Is Fidelity to Treatment Important?

Yes. Treatment fidelity is considered one of the gold standards of clinical research (Foa & Meadows, 1997). Clearly, if the treatment being tested does not adhere to the standard protocol, then the treatment being examined is not the standard treatment; the study will have poor internal validity and the results may not be informative about the actual treatment. Treatment fidelity has been a subject of much controversy (Greenwald, 1996; Rosen, 1999). There is evidence that EMDR is a robust treatment, not affected by some changes to protocol; for example, variations in the eye movement or stimulus component do not appear to interfere with outcome (Renfrey & Spates, 1994). On the other hand, there is evidence that truncating the procedure may result in poor outcomes; for example, an analysis (Shapiro, 1999) of the procedures used in the EMDR phobia studies found that those omitting more than half of the EMDR phases, achieved poor outcomes compared to those using the full protocol. In a methodological meta-analysis, Maxfield and Hyer (2002) found a significant positive correlation between pre-post effect size and assessments of fidelity. Specifically those studies with fidelity that was assessed as adequate, tended to have larger effects than those with fidelity that was assessed as variable or poor, or not assessed.