Costa 2017 PRECISE-DAPT nomogram after coronary stenting. Truncated linear points: age (0 at <=50, 19 at >=90), caller CrCl mL/min (0 at >=100, 25 at 0), hemoglobin g/dL (0 at >=12, 15 at <=10), WBC x10^3/uL (0 at <=5, 15 at >=20), plus 26 for previous spontaneous bleeding. Integer score 0-100; high_bleeding_risk if score >=25. Not a DAPT duration order. magent does not compute Cockcroft-Gault.

Pregnancy estimated due date from a certain LMP using ACOG Committee Opinion 700 Naegele: edd_days_from_lmp = 280 + (cycle_length_days - 28). Inputs are numeric days_since_lmp (0-294) and cycle_length_days (21-35); those bands are magent clamps. Reports gestational_age_days and days_until_edd (negative if past EDD). Not a dating ultrasound. magent does not interpret imaging. Not a delivery date order.

Hilder 2017 PRESET-Score: five Table 4 items (MAP mm Hg, lactate mmol/L, arterial pH, platelets k/uL, hospital days pre-ECMO) summed 0-15. Returns class I (0-5, 26% ICU mortality), II (6-9, 68%), or III (10-15, 93%) from the derivation cohort. Deterministic published arithmetic with citation; not a cannulation order; magent does not assay labs. Not RESP or SAVE.

Armstrong 2018 PREVAIL Cox linear predictor for overall survival in chemotherapy-naïve metastatic castration-resistant prostate cancer from albumin, ALP versus ULN, bone metastases under 10, hemoglobin, LDH versus ULN, NLR under 2.5, pain 0-1, liver metastases, ln(PSA), months from diagnosis, and enzalutamide versus placebo. Table 1 HRs as ln(HR). Not a 1-/2-/3-year OS percent (no published S0). magent does not read bone scans. Not a dose or treatment order.

Khan 2024 AHA PREVENT base 10-year total CVD percent (ASCVD plus heart failure) for age 30-79 from sex, lipids, treated systolic BP, diabetes, smoking, eGFR, BMI, antihypertensive, and statin. Sex-specific, race-free Table S12 coefficients. Not 30-year, not ASCVD-only, not HF-only, not 2013 PCE, not SCORE2, not a diagnosis or statin order. magent does not assay lipids.

Goodacre 2021 PRIEST: NEWS2 Chart 1 Scale 1 vitals (RR, SpO2 Scale 1 only, oxygen +2, temperature, SBP, HR) + alertness (alert 0; confused_or_not_alert 3) + sex (male 1) + age (16–49→0, 50–65→2, 66–80→3, >80→4) + performance 0–4. Max 29. low_adverse_outcome_risk iff score ≤4. Not NEWS2, not 4C, not an admission order.

Psallidas 2018 Table 2A clinical PROMISE for malignant pleural effusion: chemo 0/4, radiotherapy 0/2, haemoglobin g/dL 0-4, WBC 10^9/L 0/2/4/7/10, CRP mg/L 0/3/5/8/11, ECOG 0-1=0 or 2-4=7, cancer mesothelioma 0 / other 4 / lung 5. Groups A 0-20 <25%, B 21-27, C 28-35, D >35 >=75% 3-month mortality. Not TIMP1. Not a treatment order.

Lipsett 2022 Pneumonia Risk Score for children 3 months to 18 years: logistic probability of radiographic pneumonia from age, triage oxygen saturation, fever, rales, and wheeze, plus the published supplement integer score 1–6. Caller-assigned findings. Not a diagnosis, not a CXR order, and not adult CURB-65.

Benson 1992 PSA density: psa_ng_ml / prostate volume (ml). Volume is length_cm × width_cm × depth_cm × 0.5 (D1×D2×D3×1/2), not π/6. Provide PSA with prostate_volume_ml (1-500) or three dimensions (1-20 cm each), not both. Reports dimensionless psad to 3 decimals and retrospective bands <=0.05, 0.05-0.10, >=0.10. magent does not measure the gland or diagnose cancer. Not a medical device.

Pound 1999 PSA doubling time from two rising PSA values: psadt_months = ln(2) x interval_months / ln(psa2_ng_ml / psa1_ng_ml), 1 decimal. PSA each 0.2-10000 ng/mL; psa2 must exceed psa1; interval_months 0.25-120. Reports the Pound split <=10 vs >10 months (exactly 10.0 is <=10). Two-point first-order kinetics only. magent does not measure PSA. Not a medical device.

