Generalizability of an Electronic Health Record (EHR)-network dataset to the United States for cardiovascular disease conditions: Comparison of Cerner real world data with the national inpatient sample
Authors' conclusion
On aggregate, characteristics of hospitalizations for MI, CHF, and stroke using EHR data from one nationwide EHR-derived database, CRWD, appears similar to characteristics of hospitalizations in the nationally representative NIS.
Study context
- Population
- Adults (≥18 years) hospitalised with myocardial infarction (MI), congestive heart failure (CHF) or stroke, 2017–201853 health systems analysed
- Exposure
- Cerner RealWorldData (CRWD) vs National Inpatient Sample (NIS)
- Outcome
- Representativeness of CRWD vs NIS: demographics, comorbidities, procedures, length of stay, in-hospital mortality, hospital type
Methodological detailscomparison of EHR-network data with a national administrative sample
- Comparison of EHR-network data with a national administrative sample
Main finding
As reported in the abstract
Hospitalisation characteristics, CRWD vs NIS
| Group | MI hospitalisations | CHF hospitalisations | Stroke hospitalisations |
|---|---|---|---|
| CRWD | 116,956 | 188,107 | 93,968 |
| NIS | 2,245,300 | 4,310,745 | 1,333,480 |
MI hospitalisations
CRWD116,956
NIS2,245,300
CHF hospitalisations
CRWD188,107
NIS4,310,745
Stroke hospitalisations
CRWD93,968
NIS1,333,480
Observational data: associations, not effects.
Original publication
Generalizability of an EHR-network dataset to the United States for cardiovascular disease conditions: Comparison of Cerner real world data with the national inpatient sample. Am Heart J. Published 15 May 2023. doi:10.
Summarised from the published abstract; every number is checked against it