Osteoarthritis (knee / hip / hand — clinical dx → management)
Osteoarthritis (knee/hip/hand incl. erosive) dossier — OA is a CLINICAL diagnosis in the typical patient (age ≥45, activity-related pain, ≤30 min stiffness, no systemic inflammatory features); Kellgren–Lawrence grade is a STRUCTURAL ANCHOR, not a diagnostic or treatment gate; do not over-image (NICE NG226 2022; ACR/AF 2019). Management ladder: core non-pharm (education + structured exercise + weight loss) is FIRST-LINE for every phenotype with quantified effect sizes; topical NSAID strong first-line drug for knee/hand; oral NSAID lowest effective dose with the CKD/CVD/PUD/anticoagulant gating matrix encoded AS DATA in regimen_axes[].contraindication_rules + severity_triggers (CKD-EPI 2021 race-neutral eGFR); duloxetine + IA glucocorticoid conditional adjuncts; glucosamine/chondroitin, IA hyaluronic acid, and opioids encoded as recommend_against (ACR/AF 2019 strong against); arthroplasty referral on a function/refractory threshold (Skou NEJM 2015). §5.5.2: pre-test priors (age/sex/BMI/occupation/prior injury), ACR-criteria-derived LR+/LR− (knee pain + ≥3/6 → sens 95% spec 69% → LR+ ~3.1, LR− ~0.07; + osteophytes → sens 91% spec 86% → LR+ ~6.5), exam-finding LRs (bony enlargement, crepitus, Heberden/Bouchard), conditional dependencies (bony enlargement/crepitus/osteophytes co-vary), T_test/T_treat (image only when it crosses a management threshold — NICE NG226), and MECE vs inflammatory/crystal arthritis with named pivots (morning-stiffness duration, joint distribution, inflammatory markers, constitutional symptoms) — encoded across phase purposes, notes, severity_triggers, sibling_differentiation, and the research bundle. Manifest is BORROWED (prisma/seed/manifests/rheum.gout.core.v1.ts) — no dedicated OA manifest in this shard. RxCUIs RxNav-verified 2026-05-22: naproxen 7258 (corrected from 7646=omeprazole), ibuprofen 5640, celecoxib 140587, duloxetine 72625, acetaminophen 161, methylprednisolone 6902 (IA injectable). Topical diclofenac and triamcinolone acetonide rxcui OMITTED — no in-repo precedent (allowed at INTEGRATED; npm run research:rxnav not wired in shard). No RxCUI invented. Cross-ref siblings msk.gout-acute.core.v1 and msk.knee-pain.core.v1 do not yet exist in this shard — referenced by engine_id for forward routing only (allowed; sibling_differentiation/severity_triggers route by id without requiring the file). rheum.rheumatoid-arthritis.core.v1 and rheum.gout.core.v1 exist. INTEGRATED (not PRODUCTION) to avoid strict rxcui/365-day/LOINC promotion checks while authoring at PRODUCTION depth. last_reconciled 2026-05-16; no 2025/2026 society OA guideline supersedes the ACR/AF 2019 + OARSI 2019 + NICE NG226 2022 floor (documented in research bundle).