Kim 2000 revised natural history model for primary sclerosing cholangitis (no histologic stage). R = 0.03*age_years + 0.54*ln(bilirubin_mg_dL) + 0.54*ln(AST_U_L) + 1.24*variceal_bleeding - 0.84*albumin_g_dL. Natural log. Returns R only; does not look up unpublished survival percents or invent risk bands. Not a transplant listing, biopsy order, PSC diagnosis, or lab assay. Not the original Mayo PSC biopsy model.

Fine 1997 Pneumonia Severity Index (PSI/PORT): Class I if age ≤50 and no neoplastic, liver, CHF, cerebrovascular, or renal disease, no AMS, HR <125, RR <30, SBP >=90, and temperature 35 to <40 C. Otherwise Table 2 points (age, or age-10 if female, plus comorbidity, vital-sign, and laboratory weights) map to classes II-V. Always returns the Table 2 total. Not a medical device and not a site-of-care decision.

Boudreaux 2015 PSS-3 from four caller-assigned flags: depressed_mood (2-week segue, not in the screen), ideation_2_weeks, lifetime_attempt, and attempt_within_6_months. screen_positive if 2-week ideation or a lifetime attempt within 6 months (ED-SAFE / Boudreaux 2016 Am J Prev Med). Depression is not factored into screening results. magent does not interview the patient. Not a diagnosis. Not C-SSRS. Not a medical device.

Pediatric Trauma Score from Tepas 1987: six caller-assigned items (weight, airway, systolic BP, CNS, skeletal, wound), each +2 / +1 / −1, total −6 to +12 (higher is better). Returns score_le_8 because scores ≤8 were associated with significant mortality and trauma-center care. Deterministic published arithmetic with citation; not ISS or RTS; not a triage order. magent does not examine the child.

Turner 2007 pediatric ulcerative colitis activity index (PUCAI): six caller-assigned item points (abdominal pain, rectal bleeding, stool consistency of most stools, stools per 24 hours, nocturnal stools, activity level) using that paper's published integers. magent sums only (max 85) and returns the published activity band (<10 remission, 10-34 mild, 35-64 moderate, ≥65 severe). magent does not examine the child. Not a diagnosis and not the adult Mayo score.

Koren 2002 12-hour PUQE: three caller-assigned integers 1-5 (nausea duration, vomiting episodes, and retching episodes over the last 12 hours as in that paper's table). Sum 3-15. magent does not interview the patient. Koren 2002 published no 3-6 / 7-12 / ≥13 total-score cuts; later 24-hour and trimester-long forms are not scored. Not an antiemetic order and not a medical device.

Haimovich 2020 qCSI (Table 2 bedside points): RR ≤22 → 0, 23–28 → 1, >28 → 2; SpO2 >92 → 0, 89–92 → 2, ≤88 → 5; O2 ≤2 L/min → 0, 3–4 → 4, 5–6 → 5. Max 12. Lowest SpO2 and nasal-cannula flow as assigned by the caller. Not a COVID-19 diagnosis, triage order, or the 12-variable CSI gradient-boosting model. Derivation excluded >6 L/min. magent does not measure vitals.

Rush 2003 QIDS-SR16 from 16 caller-assigned integers 0-3 mapped to nine DSM-IV domains. Sleep is the max of four sleep items; appetite/weight is the max of four appetite and weight items; psychomotor is the max of retardation and agitation; remaining items add as published. Total 0-27. Severity: 0-5 none, 6-10 mild, 11-15 moderate, 16-20 severe, 21-27 very_severe. magent does not interpret items or administer a depression interview. Not a diagnosis, not QIDS-C, not IDS-30, and not PHQ-9.

qSOFA from Seymour 2016: respiratory rate ≥22, caller-assigned altered mentation, and systolic BP ≤100 mm Hg; 1 point each, max 3. Deterministic published score with citation; not a diagnosis of sepsis.

Heart-rate-corrected QT (Bazett or Fridericia). formula is required; magent does not pick one silently. Deterministic published correction with citation.

