases such as the United States Renal Data System (USRDS) and United Network for Organ Sharing (UNOS). For example, if you are interested in the rate of end stage renal disease by age in your region, this information may be found on the USRDS website ( While we often are more concerned with published research studies, other sources for relevant information include integrated reviews or clinical expert opinions. These types of articles often can help refine the significance and the feasibility of a research problem and questions. Do not hesitate to include a medical librarian in your search for relevant literature. Experienced librarians not only have access to numerous search engines, but librarians in academic centers usually are very good at phrasing search words and topics in order to obtain the best possible results. The Research Design Experimental vs. Quasi-experimental. There are many types of research designs, but all have the same purpose to provide a framework and plan for answering the research question (LoBiondo-Wood & Haber, 2002). Most of us are familiar with the classic double blinded, randomized controlled experimental design frequently used in clinical drug trials. Experimental designs have three characteristics: Randomization. Subjects are distributed to either the experimental or the control group by a system that gives each subject an equal chance of landing in one group or the other. Control. A control group is formed where no intervention is conducted. Manipulation. An intervention is given to only some of the subjects. Most nursing studies are not true experimental studies because of the nature of the clinical questions and the inability of having a true controlled environment. Frequently nurses employ quasi-experimental designs. Quasi-experimental designs can be similar to experimental designs, however, not all three characteristics are present. For instance, there may be no control group or the subjects are not randomized to study groups. Limitations of experimental designs are that they may be expensive to carry out and can involve many months of follow-up (Hulley et al., 2001). However, sampling techniques are important even in quasiexperimental studies. Outcomes are strengthened by the sampling strategy (Winsett et al., 2007). Non-experimental designs. Nonexperimental research designs are frequently used by practicing nurses when they need to document that a problem exists. Types of non-experimental designs include descriptive, comparative, or correlational designs. For example, many of us have read descriptive or exploratory studies where the purpose was to describe a phenomenon such as the depression level of a patient group. We may also be exploring the association between two clinical outcomes, such as the relationship between blood pressure and depression in hemodialysis patients. Usually, once the phenomenon is described and there is an agreement that a problem exists, an intervention study can be conducted to examine ways to alleviate the problem and provide better health care. In this case, the research question is often best served by a quasi-experimental study design. Non-experimental designs can be prospective (following subjects forward) or retrospective (reviewing historical records). Using retrospective data can be effective and less threatening to a novice investigator. Descriptive or correlational designs using retrospective methodology would use information already collected in a database (or patient medical record) and characterize a problem or determine if a relationship exists. A significant limitation of retrospective research studies is that the researcher is restricted to the data that was collected at that time. For example, if weight or blood pressure was not entered into the record or a creatinine level was not drawn and you needed that data, the nurse researcher cannot go back and recreate those data points. What Design Should You Use? While this chapter is not considered to be an exhaustive discussion of research design, the most important determinant of your research design is your research question. The question itself drives the design. Let s take the example of our hemodialysis depression/blood pressure phenomenon. Suppose you ask the question: What is the blood pressure of hemodialysis patients who are being treated for depression? What type of study design would be used? This could be a retrospective non-experimental design where you review medical records and explore the phenomenon of blood pressure in patients who are or are not on antidepressants. Suppose you ask the question: What is the relationship between depression treatment and blood pressure during dialysis for patients who are treated with anti-depressants? This study could be prospective nonexperimental where you assess blood pressure during dialysis (at points you determine are important) and compare the blood pressures between those treated with anti-depressants with a group that is not on antidepressants. Let s take that question one step further. You know that there is a difference in blood pressure for those on 638 NEPHROLOGY NURSING JOURNAL November-December 2007 Vol. 34, No. 6

5 Table 2 Threats to Internal Validity Internal Validity - The extent that the effects of the study are related to the study and not extraneous events Historical events Maturation Testing effects Instrumentation Statistical regression Selection bias Mortality Diffusion of treatment Compensatory equalization/rivalry Events or interactions with life outside of the study environment that may influence subjects responses Effects of learning during the course of the study Does history or events impact one group over another? Subjects altering responses based on previous testing or providing socially acceptable responses Are the instruments or questionnaires used in a study reliable and valid for the population? Is data collected the same across the entire study period? Did the treatment effect the change (Type 1 error)? Was there sufficient variability in the data and adequate sample size to detect a change if present (Type 2 error)? Did the sampling plan assure equal opportunity to participate in the study? If nonprobability sample was used, are