Measuring case severity: a novel tool for benchmarking and clinical documentation improvement

Area of Science:

• Health Services Research
• Medical Informatics
• Healthcare Quality Improvement

Background:

• Severity of Illness (SOI) is an All Patients Refined Diagnosis Related Groups (APR DRG) modifier crucial for hospital performance tracking and resource distribution.
• Current SOI systems rely on annual updates of the 3M APR DRG grouper, hindering longitudinal tracking and benchmarking across diverse patient populations.
• Quality Improvement (QI) initiatives, including Clinical Documentation Improvement (CDI) programs, benefit significantly from accurate SOI benchmarking.

Purpose of the Study:

• To develop and evaluate an alternative, grouper-independent Severity of Illness (SOI) scoring system for longitudinal tracking and benchmarking.
• To assess the predictive performance of models utilizing Elixhauser comorbidities and Healthcare Cost and Utilization Project (HCUP) body system indicators for SOI.
• To identify an optimal and efficient model for measuring patient illness severity.

Main Methods:

• Orthogonal polynomial regression models were developed using admission data from U.S. News and World Report Honor Roll facilities (2019-2020).
• Predictors included Elixhauser comorbidities, HCUP body system indicators, and ICD-10-CM complication and comorbidity (CC/MCC) indicators.
• Model performance was evaluated using Receiver Operating Characteristic (ROC) and Precision-Recall (PR) analysis, and prediction accuracy.

Main Results:

• The model incorporating both Elixhauser comorbidities and body system CC/MCC indicators demonstrated the highest accuracy for predicting admission and discharge SOI.
• A simplified model using only body system CC/MCC indicators achieved comparable performance with greater efficiency and less complexity.
• The developed J_Score and J_Score_POA demonstrated practical utility in assessing CDI performance and measuring SOI.

Conclusions:

• J_Scores derived from the body system model effectively evaluate admission and discharge Severity of Illness (SOI).
• This novel scoring system offers a valuable tool for healthcare institutions to benchmark patient illness severity.
• The J_Score system is poised to augment Quality Improvement (QI) efforts and enhance healthcare performance measurement.