Secondary Logo

A guide to critical appraisal of evidence

Fineout-Overholt, Ellen, PhD, RN, FNAP, FAAN

doi: 10.1097/01.CCN.0000554830.12833.2f
Feature
Free

Abstract: Critical appraisal is the assessment of research studies' worth to clinical practice. Critical appraisal—the heart of evidence-based practice—involves four phases: rapid critical appraisal, evaluation, synthesis, and recommendation. This article reviews each phase and provides examples, tips, and caveats to help evidence appraisers successfully determine what is known about a clinical issue. Patient outcomes are improved when clinicians apply a body of evidence to daily practice.

How do nurses assess the quality of clinical research? This article outlines a stepwise approach to critical appraisal of research studies' worth to clinical practice: rapid critical appraisal, evaluation, synthesis, and recommendation. When critical care nurses apply a body of valid, reliable, and applicable evidence to daily practice, patient outcomes are improved.

Ellen Fineout-Overholt is the Mary Coulter Dowdy Distinguished Professor of Nursing at the University of Texas at Tyler School of Nursing, Tyler, Tex.

The author has disclosed no financial relationships related to this article.

Figure

Figure

Critical care nurses can best explain the reasoning for their clinical actions when they understand the worth of the research supporting their practices. In critical appraisal, clinicians assess the worth of research studies to clinical practice. Given that achieving improved patient outcomes is the reason patients enter the healthcare system, nurses must be confident their care techniques will reliably achieve best outcomes.

Nurses must verify that the information supporting their clinical care is valid, reliable, and applicable. Validity of research refers to the quality of research methods used, or how good of a job researchers did conducting a study. Reliability of research means similar outcomes can be achieved when the care techniques of a study are replicated by clinicians. Applicability of research means it was conducted in a similar sample to the patients for whom the findings will be applied. These three criteria determine a study's worth in clinical practice.

Appraising the worth of research requires a standardized approach. This approach applies to both quantitative research (research that deals with counting things and comparing those counts) and qualitative research (research that describes experiences and perceptions). The word critique has a negative connotation. In the past, some clinicians were taught that studies with flaws should be discarded. Today, it is important to consider all valid and reliable research informative to what we understand as best practice. Therefore, the author developed the critical appraisal methodology that enables clinicians to determine quickly which evidence is worth keeping and which must be discarded because of poor validity, reliability, or applicability.

Back to Top | Article Outline

Evidence-based practice process

The evidence-based practice (EBP) process is a seven-step problem-solving approach that begins with data gathering (see Seven steps to EBP). During daily practice, clinicians gather data supporting inquiry into a particular clinical issue (Step 0). The description is then framed as an answerable question (Step 1) using the PICOT question format (Population of interest; Issue of interest or intervention; Comparison to the intervention; desired Outcome; and Time for the outcome to be achieved).1 Consistently using the PICOT format helps ensure that all elements of the clinical issue are covered. Next, clinicians conduct a systematic search to gather data answering the PICOT question (Step 2). Using the PICOT framework, clinicians can systematically search multiple databases to find available studies to help determine the best practice to achieve the desired outcome for their patients. When the systematic search is completed, the work of critical appraisal begins (Step 3). The known group of valid and reliable studies that answers the PICOT question is called the body of evidence and is the foundation for the best practice implementation (Step 4). Next, clinicians evaluate integration of best evidence with clinical expertise and patient preferences and values to determine if the outcomes in the studies are realized in practice (Step 5). Because healthcare is a community of practice, it is important that experiences with evidence implementation be shared, whether the outcome is what was expected or not. This enables critical care nurses concerned with similar care issues to better understand what has been successful and what has not (Step 6).

Back to Top | Article Outline

Critical appraisal of evidence

The first phase of critical appraisal, rapid critical appraisal, begins with determining which studies will be kept in the body of evidence. All valid, reliable, and applicable studies on the topic should be included. This is accomplished using design-specific checklists with key markers of good research. When clinicians determine a study is one they want to keep (a “keeper” study) and that it belongs in the body of evidence, they move on to phase 2, evaluation.2

In the evaluation phase, the keeper studies are put together in a table so that they can be compared as a body of evidence, rather than individual studies. This phase of critical appraisal helps clinicians identify what is already known about a clinical issue. In the third phase, synthesis, certain data that provide a snapshot of a particular aspect of the clinical issue are pulled out of the evaluation table to showcase what is known. These snapshots of information underpin clinicians' decision-making and lead to phase 4, recommendation. A recommendation is a specific statement based on the body of evidence indicating what should be done—best practice. Critical appraisal is not complete without a specific recommendation. Each of the phases is explained in more detail below.