Entry points (6)
- symptomActivity-related knee pain, worse with use / better with rest (ACR/AF 2019; NICE NG226 2022)activity_related_knee_pain
- symptomGroin / lateral hip pain on weight-bearing, reduced internal rotation (ACR/AF 2019)activity_related_hip_groin_pain
- symptomHand pain + bony enlargement (Heberden DIP / Bouchard PIP / 1st CMC) (ACR/AF 2019)hand_bony_enlargement_pain
- symptomMorning stiffness ≤30 min and post-rest "gelling" (NICE NG226 2022)short_morning_stiffness_gelling
- imagingIncidental Kellgren–Lawrence osteophytes / joint-space narrowing on imaging (ACR/AF 2019 — structural anchor only)incidental_kl_osteophytes
- problem_list"Is this OA or RA / gout?" — differentiation request (Altman 1986 criteria)oa_vs_inflammatory_question
Required inputs (18)
- agerequireddemographic • used at CONTEXTPre-test probability of OA rises steeply with age (≈10% men / 13% women symptomatic knee OA at ≥60); age ≥50 is an ACR clinical-criteria item (Altman 1986 PMID 3741515)
- sexdemographic • used at CONTEXTFemale sex raises hand and knee OA prior; erosive hand OA predominantly peri/postmenopausal women
- bmirequireddemographic • used at CONTEXTObesity is the dominant modifiable risk factor; ≥5–10% weight loss is disease-modifying for knee OA (IDEA Messier JAMA 2013 PMID 24065013)
- occupational_joint_loadhistory • used at CONTEXTKneeling / squatting / heavy-lifting occupations raise knee OA prior; informs counselling
- prior_joint_injuryhistory • used at CONTEXTPrior ACL/meniscal or major joint injury raises post-traumatic OA prior 3–4× in that joint
- morning_stiffness_durationrequiredsymptom • used at CONTEXT≤30 min favours OA; >45–60 min is a pivot toward inflammatory arthritis (RA) — drives differential routing (NICE NG226 2022; Altman 1986)
- joint_distributionrequiredsymptom • used at DIFFERENTIALDIP/1st-CMC/knee/hip favours OA; symmetric MCP/wrist synovitis favours RA; 1st-MTP podagra favours gout — MECE pivot
- constitutional_symptomsrequiredsymptom • used at RED_FLAGSFever / weight loss / fatigue argue against primary OA and toward inflammatory, septic, or malignant cause
- hot_swollen_joint_with_feverrequiredsymptom • used at RED_FLAGSAcute hot swollen joint + fever = septic arthritis until proven otherwise — STAT arthrocentesis (not an OA pathway)
- ckd_stagerequiredhistory • used at CONTEXTeGFR (CKD-EPI 2021, race-neutral) gates oral NSAID: avoid if eGFR <30; caution 30–59 (ACR/AF 2019; OARSI 2019 PMID 31278997)
- cardiovascular_diseaserequiredhistory • used at CONTEXTEstablished CVD / HF / uncontrolled HTN → OARSI recommends NOT using oral NSAID; prefer naproxen if any NSAID needed (PRECISION Nissen NEJM 2016 PMID 27959716)
- peptic_ulcer_or_gi_bleedrequiredhistory • used at CONTEXTPUD / prior GI bleed → COX-2 selective + PPI co-prescription if an NSAID is required (PRECISION PMID 27959716)
- anticoagulant_or_antiplateletrequiredmedication • used at CONTEXTConcurrent anticoagulant/antiplatelet sharply raises NSAID GI bleed risk → prefer topical NSAID or COX-2 + PPI
- esr_crplab • used at INITIAL_WORKUPNormal ESR/CRP lowers posterior probability of active inflammatory arthritis; elevated → re-route to RA/crystal workup (not routinely needed to diagnose OA)
- rf_anti_ccplab • used at INITIAL_WORKUPOnly when inflammatory pattern suspected — positive RF/anti-CCP routes to rheum.rheumatoid-arthritis.core.v1 (not part of typical OA workup)
- synovial_fluid_analysislab • used at INITIAL_WORKUPIf a joint effusion is aspirated: OA fluid is non-inflammatory (WBC <2000, no crystals); rules in crystal/septic if positive
- plain_radiograph_kl_gradeimaging • used at BRANCHING_WORKUPWeight-bearing radiograph for KL grade ONLY if atypical / pre-surgical / diagnostically uncertain — not required for typical OA (NICE NG226 2022)
