What is the clinical evidence from this study?

Study design: Cohort. Population: Adult medicine patients (n=39604). Intervention: Electronic multicondition model (discharge e-model) vs. LACE model. Primary outcome: Discrimination (C statistic) for 30-day mortality or readmission (Difference 0.06, 95% CI 0.05-0.07, p=0.02).

What does this research mean for the field?

An electronic multicondition model using EMR data available at discharge predicts 30-day mortality or readmission with a C statistic of 0.71, significantly outperforming existing claims-based models like LACE and CMS. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

May 19, 2015Open Access

Electronic medical record-based multicondition models to predict the risk of 30 day readmission or death among adult medicine patients: validation and comparison to existing models

Key Result

An electronic multicondition model using EMR data available at discharge predicted 30-day mortality or readmission with a C statistic of 0.71, significantly outperforming the LACE model.

Study Design

Type

Cohort (n=39,604)

Multicenter

Yes

Structured PICO

Does an EMR-based multicondition model improve prediction of 30-day readmission or death in adult medicine patients compared to claims-based models?

Population

39,604 consecutive adult patients admitted to internal medicine services at 7 large hospitals belonging to 3 health systems in Dallas/Fort Worth, mean age 61.3, 46.4% male.

Intervention

Electronic medical record (EMR) based multicondition models using data available within 24 hours of admission or at discharge

Comparator

Canadian LACE mortality or readmission model and CMS Hospital Wide Readmission model

Outcome

Composite of readmission or death from any cause within 30 days of dischargecomposite

An EMR-based multicondition model using data available within 24 hours of admission or at discharge accurately predicts 30-day readmission or mortality and outperforms existing claims-based models.

Main Result

Effect estimate: Difference 0.06 (95% CI 0.05-0.07)

Absolute Event Rate: 0.71% vs 0.65%

p-value: p=0.02

Limitations

Social and financial factors across the 7 hospitals may not be fully representative of hospitals and patients in other regions.
All 7 hospitals operated using the same EMR platform (Epic), limiting generalizability to hospitals with different systems or EMR adoption stages.
The model was derived and validated retrospectively, lacking prospective independent validation.
The CMS model used for comparison was developed to evaluate hospital performance using claims data and was not optimized for individual-level risk stratification.

Abstract

BACKGROUND: There is increasing interest in using prediction models to identify patients at risk of readmission or death after hospital discharge, but existing models have significant limitations. Electronic medical record (EMR) based models that can be used to predict risk on multiple disease conditions among a wide range of patient demographics early in the hospitalization are needed. The objective of this study was to evaluate the degree to which EMR-based risk models for 30-day readmission or mortality accurately identify high risk patients and to compare these models with published claims-based models. METHODS: Data were analyzed from all consecutive adult patients admitted to internal medicine services at 7 large hospitals belonging to 3 health systems in Dallas/Fort Worth between November 2009 and October 2010 and split randomly into derivation and validation cohorts. Performance of the model was evaluated against the Canadian LACE mortality or readmission model and the Centers for Medicare and Medicaid Services (CMS) Hospital Wide Readmission model. RESULTS: Among the 39,604 adults hospitalized for a broad range of medical reasons, 2.8% of patients died, 12.7% were readmitted, and 14.7% were readmitted or died within 30 days after discharge. The electronic multicondition models for the composite outcome of 30-day mortality or readmission had good discrimination using data available within 24 h of admission (C statistic 0.69; 95% CI, 0.68-0.70), or at discharge (0.71; 95% CI, 0.70-0.72), and were significantly better than the LACE model (0.65; 95% CI, 0.64-0.66; P =0.02) with significant NRI (0.16) and IDI (0.039, 95% CI, 0.035-0.044). The electronic multicondition model for 30-day readmission alone had good discrimination using data available within 24 h of admission (C statistic 0.66; 95% CI, 0.65-0.67) or at discharge (0.68; 95% CI, 0.67-0.69), and performed significantly better than the CMS model (0.61; 95% CI, 0.59-0.62; P < 0.01) with significant NRI (0.20) and IDI (0.037, 95% CI, 0.033-0.041). CONCLUSIONS: A new electronic multicondition model based on information derived from the EMR predicted mortality and readmission at 30 days, and was superior to previously published claims-based models.

Bookmark

View Full Paper