Application of statistical inference techniques in health information management

We have demonstrated that objective comparisons can be made using accepted statistical techniques. We have also shown that you can apply tests which don't meet the basic assumptions and still obtain valid results, in most cases. This robustness of statistics tests is particularly helpful with the type of data and analysis that health information management professionals typically deal with, where exactness of the results is not crucial. You can perform a quick analysis using simple statistical tools and obtain a P value that is fairly close to what it would be if you selected the tests more stringently. The examples of inferential statistics in this article demonstrate how to select tests based on characteristics of the data and how to interpret the results. The kinds of statistical analysis that can be performed in health information management are numerous. Below are some other ideas on how to use inferential statistics in HIM practice. 1. Set up an ordinal scale to evaluate coding accuracy to evaluate coders: Score 1 means the correct code was assigned for the principal diagnosis and only minor errors in coding among secondary diagnoses. Score 2 means the correct code was assigned for the principal diagnosis, but there are omissions or major errors among secondary diagnoses. Score 3 means a minor error in coding the principal diagnosis and only minor errors in secondary diagnoses. Score 4 means a minor error in coding the principal diagnosis and major errors or omissions in secondary diagnoses.(ABSTRACT TRUNCATED AT 250 WORDS)