If requested in the QRP using the
qrp.qrp_parameters input file, users can output tables in the QRP Reporting Tool to show the distribution of clinical codes that were used to define the exposure(s) and/or outcome(s) of interest. Using parameters in the QRP Report Groups File, users can specify which cohort groups to include in these tables and whether the distribution of index-defining and/or outcome-defining codes should be output. Additionally, users have the option of restricting the code distribution to a user-specified top number of most frequently appearing codes (“top N”).
For each cohort group, the QRP Reporting Tool can output two types of Code Distribution tables:
The first table shows the full distribution of codes used to define the exposure(s) and/or outcome(s), stratified by code category, code type, and encounter care setting such that each reported frequency represents a unique combination of code, code category, code type, and encounter care setting. For drug codes, the “code” column is displayed as the stockpiling group for the codes (users set stockpiling groups in
qrp.StockpilingFile; for more information on stockpiling, see QRP documentation). Code type and encounter care setting are not applicable to drug codes and as such, these columns are displayed as
N/A for drug codes. If multiple codes occur on the same day, all the codes from that day will be compiled together and that compilation will be considered as one code count for the purpose of these tables. For example, if diagnosis code A and diagnosis code B occur on the same day, they will be counted as combination code “diagnosis code A and diagnosis code B” rather than contributing individual counts to code A and code B.
The second table is a non-stratified version of the first code distribution table. As such, encounter care setting is not included in this table and codes occurring on the same day are individually counted in the overall code count for each code. Separate tables are output for index-defining codes and outcome-defining codes such that up to four tables can be output per cohort group.
To improve readability of Code Distribution Tables, users can provide a description of each code using the Code Descriptions File. If not provided, the code itself will be listed in the code description column.