CIOMS R factor from ALT, ALP, and laboratory ULNs (Bénichou 1990). R = (ALT/ALT_ULN)/(ALP/ALP_ULN), rounded to 1 decimal, then hepatocellular (R ≥ 5), cholestatic (R ≤ 2), or mixed (2 < R < 5). Deterministic published arithmetic with citation. Not a medical device, not a DILI diagnosis or RUCAM score; magent does not assay enzymes. R is the enzyme pattern only.

Palumbo 2015 R-ISS for multiple myeloma. ISS from beta-2 microglobulin and albumin (Greipp 2005). R-ISS I: ISS I, standard-risk iFISH CA, and normal LDH. R-ISS III: ISS III plus high-risk CA (del(17p) and/or t(4;14) and/or t(14;16)) or LDH > ULN. R-ISS II: all other combinations. Not a myeloma diagnosis or treatment choice; not R2-ISS; not Durie-Salmon. magent does not assay beta-2 microglobulin, albumin, LDH, or cytogenetics.

Pérez de la Ossa 2014 Rapid Arterial oCclusion Evaluation (RACE) scale: facial_palsy 0 absent / 1 mild / 2 moderate to severe + arm_motor 0 normal to mild / 1 moderate / 2 severe + leg_motor same + gaze_deviation 0 absent / 1 present + cortical 0-2 (aphasia if right hemiparesis; agnosia if left). Max 9. lvo_suggested true iff score ≥5. Not a stroke diagnosis or transfer order and not a race-based equation. magent does not examine the patient.

Oh 2012 radiologist's score (RADS) on one 1-cm axial chest CT slice. Four quadrants (not eight). Each quadrant worst finding: normal 0, interstitial 1, ground_glass 2, consolidation 3. Score 0-12; high when score > 8. Caller-assigned findings only; magent does not interpret CT. Not a diagnosis of inhalation injury. Not a bronchoscopy substitute. Not Endorf AIS inhalation grade. Not Yamamura bronchial wall thickness.

Rai 1975 original CLL staging (stages 0-IV). Returns rai_stage 0-4 from caller-assigned lymphadenopathy, hepatosplenomegaly (spleen or liver or both), anemia (Hb <11 g/dL), and thrombocytopenia (platelets <100,000/mm3). Highest applicable: thrombocytopenia 4, else anemia 3, else organomegaly 2, else nodes 1, else 0 (lymphocytosis only). Not a CLL diagnosis; magent does not examine the patient or read a CBC. Not modified Rai. Not Binet (Binet C is Hb <10).

Ranson 1974 pancreatitis score from eleven published signs (five at admission, six at 48 hours) as boolean flags. Returns points 0-11 and mortality band 0-2 / 3-4 / 5-6 / 7-11. Deterministic published arithmetic with citation; not an ICU order.

Rahman 2014 RAPID score for pleural infection: urea (<5=0, 5-8=1, >8=2 mmol/L), age (<50=0, 50-70=1, >70=2), non-purulent fluid 1, hospital-acquired 1, albumin <27 g/L 1; total 0-7 with low/medium/high bands. Deterministic published arithmetic with citation; not a drain or surgery decision and not Light's criteria.

Hevesi 2018 hip RAPID: sex (female 0, male 1) + Tonnis 0/1/2 (0, 2, 3 points; grade 3 invalid) + caller-assigned cam (1 if present). Max 5. Bands 0-1 low, 2-3 intermediate, 4-5 high. High-grade (ALAD 3-4) damage percent only at score 0 (10.5) and 5 (88.0). Not Rahman 2014 pleural RAPID. Not a diagnosis or operative plan. magent does not read radiographs or measure alpha angle.

Vertesi 2003 RASP multiplies exclusive source (A), inoculum type (B), and transmission (C) to Z, then total_risk = (1/Z) × volume (D). Table 2: <1/1000 definitely indicated; 1/1000–1/10 000 recommended but optional; 1/10 001–1/100 000 optional but not recommended; >1/100 000 not indicated. Average Canadian community prevalence is baked into the scores. Caller-assigned tokens. Not a PEP order. magent does not interpret the source patient.

Richmond Agitation-Sedation Scale (RASS, Sessler 2002) as assigned by the caller: integer -5 to +4 with the published term and description (Combative through Unarousable). magent does not observe the patient, administer voice or physical stimulation, or titrate sedation. Not a CAM-ICU feature, not a CAM-ICU administration, and not a sedation order or dose.