Phase 1: Rapid critical appraisal. Rapid critical appraisal involves using two tools that help clinicians determine if a research study is worthy of keeping in the body of evidence. The first tool, General Appraisal Overview for All Studies (GAO), covers the basics of all research studies (see Elements of the General Appraisal Overview for All Studies). Sometimes, clinicians find gaps in knowledge about certain elements of research studies (for example, sampling or statistics) and need to review some content. Conducting an internet search for resources that explain how to read a research paper, such as an instructional video or step-by-step guide, can be helpful. Finding basic definitions of research methods often helps resolve identified gaps.

To accomplish the GAO, it is best to begin with finding out why the study was conducted and how it answers the PICOT question (for example, does it provide information critical care nurses want to know from the literature). If the study purpose helps answer the PICOT question, then the type of study design is evaluated. The study design is compared with the hierarchy of evidence for the type of PICOT question. The higher the design falls within the hierarchy or levels of evidence, the more confidence nurses can have in its finding, if the study was conducted well.3,4 Next, find out what the researchers wanted to learn from their study. These are called the research questions or hypotheses. Research questions are just what they imply; insufficient information from theories or the literature are available to guide an educated guess, so a question is asked. Hypotheses are reasonable expectations guided by understanding from theory and other research that predicts what will be found when the research is conducted. The research questions or hypotheses provide the purpose of the study.

Next, the sample size is evaluated. Expectations of sample size are present for every study design. As an example, consider as a rule that quantitative study designs operate best when there is a sample size large enough to establish that relationships do not exist by chance. In general, the more participants in a study, the more confidence in the findings. Qualitative designs operate best with fewer people in the sample because these designs represent a deeper dive into the understanding or experience of each person in the study.5 It is always important to describe the sample, as clinicians need to know if the study sample resembles their patients. It is equally important to identify the major variables in the study and how they are defined because this helps clinicians best understand what the study is about.

The final step in the GAO is to consider the analyses that answer the study research questions or confirm the study hypothesis. This is another opportunity for clinicians to learn, as learning about statistics in healthcare education has traditionally focused on conducting statistical tests as opposed to interpreting statistical tests. Understanding what the statistics indicate about the study findings is an imperative of critical appraisal of quantitative evidence.

The second tool is one of the variety of rapid critical appraisal checklists that speak to validity, reliability, and applicability of specific study designs, which are available at varying locations (see Critical appraisal resources). When choosing a checklist to implement with a group of critical care nurses, it is important to verify that the checklist is complete and simple to use. Be sure to check that the checklist has answers to three key questions. The first question is: Are the results of the study valid? Related subquestions should help nurses discern if certain markers of good research design are present within the study. For example, identifying that study participants were randomly assigned to study groups is an essential marker of good research for a randomized controlled trial. Checking these essential markers helps clinicians quickly review a study to check off these important requirements. Clinical judgment is required when the study lacks any of the identified quality markers. Clinicians must discern whether the absence of any of the essential markers negates the usefulness of the study findings.6-9

Table

Table

The second question is: What are the study results? This is answered by reviewing whether the study found what it was expecting to and if those findings were meaningful to clinical practice. Basic knowledge of how to interpret statistics is important for understanding quantitative studies, and basic knowledge of qualitative analysis greatly facilitates understanding those results.6-9

The third question is: Are the results applicable to my patients? Answering this question involves consideration of the feasibility of implementing the study findings into the clinicians' environment as well as any contraindication within the clinicians' patient populations. Consider issues such as organizational politics, financial feasibility, and patient preferences.6-9

When these questions have been answered, clinicians must decide about whether to keep the particular study in the body of evidence. Once the final group of keeper studies is identified, clinicians are ready to move into the phase of critical appraisal.6-9

Phase 2: Evaluation. The goal of evaluation is to determine how studies within the body of evidence agree or disagree by identifying common patterns of information across studies. For example, an evaluator may compare whether the same intervention is used or if the outcomes are measured in the same way across all studies. A useful tool to help clinicians accomplish this is an evaluation table. This table serves two purposes: first, it enables clinicians to extract data from the studies and place the information in one table for easy comparison with other studies; and second, it eliminates the need for further searching through piles of periodicals for the information. (See Bonus Content: Evaluation table headings.) Although the information for each of the columns may not be what clinicians consider as part of their daily work, the information is important for them to understand about the body of evidence so that they can explain the patterns of agreement or disagreement they identify across studies. Further, the in-depth understanding of the body of evidence from the evaluation table helps with discussing the relevant clinical issue to facilitate best practice. Their discussion comes from a place of knowledge and experience, which affords the most confidence. The patterns and in-depth understanding are what lead to the synthesis phase of critical appraisal.

