Konsultasyon Log
The consultation tracking log covering both intradepartmental and externally-referred cases, with a source field separating the two streams.
What it is
An operational log maintained in the konsultasyon repo and integrated with the LIS. It carries a source field; the external-consult analyses filter on source = external.
Recorded content:
- Case-level consultation records, both within-department and referred-in.
- Organ, specimen type, and originating institution type for external cases.
- Timing data supporting turnaround from receipt to final report.
- The final issued diagnosis, taken from Kris reports.
- The referring diagnosis, extracted from outside-lab narrative text via Report Text Extraction.
Two governance and analysis rules attach to it. Institution identity is sensitive: analyses aggregate at institution-type level — academic, private lab, community hospital — and never by name. And “disagreement” is not one category: it bundles refinement, reclassification, and true discrepancy, which the chapter is explicit must be kept separate in every table because collapsing them hides the meaningful signal.
Cohort size, date range, and the proportion external versus internal are not recorded. [unverified]
Why it matters for my work
It powers both quality-research consultation studies and, per the ParaPathology notes, workload and communication analyses as well. External consults are described as a distinctive workload — harder cases, incomplete information, slower turnaround, higher variability — which makes this log the evidence base for characterising a part of the department’s work that is otherwise invisible in routine statistics.
How it connects
Konsultasyon — the project page for both consultation workstreams.
Interobserver Agreement — agreement between referring and Memorial diagnosis is the central measure, though the three-way disagreement structure does not fit standard kappa cleanly.
Turnaround Time — receipt-to-report timing for a workload where slowness is a defining feature.
Memorial LIS Extracts — the LIS integration this log depends on.
Report Text Extraction — supplies the referring diagnosis by parsing outside-lab reports.
Open questions
- How large is the cohort and over what period? Not recorded. [unverified]
- The three-category disagreement structure needs a coding scheme that standard kappa cannot represent — is an ordinal or hierarchical approach planned?
- The agreement coding rubric is drafted but the analysis is still at planning stage; what is the blocker?