Comparison and validity of procedures coded With ICD-9-CM and ICD-10-CA/CCI

Area of Science:

• Health Services Research
• Medical Informatics
• Data Quality Assessment

Background:

• Standardized classification systems like the International Classification of Diseases (ICD) are crucial for health administrative data research.
• Canada's recent adoption of ICD-10 and the Canadian Classification of Interventions (CCI) necessitates evaluating potential coding errors.
• Assessing data validity is imperative following the implementation of new coding systems.

Purpose of the Study:

• To evaluate the accuracy of procedure coding in hospital discharge data.
• To compare the validity of the Canadian Classification of Interventions (CCI) against ICD-9-CM.

Main Methods:

• A chart review of 4008 randomly selected hospital records from four Alberta teaching hospitals was conducted.
• Trained reviewers identified 30 specific procedures, comparing existing CCI codes with newly assigned ICD-9-CM codes.
• Statistical measures including sensitivity, specificity, positive/negative predictive values, and kappa scores were calculated against chart review data.

Main Results:

• Both ICD-9-CM and CCI demonstrated reasonable accuracy for major or invasive procedures.
• Both systems showed limitations in accurately coding less invasive or minor procedures.
• Kappa values indicated similar performance between ICD-9-CM and CCI for 14 procedures.

Conclusions:

• The Canadian Classification of Interventions (CCI) demonstrates comparable validity to ICD-9-CM for procedure coding in hospital discharge data.
• Health services and population health researchers can utilize CCI with confidence similar to ICD-9-CM.
• Further validation is recommended for coding less invasive procedures within both systems.