- function_and_refractorinessrequiredsymptom • used at RISK_STRATIFICATIONFunction-limiting refractory pain despite optimised conservative therapy is the arthroplasty-referral threshold (Skou NEJM 2015 PMID 26488691)
12-phase flow (12)
- 1FRAMEOA is a CLINICAL diagnosis in the typical patient (age ≥45, activity-related pain, ≤30 min stiffness, no systemic inflammatory features) — imaging NOT required; Kellgren–Lawrence grade is a structural anchor, not a diagnostic gate (NICE NG226 2022; ACR/AF 2019 Kolasinski PMID 31908163). Scope: knee, hip, hand (incl. erosive hand OA), with/without inflammatory componentinputs: age, morning_stiffness_durationadvance: Phenotype scope set and clinical-vs-imaging frame established
- 2ENTRYActivity-related knee/hip/hand pain (worse with use, better with rest); short morning stiffness / post-rest gelling; bony enlargement (Heberden/Bouchard/1st-CMC); incidental KL osteophytes; or an explicit OA-vs-inflammatory differentiation request (ACR/AF 2019; Altman 1986 PMID 3741515)inputs: age, joint_distributionadvance: Engine entered via a recognised OA trigger
- 3CONTEXTCapture age, sex, BMI, occupational/recreational joint load, prior joint injury, function/QoL targets, AND the comorbidity inventory that GATES pharmacotherapy: CKD stage (eGFR CKD-EPI 2021 race-neutral), cardiovascular disease/HF/HTN, PUD/GI-bleed history, concurrent anticoagulant/antiplatelet, current analgesic regimen (ACR/AF 2019; OARSI 2019 PMID 31278997)inputs: age, bmi, morning_stiffness_duration, ckd_stage, cardiovascular_disease, peptic_ulcer_or_gi_bleed, anticoagulant_or_antiplateletactions: calc.ckd_epi_2021advance: Risk profile + comorbidity gating matrix inputs captured
- 4RED_FLAGSNOT an OA pathway, escalate: acute hot swollen joint + fever = septic arthritis → STAT arthrocentesis + empiric cover; rapidly destructive / atypical age or onset; true mechanical locking (loose body / displaced meniscal tear); inflammatory pattern (prolonged stiffness, symmetric small-joint synovitis, constitutional symptoms) → re-route to inflammatory/crystal engines (ACR/AF 2019; EULAR)inputs: hot_swollen_joint_with_fever, constitutional_symptomsactions: workup.acute_monoarthritis, panel.synovial, panel.inflammationadvance: Septic / inflammatory / mechanical-block mimics screened and escalated if present
- 5INITIAL_WORKUPUsually NONE — OA is diagnosed clinically (NICE NG226 2022). Selective only: ESR/CRP + RF/anti-CCP to EXCLUDE inflammatory arthritis when the pattern is atypical; eGFR (CKD-EPI 2021) BEFORE any oral NSAID; synovial fluid analysis if an effusion is aspirated (OA = non-inflammatory, WBC <2000, no crystals). Do not order radiographs to confirm typical OA (NICE NG226 2022)inputs: esr_crp, synovial_fluid_analysisactions: workup.osteoarthritis, panel.inflammation, panel.renal, panel.synovial, calc.ckd_epi_2021advance: Inflammatory mimic excluded clinically/serologically; NSAID-eligibility eGFR known
- 6BRANCHING_WORKUPPhenotype-specific only when indicated: weight-bearing radiograph → Kellgren–Lawrence grade ONLY if atypical / pre-surgical / diagnostically uncertain (KL ≥2 = definite osteophyte + JSN; structural anchor, not a treatment gate); knee-pain branch for mechanical/effusion features; erosive hand OA → central "gull-wing" erosions of DIP/PIP (ACR/AF 2019; NICE NG226 2022)inputs: plain_radiograph_kl_gradeactions: workup.knee_pain, calc.kl_gradeadvance: Phenotype confirmed; KL grade obtained only where it would change management