AABB Standards 28th ed. Standard 5.14 RBC ABO/Rh compatibility from caller-assigned recipient and unit ABO (o, a, b, ab) and RhD (positive, negative). Group O RBCs to O/A/B/AB; A to A/AB; B to B/AB; AB to AB. RhD-negative units to either RhD; RhD-positive units only to RhD-positive recipients. Not a transfusion order. magent does not type blood. Does not encode emergency RhD-positive-to-negative exceptions, plasma, or platelets.

Lee 1999 revised cardiac risk index: six 1-point factors (high-risk surgery, ischemic heart disease, heart failure, cerebrovascular disease, insulin diabetes, creatinine >2.0 mg/dL) and class I–IV. Deterministic published arithmetic with citation; not operative clearance.

Rocha 2019 RCVS2 score from five caller-assigned items: recurrent or single thunderclap headache (+5), intracranial ICA involvement (-2), vasoconstrictive trigger (+3), female sex (+1), subarachnoid hemorrhage (+1). Range -2 to 10. Score >=5 supports RCVS; <=2 against; 3-4 equivocal. magent does not read angiograms. Not a diagnosis and not PACNS treatment advice. Not a medical device.

Campbell 2010 RDOS for patients unable to self-report dyspnea. Eight caller-assigned 0-2 integers (heart_rate, respiratory_rate, restlessness, paradoxical_breathing, accessory_muscle_use, grunting, nasal_flaring, look_of_fear). magent sums only (max 16); does not convert raw heart or respiratory rate or observe the patient. Campbell 2010 published the 0-16 instrument; magent does not apply later cutoff papers. Not a medical device and not an opioid or oxygen order.

Yang 2011 REACH-B 17-point HCC risk (non-cirrhotic REVEAL-HBV). Sex female 0, male 2; age 30-65 (30-34 0 through 60-65 6); ALT U/L <15 0, 15-44 1, >=45 2; HBeAg positive 2; HBV DNA copies/mL <10000 0, 10000-99999 3, 100000-999999 5, >=1000000 4. Returns score 0-17 and published 3/5/10-year percents. Not a diagnosis, not a biopsy, not for cirrhosis. magent does not assay ALT or HBV DNA.

Nguyen 2018 first-day AMI READMITS integer score for 30-day all-cause readmission: creatinine >2 mg/dL (6), elevated BNP >=50 or NT-proBNP >=125 pg/mL (8), 1 point per decade over 18, diabetes (4), female sex (2), no PCI within 24 hours (1), SBP <100 mm Hg (3). Table 4 quintiles with mean predicted risk. Not a medical device, not a PCI or discharge order, and not the full-stay model.

Chang 2017 three-variable rebound hyperbilirubinemia score: +15 if GA <38 weeks, −7 × age at phototherapy start in days, −4 × (caller AAP phototherapy threshold − TSB at termination), +50. Scores <20 correspond to a <4% chance of TSB returning to threshold within 72 hours. magent does not assay bilirubin or compute the AAP nomogram. Not Chang 2019. Not a phototherapy order.

Kardaun 2007 RegiSCAR DRESS scoring letter from ten caller-assigned items (sum -4 to 9). Bands: <2 none, 2-3 possible, 4-5 probable, >=6 definite. Not a DRESS diagnosis, not SCORTEN, not J-SCAR. magent does not count eosinophils or read a biopsy. Biopsy unknown is not offered. Not a medical device.

Reid 1985 four-sign colposcopic index: four caller-assigned 0-2 items (margin, color, vessels, iodine; max 8). band 0-2 low_grade (HPV or CIN I), 3-5 intermediate (CIN I or CIN II), or 6-8 high_grade (CIN II or CIN III). magent does not look through a colposcope and does not recompute iodine from the first three signs. Not a CIN diagnosis or biopsy order.

Álvarez-Twose 2012 REMA: male +1, female −1; serum baseline tryptase <15 μg/L −1, 15–25 0, >25 +2; pruritus, hives or angioedema present −2, absent +1; syncope or presyncope +3. Range −4 to +7. Clonality likely iff score ≥2. Caller assigns sex, sBt, cutaneous signs, and syncope. Not a medical device, not a diagnosis of SM, not WHO SM criteria. magent does not assay tryptase.

REMS from Olsson 2004 Table 1: MAP, heart rate, respiratory rate, SpO2, GCS (RAPS 0–4, not 15−GCS), and age summed to a 0–26 total for nonsurgical adult ED in-hospital mortality. Deterministic published arithmetic with citation; not MEWS, NEWS, or APACHE II; not an ICU-admission order.

