What it is:The most common form of arthritis. A slowly progressive, mechanical wear-and-tear disease of a single joint or a handful of joints, mainly the ones that bear weight or take repetitive load: knees, hips, hands, and spine.
The core problem:Cartilage breaks down faster than the joint can repair it. The exposed bone underneath reacts by remodeling and growing spurs (osteophytes), and the joint capsule gets a little inflamed along the way. That's it. There's no systemic autoimmune process driving this, which is the single biggest thing that separates OA from rheumatoid arthritis.
What you do about it:Nothing reverses cartilage loss. Every drug on this page treats pain, not disease. Start conservative (acetaminophen, topical NSAIDs), escalate to oral NSAIDs or intra-articular steroid if pain isn't controlled, and reserve opioids and duloxetine for patients who fail everything else or can't take NSAIDs.
OA is a local mechanical disease treated with systemic and local analgesics, not a disease you modify pharmacologically. Nothing here slows joint destruction, not glucosamine, not chondroitin, not any drug in the DiPiro table. That reframes the entire chapter: your job is symptom control with the safest agent for that specific patient, escalating only as needed. Compare that to RA, where DMARDs actually halt the disease process, this is the distractor examiners love.
Two flavors, same end result. Primary (idiopathic) OAhas no identifiable cause, it's just cumulative joint stress over time. Secondary OAhas a driver you can name: prior trauma, joint instability, obesity-driven overload, or a metabolic/endocrine/congenital problem. Same downstream pathology either way, so treatment doesn't change based on which type you're looking at.
Cartilage itself has no nerve endings, so cartilage loss alone doesn't hurt. Pain comes from the structures around it: distention of the joint capsulefrom excess synovial fluid, microfractureof subchondral bone, periosteal irritation, and damage to ligaments, synovium, or the meniscus. That's why intra-articular steroid, which calms the capsule and synovium, can work even though it does nothing to rebuild cartilage.
Risk factors are the ones you'd guess: increasing age, obesity, female sex, prior joint injury or surgery, repetitive occupational or sports loading, and genetic predisposition.
| Finding | What it looks like |
|---|---|
| Pain pattern | Deep, aching pain that comes with joint useand eases with rest. Worsens through the day, unlike inflammatory arthritis which is often worse first thing in the morning. |
| Joints affected | DIP and PIP joints of the hand, first carpometacarpal (base of thumb), knees, hips, cervical and lumbar spine, first MTP joint of the toe. Almost always asymmetric, unlike RA. |
| Morning stiffness | Under 30 minutes, resolves quickly with movement. RA stiffness runs well over 30 minutes, often an hour or more. |
| Exam | Tenderness, crepitus (that grinding/crackling with motion), possible joint enlargement. Lower-extremity disease may bring weakness or instability. |
| Heberden and Bouchard nodes | Bony enlargements (osteophytes) at the DIP and PIP joints respectively. Classic exam finding, classic exam question. |
Warm, red, tender joints point toward inflammatory synovitis, not garden-variety OA. If a joint looks hot and inflamed rather than just achy and stiff, reconsider the diagnosis or think about a superimposed process (crystal arthropathy, septic joint, or an inflammatory arthritis).
OA:asymmetric, DIP/PIP/hips/knees/spine, stiffness <30 min, worse with activity, normal ESR, negative rheumatoid factor, mild synovial leukocytosis (<2000 WBC/mm³). RA:symmetric, MCP/wrist/small joints, stiffness >30-60 min, worse in the morning, elevated inflammatory markers, positive RF/anti-CCP in most patients. If a question gives you Heberden/Bouchard nodes, it's OA. If it gives you symmetric MCP swelling with prolonged morning stiffness, it's RA.
Diagnosis leans on history, exam, radiographs, and just enough labs to rule out something else. There's no single confirmatory test, it's a clinical diagnosis supported by imaging.
| Joint | Criteria |
|---|---|
| Hip | Hip pain plus 2 of 3: (1) ESR <20 mm/h, (2) radiographic femoral or acetabular osteophytes, (3) radiographic joint space narrowing. |
| Knee | Knee pain plus radiographic osteophytes, plus at least one of: age >50, morning stiffness ≤30 min, crepitus on motion, bony enlargement, bony tenderness, or palpable joint warmth. |
| Hand | Pain with bony changes on exam, normal ESR, and radiographs showing osteophytes or joint space narrowing. |
Five goals, and none of them is "cure the disease" because you can't: (1)educate patient and caregivers, (2)relieve pain and stiffness, (3)maintain or improve joint mobility, (4)limit functional impairment, (5)maintain or improve quality of life.
