Clinical Commander

All dossiers
prev.colorectal-screening.v1

Colorectal cancer screening (adult)

general_internal_medicinechronicadultoutpatient
Start encounter →Print handoutPRODUCTION

NEW disease-specific CHILD of prev.cancer-screening.core.v1 (colorectal-only deep-dive). Authored 2026-05-16 as a Bayesian screening-as-data engine, shape copied EXACTLY from the just-deepened parent + pulm.pe.core.v1 §5.5.2 rigor. SCHEMA GAP: EngineDossier has NO first-class sensitivity/specificity/LR/PPV/pre-test-prevalence/NNS field — the modality test-characteristic table, the risk-tier→interval table, and the USMSTF 2020 surveillance-interval-by-histology table are the authoritative payload in _briefs/prev.colorectal-screening.v1.depth.md; encoded narratively here via severity_triggers, phase purpose/advance_when, workup trigger text, and regimen rationale. SCHEMA GAP: DossierDomain has no preventive_medicine value → domain general_internal_medicine (same as parent). DossierSetting has no acute value (used outpatient). EVIDENCE: 10 PubMed-verified PMIDs (all titles/journals/effect-sizes read verbatim from PubMed abstracts 2026-05-16). ≥10 effect-size numbers: NordICC RR 0.82 + risk 0.98%/1.20% + NNT-invite 455 + CRC-death RR 0.90; UKFSS incidence HR 0.77 + mortality HR 0.69 + distal −50% + NNS 191/489; PLCO incidence RR 0.79 + mortality RR 0.74 + distal-mortality RR 0.50; FIT sens 0.79/spec 0.94/LR+ 13.10/LR− 0.23; FIT-DNA CRC sens 92.3% + advanced 42.4% + spec 86.6%; CAPP2 Lynch aspirin CRC HR 0.65 + IRR 0.58. See .depth.md for the full table + DOIs. Cross-dossier engine_ids (real, Glob/Grep-confirmed on disk): prev.cancer-screening.core.v1 (parent), gi.lgib.core.v1 (symptomatic bleed pivot), gi.crohns.core.v1 + gi.ulcerative-colitis.core.v1 (IBD surveillance) — wired via workups[].branches_to + 3 sibling_differentiation blocks. Registry ids reused (all resolve in clinical-tools-registry.ts PREVENTIVE_TOOLS): workup.colorectal_screening, workup.colonoscopy_diagnostic, workup.genetic_counseling. No CRC-specific calculator/panel exists in the registry — calculators/panels intentionally [] (no invented ids; registry NOT edited per dispatch). Special-population branches (≥6) encoded as triggers + contraindication_rules: Lynch (q1-2y from 20-25 + aspirin chemoprevention), FAP (annual from 10-15), FDR-CRC (start 40 or 10y-before, q5y), IBD surveillance (8y post-pancolitis-dx, q1-3y), elderly/limited-life-expectancy STOP-screening (>85 / <10y / 76-85 individualized), pregnancy (defer non-urgent colonoscopy/CTC), prior-CRC survivor (survivorship not screening). No hand-authored RxCUIs (screening engine — non-pharmacologic). Only drug touchpoint = Lynch aspirin chemoprevention, encoded non_pharm:true as a CAPP2 trial-anchor with NO RxCUI (clinical dose per specialist); bowel-prep PEG referenced narratively, RxCUI omitted + noted (RxNav REST validation deferred — registry/script out of scope this pass). PROMOTED 2026-05-25 PLANNED->INTEGRATED: dedicated seed manifest authored at prisma/seed/manifests/prev.colorectal-screening.v1.ts (defineBatch23ScaffoldManifest, specialtyPack preventive_medicine, terminology projected 1:1 from this dossier — no new codes); all 10 evidence PMIDs PubMed-E-utilities-verified (titles/journals matched 2026-05-25); 3 workups[] confirmed resolving in clinical-tools-registry.ts. NEXT STEPS (deferred): (1) FIT/FIT-DNA result-interpretation atom; (2) USMSTF 2020 surveillance-interval-by-histology atom; (3) PLCOm2012-equivalent CRC risk-tier atom if a registry calculator id is later added. USPSTF 2021 CRC currency re-verified via uspreventiveservicestaskforce.org 2026-05-16 — Grade A 50-75 / B 45-49 / C 76-85 individualized / discontinue >85 NOT superseded 2021→2026.