Kutikov 2009 R.E.N.A.L. nephrometry. Caller-assigned radius le_4 1 / gt_4_lt_7 2 / ge_7 3 (cm); exophytic ge_50 1 / lt_50 2 / entirely_endophytic 3; nearness ge_7 1 / gt_4_lt_7 2 / le_4 3 (mm); location entirely_polar 1 / crosses_polar 2 / gt_50_midline_or_between 3. Anterior a|p|x and hilar suffix h are descriptors, not points. Sum 4-12. No published complexity bands. Not PADUA. Not STONE nephrolithometry. magent does not read CT. Not a medical device. Not a nephrectomy order.

Tonna 2022 RESCUE-IHCA Table 2 integer score for in-hospital death after adult ECPR. Age 2 + 2 per 10 years after 20. Renal +8. Time: day 0, evening +4, night +13. Illness: medical noncardiac 0, medical cardiac -2, surgical noncardiac -6, surgical cardiac -11. Rhythm: asystole 0, PEA -1, pVT -5, VF -8, palpable pulse -5. Duration +2 per 10 min. Table 2 min -17. Integer score only. Not an ECPR order.

Taylor 2021 RESECT-90 logistic 90-day mortality after lung resection from age, sex, ECOG, percent predicted DLCO, BMI, creatinine (umol/L), anaemia, arrhythmia, laterality, thoracotomy vs VATS, diagnosis, and resected segments. Returns logit and estimated_mortality_percent. Not a medical device, not an operative go/no-go, and not a published cutoff. magent does not assign ECOG or read spirometry.

Schmidt 2014 RESP Table 3 (-22 to 15): age 18-49 0, 50-59 -2, >=60 -3; immunocompromised -2; MV lt_48h +3, h48_to_7d +1, gt_7d 0; one diagnosis (viral/bacterial pneumonia +3, asthma +11, trauma_burn +3, aspiration_pneumonitis +5, other acute +1, nonrespiratory 0); CNS -7; nonpulmonary infection -3; NMB +1; NO -1; bicarbonate -2; arrest -2; PaCO2 >=75 -1; PIP >=42 -1. Class I >=6 92%, II 3-5 76%, III -1 to 2 57%, IV -5 to -2 33%, V <=-6 18%. Not an ECMO order.

Wolf 2024 REST-LGS: four major criteria (+3 each) and four minor (+1 each). Max 16. likely if score > 11, possible if 8-11, unlikely if < 8. Caller-assigned flags only; magent does not read EEG. Screening, not a diagnosis of Lennox-Gastaut syndrome. Not the 2019 CRF count threshold.

Corrected reticulocyte count (retic% × Hct/45), Hillman 1969 maturation-shift reticulocyte index, and optional absolute reticulocyte count. Deterministic published arithmetic with citation; not a marrow diagnosis.

Benza 2019 REVEAL 2.0: start at 6; add published points for etiology (CTD +1, PoPH +3, heritable +2), male >60, renal insufficiency, WHO FC, 6-month hospitalization, SBP <110, HR >96, 6MWD, BNP/NT-proBNP, pericardial effusion, DLCO <40%, mRAP >20, and PVR <5. Returns low (0-6), intermediate (7-8), or high (>=9) 12-month mortality. Caller-assigned; not a PAH diagnosis or dose. Not REVEAL 1.0 or Lite 2.

Revised Geneva score (Le Gal 2006) for suspected PE in the emergency department: age >65 (1), previous DVT or PE (3), surgery or lower-limb fracture within 1 month (2), active malignancy (2), unilateral lower-limb pain (3), hemoptysis (2), heart rate 75-94 (3) or >=95 (5), pain on deep-vein palpation and unilateral edema (4). Max 22. Bands 0-3, 4-10, and 11-22. Not a diagnosis, not simplified Geneva, and not Wells.

Ridker JAMA 2007 Reynolds Risk Score, women Model B only: 10-year CVD percent from age, SBP, hsCRP, total cholesterol, HDL-C, HbA1c if diabetic, current smoking, and parental MI before 60. Natural-log Box formula. Not men (Ridker 2008), not ATP-III, not 2013 PCE, not a diagnosis or statin order. magent does not assay lipids or hsCRP.