Nonpharmacologic therapy is not a footnote here, it's foundational and continues alongside whatever drug you add:
Drug therapy targets pain, full stop, and the approach stays conservative because most OA patients are older adults carrying other conditions and other medications. Continue nonpharmacologic therapy at every step below rather than swapping it out for drugs.
Preferred first-line. Less effective than oral NSAIDs but far safer GI/CV profile.
Preferred over oral NSAID first, especially age >75 or high CV/GI risk.
Reasonable alternative if age <75 and low CV/GI risk.
For inadequate response to first- and second-line therapy.
Acetaminophen isn't contraindicated? Start there, max 4 g/day, and counsel on every other acetaminophen-containing product they might be taking (combination cold meds, hydrocodone/APAP, etc.) so they don't stack doses without realizing it. Not working? Move to the alternative first-line bundle: topical NSAID (knee only) and/or IA corticosteroid and/or tramadol, or oral NSAIDs if the patient is under 75 with low CV and GI risk. Still not working? That's when opioids, duloxetine (knee only), IA hyaluronate (knee only), or surgery enter the conversation.
It's not that acetaminophen works better, oral NSAIDs generally provide more pain relief. It's that the risk-benefit favors acetaminophen as a starting point in a population where GI bleeds and CV events from NSAIDs carry real consequences. Acetaminophen is well tolerated at appropriate doses; its risk is dose-dependent hepatotoxicity, which is why you avoid it or cap it lower in chronic alcohol users and liver disease.
| Risk factor | Why it matters |
|---|---|
| Age >60, especially >75 | Ulcer and GI bleed risk climbs; topical NSAID preferred over oral past 75 |
| Longer NSAID duration, higher dose | Direct dose/time relationship with ulcer complications |
| Prior PUD of any cause | Strongest single predictor of NSAID-related GI bleed |
| Alcohol use | Compounds GI mucosal injury |
| Concomitant glucocorticoids or anticoagulants | Multiplies bleeding risk, doesn't just add to it |
| CV disease / CV risk factors | Both nonselective NSAIDs and COX-2 inhibitors raise HTN, MI, stroke, and death risk |
| CKD or CrCl/eGFR reduction | NSAIDs reduce renal prostaglandins that maintain GFR; avoid most NSAIDs under CrCl 30, meloxicam avoided under eGFR 60 |
ACEi/ARB + diuretic + NSAIDtogether raises acute kidney injury risk by roughly 36%. The ACEi/ARB drops efferent arteriolar tone, the diuretic drops volume, and the NSAID knocks out the prostaglandin-mediated afferent arteriolar dilation that was compensating for both. All three together can tank GFR fast. This combination shows up constantly in OA patients because so many also have hypertension, so screen for it at every NSAID recommendation.
Hand OA gets its own algorithm because topical therapy works unusually well on small, accessible joints. First-line options are topical NSAIDs, topical capsaicin, and/or tramadol, not oral NSAIDs.
Never combine topical and oral NSAIDs.Patients on a topical NSAID should skip oral NSAIDs entirely, the additive systemic exposure isn't worth it and defeats the purpose of choosing topical in the first place.
symptom reliefmarks every regimen on this page. Nothing here carries a mortality or disease-modifying benefit, that's the whole point of OA pharmacotherapy.
| Class / Drug | Starting dose | Usual range |
|---|---|---|
| Oral analgesics | ||
| Acetaminophen | 325-500 mg TID | 325-650 mg q4-6h, or 1 g TID-QID max 3-4 g/day |
| Tramadol | 25 mg every morning | Titrate by 25 mg to maintenance 50-100 mg TID symptoms |
| Tramadol ER | 100 mg daily | Titrate to 200-300 mg daily |
| Hydrocodone/acetaminophen | 5/325 mg TID | 2.5-10 mg / 325-650 mg 3-5x daily |
| Oxycodone/acetaminophen | 5/325 mg TID | 2.5-10 mg / 325-650 mg 3-5x daily |
| Topical analgesics | ||
| Capsaicin 0.025-0.15% | Apply to affected joint 3-4x daily | |
| Diclofenac 1% gel | 2 or 4 g per site, 4x daily | |
| Diclofenac 1.3% patch | 1 patch to site, twice daily | |
| Diclofenac 2% solution | 40 mg (2 pump actuations, 10 drops) to knee, twice daily | |
