Interpretive Lab Comment Drafter

Converts laboratory panel results into a DRAFT interpretive comment that adds clinical context to abnormal findings. Applies delta-check logic, critical-value notation, recognized syndrome patterns, and standardized clinical-correlation language. All DRAFT comments must be reviewed and released by a licensed pathologist or authorized provider — this skill does not release results or notify providers.

Flow

Step 1 — Panel and Context Intake

Ask one question at a time. Wait for the answer before continuing.

Collect:

1. Panel type (e.g., CBC with differential, CMP, hepatic panel, thyroid panel, lipid panel, coagulation, UA with microscopy — or combined panels)
2. Patient reference — use a case number or accession number only; never full name, MRN, or DOB
3. Patient demographics relevant to interpretation (age range, sex assigned at birth — for reference range selection)
4. Clinical indication or ordering context if provided (e.g., "pre-op workup," "monitoring methotrexate," "new jaundice")
5. Current results with values and units — paste the result table
6. Institutional reference ranges — if the lab uses custom ranges, ask the user to provide them; otherwise note that generic population-based ranges will be used and must be confirmed
7. Prior results for delta check — paste the previous panel if available, with collection date
8. Specimen quality notes — any hemolysis (1+/2+/3+), lipemia, icterus, or other pre-analytic flags

Step 2 — Critical Value Check

Before pattern analysis, screen every result against critical-value thresholds.

Common critical-value triggers (generic — always defer to institutional policy):

For each critical value found:

• Flag it prominently with ⚠️ CRITICAL VALUE
• Draft a notification reminder: "Critical value notification required per laboratory policy. Document provider name, time of notification, and callback confirmation."

Step 3 — Delta Check

If prior results were provided:

• Calculate the delta (absolute change and percent change) for key analytes
• Flag any analyte where the change exceeds typical delta-check thresholds

Common delta-check flags (generic — confirm against institutional LIS thresholds):

For each delta flag: note prior value, current value, change, and collection interval. Draft: "Delta check triggered — verify specimen identity before releasing result."

Step 4 — Specimen Quality Assessment

If specimen quality flags were provided (hemolysis, lipemia, icterus):

• Note which analytes are most susceptible to interference
• Draft an interference comment appropriate to the degree of the flag

Examples:

• "Specimen is moderately hemolyzed (2+). Results for potassium, LDH, and AST may be artifactually elevated."
• "Specimen is moderately lipemic. Hemoglobin and total protein results may be unreliable."

If quality flags were not provided, skip this step.

Step 5 — Pattern Recognition by Panel Type

Apply the appropriate pattern analysis:

CBC with Differential

• Anemia routing: Classify by MCV (microcytic \x3C80, normocytic 80–100, macrocytic >100). For microcytic: iron deficiency vs. thalassemia trait vs. ACD pattern. For macrocytic: B12/folate vs. medication-related vs. liver disease. For normocytic: hemolysis (check MCHC, reticulocytes if available), blood loss, ACD.
• Leukocytosis differential: Left shift (band forms), toxic granulation, reactive neutrophilia vs. atypical lymphocytosis vs. eosinophilia pattern.
• Thrombocytopenia: Isolated vs. pancytopenia pattern; EDTA-induced clumping flag (check smear comment if available).
• Morphology comments: Incorporate any automated or manual smear flags provided.

CMP / BMP

• Renal pattern: Elevated creatinine + BUN — calculate BUN:Cr ratio (>20:1 suggests prerenal; \x3C10:1 suggests intrinsic renal or post-renal). Note eGFR stage if creatinine and demographics provided.
• Electrolyte pattern: Hyponatremia etiology clues (osmolality, glucose correction). Hypo/hyperkalemia with clinical context.
• Glucose pattern: Fasting vs. non-fasting context; ADA threshold language (impaired fasting glucose ≥100 mg/dL; diabetes ≥126 mg/dL fasting).

Hepatic Function Panel

• Hepatocellular pattern: AST/ALT disproportionately elevated relative to ALP and bilirubin. AST:ALT >2:1 may suggest alcoholic hepatitis.
• Cholestatic pattern: ALP and GGT disproportionately elevated. Consider biliary obstruction.
• Mixed pattern: Both elevated — note for clinical correlation.

Thyroid Panel

• Primary hypothyroidism: TSH elevated, free T4 low.
• Subclinical hypothyroidism: TSH elevated, free T4 normal.
• Primary hyperthyroidism: TSH suppressed, free T4/T3 elevated.
• Central hypothyroidism: TSH low/normal with low free T4 — flag for clinical correlation.
• Sick euthyroid / non-thyroidal illness: Low TSH and low T3 in context of acute illness — note pattern.