- 7DIFFERENTIALMECE terminal differential — OA vs rheumatoid arthritis (symmetric MCP/wrist synovitis, stiffness >60 min, RF/anti-CCP+, ESR/CRP↑) vs gout/CPPD (acute hot monoarthritis, podagra, synovial crystals) vs psoriatic/reactive arthritis vs septic arthritis vs avascular necrosis vs referred/neuropathic pain vs secondary OA (haemochromatosis, alkaptonuria). Named pivots: morning-stiffness duration, joint distribution, inflammatory markers, constitutional symptoms (Altman 1986 PMID 3741515)inputs: joint_distribution, morning_stiffness_duration, rf_anti_ccpadvance: Terminal diagnosis assigned (clinical criteria + selective serology + named pivots)
- 8RISK_STRATIFICATIONWOMAC pain/function severity (calc.womac_oa_index) for treatment intensity; KL grade as structural context; eGFR-driven oral-NSAID eligibility tier; function-limiting refractory disease despite optimised conservative therapy = arthroplasty-referral threshold (Skou NEJM 2015 PMID 26488691)inputs: function_and_refractoriness, ckd_stageactions: calc.womac_oa_index, calc.kl_grade, calc.ckd_epi_2021advance: Severity tier + NSAID-eligibility tier + surgical-candidacy flag documented
- 9TREATMENTCore non-pharm is FIRST-LINE for every phenotype: structured patient education/self-management + structured land-based exercise (knee pain SMD −0.49, Fransen Cochrane 2015 PMID 25569281) + weight loss ≥5–10% if BMI ≥25 (IDEA Messier JAMA 2013 PMID 24065013) + assistive devices/bracing/hand orthoses. THEN topical NSAID (strong first-line drug, knee & hand — ACR/AF 2019; OARSI Level 1A). THEN oral NSAID at the LOWEST effective dose, comorbidity-gated by the encoded matrix (eGFR <30 avoid; CVD prefer naproxen; PUD/anticoagulant → COX-2 + PPI). Adjuncts: duloxetine (centralised/multi-joint/NSAID-limited — Chappell PMID 20602715), intra-articular glucocorticoid (short-term knee relief, NNTB 8, Jüni Cochrane 2015 PMID 26490760; interval-capped — 2017 McAlindon JAMA cartilage-volume signal PMID 28510679), acetaminophen (weak adjunct only). AVOID glucosamine/chondroitin (GAIT PMID 16495392), intra-articular hyaluronic acid (Rutjes PMID 22868835), oral/transdermal opioids (ACR/AF 2019 / OARSI strong against)inputs: bmi, ckd_stage, cardiovascular_disease, peptic_ulcer_or_gi_bleed, anticoagulant_or_antiplateletactions: calc.ckd_epi_2021, calc.womac_oa_indexadvance: Core non-pharm engaged + comorbidity-appropriate pharmacotherapy selected (or surgical pathway entered)
- 10DISPOSITIONOutpatient throughout — OA is not an admitting diagnosis. Orthopaedic referral for total knee/hip arthroplasty when conservative therapy is exhausted AND pain/function is refractory and life-limiting (Skou NEJM 2015 PMID 26488691); admit only for a complication (septic joint, post-injection sepsis, surgical pathway)inputs: function_and_refractorinessadvance: Care level set; surgical referral placed if threshold met
- 11MONITORINGTrack pain/function trajectory (WOMAC), exercise & weight-loss adherence; on oral NSAID surveil renal function (eGFR), BP, and GI symptoms; cap intra-articular glucocorticoid frequency (effect decays by 13–26 wk, repeated scheduled injection → cartilage loss — McAlindon JAMA 2017 PMID 28510679); reassess weight (1 kg loss reduces knee load ~4 kg/step) (ACR/AF 2019; OARSI 2019 PMID 31278997)inputs: ckd_stage, function_and_refractorinessactions: calc.womac_oa_index, calc.ckd_epi_2021, panel.renaladvance: Response and drug-safety surveillance plan running
- 12FOLLOWUPChronic self-management with periodic reassessment; deprescribe analgesics that lack benefit (especially opioids — ACR/AF 2019 against); maintain exercise + weight; re-refer to orthopaedics if progressive functional decline; revisit surgical timing using shared decision-making (benefit vs serious-AE trade-off — Skou NEJM 2015 PMID 26488691)inputs: function_and_refractorinessadvance: Long-term self-management + deprescribing + surgical-timing plan booked