Network Analysis of Chronic Disease Diagnosis Co-documentation in Insurance Claims Among Older Arkansans

Multimorbidity among older adults presents growing clinical and economic challenges, particularly in medically underserved states such as Arkansas. However, it remains unclear whether chronic conditions documented in routine healthcare encounters exhibit structured patterns of diagnosis co-documentation or primarily reflect incidental co-documentation. Understanding these patterns may inform future studies of multimorbidity and healthcare utilization.

To characterize statewide, sex-stratified patterns of claim-level diagnosis co-documentation among Arkansas adults aged ≥65 years using administrative claims data.

We analyzed 2024 IQVIA administrative claims for Arkansas residents aged ≥65 years (1,408,707 claims from 35,253 unique patients). Analyses were conducted at the claim level; therefore, individuals could contribute multiple claims, and results reflect utilization-weighted diagnosis co-documentation rather than patient-level disease prevalence or multimorbidity. Twenty-five chronic condition categories were defined using ICD-10-CM codes. Pairwise diagnosis co-documentation was quantified using Lewontin's standardized disequilibrium coefficient (D'), which compares observed joint documentation with the expectation under statistical independence. Sex-stratified D' matrices, disease connectedness metrics, and hierarchical clustering were used to characterize patterns of diagnosis co-documentation.

Hypertension was the most frequently documented chronic condition among claims for both females (22.2%) and males (23.7%) and exhibited the highest positive connectedness in both sex-stratified analyses. Among females, the strongest diagnosis co-documentation patterns involved hypertension with dyslipidemia, coronary artery disease (CAD), obesity, thyroid disorders, and heart failure (D' = 0.46-0.69). Among males, the strongest diagnosis co-documentation patterns involved hypertension with dyslipidemia (D' = 0.73), obesity (D' = 0.64), CAD (D' = 0.57), and thyroid disorders (D' = 0.56). Females demonstrated relatively greater diagnosis co-documentation among musculoskeletal, endocrine, and mental health conditions, whereas males showed more concentrated diagnosis co-documentation among cardiometabolic and organ dysfunction conditions. Negative D' values identified disease pairs that were infrequently co-documented within the same claims (e.g., autoimmune disease-chronic gastrointestinal disease among males, D' = -1.00), reflecting absent or infrequent claim-level joint documentation.

In this claim-level analysis of Arkansas adults aged ≥65 years represented in the IQVIA administrative claims database, diagnosis co-documentation exhibited structured, sex-specific patterns, with hypertension demonstrating the highest positive connectedness among the 25 chronic conditions. These utilization-weighted findings provide a descriptive basis for hypothesis generation and support future longitudinal, patient-level studies to evaluate temporal disease relationships and the potential role of cardiovascular-renal-metabolic (CVRM) risk management.