Evidence-Based Integrative Medicine

, Volume 1, Issue 4, pp 225–231 | Cite as

Separation Tests for Early-Phase Complementary and Alternative Medicine Comparative Trials

  • Mikel Aickin


There are some substantial ways in which the current approach to establishing an evidence base for a treatment effect works against complementary and alternative medicine (CAM) research. The standard statistical method in comparative trials is the null hypothesis test, which requires a decision that the ‘no treatment effect’ hypothesis is confirmed, or it is rejected. While this approach is appropriate for evidence-based medicine, the requirement for a decisive result drives sample sizes upward, or leaves small trials subject to being criticised for being underpowered. Employing the fundamental theory of hypothesis tests, as developed by Neyman and Pearson, it is possible to define a ‘separation test’ that avoids this problem, by invoking a different inferential rationale. The purpose of a separation test is to find an indication that further research is justified, or that it is not. This change in strategy can considerably lower the required sample size. Since many CAM comparative trials are in early phases of research, where both budgetary and safety considerations argue for a small sample size, separation tests may be especially of interest to CAM researchers. There is, therefore, an opportunity for evidence-based medicine to generate a new kind of study, in which the purpose is to assess whether it is worthwhile to pursue research on an alternative treatment, rather than to determine whether it has been proved effective.


Fibromyalgia Fibromyalgia Impact Questionnaire Massage Therapy Null Hypothesis Significance Test Separation Test 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.



The authors have provided no information on sources of funding or on conflicts of interest directly relevant to the content of this article.


  1. 1.
    Neyman J, Pearson ES. On the problem of the most efficient tests of statistical hypotheses. Philos Trans R Soc Lond A 1933; 236: 333–80CrossRefGoogle Scholar
  2. 2.
    Alfano A, Taylor AG, Foresman PA, et al. Static magnetic fields for treatment of fibromyalgia: a randomized controlled trial. J Altern Complement Med 2001; 7(1): 53–64PubMedCrossRefGoogle Scholar
  3. 3.
    Cohen J. Statistical power analysis for the behavioral sciences. Hillsdate (NJ): Lawrence Erlbaum, 1987Google Scholar
  4. 4.
    Okvat HA, Oz MC, Ting W, et al. Massage therapy for patients undergoing cardiac catheterization. Altern Ther Health Med 2002; 8(3): 68–75PubMedGoogle Scholar
  5. 5.
    Smith MC, Reeder F, Daniel L, et al. Outcomes of touch therapies during bone marrow transplant. Altern Ther Health Med 2003; 9(1): 40–9PubMedGoogle Scholar
  6. 6.
    Wellek S. Testing statistical hypotheses of equivalence. London: Chapman & Hall, 2002CrossRefGoogle Scholar

Copyright information

© Adis Data Information BV 2004

Authors and Affiliations

  1. 1.Program in Integrative MedicineUniversity of ArizonaTucsonUSA
  2. 2.Helfgott Research InstituteNational College of Naturopathic MedicinePortlandUSA

Personalised recommendations