Arthroplasty Landmark Studies

FRCS Level

What the Question Was

Knee arthroplasty generated two literatures that answer different questions and are frequently confused.

The design literature asks what a prosthesis should do mechanically. The outcome literature asks whether patients are better. Both have good answers and they do not agree as well as they should.

What does a normal knee do that a prosthesis must reproduce. What does an implanted knee actually do. How should the components be aligned. And why are some patients dissatisfied with a technically sound result.

What Was Done

Walker measured knee kinematics and applied the results to prosthesis design across several decades, which connected the two literatures more directly than most work did.

Freeman and Swanson designed and reported an early total knee replacement in the 1970s, and argued the case for a particular balance between constraint and fixation.

Insall and colleagues developed the total condylar prosthesis in the same period, which became the pattern most modern designs descend from.

Goodfellow and O'Connor designed and reported the meniscal bearing unicompartmental knee, arguing from contact mechanics for a mobile bearing.

Banks and Hodge measured implanted knee kinematics with fluoroscopy, which made it possible to see what a prosthesis does in a living loaded knee rather than on a rig.

Dennis and Komistek used in vivo fluoroscopic measurement to describe paradoxical anterior translation of the femur during flexion in many implanted knees.

Ritter and colleagues related component alignment to survival from a large series, which is the basis of the mechanical alignment target.

Bellemans and colleagues measured alignment in a healthy population in the 2010s and described constitutional varus, which challenged that target. 3.13.5 covers it.

Bourne and colleagues reported patient satisfaction after knee arthroplasty from a large cohort, which separated satisfaction from surgical success as outcomes.

What It Showed

A replaced knee does not move like a normal knee. Roll back is reduced or reversed, and paradoxical anterior translation is common rather than exceptional. 3.3.2 covers roll and glide.

Deep flexion kinematics are the worst reproduced, which matches the activities patients report most difficulty with.

Component alignment relates to survival in large series, which is why a mechanical alignment target became standard.

A substantial minority of the healthy population is not in neutral alignment, so restoring neutral alignment means changing some knees from their own normal.

About one patient in five is less than satisfied after a technically successful knee replacement, and that figure has proved stable across cohorts.

Satisfaction correlates poorly with radiographic result and better with expectation and with pain elsewhere.

What It Changed

Design moved towards reproducing kinematics rather than only providing a stable surface, and that aim has been pursued for forty years without being achieved.

In vivo fluoroscopic measurement became the standard way to evaluate a design, replacing rig testing as the final word.

The alignment target is now argued rather than assumed, and kinematic alignment exists as an alternative because of the constitutional varus work.

Patient reported outcome and satisfaction entered the field as endpoints, and the satisfaction figure changed how consent is taken more than it changed how the operation is done.

Fellowship Level

What Was Done

Fluoroscopic kinematic studies image a small number of knees doing selected activities under supervision, and the activities are not the ones patients complain about.

Survival series relate alignment to revision, and revision is a decision as well as an event, so the endpoint carries surgeon behaviour inside it.

The alignment measurements in those series come from long leg radiographs with their own rotational error, which is large relative to the angular differences being compared.

Satisfaction studies depend on the instrument used, and different instruments give different proportions dissatisfied from the same population.

What It Showed

Where the literature is strong: that implanted kinematics differ from normal, and that a stable proportion of patients are dissatisfied.

Where it is weak: the link between kinematics and satisfaction. Both are well measured and the connection between them is not established, which is the field's central unresolved problem.

Where it is contested: kinematic against mechanical alignment. Each cites a different kind of evidence, survival series against restoration of individual anatomy, and the trials so far are short.

Unicompartmental against total replacement is argued on the same split: better kinematics and function against higher revision rates in registry data.

What to Read Now

Read a kinematic study for the activity it measured, since deep flexion and stair descent behave differently from level walking.

Read revision as an endpoint that includes surgeon decision making, not as a pure measure of failure.

Check which satisfaction instrument a study used before comparing its proportion with another study's.

Read the alignment argument as a difference in evidence type rather than as a difference in evidence quality.

Consultant Perspective

What It Changed

The decision this literature informs is alignment strategy, and it is a balance between two kinds of evidence rather than between a right and a wrong answer.

Arguing for mechanical alignment: the survival evidence is long, large and consistent, and a target that has produced reliable implant survival for decades should not be abandoned for a short term functional signal.

Arguing for kinematic alignment: a large minority of knees were never neutral, restoring an individual's own alignment produces better early function in several trials, and survival evidence for the new approach is absent rather than unfavourable.

Take age and physiology into account, because a thirty year implant horizon weighs survival far more heavily than a fifteen year one. Take comorbidities into account, since several of them change both revision risk and the value of early function. Take the demands on the knee into account, because the functional argument is strongest in a patient who kneels, squats or works. Take their psychological state and expectations into account, since dissatisfaction tracks expectation more closely than it tracks radiographs. And elicit their ideas, concerns and expectations explicitly before the operation, because this is the one factor in this chapter that is known to change the outcome being measured.

Choose on the balance, say which evidence you weighted and why, and record it.

What to Read Now

What a trainee should take from this chapter is that an operation can be reliable and mechanically unsuccessful at the same time, and that both statements are supported.

Read the kinematic literature and the satisfaction literature together, because the gap between them is the field's live question rather than an inconsistency to explain away.

Teach the satisfaction figure during consent teaching, since a patient told beforehand that one in five is less than satisfied is a different patient afterwards.