Diagnostic coding combination patterns, CHS-DRG severity stratification, and hospitalization expenditure in pneumonia with coexisting hypertension or diabetes mellitus: a single-center cautionary case study of definitional circularity.

Authors

Tingting Liu, Yan Li, Yueyue Liang, XiaoJie Lü, Ning Xu

Published in

BMC health services research. Jul 04, 2026. Epub Jul 04, 2026.

Abstract

China's Healthcare Security Diagnosis-Related Groups (CHS-DRG) framework anchors inpatient payment to a relative weight (RW) set by each hospitalization's severity stratum. That stratum depends on the complications and comorbidities (CC) and major complications and comorbidities (MCC) identified among recorded secondary diagnoses, so ICD-10 documentation carries direct payment consequences. Evidence on diagnostic coding combination patterns in clinically complex populations, and on the methodological hazards of analyzing an exposure defined by grouper logic, remains limited.
This single-center retrospective study used medical record front page and CHS-DRG grouping data (locally implemented national grouper, v2.0) for pneumonia hospitalizations discharged from Anhui Chest Hospital in 2025. The hospitalization was the unit of analysis; 131 met pre-specified criteria and were classified by recorded ICD-10 secondary diagnoses into three mutually exclusive coding patterns (basic-disease-only; basic disease with CC; enhanced complication). Because the exposure shares the grouper's own logic, its correspondence with the severity stratum is treated as definitional and the regressions as exploratory. Reporting follows STROBE and RECORD.
Among 131 hospitalizations (mean age 70.4 ± 10.7 years; 91 of 131 male, 69.5%), 15 (11.45%), 20 (15.27%), and 96 (73.28%) fell into the basic-disease-only, basic disease with CC, and enhanced complication patterns. By construction each pattern mapped one-to-one onto its severity stratum. Total hospitalization expenditure, the only non-definitional outcome, rose across patterns (H = 43.57, P < 0.001): medians 5,938.62, 6,769.91, and 25,501.36 yuan (approximately US$827, US$943, and US$3,552); RW followed a parallel gradient (H = 42.96, P < 0.001). In exploratory regression, length of stay and the count of recorded chronic conditions co-varied with MCC entry (adjusted OR 1.234, 95% CI 1.101 to 1.384, and 1.411, 95% CI 1.144 to 1.741, respectively), both contemporaneous and partially coding-dependent rather than independent predictors.
The structure of recorded ICD-10 diagnoses corresponded to CHS-DRG severity stratification and RW by definition rather than through an independent association, while realized expenditure rose in parallel. The study is best read as a cautionary illustration of definitional circularity: rule-conformant ICD-10 coding remains necessary for accurate stratification and equitable payment, but coding-pattern data alone cannot establish independent clinical or economic effects.
Not applicable.

PMID:
42401880
Bibliographic data and abstract were imported from PubMed on 05 Jul 2026.

Read full publication at:
Please sign in to see all details.

Stats

1-terrible, 9-excellent. How would you rate this publication? Sign in in to submit your rating.

Post a comment

You need to be signed in to post comments. You can sign in here.

Comments

There are no comments yet.