Table 3 Algorithms implemented in VIADS with examples of their usage
Filter | Definition | Usage example |
|---|---|---|
NC (node counts) | NC = usage frequency of a node (ICD code or MeSH term) | Displaying a summary view: the most frequently used MeSH terms and their ancestors in 2011 |
CC (class counts) | CCnode = NCdescendant1 + NCdescendant2 + NCdescendant 3… | Displaying a summary view: the most frequently used ICD9-CM codes in 2011 in a selected institution |
Ratio | Ratio = CCchild node/ CCparent node | Identifying the largest MeSH contributors to upper-level MeSH terms and their ancestors in 2011 |
Top nodes | Top NC nodes (numbers) Top CC nodes (numbers) Top NC% (percentages) Top CC% (percentages) | Displaying the top 50 or (top 5%) of ICD9-CM codes that have the highest NC (or CC) in 2018 in a selected institution. This algorithm can show the most important nodes in a date set |
Systematic comparison (data set1 vs data set2) | CCnode1 vs CCnode1; CCnode2 vs CCnode2; CCnode3 vs CCnode3; … … | Displaying a comparison view: the most significant different ICD9-CM codes between pioglitazone(data set1) and rosiglitazone (data set2) groups after systematical comparison |
Combination | NC + Ratio | Displaying a summary view: the most frequently used MeSH terms and the largest MeSH contributors and their ancestors in 2011 |
CC + Ratio | Displaying a summary view: the most frequently used ICD9-CM codes and the largest ICD9-CM contributors and their ancestors in 2011 |