| Intra-articular corticosteroids (per joint, not more than q3 months) | ||
| Triamcinolone | 5-15 mg small joint | 10-40 mg large joint (knee, hip, shoulder) |
| Methylprednisolone acetate | 10-20 mg small joint | 20-80 mg large joint |
| Nonselective NSAIDs | ||
| Ibuprofen | 200 mg TID | 1200-3200 mg/day divided TID-QID avoid CrCl <30 |
| Naproxen | 250 mg BID | 500 mg BID avoid CrCl <30 |
| Naproxen sodium | 220 mg BID | 220-550 mg BID |
| Diclofenac IR / XR | 50 mg BID / 100 mg daily | 50-75 mg BID / 100-200 mg daily |
| Meloxicam | 7.5 mg daily | 15 mg daily avoid eGFR <60 |
| Indomethacin | 25 mg BID | titrate to max 50 mg TID avoid CrCl <30 |
| Nabumetone | 500 mg daily | 500-1000 mg 1-2x daily |
| Piroxicam | 10 mg daily | 20 mg daily |
| Aspirin | 325 mg TID | 325-650 mg QID |
| COX-2 selective | ||
| Celecoxib | 100 mg daily | OA: 100 mg BID or 200 mg daily avoid severe sulfa allergy |
| Adjunct (knee OA only) | ||
| Duloxetine | Standard neuropathic/musculoskeletal pain dosing; effect at ~4 weeks | |
Mechanism isn't fully worked out: weak COX inhibition, serotonergic agonism in descending CNS pathways, and TRPV1/cannabinoid-1 receptor activity all seem to contribute. Oral bioavailability ~88%, peak effect around 60 minutes, half-life ~2.5 hours (extends to 4-8 hours if hepatic metabolism is impaired).
The metabolism that matters clinically:at therapeutic doses, ~90% of acetaminophen is inactivated by glucuronidation and sulfation and excreted. The remaining 5-10% gets oxidized by CYP2E1 into NAPQI, a toxic metabolite that's normally neutralized by binding glutathione. At supratherapeutic doses (above roughly 4 g/day), glucuronidation and sulfation pathways saturate, more drug gets shunted to NAPQI, glutathione stores run out, and NAPQI accumulates and destroys hepatocytes.
Chronic alcohol use and liver disease deplete glutathione reserves and induce CYP2E1, both of which push more acetaminophen toward NAPQI at lower total doses. Cap these patients at 2 g/dayrather than the usual 3-4 g/day max, and always ask about every other source: combination cold/flu products, and opioid/APAP combinations like hydrocodone/APAP, are the ones patients forget to count. Antidote is N-acetylcysteine, which replenishes glutathione.
NSAIDs inhibit cyclooxygenase. COX-1is constitutive, protecting GI mucosa, supporting renal blood flow, and driving platelet thromboxane production. COX-2is inducible, upregulated during inflammation, which is the target you actually want to hit. Nonselective NSAIDs block both; that's why they help pain but also cause GI ulcers and impair platelet function. Celecoxib is COX-2 selective, which spares platelets and lowers (but doesn't eliminate) GI risk.
| GI risk | CV risk | Platelet effect | |
|---|---|---|---|
| Nonselective NSAIDs | Higher | Increased (HTN, MI, stroke, death) | Reversibly inhibit thromboxane; normalizes 1-3 days after stopping |
| Celecoxib (COX-2 selective) | Lower, but advantage shrinks if patient also takes aspirin | Still increased | Spared |
GI risk reduction strategieswhen an NSAID is necessary: lowest effective dose for the shortest time, add misoprostol four times daily, or add a PPI or full-dose H2RA.
Notable drug interactions:lithium (raises levels), warfarin (bleeding), oral hypoglycemics, methotrexate (raises levels, toxicity risk), antihypertensives and ACEi/beta-blockers/diuretics (blunted BP control, and don't forget the Triple Whammy above). Avoid in the third trimester of pregnancy due to risk of premature ductus arteriosus closure.
Topical diclofenac still achieves meaningful systemic absorption (roughly 6%), but far less than an oral dose, so it delivers similar local pain relief to oral NSAIDs with dramatically fewer GI events. Application-site reactions (dry skin, pruritus, rash) are the tradeoff. Topical NSAIDs haven't been linked to increased CV event risk the way oral NSAIDs have.
The handbook flags topical NSAIDs as preferred to oral NSAIDs specifically past age 75, and as the step to try for knee OA after acetaminophen fails, before jumping to an oral NSAID. Don't expect immediate relief, full effect can take up to 7 days of consistent use.
