Lateral Elbow Extensor Tendinopathy (Tennis Elbow)

Who: Working-age adults. Manual workers, tradies, keyboard-and-mouse jobs, racquet players, anyone whose grip volume stepped up. Peak 45–54.
What: Gradual-onset pain right on the common extensor origin, no trauma.
How it behaves: Aggs: gripping, lifting palm-down, wrist extension with the elbow extended (that's ECRB on stretch), kettles, doorknobs, tools, a keyboard, a racquet. Eases: putting the thing down, forearm strap, elbow flexed.

The illness script in one line: lateral elbow pain in someone whose gripping load recently changed, reproduced by grip and resisted wrist extension, with full elbow range and normal neurology.

The one thing to take from this page — and it goes first, not last:

Most of these people get better without you. Wait-and-see success was 83% at 52 weeks [1]. 96% of the placebo-injection arm reported complete recovery or much improvement at one year [2]. 83% improved at one year across 266 consecutive GP cases with no relation to treatment type [3]. Every effect size on this page sits on top of that number. Say it to the patient in week one.

([4] [5]; Smidt et al. 2002; Coombes et al. 2013 [6].)

⚠️ No Australian guideline for this presentation; anchored to the JOSPT 2022 Clinical Practice Guideline — Lateral Elbow Pain and Muscle Function Impairments [7]. Neither ACSQHC nor RACGP publishes one. And here's the irony worth carrying into a tutorial: the two strongest primary trials on this pattern anywhere in the world are Australian — Coombes 2013 in JAMA and Bisset et al. 2006 in the BMJ [8], both out of Brisbane — and nobody here has built a guideline on them. The CPG is 111 pages; only its abstract has been read, so no graded recommendation is quoted from it.

Patterns > Labels. "Lateral epicondylitis" names an inflammation that isn't there. A grip-load tolerance problem at the common extensor origin names what you're actually treating.

Five stages, in order. On this pattern the S/E does nearly all of it — the P/E confirms a hypothesis you already hold, and on the numbers below it barely earns the right to do that.

MAIN PROBLEM

"It's this spot on the outside of my elbow. I can't hold the kettle."

Onset gradual, no trauma, and there is almost always a load story sitting behind it. Ask about the job before you ask about the elbow.

BODY CHART

Chart interpretation: The primary distribution is lateral epicondylar/proximal extensor-region pain. Distal tingling or finger/thumb-extension weakness suggests an alternative explanation requiring separate assessment.

• P1: a fingertip-sized area directly on the lateral epicondyle / common extensor origin. Ask them to point with one finger, not to wave at it.
• P2: a spread down the extensor mass of the proximal forearm. Non-dermatomal, no distal march.
• Mark 3–5 cm distal separately. Maximal tenderness there, along the radial tunnel, is a different pattern (see §3).
• φp elsewhere. No pins and needles, no numbness, no neck symptoms — and check that rather than assume it.
• Note the side. Dominant side carried OR 3.1 (95% CI 1.4–6.8) for a <50% pain reduction at one year (Haahr & Andersen 2003).

BEHAVIOUR

• Coding: In early — intermittent, load-linked, provoked by grip. It becomes P (present) as it drags on, and that shift is a prognosis signal, not a new diagnosis.
• Aggs: grip, lifting palm-down, screwdriver, kettle, doorknob, mouse, racquet, carrying shopping. Worse with the elbow extended.
• Eases: stopping the task, elbow flexed, forearm strap during gripping.
• 24-hr pattern: morning stiffness or an ache after activity, usually M at under 30 minutes; grip through the day is the Aggs.
• Severity: VAS 0–10 at best / at worst / right now.
• Morning stiffness or ache after activity. Under UniSA coding, <30 min = M; >30 min = In. This pattern is usually M.
• Irritability, all three parts: resting symptoms present or not · how vigorous the provoking task has to be · how long it takes to settle. That triad sets your starting load, and on this pattern the third part is the one that decides whether you load today.

HISTORY

The work questions are the strongest thing in the whole subjective. Prospective cohorts, not recall.

Combined repetitive and forceful arm activity carries OR 5.6 against neither (Shiri et al. 2006); the prospective pooled figure is OR 2.6 (Descatha et al. 2016). Current smoking OR 3.4 (1.4–8.3) in the same population sample.

SPECIAL / SCREENING QUESTIONS

• Neurological screen in words first: any weakness opening the hand or extending the fingers? Any numbness or pins and needles? Any night symptoms? Then test it (§5).
• Screen the neck on everybody, even when they don't mention it. 36% of 165 people with unilateral lateral elbow pain and NO neck or upper-limb symptoms had cervical palpation dysfunction, and 41% had a positive radial neurodynamic test [10].
• The subgrouping prompts — neck pain, tendon tears, features of central sensitisation — come from Coombes et al. 2015 [11], whose own authors call the algorithm preliminary and untested. Use them as questions to ask, not as a pathway to follow.
• Screening questions as taught: general health, unexplained weight change, medications including steroids, previous surgery, imaging already done.

Jones' hypothesis categories that actually change your plan here: contributing factors (the occupational load — it is the strongest evidence for this pattern), prognosis (manual work, baseline pain, baseline distress, duration), and precautions (motor weakness, and the skeletally immature elbow).

70/30 is a working heuristic, not a statistic. On this pattern it's close to literal: the S/E gives you the pattern, and the P/E has almost no measured accuracy to add.

Top three, one discriminating feature each. Rank them, then rule out the dangerous one first.

And the variant that isn't a rehab plan at all: PIN entrapment with motor involvement — weak finger and thumb extension. PIN is a motor branch, so sensation stays intact. Normal sensation does not rule it out. Any true extensor lag is a referral.

Where's the number for the top differential? There is essentially one, and it needs a warning. In a prospective double-blinded pilot [14], palpation separating radial tunnel syndrome from lateral epicondylitis ran sensitivity 75.7%, specificity 50.0%; neuromuscular ultrasound ran 92.6% / 80.0%. 68 elbows in 34 people, and the published results paragraph describes the same 68 nerves as both patients and healthy controls. Take the direction — your thumb is close to a coin-toss at telling these two apart — and treat every digit as provisional.

Consistency = confidence. Three or four features pointing the same way beats any single test on this pattern, because the single tests here have almost never been measured.