ICD coding quality for outpatient cancer diagnoses in SHI claims data

Background: Despite their importance for the morbidity-oriented risk structure compensation scheme, the ICD coding quality for outpatient diagnoses is often criticized. However, the coding quality for outpatient cancer diagnoses remains unclear. In this study, we examined the ICD-10 coding quality for outpatient breast, prostate and colorectal cancer diagnoses in statutory health insurance (SHI) claims data. Methods: The cohort study comprised 34,756 members of the BARMER GEK statutory health insurance fund diagnosed with breast, prostate or colorectal cancer in 2012 and with no respective diagnosis in the previous 4 quarters. Using internal validation approaches, we examined the proportion of diagnoses confirmed by persistence (further diagnoses or death) and/or congruence (EBM, OPS and ATC codes). Results: The proportion of diagnoses confirmed by a further diagnosis or death within one quarter of follow-up was 69.8 % for breast cancer, 73.0 % for prostate cancer and 64.6% for colorectal cancer, increasing to 71.8 %, 74.6 % and 68.0 %, respectively, when considering provision of care using EBM, OPS and ATC codes. By extending the diagnosis-free lead time to 16 quarters and the follow-up time to 4 quarters, the proportion rose to 83.9 %, 85.1 % and 75.8 %, respectively. Results varied by sex, age, the Associations of SHI Physicians and settlement pattern. Conclusions: Because there were no further indications confirming the incident outpatient cancer diagnoses a proportion of diagnoses remain unconfirmed. When conducting studies based on SHI claims data, these limitations in the ICD coding quality for outpatient cancer diagnoses should be taken into account by using internal validation approaches.