The key to a successful evaluation table is simplicity. Entering data into the table in a simple, consistent manner offers more opportunity for comparing studies.6-9 For example, using abbreviations versus complete sentences in all columns except the final one allows for ease of comparison. An example might be the dependent variable of depression defined as “feelings of severe despondency and dejection” in one study and as “feeling sad and lonely” in another study.10 Because these are two different definitions, they need to be different dependent variables. Clinicians must use their clinical judgment to discern that these different dependent variables require different names and abbreviations and how these further their comparison across studies.

Table

Table

Sample and theoretical or conceptual underpinnings are important to understanding how studies compare. Similar samples and settings across studies increase agreement. Several studies with the same conceptual framework increase the likelihood of common independent variables and dependent variables. The findings of a study are dependent on the analyses conducted. That is why an analysis column is dedicated to recording the kind of analysis used (for example, the name of the statistical analyses for quantitative studies). Only statistics that help answer the clinical question belong in this column. The findings column must have a result for each of the analyses listed; however, in the actual results, not in words. For example, a clinician lists a t-test as a statistic in the analysis column, so a t-value should reflect whether the groups are different as well as probability (P-value or confidence interval) that reflects statistical significance. The explanation for these results would go in the last column that describes worth of the research to practice. This column is much more flexible and contains other information such as the level of evidence, the studies' strengths and limitations, any caveats about the methodology, or other aspects of the study that would be helpful to its use in practice. The final piece of information in this column is a recommendation for how this study would be used in practice. Each of the studies in the body of evidence that addresses the clinical question is placed in one evaluation table to facilitate the ease of comparing across the studies. This comparison sets the stage for synthesis.

Phase 3: Synthesis. In the synthesis phase, clinicians pull out key information from the evaluation table to produce a snapshot of the body of evidence. A table also is used here to feature what is known and help all those viewing the synthesis table to come to the same conclusion. A hypothetical example table included here demonstrates that a music therapy intervention is effective in reducing the outcome of oxygen saturation (SaO2) in six of the eight studies in the body of evidence that evaluated that outcome (see Sample synthesis table: Impact on outcomes). Simply using arrows to indicate effect offers readers a collective view of the agreement across studies that prompts action. Action may be to change practice, affirm current practice, or conduct research to strengthen the body of evidence by collaborating with nurse scientists.

When synthesizing evidence, there are at least two recommended synthesis tables, including the level-of-evidence table and the impact-on-outcomes table for quantitative questions, such as therapy or relevant themes table for “meaning” questions about human experience. (See Bonus Content: Level of evidence for intervention studies: Synthesis of type.) The sample synthesis table also demonstrates that a final column labeled synthesis indicates agreement across the studies. Of the three outcomes, the most reliable for clinicians to see with music therapy is SaO2, with positive results in six out of eight studies. The second most reliable outcome would be reducing increased respiratory rate (RR). Parental engagement has the least support as a reliable outcome, with only two of five studies showing positive results. Synthesis tables make the recommendation clear to all those who are involved in caring for that patient population. Although the two synthesis tables mentioned are a great start, the evidence may require more synthesis tables to adequately explain what is known. These tables are the foundation that supports clinically meaningful recommendations.

Phase 4: Recommendation. Recommendations are definitive statements based on what is known from the body of evidence. For example, with an intervention question, clinicians should be able to discern from the evidence if they will reliably get the desired outcome when they deliver the intervention as it was in the studies. In the sample synthesis table, the recommendation would be to implement the music therapy intervention across all settings with the population, and measure SaO2 and RR, with the expectation that both would be optimally improved with the intervention. When the synthesis demonstrates that studies consistently verify an outcome occurs as a result of an intervention, however that intervention is not currently practiced, care is not best practice. Therefore, a firm recommendation to deliver the intervention and measure the appropriate outcomes must be made, which concludes critical appraisal of the evidence.

A recommendation that is off limits is conducting more research, as this is not the focus of clinicians' critical appraisal. In the case of insufficient evidence to make a recommendation for practice change, the recommendation would be to continue current practice and monitor outcomes and processes until there are more reliable studies to be added to the body of evidence. Researchers who use the critical appraisal process may indeed identify gaps in knowledge, research methods, or analyses, for example, that they then recommend studies that would fill in the identified gaps. In this way, clinicians and nurse scientists work together to build relevant, efficient bodies of evidence that guide clinical practice.

