Case-log corpus · nephrology · e-log archive · nested analysis
156 chronic kidney disease case logs, read at three resolutions: individual case profiles → group central tendency → cluster-level synthesis explaining why sub-groups diverge from the cohort average.
Note on method: unlike a hand-curated 6-case narrative review, this analysis is generated programmatically from structured fields (age, sex, diagnosis text, clinical features, labs) extracted from 156 real student e-log case reports. Layer 3 clusters and "diverges on" statements are computed from real prevalence differences within this dataset, not individually verified clinical narratives. Verify against source logs before any clinical or academic use.
All 156 cases, unfiltered. Search or filter by cluster to zoom into a subset — click any card for the full extracted record.
What does this group of 156 CKD cases share? Where is the central tendency, and how much does the cohort actually vary?
Cases are grouped into clinical subtypes by diagnosis text and feature pattern. For each cluster, the divergence line names the features whose prevalence differs most from the cohort baseline — the actual computed signal, not narrative inference.