Johns Hopkins Nursing Evidence-Based Practice

The Johns Hopkins EBP Model includes five steps in the searching for evidence phase:

Step 7: Conduct internal and external search for evidence

Step 8: Appraise the level and quality of each piece of evidence

Step 9: Summarize the individual evidence

Step 10: Synthesize overall strength and quality of evidence

Step 11: Develop recommendations for change based on evidence synthesis

Steps to Effective Literature Searching:

 1: Identify your topic - use your PICO

Identify the major concepts of your PICO question. These concepts will serve as search terms.

Do:	Don't:
write down and examine your PICO question	use every element of your PICO
think of all possible synonyms, acronyms, and plurals for major concepts	search with soft terms like "roles" or "best practice"
write down all the concepts you can think of	search on biasing terms like "outcomes"

2. Identify the resources to search

To find evidence that answers your question you will need to use a database. The Upstate Health Sciences Library provides access to several core databases that will help you to locate articles related to your search. These include: CINAHL, PubMed, Scopus, JBI, TRIP, and many more... 

3. Assemble and sort your search terms

• Create lists of words or phrases which are synonyms or acronyms for the major concepts identified.
• Use words and phrases likely to appear in the title, abstract or full-text of literature you are attempting to retrieve. 
• Use truncation if appropriate. Examine each keyword to see if it can be truncated with an asterisk (*) symbol to retrieve all variant forms after its "root" form, such as plurals. 
• Put quotation marks (" ") around exact phrases

4. Run and refine your search

• Use Boolean operators to combine your search terms
  • Use OR with similar terms in a concept - makes search broader
  • Use AND with opposing concepts - makes search narrower
  • Use NOT to exclude specific terms
• Use database filters to limit to a reasonable set of literature
  • Apply a publication date range of 5-10 years
  • Apply a language filter for the languages you read
  • Use publication type filters to limit to clinical trials, systematic reviews and more

5 . Record and evaluate your findings

• Save your search in a document and/or the database
  • Your search will be reproducible
  • It will save you time later

6. Critically appraise the information

• There are several resources for evaluating evidence. The Johns Hopkins Model provides tools for evaluating research and non-research evidence.

Evidence Appraisal: Three Basic Steps

• Determine type of evidence
• Determine evidence level and evaluate strength of evidence
• Evaluate quality of evidence (high to low)

Levels of Evidence

Evidence incorporates both research and non-research.

When searching for information, you want to select articles or studies with the highest evidence level possible.  

LEVEL 1

Randomized Control Trials

In Randomized Control Trials (RCTs) study subjects are randomly assigned to intervention or control groups.  Each subject has the same probability of being selected for either group.  The intervention group receive a treatment/ intervention. The comparison group receives "usual care," i.e. no intervention. The clinician conducting the study is blinded to which participants will be assigned throughout the trial so results are unbiased. 

Systematic Review of RCTs (with or without Meta-Analysis)

Systematic reviews are a comprehensive review of the existing medical literature meeting a set of eligibility criteria as it pertains to a pre-defined research question.  What this means is that researchers create a systematic, reproducible, search strategy to uncover all related articles.  They then analyze all of the articles related to the question and that meet the criteria for inclusion and summarize the findings.  Researchers then make recommendations for clinical practice based on the strength of the evidence they find.

A meta-analysis systematically synthesizes and merges the findings of single, independent studies, using statistical methods to calculate an overall or "absolute" effect.  All meta-analyses are based on systematic review, but not all systematic reviews become meta-analyses.

Mixed methods Design that includes only a Level 1 Quantitative study

Mixed methods is a research approach whereby researchers collect and analyze both quantitative and qualitative data within the same study,

LEVEL 2

Quasi-Experimental Studies

Quasi-Experimental research tries to demonstrate that a specific intervention causes a particular outcome.  Quasi-experimental studies do not include randomization, however, they may have control or comparison groups. Quasi-experimental studies are often conducted when it is not practical, ethical, or possible to randomize subjects to experimental and control groups. Study designs include pretest-posttest or posttest only with non-equivalent comparison groups, pretest-posttest or posttest only with a single group, and time series with untreated control groups with repeated measures, or repeat treatment with subjects acting as their own control.

Systematic Review of a combination of RCTs and quasi-experimental studies (with or without meta-analyses)

Mixed Methods Design that includes only a Level 2 study

LEVEL 3

Non-Experimental Studies

Non-Experimental research studies natural occurring phenomena without introducing an intervention.   Study designs include exploratory, survey (cross-sectional or longitudinal), and correlational (descriptive, predictive, model testing).

Qualitative Studies

Qualitative research is used when there is very little known on the subject matter.  It studies human phenomena, usually in a naturalistic setting.  Study designs include historical research, grounded theory, ethnography, and phenomenological.