there safeguards in place to generalize from the sample? The effect of drop out in the study (by any means) and how does the investigator address study mortality? The controls over the intervention group and how the investigator assures that the control group and the intervention group are not contaminated Acts from others or from subjects that alter the participation in the assigned group. Resentful demoralization Subjects perceived that they are receiving less than desirable treatment and opt to withdraw from the study. Note: Used with permission from Progress in Transplantation (Winsett et al., 2007). anti-depressants and now you want to know if you can impact the blood pressure. So you ask the question, What is the impact of cognitive therapy on blood pressure during hemodialysis in patients who are treated for depression with antidepressants? This study could be quasi-experimental where you randomize those under treatment for depression into control and experimental groups. The control group receives standard care and the experimental group receives cognitive therapy at some point during (or before) hemodialysis. As you can see, the question itself defines what type of design that is used. Internal and External Validity Two important concepts when discussing research design are internal and external validity. Internal validity is the extent to which study results are truly the results of the study and not a product of influences outside the study. External validity is the extent to which study findings can be generalized to other groups of patients (Burns & Grove, 2005). Of course, the researcher would want few, if any, external influences on the study; and the more the study results can be generalized to other patient groups the more relevant and important the study. Tables 2 and 3 provide a list of common threats to internal and external validity. Study Methodology Once the problem has been identified and described, the research questions conceived and the design chosen, it is time to outline your methodology. The study design is the map of how the study will be conducted, while the study methodology is the specific instructions on how you will go about completing your specific study. The study methodology has a somewhat standard format that includes sections describing the study design, methods for selecting subjects, procedures to the study, instruments to be used, and the analysis plan. The study methodology is often referred to as the study protocol. You should initially focus on the procedures, including all of those little steps that the investigator and the subject will do in order to collect the data. This is an important step in study validity, so it is critical to be very detailed in identifying each step of the procedures. It is beneficial to break down the recruitment of subjects into multiple steps, such as identifying potential subjects from chart review, approaching potential subjects prior to initiating dialysis (for example), and following-up with two phone calls during the evening hours made by their assigned nurse, etc. As you create a list of all the tasks to be completed in the study protocol, NEPHROLOGY NURSING JOURNAL November-December 2007 Vol. 34, No

6 Table 3 Threats to External Validity External Validity - The extent in which the outcomes are generalizable beyond the study sample Interaction of selection and treatment The impact of entering subjects for whom it is easy to become a subject. Interaction of setting and treatment Interaction of history and treatment The impact of environmental culture in the study setting. For example, is the study using additional time of personnel; does it represent a burden to the staff; is it controversial? Does the environment support research? Is there an event that influences study results, making the study less generalizable or replicable? For example, changes in health policies, or natural disasters. Note: Used with permission from Progress in Transplantation (Winsett et al., 2007). threats to study validity are often identified. Planned sample size and subject inclusion and exclusion criteria are described in the methodology. Determining appropriate sample size is an important part of the study. A procedure known as power analysis is conducted in order to determine sample size. A power analysis tells the investigator how many subjects are needed in order to detect if a significant relationship exists between the study variables. A biostatistician may be needed to determine a power analysis. We are often reminded as we begin to develop our inclusion criteria that there is a phenomenon known as the disappearing subject. This means that we created the perfect subject inclusion criteria so that we can minimize threats to study validity, but find once the study is started, all of a sudden there are no subjects that meet the inclusion criteria. So, our advice is to be as broad as possible when developing inclusion criteria. It may be possible that the threats you had identified earlier can be controlled in other ways. Other important items to be addressed in the methodology include identifying and describing proposed instruments such as surveys, questionnaires (i.e., depression or quality of life) and clinical tools (i.e., thermometers, blood pressure cuffs). Using previously validated tools can save a tremendous amount of time and also gives your study credibility. Once you have the procedure clearly outlined, take the time to map out a timeline for completion. Planning for each part of your study can help you set realistic goals. Your study methodology will also describe your intended data analysis plan, including the variables to be collected and how you will protect the data. Data analysis. The type of design chosen for the study will determine the statistical analyses conducted. For example, a study characterizing depression levels may use descriptive statistics to answer the research question, while a study with a quasi-experimental design may use inferential statistics to examine associations and identify predictors of risk among the variables. Deciding the data analysis plan is important before you begin to collect data. If you do not have access to a biostatistician, at least discuss this very important issue with a veteran nurse researcher. Your quality department in your hospital may have someone experienced in data analysis. While they may not have research experience, they may be able to give you tips on how to collect data or how to use a spreadsheet to enter your data for it to be analyzed. Having the input of a biostatistician during the initial design phase of your study can be incredibly valuable. Depending upon the complexity of your study, you may want to consider hiring a biostatistician. This is just one more reason to seek funding. From the initial design phase to development of conclusions, the biostatistician s input is indispensible. Protecting human subjects. Institutional Review Boards (IRBs) were established to insure that the three basic ethical principles of autonomy, beneficence and justice are followed throughout the research protocol (Ackerman & Winsett, 2002; National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1979). To conduct research, the investigator must provide a plan for subject protection. Often, one hears about the informed consent, but this is only the end product of a process that lays out how subjects will be protected in the study, how data will be used, and how the subjects will be informed of the study. This process begins when the written study methodology moves on to review by the IRB and culminates with the investigator performing an informed consent process with the potential subject understanding the process and voluntarily signing a consent form. It is federally mandated that all research studies undergo a review process by an IRB prior to starting the study (Department of Health & Human Services, 1991). Even if you are just doing a chart review and there is no real or perceived threat to the patient, protecting the subject s privacy is critical. Any identifying information, such as might be contained in the subject s demographic data, must be handled with a great deal of scrutiny and in accordance with the review board s policies. If you are unaware of how to go about obtaining IRB oversight, begin 640 NEPHROLOGY NURSING JOURNAL November-December 2007 Vol. 34, No. 6

7 with the administrator of your institution. Many times, if your institution does not have an IRB, they collaborate with other institutions that have an established IRB so you will have access to submitting your proposal to a collaborating institution. It is within your rights and responsibilities to question whether a clinical trial occurring within your unit has undergone IRB oversight. The clinical research nurse coordinator or the principal investigator will be able to answer any of your questions regarding the informed consent process or scientific merit of the study. Questions regarding the ethical conduct of a research study or research investigator can and should be directed to the IRB. Funding Opportunities Funding for research studies in the clinical setting can be obtained from a variety of sources and range from hundreds to millions of dollars. While some organizations require an indepth research application that may be 25 pages or more, others require only a synopsis of your study in 5 pages or less. Our recommendation is that you discuss potential funding sources with other researchers and clinicians in your organization. Potential sources for funding include private foundations, hospitals, all levels of government, schools, nursing organizations, and pharmaceutical companies. Professional obligations. Once the study is completed and the findings are summarized, it is important to disseminate findings through presentations and publications. This can be accomplished on the local level, but it is also appropriate to present on the national or international level. If your results are relevant to those living in community settings, it may be appropriate to publish in general consumer magazines. Frequently Asked Questions Although the intent of this chapter was to provide you with an overview of the nursing research process, the content may have sparked more questions and we hope that you will pursue those questions with a veteran nurse researcher. Below are some commonly asked questions that have been posed to the authors at nursing research conferences: I think I have a research question. I work on a hospital renal floor and we have never done a nursing research project. I ve talked to the unit manager and she agrees that I can do this. How do I get this project off the ground? While it seems as if the answer is complicated, let s divide it up into 4 levels: developing the design, developing the institutional approval, developing the co-worker buy-in, and developing collaboration with other disciplines. Level 1. Developing the design. As suggested in the chapter, it is time to team up with someone who has the expertise to guide and advise you. Hopefully, you can work with someone within your institution, particularly if you are in a Magnet hospital or one who is seeking Magnet status (American Nurses Credentialing Center, 2005). Although having someone at your institution is ideal, you can also develop relationships with local university nursing faculty or seek resources from your professional organization. With the internet, you have an additional resource to find the expertise that you are seeking. It may be important for you to identify someone who is interested in your topic; although if you are unable to find someone with the same interests, don t let that deter you from developing a relationship. You can provide the content expertise, and the researcher can assist with the study design. What you are looking for in this clinician/researcher relationship is one that provides mutually beneficial outcomes. A nurse researcher may only be able to provide you with good advice and be available to answer questions you have along the way. Depending upon your goals and the researcher s interest, the researcher may be available to assist with literature searches or provide writing or editing services. Whatever the level of support you received in your research endeavor, it is important to give credit by including these individuals names on posters, abstracts, or manuscripts that come from a completed project. Discuss your long-term goals with the researcher up front. Identifying your expectations before you get started can clarify tasks for both parties. Nurse researchers understand that novice investigators are an investment into our nursing future and in most instances are very willing to help. Establishing this relationship is a commitment for scholarly achievement. Both you and your nurse research colleague can benefit from this commitment. The critical development of your study is the research question. Spend time on constructing a solid researchable question. Remember, the research question drives the study design. Often, new investigators have lots of questions to ask and the study gets larger and larger.what we chidingly call trying to cure world hunger. Focus on the question you want answered and create the strongest design to answer that question. Early in your research endeavors, do not try to make your project complex. Some of the simplest research questions have been of fundamental importance with global implications. No well-done study is insignificant and even negative findings are worthy of dissemination. Level 2. Developing institutional approval. Most hospitals have an ethics committee or institutional review board in which research projects are submitted for approval. It is important that you follow your institutional guidelines in order to obtain authorization for you to do research on your floor or dialysis unit. Free-standing dialysis units are not exempt from IRB oversight so that if you wish to conduct nursing research in a free standing unit, first investigate your human subject protection obligations. Submitting for institutional approval is only done after you have developed a strong research design. The study does not have to be com- NEPHROLOGY NURSING JOURNAL November-December 2007 Vol. 34, No

8 plicated; just have the assurances of valid research principles, protection of human subjects, and protection of research data. Most review boards have a system in place with guidelines for completion of forms, procedure for obtaining appropriate signatures, and deadlines for submission. Contact the ethics committee or review board to obtain the instructions that you will need to submit your study for approval. Following the instructions is very important. Level 3. Developing co-worker buy-in. This may be the hardest task of all, particularly if your unit has never been exposed to nursing research. Before you begin developing a research project for your unit, approval from your unit manager will certainly help as the unit manager can pave the way for your project. Consider the level of knowledge about research of your co-workers. Do they have a preconceived notion that all research is considered guinea pig research? Spend time informing your co-workers of your idea along the way so that you will have their support as you begin the project. Often, one or two nurses teaming up to do the research project help soften the idea of conducting research on the floor. Keep in mind that nurses often consider clinical care a high priority, and a research project might be perceived as interference with their care. Use your unit meetings to keep the staff informed of your research questions, the procedures, and the steps that you are taking to ensure protection of your subjects as well as safeguarding the nursing care on the floor. Take the time to keep your co-workers informed of what is going on, the best way to avoid the subtle unintentional sabotage of your project. Level 4. Developing collaboration with other disciplines. Your first study may not involve other disciplines, but working on a collaborative, interdisciplinary team helps bring different perspectives to the table. This process can challenge as well as enlighten your own thoughts. Team work does take extra effort, but working within a team adds richness to the interpretation of your findings. Your team does not have to be large; adding someone from pharmacy on a study can bring additional talents to the table. Your team of three (you, your research mentor, and the pharmacist) can produce outcomes that impact several arenas. Nursing research is not a single adventure, nor does it need to involve only nursing. I have collected a lot of data for a project, but how do I do a statistical analysis such as a t-test? Confused about what to do with your data? This is common. If you have never had a statistical course, or if it has been many years since you had to deal with data analysis, the time is now to find someone to guide you. In a research study, the data analysis plan is made during the development of the project. The question drives the design, the design provides you the structure for data collection, and the type of data determines the type of statistical analysis. But sometimes you are in a situation where you have data and no plan. Use your nurse researcher to help you decide what questions you want your data to answer, and formulate an analysis plan. If you have access to a biostatistician, discuss the data analysis once you have the research question. The statistician will help to ensure statistical power, or help you determine how many subjects are needed to answer the research question or determine a difference. If there are funds available, having a qualified biostatistician to perform analyses may alleviate your learning curve. However, performing statistical tests are the easy part.setting up the data, choosing the appropriate test, and then interpreting the results are the keys to data analysis. Basic statistical functions are available in Excel. Powerful statistical software is available for your computer if you have an interest in learning to analyze data. SPSS (George & Mallery, 2006) is an expensive statistical software package, but is comprehensive, and once learned is easier to use than the statistical functions of Excel. Investing in an expensive software package is not recommended unless you are planning to pursue multiple studies. Explore options with your institution or nurse researcher as there may be access to statistical software without having to purchase expensive software. Where do you learn about all the threats to study validity? Experience has taught these investigators that your best ally is a good research textbook. As you develop your study design, use the textbook to identify each type of threat. Make a list of each of the threats and the strategies that you will follow to prevent, avoid or control for the threat. Creating an outline is the best way to protect the integrity of your outcomes. The tables included with this chapter provide you with the most common types of threats. If you are new to research, discuss each of these threats with an experienced nurse researcher who may have some suggestions for strategies. At first, the threats can appear to be overwhelming, but as you develop a sound study design and sampling strategy, you will have the structure in place to address each of the threats. We did a project evaluation and came away with some interesting findings. Is this publishable in a research journal? We have to go back to the definition of research: Systematic collection of information to increase universal knowledge. Often we have serendipitous findings that we feel are important enough that dissemination to our colleagues is essential. Research journals have expectations of systematic data collection, control of internal and external threats to validity, along with human subjects protection (IRB review). Project evaluation may have all these aspects, but we must proceed with caution as many times, the project did not proceed under the structure of the research process. The data may provide pilot information that will lead to a study to systematically evaluate the question and by all means this pilot data may be publishable. A research journal may not be the place for this type of data, but we can often find a journal that accepts exploratory data analyses. There are also journals that regularly publish program evaluation or process improvement projects. Read journal author guidelines for the 642 NEPHROLOGY NURSING JOURNAL November-December 2007 Vol. 34, No. 6

9 type of manuscripts accepted by the journal and prepare your manuscript according to the guidelines. It is perfectly acceptable to query an editor about your proposed manuscript to see if it will fit within the journal s philosophy and publishing guidelines. Think about your audience and who you want to read your findings. Choose a journal with this type of audience. Summary The objective of this short chapter was to clarify the differences among process improvement, evidence-based practice, and nursing research. The nursing research process was outlined with practical examples to help a novice researcher get involved with nursing research. Nursing research, at whatever level you are able to participate, is a professional obligation to advance our profession. Hulley, S., Cummings, S., Browner, W., Grady, D., Hearst, N., & Newman, T. (2001). Designing clinical research (2nd ed.). Philadelphia: Lippincott Williams & Wilkins. LoBiondo-Wood, G., & Haber, J. (2002). Nursing research: Methods, critical appraisal, and utilization (5th ed.). St. Louis: Mosby, Inc. Malloch, K., & Porter-O Grady, T. (2006). Introduction to evidence-based practice in nursing and health care. Sudbury, MA: Jones and Bartlett Publishers. Medical Risk Management Associates. (2006). Employing tools to enhance patient safety. Retrieved November 8, 2006 from Melnyk, B.M., & Fineout-Overholt, E. (2005). Evidence-based practice in nursing and healthcare. Philadelphia: Lippincott, Williams & Wilkins. National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. (1979). The Belmont report: Ethical principles and guidelines for the protection of human subjects of research (No. GPO ). Washington, DC: U.S. Department of Health, Education, and Welfare. Polit, D.F., & Beck, C.T. (Eds.). (2004). Nursing research: Principles and methods (7th ed.). Philadelphia: Lippincott Williams & Wilkins. Pyzdek, T. (2003). Six sigma handbook revised and expanded. New York: McGraw- Hill. Sackett, D.L., Rosenberg, W.M., Gray, J.A., Haynes, R.B., & Richardson, W.S. (1996). Evidence-based medicine: What it is and what it isn t. British Medical Journal, 312(7023), Winsett, R.P., Thompson, D.A., Gower, G., Leimig, R., Loyd, M., & Hathaway, D.K. (2007). Implementing randomized assignment in a nursing clinical randomized control trial. Progress in Transplantation, 17(2), Zidel, T. (2006). A lean guide to transforming healthcare: How to implement lean principles in hospitals, medical offices, clinics, and other healthcare organizations. Milwaukee, WI: ASQ Quality Press. References Ackerman, T., & Winsett, R.P. (2002). Ethics and regulation in organ procurement research. Progress in Transplantation, 12(4), American Nurses Credentialing Center. (2005). Magnet instructions and application process manual Washington, DC: American Nurses Association. Burns, N., & Grove, S. (2005). The practice of nursing research: Conduct, critique, and utilization (5th ed.). St. Louis, MO: Elsevier Saunders. Council of Nurse Researchers & Council of Nursing Practice. (1993, April, 1994). Education for participation in nursing research. Retrieved November 10, 2006 from readroom/position/research/rsedu cat.htm Davidoff, F., & Batalden, P. (2005). Toward stronger evidence on quality improvement. Draft publication guidelines: The beginning of a consensus project. Quality and Safety in Health Care, 14, Department of Health and Human Services. (1991). Common rule, 56 federal register (45 CFR part 46 subpart a) (Vol. 56, pp ). George, D., & Mallery, P. (2006). SPSS for windows step-by-step: A simple guide and reference, 14.0 update (7th ed.). Boston: Allyn & Bacon. a nurse LIKE NO OTHER Enjoy more mobility. Greater flexibility. A more dynamic career. For the Nephrology Nurse, Fresenius Medical Care NA presents a world of options and possibilities. Working for the world leader in nephrology care, you'll have the opportunity to be part of a great organization with over 1,600 clinics in cities from coast to coast. No matter where you are in your career, this is an open invitation to discover the difference between a job and real job satisfaction. Fresenius Medical Care North America is an equal opportunity employer. NEPHROLOGY NURSING JOURNAL November-December 2007 Vol. 34, No