The_FD

Study summary
The aim of the study is to evaluate the reliability of a total test of cranial density by catching with eight fingers. Two students in 5th year at the Swiss school of osteopathy tested 30 unknown patients. Prior to the two cranian tests (the “total test of cranial density” or the “test of density according to 4 various cranial stages”) we made a gauging between the two students to obtain a better agreement between them.
Each result found, either rigid (absence of rebound) is noted by a plus (+) as positive, or not rigid (rebound) is noted by a minus (-) as negative.. The results were then analyzed using three statistical methods : the Kappa of Cohen Coefficient, the test of χ2 from independence and the test of Mac Nemar for paired series.
We obtain, using the Kappa of Cohen Coefficient test, mediocre results for the first and third “quadrant” when referring to the Landis and Koch classification, bad results for the second “quadrant” and very bad results for the last “quadrant”.
When referring to the smallest risk in the Kappa of Cohen Coefficient test, we set at 5%, we have a 30.12% and 41.27% result for the second and fourth “quadrant” which is to say that the total test of cranial density would be weakly reliable. It is the same for the first “quadrant” which result is 7.42%. At the opposite, we obtain reliable results in the third “quadrant” with 4.33%. So, thanks to the Fleiss “écart-type”, the results become more interesting. For the first and third “quadrant” the tests would be easily reproducible if done to a whole population, not just to a sample.
The χ2 test result is 36.35%, we may say then that the two students had reliable results in the four “quadrant” at risk 5%.
Analyzing the results from the test of Mac Nemar for paired series with an α threshold at 5%, we had very good result of 28.91% and 58.11% for the first and second “quadrant”. Which let us say we have a better reliability for the first two “quadrant”. On the other hand, for the last two “quadrant” there would be a too large risk for errors, the results of 2.15% and 5.74% are too next to the threshold.
The results of this study are unfortunately disappointing, considering that our test is not reliable and reproducible. Though we should not forget the lack of practicing and sensitiveness of the two students and the fact that we tested an asymptomatic sample. For these reasons we came faster to a major disagreement.
For our profession, different elements can be considered. Notably the perspective that we should not give such an importance to our feelings touching people. This study also taught us that, practicing a “preventive” osteopathy, we can increase the number of wrong diagnostics.



Comme je l'ai dit il y a surement qq fautes,
a vous trouver les mots quadrant et écart type
Voilà, j'éspère que jackspeese ne va pas se sentir volé de 30 US dollars. Mais de toute façon on ne joue pas dans la même catégorie lui et moi. C'est sur que vu la qualité de ses trad il pourrait faire payer tt le monde... enfin bon c'est un sujet épineux sur lequel je ne m'étends pas...