Entry points (6)

  • history
    Routine preventive/wellness visit — CRC screening due age 45-75 (USPSTF 2021 Grade A 50-75 / B 45-49 PMID 34003218)
    routine_preventive_visit
  • history
    Overdue or never-screened patient identified at any encounter — health-equity recall (USPSTF 2021)
    overdue_or_never_screened
  • history
    FDR with CRC, or known Lynch/FAP — earlier + more frequent surveillance (NCCN 2024; USMSTF 2017)
    family_history_or_syndrome_referral
  • history
    IBD (Crohn colitis / ulcerative colitis) ≥8 yr from dx — dysplasia surveillance colonoscopy (USMSTF)
    ibd_surveillance_due
  • lab_abnormality
    Follow-up of a positive non-colonoscopy test (FIT / FIT-DNA / flex-sig) — apply PPV reasoning, route to diagnostic colonoscopy (USPSTF 2021)
    positive_noncolonoscopy_test_followup
  • history
    Post-polypectomy surveillance interval reached — interval set by index polyp histology (USMSTF 2020 PMID 32044092)
    post_polypectomy_surveillance_due

Required inputs (11)

  • agerequired
    demographic • used at CONTEXT
    Age sets the pre-test prevalence (the Bayesian prior) and the eligibility window: 45-75 screen, 76-85 individualized, >85 discontinue (USPSTF 2021 PMID 34003218)
  • family_hx_crcrequired
    history • used at RISK_STRATIFICATION
    First-degree relative with CRC/advanced adenoma raises the prior and shifts start age/interval (start 40 or 10y before youngest dx, colonoscopy q5y) (USMSTF 2017)
  • hereditary_crc_syndromerequired
    history • used at RED_FLAGS
    Lynch (MMR — CRC lifetime ~50-80%) / FAP (APC near-100%) set a very high prior → earlier + far more frequent colonoscopy, different protocol (NCCN 2024)
  • ibd_status_duration_extentrequired
    history • used at RISK_STRATIFICATION
    IBD colitis duration/extent sets the dysplasia-surveillance prior (surveillance from 8 yr post-pancolitis dx, q1-3y) — NOT the average-risk interval (USMSTF)
  • prior_colonoscopy_resultrequired
    history • used at CONTEXT
    Prior colonoscopy date + polyp histology/size/number set the next interval and the conditional post-test prior (USMSTF 2020 PMID 32044092)
  • life_expectancy_estimaterequired
    history • used at FRAME
    Screening benefit requires ≥10-yr life expectancy; below that lead-time + procedure harm + overdiagnosis exceed benefit → STOP-screening (USPSTF 2021)
  • alarm_gi_symptomsrequired
    symptom • used at RED_FLAGS
    Hematochezia, iron-deficiency anemia, unexplained weight loss, change in bowel habit = NOT screening — exit to diagnostic workup / gi.lgib.core.v1 (test-characteristic context changes entirely)
  • fit_result
    lab • used at BRANCHING_WORKUP
    FIT pooled sens ~0.79 / spec ~0.94, LR+ 13.10; a positive FIT raises post-test probability enough to mandate diagnostic colonoscopy (Lee Ann Intern Med 2014 PMID 24658694)
  • fit_dna_result
    lab • used at BRANCHING_WORKUP
    Multitarget stool DNA CRC sens 92.3% but spec 86.6% — positive → diagnostic colonoscopy; lower spec = more false-positive colonoscopies (Imperiale NEJM 2014 PMID 24645800)
  • pregnancy_status
    history • used at CONTEXT
    Pregnancy defers non-urgent colonoscopy/CTC (radiation) — special-population branch; FIT may be deferred and resumed postpartum
  • prior_crc_survivor
    history • used at CONTEXT
    Prior-CRC survivors follow survivorship surveillance, not average-risk screening — different prior + protocol (route out)

