Reallocating diabetes-related garbage codes to improve mortality estimates: a case study in Weifang, China
Abstract Effective identification and correction of diabetes mellitus (DM)-related garbage codes (GCs) in mortality surveillance data is crucial for accurately estimating regional DM mortality rates. This study applied a structured, three-step approach—using standard WHO ICD-10 mortality coding rule...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2025-07-01
|
| Series: | Population Health Metrics |
| Online Access: | https://doi.org/10.1186/s12963-025-00399-5 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract Effective identification and correction of diabetes mellitus (DM)-related garbage codes (GCs) in mortality surveillance data is crucial for accurately estimating regional DM mortality rates. This study applied a structured, three-step approach—using standard WHO ICD-10 mortality coding rules, coarsened exact matching (CEMM), and fixed proportion reassignment (FPRM)—to redistribute diabetes-related GCs in Weifang’s mortality data (2010–2022). Using ICD-10 coding rules, we reclassified 29 deaths originally assigned to DM as the underlying cause of death (UCD) to other causes, and reassigned 1,945 records previously not attributed to DM to DM as the UCD. CEMM then reclassified 283 DM-related GC records to DM, followed by FPRM, which reassigned 160 “unknown cause” records to DM. Together, these steps increased the number of DM deaths by 22.82%. Based on the reallocated data, crude DM mortality rates rose from 7.64 to 17.75 per 100,000 between 2010 and 2022, with males experiencing a greater overall increase than females. While no new algorithms were developed, this study demonstrates how internationally recommended coding standards—often neglected in routine subnational settings—can be systematically and rigorously applied to improve DM mortality surveillance. This work highlights operational gaps in local death certification and presents a replicable protocol for enhancing mortality data reliability using existing tools. |
|---|---|
| ISSN: | 1478-7954 |