Did you not notice where the second study said that the first one you're hanging onto so desperately *was not accurate*, and even if there are some measurements that can approximate potential pressures needed, ...... a larger sample size would be needed to validate it?TedVPAP wrote:You clearly have no understanding of statistics - yet you make silly statements.palerider wrote:However, it is such an incredibly small sample as to be statistically insignificant.TedVPAP wrote:One study showed pressure predictions based on BMI, neck size, and AHI were accurate within +/- 1 cm-h20 for 20 out of 26 patients. That is not COMPLETELY USELESS.
I can only read the paper's abstract so I can't quote their statistical analysis.
However I can perform my own calculation given the data in the abstract.
A sample size of 26, with binomial outcomes of 20 and 6 means that the two methods are equivalent (defined as within +/- 1cm-H2O) with a confidence of 99.5%
It's right up thread where I quoted that bit.