12-phase flow (12)

  1. 1FRAME
    Confirm CRC screening scope: asymptomatic adult age 45-75 (76-85 individualized Grade C; discontinue >85), life expectancy ≥10 yr (below which lead-time + colonoscopy harm + non-advanced-adenoma overdiagnosis > benefit → STOP-screening). Excludes symptomatic workup (hematochezia/iron-def anemia/weight loss → diagnostic, gi.lgib.core.v1) and prior-CRC survivorship surveillance (USPSTF 2021 PMID 34003218)
    inputs: age, life_expectancy_estimate, alarm_gi_symptoms
    advance: Asymptomatic, age-eligible, ≥10-yr life expectancy confirmed
  2. 2ENTRY
    Trigger: routine wellness visit, overdue/never-screened flag, FDR-CRC or hereditary-syndrome referral, IBD surveillance due, post-polypectomy interval reached, or positive non-colonoscopy test follow-up (USPSTF 2021)
    advance: Entry trigger captured
  3. 3CONTEXT
    Establish the pre-test prevalence prior: age, FDR-CRC count + ages of onset, Lynch/FAP status, IBD duration/extent, prior colonoscopy date + polyp histology/size/number, pregnancy, prior-CRC status (USPSTF 2021; USMSTF 2017/2020). This phase SETS the Bayesian prior — modality and interval flow from it
    inputs: age, family_hx_crc, prior_colonoscopy_result, ibd_status_duration_extent, pregnancy_status, prior_crc_survivor
    advance: Pre-test prevalence tier assignable (average vs high-risk)
  4. 4RED_FLAGS
    Very-high-prior / non-screening exits: Lynch (MLH1/MSH2/MSH6/PMS2 — CRC lifetime ~50-80%) and FAP (APC, near-100% CRC) → genetics referral + NCCN intensified colonoscopy protocol + cascade family testing; ALARM symptoms (hematochezia, iron-deficiency anemia, unexplained weight loss, obstructive change in bowel habit) are NOT screening — exit to diagnostic colonoscopy / gi.lgib.core.v1 (NCCN 2024; ACG)
    inputs: hereditary_crc_syndrome, alarm_gi_symptoms
    actions: workup.genetic_counseling
    advance: Hereditary syndrome flagged/excluded AND alarm symptoms excluded (else exit to dx)
  5. 5INITIAL_WORKUP
    Order the age/risk-appropriate screen — the test with its fixed characteristics: colonoscopy q10y (adenoma sens ~95%; NordICC RR 0.82), FIT annual (sens ~0.79 / spec ~0.94, LR+ 13.10), FIT-DNA q1-3y (CRC sens 92.3%, spec 86.6%), CT colonography q5y, or flexible sigmoidoscopy q5y / q10y + annual FIT (UKFSS HR 0.77/0.69). The order IS the "treatment"; document the shared-decision modality choice (USPSTF 2021 PMID 34003218)
    inputs: age
    actions: workup.colorectal_screening
    advance: Indicated screening test ordered with shared-decision modality documented
  6. 6BRANCHING_WORKUP
    Positive-screen Bayesian follow-up — apply PPV: a positive FIT (LR+ ~13 — Lee PMID 24658694), positive FIT-DNA (Imperiale PMID 24645800), or positive flexible sigmoidoscopy raises the post-test probability above the work-up threshold → DIAGNOSTIC COLONOSCOPY. Incomplete/positive CTC → colonoscopy. Colonoscopy is itself the diagnostic gold standard (no further test) (USPSTF 2021; USMSTF 2020)
    inputs: fit_result, fit_dna_result
    actions: workup.colonoscopy_diagnostic
    advance: Every positive non-colonoscopy screen routed to diagnostic colonoscopy
  7. 7DIFFERENTIAL
