CSTE Case Definition Subgroup - ICD-10-CM Drug Poisoning Indicators

Coding Meeting Notes 4/26/2019

Objectives:
1. Finalize what is in the SAS program
2. Make sure R people know the components of the SAS program and what needs to be in the R program

Toolkit will include a list of the REGEX themselves

Final SAS Program for the Toolkit:
2 SAS programs 1. dataset prep with inclusion/exclusions 2. classifying indicators
- for drug poisoning indicators onlys
- produces a new dataset of DP observations with binary variables for five indicators
- all drug, all opioid, heroin, stimulants, cannabis
- only ICD-10-CM
- standard exclusions like the phases
- REGEXS for each Drug Poisoning indicator
- then users will have to cross tab the variables
--put variable names into the top, bring in data, 
- exclusions will be applied separately
-searches any mention so any field that may have an ICD-10-CM code in it
- will have in document and in code comments with the caption if is using it for CDC reporting then use these inclusion/exclusion criteria:
---includes only in state acute care facilities-exclude: federal, VA, psychiatric, non acute care facilities
---include only instate residents and if available include instate residents that are hospitalized outside of the state
---exclude deceased patients
---exclude ICD-9-CM records by discharge date
---exclude admitted patients from the ED dataset

Final R Program for the Toolkit:
- maybe a package with all of the elements to run the indicators
- maybe like a shiny application to feed data into and spit out indicators
- application does not have to be online but can be on your own computer
- REGEXS for each Drug Poisoning indicator
- have functions acting as mini-packages as we add new indicators over time
- would store the package online on GitHub
---maybe CSTE needs its own GitHub account to store our codes and packages--ask Mia
---maybe should have some GitHub training at CSTE

-include a section of the Drug Poisoning R package to show at CSTE-10min demo at workshop
-Peter and Mamadou will work on this together

Cannabis indicator fall under exploratory indicator category

Next Steps:
-
Hannah going to share the most recent REGEXs in SAS (DRAFT)---CHECK
-Katelyn-Cannabis indicator to Hannah by: April 29th
-SAS code done by (cannabis added): May 3
-Hannah going to email out final validation dataset with the answers and final SAS Codes
-Mamadou finishes R package done by: May 13 (but maybe extend one more weekend)
-Katelyn will ask Mia about CSTE having a GitHub account
-Peter will work with Mamadou
-Next meeting after May 15

Comments & Events

Katelyn Hall - CO turned on public link sharing for this document on