TABLE 10.3.

Levels of Evidence and SNL Research

Level of Evidence	SNL Research
Level 1: Evidence from systematic review meta-analysis of all relevant control trials or evidence-based practice guidelines based upon RCTs	None identified
Level 2: Evidence obtained from at least one well-designed randomized clinical trial (RCT)	Studies of teaching SNLs and accuracy of nursing diagnoses (Levin, Lunney, & Krainovich-Miller, 2004; Mueller-Staub et al., 2007; Paans et al., 2010, May) and accuracy of documentation (Paans et al., 2010b)
Level 3: Evidence obtained from well controlled clinical trials without randomization (quasi experimental)	Studies of the effect of teaching critical thinking (e.g., Cruz et al., 2009), and of implementing policy (e.g., Thoroddsen & Enfors, 2007)
Level 4: Evidence from nonexperimental studies, for example, case control or cohort studies	Extensive numbers of nonexperimental studies (e.g., del Bueno, 2005; Gordon, 1987; Sparks, 1990). Measurement (Hoskins, 1989; Kim, 1990)
Level 5: Evidence from systematic reviews of descriptive/qualitative studies	Epidemiological studies on occurrence or frequency (e.g., Schroeder, 1990); testing and refinement, and some systematic reviews related to generation of diagnoses
Level 6: Evidence from single descriptive/qualitative studies	Many studies that focused on populations or groups to identify high frequency or commonly occurring nursing diagnoses. The literature continues to report these studies: Flanagan and Jones (2009), Jeffries, Cox, et al. (2010, in press), Gordon, 1987, Gordon and Sweeney (1979) Fehring’s validation model (1987) to estimate content and construct validity of the concepts (Whitley, 1996)
Level 7: Evidence from opinion of authority or experts	Much of early development used expert opinion (e.g., Gebbie & Lavin, 1975), including Delphi methods, for concept development, testing, and refinement