Recommended for hip and knee OA when acetaminophen or NSAIDs aren't cutting it, and they shine when there's a visible joint effusion. Technique: aseptic aspiration of the effusion, then injection, often combined with a local anesthetic (lidocaine or bupivacaine) for immediate relief while the steroid takes effect.
| Timepoint | What happens |
|---|---|
| 24-72 hours | Initial pain relief begins |
| 7-10 days | Peak relief |
| 4-8 weeks | Duration of benefit before it wanes |
No more often than once every 3 monthsper joint, to limit local complications: infection, osteonecrosis, tendon rupture, skin atrophy at the injection site. Systemic (oral/IV) corticosteroids are not recommendedfor OA at all, no proven benefit and the well-known long-term harms of steroid exposure aren't justified for a non-inflammatory disease.
Tramadolis positioned for patients who've failed scheduled full-dose acetaminophen and topical NSAIDs, aren't oral NSAID candidates, and aren't candidates for IA corticosteroids. It can also be layered onto partially effective acetaminophen or NSAID therapy. It's a Schedule IV controlled substance (dependence/diversion potential lower than full opioids but real), and it has a signature risk full opioids don't: seizures, its most serious adverse effect. It's also serotonergic, so combining with duloxetine or other serotonergic drugs raises serotonin syndrome risk. Standard opioid AEs apply too: nausea, vomiting, dizziness, constipation, headache, somnolence.
Full opioidscome in only after nonpharmacologic and first-line pharmacologic therapy fail, or in patients who are poor surgical candidates and can't get joint arthroplasty. Lowest effective dose, smallest quantity, avoid stacking with other sedating drugs, and counsel on safe use/storage/disposal. Sustained-release formulations give steadier control through the day. Reassess use at least every 3 months for progress toward functional goals, harms, and adverse effects, since dependence, tolerance, hyperalgesia, and diversion are real risks with long-term use.
Adjunctive therapy for knee OA only(not hip), in patients with partial response to first-line acetaminophen/oral NSAIDs. It's an SNRI, so it's a particularly good second-line pick if the patient also has a neuropathic pain component overlapping their musculoskeletal pain. Pain reduction takes about 4 weeks to show up, so set that expectation upfront. Adverse effects: nausea, dry mouth, constipation, anorexia, fatigue, somnolence, dizziness; rare but serious risks include Stevens-Johnson syndrome and liver failure.
Same serotonin syndrome interaction flag as above: don't combine with tramadol or other serotonergic agents without watching for it.
Sodium hyaluronate injections are meant to supplement synovial fluid viscosity ("viscosupplementation"), but the evidence shows only limited benefit for knee OA and none demonstrated for hip OA, so this is not routinely recommended. Generally well tolerated; watch for acute joint swelling, effusion, stiffness, or local skin reactions (rash, ecchymoses, pruritus).
Knee-only regimens you should be able to recognize:Gel-One (single injection), Orthovisc (weekly x3), Monovisc (single), Synvisc (weekly x3), Synvisc-One (single), Hyalgan (weekly x5), Euflexxa (weekly x3), Supartz FX (weekly x5), GenVisc 850 (weekly x5). You don't need to memorize every dose, just recognize this is the hyaluronate family and that dosing schedules range from a single shot to five weekly injections.
| Intervention | Why it's out |
|---|---|
| Glucosamine and/or chondroitin | No uniform, reproducible improvement in pain or function for hip/knee OA across trials. Adverse effects are mild (flatulence, bloating, cramps with glucosamine; nausea with chondroitin), so the issue is lack of efficacy, not safety. |
| Topical rubefacients(methyl salicylate, trolamine salicylate) | Same story, not a preferred option due to inconsistent benefit. |
| Systemic corticosteroids | No proven benefit in OA, and long-term steroid harms aren't justified for a disease that isn't primarily inflammatory. |
| Disease-modifying agents(IL-1 receptor antagonists, bisphosphonates, methotrexate, hydroxychloroquine) | These are RA-world drugs. OA has no autoimmune process for them to modify, so they don't belong here, a favorite distractor when a question mixes OA and RA drug lists. |
| Parameter | When | Watching for |
|---|---|---|
| Pain (VAS) and ROM | Baseline, then every follow-up | Whether current therapy is actually working, flexion/extension/abduction/adduction of affected joints |
| Grip strength | Hand OA follow-up | Functional decline specific to hand involvement |
| 50-ft walking time | Hip/knee OA follow-up | Functional decline in weight-bearing joints |
| Radiographs | Baseline, then periodically | Extent of joint involvement and progression over time |
| WOMAC / Stanford HAQ / clinician global assessment | Periodically | Composite function and disability tracking beyond just pain scores |
| Serum creatinine, CBC, LFTs | Baseline, then every 6-12 months on chronic NSAID/acetaminophen therapy | Renal, hepatic, GI, or bone marrow toxicity building silently over time |
| Adverse effect check-in | Every visit | Rash, headache, drowsiness, weight gain, or new hypertension from NSAIDs |