[Model project on coding accuracy]

Area of Science:

• Health Services Research
• Medical Informatics
• Hospital Administration

Background:

• The German Diagnosis Related Groups (G-DRG) system relies on accurate coding for prospective payment.
• External evaluation of coding accuracy is essential to ensure data integrity and financial stability in hospitals.
• Previous studies have highlighted variability in coding practices, necessitating systematic assessment.

Purpose of the Study:

• To report on the initial experiences with external evaluation of coding accuracy within the German G-DRG system.
• To assess the impact of coding accuracy on case-mix index and DRG assignment.
• To identify areas for improvement in coding practices at the departmental level.

Main Methods:

• Evaluation of 387 randomized inpatient cases from three hospital departments.
• Analysis of 1,648 diagnosis codes and 946 procedure codes against German coding standards.
• Recalculation of DRG grouping and case-mix index before and after external review and correction.

Main Results:

• Coding accuracy varied significantly: primary diagnoses (45.9%-56.7%), secondary diagnoses (25.2%-37.5%), and procedure codes (54.2%-65.7%) required correction or completion.
• A substantial percentage of secondary diagnoses (49.2%-60.5%) and procedure codes (23.1%-39.9%) were deemed not relevant according to German coding standards.
• Post-review, the case-mix index increased by an average of 6.9%, indicating a significant shift in DRG assignment due to improved coding.

Conclusions:

• External coding accuracy evaluation is vital for identifying and rectifying errors within the G-DRG system.
• Coding inaccuracies have direct financial implications, affecting reimbursement under prospective payment systems.
• The methodology employed can serve as a basis for future research into coding quality and its impact on hospital revenue.