Evaluating Consistency of Documented ICD Codes for Hospitalized Patients With Sickle Cell Disease

Abstract

OBJECTIVES Conducting health services research relies on consistent diagnosis code documentation; however, it is unknown if consistent documentation in claims data occurs among patients with sickle cell disease (SCD) and/or trait (SCT). The objective of this study was to examine the consistency of International Classification of Diseases (ICD) code documentation for SCD/SCT and identify coding discrepancies between patients’ hospitalizations. PATIENTS A total of 80 031 hospitalization records across 528 hospitals belonging to 15 380 unique patients who had at least 1 documentation of SCD/SCT and 2 or more hospitalizations during the study period (April 2015–December 2016). METHODS Secondary analysis of patient discharge abstracts in California, Florida, New Jersey, and Pennsylvania. ICD 9 and ICD 10 codes identified patients with SCD/SCT. Variations in documentation consistency across hospitals were examined. RESULTS Only 51% of patients were consistently documented. There were statistically significant differences in whether a patient was or was not consistently documented based on: age, race/ethnicity, sex, insurer, and disease type. Twenty-five percent of hospitalization records were not consistently documented with an SCD code. Hospitalization records, for patients not consistently documented (49%), often included primary admitting diagnoses for conditions associated with SCD. Few hospitals (18%) were above average in consistently documenting SCD/SCT. CONCLUSIONS Not consistent documentation for SCD/SCT occurs with variation among patients and across disease type and hospitals. These findings signal to researchers the importance of thoroughly identifying all hospitalizations when studying populations with chronic disease. Without accurate documentation, research relying on claims data may produce inaccurate findings.