Objective: The objective of the work was to test whether differently formulated hypotheses, when using correct statistical methods, and the same data lead to the same results. The work pertains to the interpretation of the prevalence of antiborrelia antibody levels in more or less high risk populations.
Group: A group of 200 blood donors from Prague (100 men, mean age 33.5 years and 100 women mean age 24.4 years) and 71 subjects with a higher risk of contact with ticks infected with Borrelia burgdorferi--forestry workers from the Sumava mountains (39 men, mean age 40.6 years and 32 women, mean age 44.5 years).
Method: In all participants antiborrelia antibodies class IgM and IgG were assessed and using the Western blot method the presence of specific antigens was examined. The blood samples of all probands were analyzed in the national reference laboratory for Lyme borreliosis CEM in the National Institute of Public Health in Prague. To test the hypothesis on the difference of groups as regards the frequency of positive antibody levels the chi square test was used. For testing the hypothesis on the difference of the groups as regards antibodies irrespective of normal laboratory values the non-parametric method--the Mann-Whitney test was used.
Results: The group of Prague blood donors and forestry workers from the Sumava mountains do not differ as to the individual frequency of positive antibody values assessed by the ELISA method in class IgM and IgG but they differ in class IgM assessed by the Western blot method. Both groups differ moreover markedly in the general distribution of antibodies unrelated to normal laboratory values assessed by the ELISA method in class IgM, in class IgG the distribution does not differ.
Conclusion: We provided evidence that when using a different formulation of the hypothesis while using correct statistical methods identical data can be differently interpreted. Thus in the interpretation of the prevalence of antiborrelia antibody levels in healthy populations with a different risk--even if the two groups do not differs as to the frequency of positive findings, there is a significant statistical difference in their distribution.