Meta-Synthesis

Meta-Synthesis analyzes and then synthesizes concepts and themes found in multiple qualitative studies.  Meta-synthesis does not try to produce a summary statistic, but rather interprets and translates findings.

Systematic Reviews of a combination of RCTs, quasi-experimental studies, and non-experimental studies (or non-experimental studies only)

LEVEL 4

Opinion of Respected Authorities and/or Nationally Recognized Experts 

Opinions of respected authorities and/or nationally recongnized experts includes clinical practice guidelines, consensus statements, position statements, and regulatory standards

Clinical Practice Guidelines, Consensus Statements, and Position Statements combine research and non-research evidence.  That evidence may come from scientific findings, clinician expertise, and patient preferences.  They are often issued by professional organizations and societies (i.e. American Academy of Pediatrics, American College of Surgeons, American Heart Association) and healthcare organizations (i.e. CDC, WHO, NIH). Clinical practice guidelines, consensus statements, and position statements aim to guide the practitioner about appropriate care for specific conditions.  

Regulatory Standards are issued by accreditation, and regulating agencies including CMS, DNV, Joint Commission, and Agency for Healthcare Quality. There is not one database for regulatory standards and you often have to visit individual websites to obtain them.

LEVEL 5

Literature Review

Literature Reviews are unsystematic narratives that refer to research studies, which support the author's view.  They do not critically appraise, evaluate, or summarize findings.

Case Studies

Case Studies are in-depth narratives of a single patient, group, or unit.

Quality Improvement Programs

Quality Improvement (QI) programs are intended to improve systems and processes.  QI is a cyclical process designed to evaluate workflow and work processes.  QI programs ultimately seek to improve patient care and outcomes through good scientific methods and rigor.

Expert Opinion

Expert Opinion can be written or spoken and is based on extensive personal (practitioner) experience or expertise, organizational experience, or economic evaluation.

 

Johns Hopkins Toolkit Resources for Step 8

• The Johns Hopkins Toolkit provides an Evidence Level and Guide (Appendix D) which outlines three levels of evidence with quality ratings and describes each in a rubric.
• The Research Evidence Appraisal Tool (Appendix E) helps you decide if the evidence is quantitative or qualitative, and how to use that evidence to support your topic. 
• Sometimes you'll find literature that is not primary research. Appendix F walks you through the steps of grading non-research evidence with the Non-Research Evidence Appraisal Tool.

The Synthesis Process - Step One

After searching the databases for studies that represent the highest level of evidence for your clinical question you need to document the results of evidence appraisal in preparation for evidence synthesis. 

The Individual Evidence Summary Tool (Appendix G) provides the EBP with documentation of the sources of evidence used, the year the evidence was published or otherwise communicated, the information gathered from each evidence source that helps the team answer the EBP question, and the level and quality of each source of evidence.

The Synthesis Process - Step Two

Evidence synthesis is best done through group discussion. All team members share their perspectives, and the team uses critical thinking to arrive at a judgment based on consensus during the synthesis process. The synthesis process involves both subjective and objective reasoning by the full EBP team. Through reasoning, the team:

• Reviews the quality appraisal of the individual pieces of evidence
• Assesses and assimilates consistencies in findings
• Evaluates the meaning and relevance of the findings
• Merges findings that may either enhance the team’s knowledge or generate new insights, perspectives, and understandings
• Highlights inconsistencies in findings

When evidence includes multiple studies of Level I and Level II evidence, there is a similar population or setting of interest, and there is consistency across findings, EBP teams can have greater confidence in recommending a practice change. However, with a majority of Level II and Level III evidence, the team should proceed cautiously in making practice changes. In this instance, recommendation(s) typically include completing a pilot before deciding to implement a full-scale change. 

Generally, practice changes are not made on Level IV or Level V evidence alone. Nonetheless, teams have a variety of options for actions that include, but are not limited to: creating awareness campaigns, conducting informational and educational updates, monitoring evidence sources for new information, and designing research studies. 

The quality rating (see Appendix D) is used to appraise both individual quality of evidence and overall quality of evidence.

The Synthesis Process and Recommendations Tool (Appendix H) helps you make sense of the strength of the evidence toward a particular recommendation.

The Synthesis Process - Step Three

Develop recommendations based on evidence synthesis and the selected translation pathway Review the synthesis of findings and determine which of the following four pathways to translation represents the overall strength of the evidence:  

• Strong, compelling evidence, consistent results: Solid indication for a practice change.
• Good and consistent evidence: Consider pilot of change or further investigation.
• Good but conflicting evidence: No indication for practice change; consider further investigation for new evidence or develop a research study.
• Little or no evidence: No indication for practice change; consider further investigation for new evidence, develop a research study, or discontinue the project.

The Synthesis Process and Recommendations Tool (Appendix H) helps you make sense of the strength of the evidence toward a particular recommendation.