Less hunting.
More reviewing.

Daily shortcuts

Workflow targets retain your existing configuration. These shortcuts do not select a new facility destination automatically.

After capture

• Assessments: allergy/priority findings, Med Rec discrepancies, outpatient supply, targeted interactions, renal results and coverage.
• Medications: profile, status and captured order. Select a row → Selected RX days supply.
• Changes / coverage: changed inputs and missing evidence. Mark reviewed applies only to the current assessment; findings remain visible.
• Evidence / tools: source text, patient corrections, order review, renal library, all workbenches and the guide.

Assessments refresh locally about every 1.5 seconds when inputs change. Review marks expire after relevant patient, medication, lab, capture, renal-context or tracked-reference changes. This is a local rule engine, not a remote AI or comprehensive interaction service.

Medication reconciliation

1. Capture the ambulatory (outpatient/home) profile and the acute (inpatient) profile for the same patient and encounter. Capture all pages and relevant status sections. Use Assign home / Assign acute to correct source assignment.

2. Confirm each completeness checkbox only after checking the entire list. An ambulatory prescription or fill record does not prove what the patient currently takes; verify actual use against your history sources. Partial captures are labeled as partial comparisons.

3. Review Discrepancies / evidence gaps first. Exact regimens are matched before changed regimens. Multiple remaining same-ingredient candidates stay separate; inspect every candidate and record your determination. Therapeutic-class alternatives are suggestions, never automatic equivalents.

4. Select a row for dose, route, frequency, form, concentration, status and dates alongside both original source records. Missing fields, modified release, PRN differences, held/stopped/pending orders and repeated ingredients remain visible. When full SIG instructions are present, verify administered dose versus product strength and any hold parameters or timing instructions.

5. Verify / correct fixes captured text and retains prior text in the session. Review / decision records a timestamped local decision: intentional change/hold, confirmed continuation, substitution accepted, clarification needed, or another documented plan. It does not change an EHR order. Changed inputs invalidate review marks.

6. Filter ambulatory-only, acute-only, matched, reviewed or unreviewed rows as needed. Copy comparison always includes the entire comparison with completeness and decisions, regardless of the current filter. Days supply opens the ambulatory order when present.

Same-ingredient repeat flags may represent a split regimen, PRN order, duplicate capture or duplicate therapy. Resolve the source evidence rather than assuming a duplicate is an error. The background LSV comparison uses the same matching rules but remains a partial retrieved profile; verify current encounter, status and live chart completeness.

For automatic outpatient extraction, preserve explicit labels, for example: Lantus SoloStar U-100 pen SIG: Inject 20 units nightly Qty: 15 mL. A product strength is never treated as a dispense quantity.

Renal assessment and library

Enter renal inputs accepts verified adult age, sex, height (cm), weight (kg), SCr (mg/dL) and optional indexed eGFR. The result shows CG's weight method and distinguishes CG from indexed/de-indexed eGFR. Accepted SCr values are not silently decimal-shifted.

Select stability, RRT status and the matching regimen context. Six source-linked adult oral levofloxacin branches are included. Metformin screening uses indexed eGFR. Other bundled renal rows remain screening prompts, not complete dose monographs.

The library provides offline search, direct DailyMed product searches, and add/edit/delete for local dose rules. Every local rule stores ingredient, metric, lower-inclusive/upper-exclusive range, exact indication/context, guidance, source and version. Records live in renal-dose-rules.ini under local settings. Editing a bundled label row creates a local copy; avoid unintended overlapping rules.

Dose guidance needs its matching adult context, metric and indication, stable renal function and confirmed no RRT. Dialysis/CRRT and unstable function use separate review paths. Open the existing CRRT tool from the quick menu.

SIG and insulin days supply

Results update while typing. Automatic post-capture calculation requires a source-confirmed active outpatient order, recognizable SIG and explicit quantity. Insulin additionally needs an unambiguous product, concentration, device and quantity unit.

1. Select Insulin pens or vials and the exact preset or custom product.

2. Enter quantity and choose mL or containers.

3. Check concentration, container volume, prime units, in-use days, maximum units/injection and dose increment.

4. Enter the SIG. Variable/sliding-scale/titration directions need documented maximum units/day and injections/day; results are labeled estimates.

5. Compare device-use supply and medication-only arithmetic. Edit the final billing value as needed, then Copy days.

Presets cover Lantus SoloStar/vial; Tresiba U-100/U-200 FlexTouch and U-100 vial; Toujeo SoloStar/Max; Humalog U-100/U-200 KwikPen, Junior KwikPen, and 3 mL/10 mL U-100 vials. Other products use custom label values.

The fixed-schedule simulation accounts for complete administrations, split injections, priming, unusable residual insulin and in-use expiration. It assumes unopened containers stay refrigerated until opened and one successful safety test per injection. Extra priming, pump reservoirs, dilution, non-label storage, early printed expiration and variable real-world schedules are not modeled. Pen dose units are never multiplied by concentration.

Example: five 3 mL U-100 pens at 20 units daily, 2 units priming and a 28-day in-use limit support 65 scheduled days under this model. Medication-only arithmetic gives 75 days. The tool shows both; payer rules are not inferred.

Workbenches and preparation BUD

Evidence / tools → All clinical workbenches contains body size, renal, vancomycin, aminoglycoside, RSI, empiric antibiotics, medication risk, supply, anticoagulation, TLS, order review, capture evidence, lab trends, infusion/MAR and nutrition/electrolyte tools. The quick menu includes conversions, emergency/reversal, DKA, CRRT, exposure prophylaxis and other calculators. Windows fit the monitor and provide native scrollbars for larger content; lists and reports retain their own scrolling.

Tray menu → Preparation BUD Library stores drug, concentration, diluent, container, storage, duration, source and version together. Calculate BUD uses the selected preparation time. The library starts without invented preparation durations. Ctrl+Up remains the quick arithmetic tool.

Tray menu → Replay de-identified capture runs a JSON OCR-line fixture locally. Four synthetic scale fixtures are included in Tests/fixtures. Replay counts missed matches and unexpected rows; it tests parsing of OCR lines, not image-to-text recognition accuracy.

Automation and validation

mCDS Continue, eye-drop submission and fill-date submission remain automatic. Fill-date Automation Setup supports native control identifiers with coordinate fallback. Background findings and dose guidance do not independently rewrite prescription doses.

Tests use synthetic controls and parser fixtures. Real EHR controls, real de-identified capture accuracy, licensed comprehensive interaction coverage and institution-approved local dosing/BUD content still require facility-specific validation or configuration. Saving a local rule does not represent institution approval.