Catalyst Advanced Large Database Analysis Research Workshop

- Longitudinal analysis with NIS: TRENDWT files ,ICD-9/10 codes adjustments, how to use a standard population to control for temporal changes across the years, how to account for inflation over the years using the consumer price index, how to generate trend p values, how to compare two trends over the years (interaction term use in regression analysis)

- How to use multiple imputation to account for missing data
- How to use propensity matching to account for confounders
- How to do survival analysis, Kaplan-Myers curves and cox regression analysis
- How to cheat STATA and shorten the databases to make calculations faster without losing accuracy
- How to use do loops to execute a set of programs over multiple years automatically
- How to change hospital charges to hospital costs, calculate hospital volumes and find the most common diagnoses in a study population
- How to append databases: pointers and pitfalls
- How to use Poisson regression and the log conversion regression for LOS and total costs and charges
- How to use the hospital readmission database (NRD): what are the limitations and assumptions in the database, how to write the do loops to calculate readmission, what are the different definitions of readmissions, how to calculate 30-day mortality, and how to calculate time to readmission for time-to-event analysis, how to perform 12-month follow up for individual patients.  

- All the above commands will be in do files that are ready to use. In addition, we will go over the complete do file of a research project based on NRD and that has been recently published in Gastroenterology, the number one GI journal in the USA by impact factor: 

Abougergi MS, Peluso H, Saltzman JR. Thirty-Day Readmission Among Patients With Non-Variceal Upper Gastrointestinal Hemorrhage and Effects on Outcomes. Gastroenterology 2018 ;155:38-46

  • Curriculum: The following topics will be covered during the workshop. A STATA do file is provided for every topic to make execution be as easy as a copy and paste maneuver:

During this workshop, advanced topics in statistics and NIS, NEDS and NRD are presented. Therefore, having attended the Catalyst Large Database Analysis Research Workshop is a pre-requisite for registeration

  • The workshop registration fee includes the following:  

                      - Access to all the lectures for 2 days

                      - Access to all the hands-on sessions for 2 days

                      - 16 AMA category 1 CME credits

                      - The complete STATA program (do file) of an already published article using NRD

                      - Free access to STATA for the duration of the workshop

                      - Free access to the NRD databases for the duration of the workshop

                      - A USB flash drive that includes all lectures, hands-on sessions handouts and extra material                                             as well as ready to use STATA do files for the commands used

                      - A binder with all the lecture handouts and hands-on sessions handouts for note taking

                      - An attendance certificate