    Average-risk vs high-risk (Lynch/FAP/FDR-CRC/IBD) pathway, AND true-positive (advanced adenoma / CRC) vs false-positive (FIT cross-reactivity, NSAID/diet, non-advanced tubular adenoma overdiagnosis). The false-positive cascade (unnecessary colonoscopy bleeding/perforation) and overdiagnosis of non-advanced polyps are encoded as harm data, not prose. Screen-positive routes to diagnostic colonoscopy; screen-negative returns to interval (USPSTF 2021; NordICC harm PMID 36214590)
    advance: Each result classified average/high-risk and true/false-positive likelihood assigned
  8. 8RISK_STRATIFICATION
    Risk-tier assignment sets the prior + interval + the action threshold: average-risk 45-75 (colonoscopy q10y / FIT q1y), Lynch (colonoscopy q1-2y from 20-25 — NCCN 2024), FAP (annual sig/colonoscopy from 10-15 — NCCN 2024), FDR-CRC (start 40 or 10y before youngest dx, q5y — USMSTF 2017), IBD (surveillance from 8 yr post-pancolitis dx, q1-3y by dysplasia — USMSTF). Post-polypectomy: USMSTF 2020 histology strata set the next interval (1-2 tubular <10mm → 7-10y; advanced adenoma/≥5/villous/HGD/≥10mm → 3y; piecemeal ≥20mm → 6mo) (Gupta PMID 32044092)
    inputs: family_hx_crc, hereditary_crc_syndrome, ibd_status_duration_extent, prior_colonoscopy_result
    advance: Risk tier + interval + action threshold assigned
  9. 9TREATMENT
    The screen IS the intervention; document the shared-decision modality choice (USPSTF 2021 — all approved modalities reduce CRC mortality; the best test is the one that gets done). Lynch chemoprevention: daily aspirin reduces CRC HR 0.65 ITT (CAPP2 Burn Lancet 2020 PMID 32534647) — discuss in confirmed Lynch. Bowel-prep PEG-electrolyte for any colonoscopy (non-pharm trial-anchor; no hand-authored RxCUI — screening engine)
    inputs: hereditary_crc_syndrome
    advance: Screen ordered, shared decision documented, Lynch chemoprevention considered where applicable
  10. 10DISPOSITION
    All screening is ambulatory; positive screens route to the specific pathway: advanced adenoma / CRC on colonoscopy → GI/oncology (no CRC-staging engine in repo — route narratively); positive FIT/FIT-DNA → diagnostic colonoscopy; alarm hematochezia / symptomatic occult bleed → gi.lgib.core.v1 (NOT screening); confirmed Lynch/FAP → NCCN high-risk protocol + cascade testing; IBD dysplasia → IBD engine (gi.crohns/gi.ulcerative-colitis) for management (USPSTF 2021; NCCN 2024)
    advance: Disposition + specialist routing documented
  11. 11MONITORING
    Interval + recall: colonoscopy q10y average-risk; FIT annual; FIT-DNA q1-3y; CTC q5y; flex-sig q5y (or q10y + annual FIT) (USPSTF 2021 PMID 34003218). Post-polypectomy surveillance per USMSTF 2020 histology strata (Gupta PMID 32044092 / GIE PMID 32044106). Lynch colonoscopy q1-2y; FAP annual; FDR-CRC q5y; IBD surveillance q1-3y from 8 yr post-dx. A positive non-colonoscopy test interrupts the interval → diagnostic colonoscopy then USMSTF interval
    advance: Next-screen / surveillance date set + recall scheduled
  12. 12FOLLOWUP
    STOP-screening logic (the deprescribing-equivalent): discontinue when life expectancy <10 yr or age >85; age 76-85 = individualized decision weighing health, prior-screening history, and patient values (USPSTF 2021 Grade C). Patient education on FIT false-positive rate, FIT-DNA lower specificity (more unnecessary colonoscopies), colonoscopy bleeding/perforation harm, non-advanced-adenoma overdiagnosis, and the value of NOT screening when harm dominates (USPSTF 2021)
    inputs: life_expectancy_estimate, age
    advance: Stop-screening decision OR continued-interval education delivered and documented