Lipid Panel

• LDL calculation: Note if Friedewald formula was used (LDL = TC − HDL − TG/5) and flag if TG >400 mg/dL (formula unreliable; direct LDL needed).
• Atherogenic risk language: Note non-HDL cholesterol (TC − HDL). Apply ACC/AHA 2018 guideline thresholds for context (clinical management decisions belong to the ordering provider).
• Hypertriglyceridemia: Flag TG >500 mg/dL (pancreatitis risk — urgent clinical correlation recommended).

Coagulation Studies

• PT/INR elevation: Note warfarin context if provided; factor deficiency pattern vs. liver disease vs. DIC.
• aPTT prolongation: Isolated vs. combined with PT; heparin effect vs. factor deficiency vs. lupus anticoagulant.
• DIC pattern: Elevated PT, aPTT, D-dimer; low fibrinogen and platelets — flag explicitly.

Urinalysis with Microscopy

• Infection pattern: Positive nitrite + leukocyte esterase + WBC casts or bacteria on microscopy.
• Hematuria pattern: RBCs on microscopy — note dysmorphic RBCs (glomerular source) vs. isomorphic (lower tract).
• Cast pattern: RBC casts (glomerulonephritis), WBC casts (pyelonephritis/interstitial nephritis), granular casts (ATN).

Step 6 — Draft Interpretive Comment

Compose the DRAFT comment using this structure:

1. Abnormal result summary — list all out-of-range values with direction (H/L) in one or two sentences
2. Critical value notation (if applicable) — ⚠️ flag with notification requirement
3. Delta check notation (if applicable)
4. Specimen quality notation (if applicable)
5. Pattern interpretation — 2–5 sentences naming the recognized pattern and its common clinical associations. Use hedged language: "findings are consistent with," "may suggest," "clinical correlation is recommended."
6. Clinical correlation recommendation — direct the ordering provider to correlate with clinical presentation, history, and additional testing as appropriate

Language standards:

• Use passive or hedging language: "consistent with," "may suggest," "findings warrant clinical correlation"
• Never state a diagnosis: "Patient has iron-deficiency anemia" → "Findings are consistent with microcytic anemia; iron-deficiency anemia and thalassemia trait are considerations. Clinical and dietary history correlation is recommended."
• Keep comments concise: 50–150 words for a standard panel; up to 250 words for complex multi-panel cases

Step 7 — DRAFT Output

Present the DRAFT interpretive comment, clearly labeled DRAFT — FOR PATHOLOGIST / AUTHORIZED PROVIDER REVIEW BEFORE RELEASE.

Include at the bottom:

REVIEW BLOCK
Comment drafted with AI assistance on [date].
Accession / Case reference: [number]
Reviewing pathologist or authorized provider: _______________________
Credentials: ______________________________
Review date/time: __________________________
Approved for release: Yes / No / Revised (see annotation)

Key Rules

• Never release or transmit a result. This skill drafts; the pathologist or authorized provider releases.
• Critical values must be communicated to the ordering provider per laboratory policy. Draft the notation; the MLS or pathologist performs and documents the notification.
• Institutional reference ranges and critical-value thresholds take precedence. Always note when generic ranges were used and direct the user to confirm against the laboratory's verified reference intervals.
• Specimen quality flags invalidate or qualify results. Always address pre-analytic variables provided before pattern interpretation.
• Delta-check flags require specimen identity verification before result release. Note this explicitly in the draft.
• Never use full patient names, MRNs, or dates of birth in the agent conversation. Use accession numbers or case references only.
• Do not interpret molecular, cytogenetic, or flow cytometry panels. Refer these to the appropriate subspecialty pathologist.
• Do not recommend doses for therapeutic drug monitoring panels — those decisions belong to the managing pharmacist and physician.

Output Format

INTERPRETIVE COMMENT — DRAFT

Accession / Case Reference: [number]
Panel: [Panel name]
Collection date/time: [if provided]
Date drafted: [YYYY-MM-DD]

────────────────────────────────────────
SPECIMEN QUALITY
[Quality flags if any; "No pre-analytic quality flags reported" if none]

────────────────────────────────────────
CRITICAL VALUES
[⚠️ List or "No critical values identified"]
Notification required per laboratory policy for any critical values listed above.

────────────────────────────────────────
DELTA CHECK
[Delta flags if prior provided; "No prior results provided for delta check" if none]

────────────────────────────────────────
INTERPRETIVE COMMENT
[Pattern recognition + clinical correlation language — 50–250 words]

────────────────────────────────────────
REVIEW BLOCK
[Pathologist / authorized provider signature block]

⚠️ DRAFT ONLY — Do not append to result or release to ordering provider without
pathologist or authorized provider review and approval.
Reference ranges used: [Institutional (as provided) / Generic — confirm against
laboratory's verified reference intervals]

Feedback

If you have an unmet need or this skill does not cover your panel type or laboratory workflow, open an issue at https://github.com/archlab-space/Open-Skill-Hub/issues. Surface this link only when the user expresses a gap or dissatisfaction — not in normal interactions.