What type of study provides the strongest evidence to support a practice change?

Levels of Evidence

Evidence-based practice is a conscientious, problem-solving approach to clinical practice that incorporates the best evidence from well-designed studies, patient values and preferences, and a clinician's expertise in making decisions about a patient's care. Unfortunately, no standard formula exists for how much these factors should be weighed in the clinical decision-making process. However, there are a variety of rating systems and hierarchies of evidence that grade the strength or quality of evidence generated from a research study or report. Being knowledgeable about evidence-based practice and levels of evidence is important to every clinician as clinicians need to be confident about how much emphasis they should place on a study, report, practice alert or clinical practice guideline when making decisions about a patient's care.

Rating System:

The levels of evidence listed here have been developed with the help of nurse experts and other industry resources. We thank those who have contributed to making our system relevant and applicable to determining the levels of evidence that support our CE publications.

Evidence-based information ranges from Level A (the strongest) to Level C (the weakest). In 2013, Level ML, multilevel, was added to identify clinical practice guidelines that contain recommendations based on more than one level of evidence:

LEVEL A: Evidence obtained from:

• Randomized control trials: the classic "gold standard" study design. In RCTs, subjects are randomly selected and randomly assigned to groups to undergo rigorously controlled experimental conditions or interventions.
• Systematic review or meta-analysis of all relevant RCTs. A systematic review is a critical assessment of existing evidence that addresses a focused clinical question, includes a comprehensive literature search, appraises the quality of studies and reports results in a systematic manner. Meta-analysis is a study design that uses statistical techniques to combine and analyze data from many RCTs.
• Clinical practice guidelines: based on systematic reviews of RCTs. Evidence-based clinical practice guidelines provide the strongest level of evidence to guide clinical practice because they are based on rigorous reviews of the best evidence on specific topics.

LEVEL B: Evidence obtained from:

• Well-designed control trials without randomization: In this type of study, random assignment is not used to assign subjects to experimental and control groups. Therefore, this type of research is less strong in internal validity because it can't be assumed the subjects in the study are equal on major demographic and clinical variables at the beginning of the trial. Frequent problems with this type of study include intentional or unintentional bias in sample enrollment; nonblinding, unclear criteria for participant selection; or unreliable or invalid tools.
• Clinical cohort study: an examination of groups of people who have common characteristics or exposure experiences to compare outcomes in those exposed vs. outcomes in those not exposed (e.g., development of heart disease after exposure or nonexposure to 10 years of secondhand smoke).
• Case-controlled study: use of an observational approach in which subjects known to have a disease or outcome are compared with subjects known not to have that disease or outcome. Subjects are matched on characteristics so that they are as similar as possible except for the disease or outcome. Case-control studies are generally designed to estimate the odds (using an odds ratio) of developing the studied condition or disease and can determine if an associated relationship exists between the condition/disease and risk factors.
• Uncontrolled study: studies that do not control participant selection or interventions (e.g., a convenience sample, such as patients on a given unit, may be studied because it's the only group reasonably available).
• Epidemiological study: studies that observe people over a long time to determine risk or likelihood of developing diseases. These studies include retrospective database searches or prospective studies that follow a population over time.
• Qualitative study/quantitative study: descriptive, word-based phenomena, such as symptoms, behaviors, culture and group dynamics. Quantitative studies use statistical methods to establish numerical relationships that are correlational or cause and effect.

LEVEL C: Evidence obtained from:

• Consensus viewpoint and expert opinion: a study that obtains agreement about specific practices from all clinical experts on a review panel. Expert opinion involves obtaining agreement from a majority of clinical experts on a review panel. Note: This level of evidence is used when there are no quantitative or qualitative studies in a particular area.
• Meta-synthesis: a systematic review that synthesizes findings from qualitative studies using an interpretive technique to bring small study findings, such as case studies, to clinical application.

LEVEL ML (multilevel): clinical practice guidelines, recommendations based on evidence obtained from:

• More than one level of evidence as defined in the rating system.

Evidence-based Practice Resources:

• Agency for Healthcare Research and Quality Evidence-based Practice Centers (https://www.ahrq.gov/professionals/clinicians-providers/guidelines-recommendations/index.html)
• The Cochrane Collaboration:
  • Cochrane Reviews (http://www.cochrane.org/what-is-cochrane-evidence)
  • Evidence-based healthcare (http://www.cochrane.org/what-is-cochrane-evidence)
• National Guideline Clearinghouse: (https://www.guideline.gov/)

References for EBP:

Alfaro-LeFevre R. Critical Thinking, Clinical Reasoning, and Clinical Judgment: A Practical Approach. 5th ed. St. Louis, MO: Elsevier-Saunders; 2013.

Ebell MH, Siwek J, Weiss BD, et al. Strength of recommendation taxonomy (SORT): a patient centered approach to grading evidence in the medical literature. Am Fam Physician. 2004;69(3):548-556. http://www.aafp.org/afp/2004/0201/p548.html. Published February 1, 2004. Accessed November 11, 2015.

Evidence-based medicine toolkit. American Academy of Family Physician Web site. http://www.aafp.org/journals/afp/authors/ebm-toolkit.html. Accessed November 11, 2015.

What is evidence based medicine? University of Illinois at Chicago University Library Web site. http://researchguides.uic.edu/ebm. Updated March 7, 2008. Accessed November 11, 2015.

Levels of evidence. Oxford Centre for Evidence-Based Medicine Web site. http://www.cebm.net/oxford-centre-evidence-based-medicine-levels-evidence-march-2009/. Published March 2009. Updated April 15, 2011. Accessed November 11, 2015.

Melnyk BM, Fineout-Overholt E. Evidence-Based Practice in Nursing & Healthcare. A Guide to Best Practice. Philadelphia, PA: Lippincott Williams & Wilkins; 2005.

Newhouse RP, Dearholt SL, Poe SS, Pugh LC, White KM. Johns Hopkins Nursing Evidence-Based Practice Model and Guidelines. Indianapolis, IN: Sigma Theta Tau International; 2007.

Strength of recommendation taxonomy (SORT). American Academy of Family Physicians Web site. http://www.aafp.org/dam/AAFP/documents/journals/afp/sortdef07.pdf. Accessed November 11, 2015.

Understanding research study designs. University of Minnesota Bio-Medical Library Web site. http://www.biomed.lib.umn.edu/guides/understanding-research-study-designs. Accessed November 11, 2015.