Back to Top | Article Outline

Evidence into action

Critical appraisal helps clinicians understand the literature so they can implement it. Critical care nurses have a professional and ethical responsibility to make sure their care is based on a solid foundation of available evidence that is carefully appraised using the phases outlined here. Critical appraisal allows for decision-making based on evidence that demonstrates reliable outcomes. Any other approach to the literature is likely haphazard and may lead to misguided care and unreliable outcomes.11 Evidence translated into practice should have the desired outcomes and their measurement defined from the body of evidence. It is also imperative that all critical care nurses carefully monitor care delivery outcomes to establish that best outcomes are sustained. With the EBP paradigm as the basis for decision-making and the EBP process as the basis for addressing clinical issues, critical care nurses can improve patient, provider, and system outcomes by providing best care.

Back to Top | Article Outline

Seven steps to EBP

Step 0–A spirit of inquiry to notice internal data that indicate an opportunity for positive change.

Step 1– Ask a clinical question using the PICOT question format.

Step 2–Conduct a systematic search to find out what is already known about a clinical issue.

Step 3–Conduct a critical appraisal (rapid critical appraisal, evaluation, synthesis, and recommendation).

Step 4–Implement best practices by blending external evidence with clinician expertise and patient preferences and values.

Step 5–Evaluate evidence implementation to see if study outcomes happened in practice and if the implementation went well.

Step 6–Share project results, good or bad, with others in healthcare.

Adapted from: Steps of the evidence-based practice (EBP) process leading to high-quality healthcare and best patient outcomes. © Melnyk & Fineout-Overholt, 2017. Used with permission.

Back to Top | Article Outline

Critical appraisal resources

A full set of critical appraisal checklists are available in the appendices.

Back to Top | Article Outline

Bonus content!

This article includes supplementary online-exclusive material. Visit the online version of this article at www.nursingcriticalcare.com to access this content.

Back to Top | Article Outline

Bonus content!

This article includes supplementary online-exclusive material. Visit the online version of this article at www.nursingcriticalcare.com to access this content.

Back to Top | Article Outline

REFERENCES

1. Fineout-Overholt E, Stillwell SB. Asking compelling clinical questions. In: Melnyk BM, Fineout-Overholt E, eds. Evidence-Based Practice in Nursing and Healthcare: A Guide to Best Practice. 3rd ed. Philadelphia, PA: Wolters Kluwer; 2015:25–39.
2. O'Mathúna DP, Fineout-Overholt E. Critically appraising quantitative evidence for clinical decision-making. In: Melnyk BM, Fineout-Overholt E, eds. Evidence-Based Practice in Nursing and Healthcare: A Guide to Best Practice. 4th ed. Philadelphia, PA: Wolters Kluwer; 2019:124–188.
3. Fineout-Overholt E, Stillwell SB, Williamson KM, Cox J, Robbins R. Teaching evidence-based practice in academic settings. In: Melnyk BM, Fineout-Overholt E, eds. Evidence-Based Practice in Nursing and Healthcare: A Guide to Best Practice. 3rd ed. Philadelphia, PA: Wolters Kluwer; 2015:330–362.
4. Balakas K, Fineout-Overholt E. Teaching evidence-based practice in clinical settings. In: Melnyk BM, Fineout-Overholt E, eds. Evidence-Based Practice in Nursing and Healthcare: A Guide to Best Practice. 3rd ed. Philadelphia, PA: Wolters Kluwer; 2015:363–375.
5. Polit DF, Beck CT. Essentials of Nursing Research: Appraising Evidence for Nursing Practice. 8th ed. Philadelphia, PA: Wolter Kluwer; 2014.
6. Fineout-Overholt E, Melnyk BM, Stillwell SB, Williamson KM. Evidence-based practice step by step: critical appraisal of the evidence: part I. Am J Nurs. 2010;110(7):47–52.
7. Fineout-Overholt E, Melnyk BM, Stillwell SB, Williamson KM. Evidence-based practice, step by step: critical appraisal of the evidence: part II: digging deeper—examining the “keeper” studies. Am J Nurs. 2010;110(9):41–48.
    8. Fineout-Overholt E, Melnyk BM, Stillwell SB, Williamson KM. Evidence-based practice, step by step: critical appraisal of the evidence: part III. Am J Nurs. 2010;110(11):43–51.
      9. O'Mathúna DP, Fineout-Overholt E. Critically appraising quantitative evidence for clinical decision-making. In: Melnyk BM, Fineout-Overholt E, eds. Evidence-Based Practice in Nursing and Healthcare: A Guide to Best Practice. 4th ed. Philadelphia, PA: Wolters Kluwer; 2019:124–188.
      11. Fineout-Overholt E, Williamson KM, Gallagher-Ford L, Melnyk BM, Stillwell SB. Following the evidence: planning for sustainable change. Am J Nurs. 2011;111(1):54–60.
      Keywords:

      critical appraisal; decision-making; evaluation of research; evidence-based practice; synthesis

      Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved