Prospective study of coding practices for cesarean deliveries.

OBJECTIVE: To assess consistency of hospital coding for patients with cesarean delivery-related admissions.
METHODS: Hospital coders from 11 participating institutions received a brief questionnaire and a standardized, inpatient record that was developed for nine hypothetical patients who delivered by cesarean delivery. They were asked to assign a diagnosis-related group (DRG) for each case and to submit the DRG, assigned International Classification of Diseases, 9th Revision (ICD-9) codes, and any relevant Physicians' Current Procedural Terminology Coding System codes used in their coding for each case. These responses were analyzed by mean analysis and analysis of variance tests to evaluate variation in coding practices submitted.
RESULTS: Each participating academic hospital has a level III nursery, takes maternal transports, and has a residency and maternal-fetal medicine staff. Consensus in DRG coding was found in only two thirds of cases. Variation in use of ICD-9 codes existed, with 13.7% of ICD-9 codes assigned by all of the coding departments and 24.2% of ICD-9 codes uniquely used by a single institution. Variation in use of Physicians' Current Procedural Terminology Coding System codes also occurred, with 16.3% of the procedure codes used in the same case by all institutions and 28.6% used by a single institution.
CONCLUSION: Documenting providers, coders, and institutions should exercise caution in the use of DRG codes, procedure codes, and ICD-9 codes for cesarean deliveries. The variability noted reflects the difficulty of the coding process and judgments that need to be made by the coders in assigning the codes. LEVEL OF EVIDENCE: III.