Physiotherapy after knee replacement – Knowledge

1. Why this article?

Six weeks after surgery, two people sit in the waiting room. Both received the same joint from the same surgeon, both are in their early seventies. One says: “The pain is gone, but the leg doesn't feel like mine.” The other says: “I'm back up the mountain – I just don't dare go down stairs.”

Both are describing the same phenomenon. A knee replacement replaces a worn joint surface. It does not replace a muscle, a sense of balance or years of movement experience. Those things do not come back with the implant – they have to be fetched back.

That is the reason for this article. Since 2020 there has been an international guideline on what physiotherapy after a knee replacement should look like; a revised version appeared in 2026 with twenty recommendations [1]. Much of it is gratifyingly clear – and some of it contradicts what is still common in practice.

This article explains what surgery can and cannot deliver, what really counts in the weeks before and after, and how you can tell whether your rehabilitation is on track. Every statement is referenced; the numbers in square brackets point to the reference list at the end.

One note up front: this article does not replace individual advice. Aftercare protocols differ according to surgical technique, bone quality and other conditions. What your surgeon specifies for your knee takes precedence.

2. What a knee replacement is – and how common it is

In a total knee replacement, the damaged bearing surfaces of the thigh bone and the shin bone are capped and a plastic insert is placed between them; often the back of the kneecap is resurfaced as well. In a partial replacement, only one compartment is treated – usually the inner one. Depending on the design, the ligaments are kept in full or in part.

In Switzerland the procedure is common. In 2023 the national implant registry SIRIS recorded 20,494 total knee replacements and 3407 partial replacements [2]. Those operated on were 69.7 years old on average, just under 60 per cent were women, and in 87.9 per cent of cases primary osteoarthritis was the reason.

One figure from the same report deserves particular attention: in 2023 the rate was 581 total knee replacements per 100,000 people aged 50 to 89 – according to the registry one of the highest rates in Europe and worldwide, with average annual growth of 4.8 per cent since 2013 [2]. Anyone having a knee replaced in Switzerland is therefore in plenty of company. But it also means the indication is set more generously here than elsewhere – one more reason to exhaust the non-surgical options seriously beforehand (section 5).

2.1 How long does such a joint last?

This question comes up in almost every first consultation, and it has a good answer. A synthesis of case series and national registries covering almost 300,000 total replacements found a 25-year survival of 82.3 per cent; for partial replacements it was 69.8 per cent [3]. In other words: four out of five total replacements last a quarter of a century.

The early years are not the problem. In Switzerland the revision rate within two years was 3.5 per cent for total replacements [2]. The most common reason for an early second operation was a problem with the kneecap (37.3 per cent), followed by infection (20.6 per cent) and instability (18.0 per cent). Joint stiffness accounted for 8.2 per cent – the one reason on that list that has anything to do with rehabilitation (section 13).

The worry about “wearing the joint out” through use is understandable, but it points in the wrong direction. Most early revisions have nothing to do with how much someone trained.

3. What you can realistically expect

This is where the article becomes uncomfortably honest – and precisely for that reason important. The answer to “how good will it get?” depends on what you ask about.

3.1 Pain: the real strength of the operation

On pain, surgery delivers. Most people have markedly less pain after a knee replacement than before, and that is the main reason the procedure is so widespread.

But not everyone. A meta-analysis of 68 studies with almost 600,000 patients tracked the proportion with persistent, relevant pain over time [4]:

• 21.9 per cent at three months
• 14.1 per cent at six months
• 12.6 per cent at twelve months
• 14.6 per cent at two years

So around one in eight people still has notable knee pain a year after surgery. That is not a fringe group. And it is important to know this beforehand: if you expect that pain at three months can still be normal, week twelve will not send you into a panic.

On satisfaction, the figure “20 per cent dissatisfied” has circulated for years. A systematic review of 21 studies recalculated it and arrived at an average dissatisfaction rate of 10 per cent; excluding cases with complications leaves 7.3 per cent [5]. The widespread figure is therefore too pessimistic – but ten per cent is still ten per cent.

3.2 Function: the awkward finding

If you take only one section away from this article, take this one.

A longitudinal meta-analysis pooled 72 high-quality studies with 19,063 patients and looked separately at two things: what people report about their function, and what is actually measured in performance tests [6]. The result is strikingly asymmetric:

• Self-reported function improved considerably – from 55.6 points before surgery to 21.1 points at three to six months (lower is better). It then deteriorated slightly to 31.0 points and stayed stable up to two years.
• Measured function – walking distance, stairs, standing up – showed no clinically meaningful improvement at any time point.

That sounds paradoxical, but it is not. The pain disappears, and that changes self-assessment enormously: you trust yourself again, everyday life feels lighter. The performance behind it does not change as a result. It has to be trained.

An older, smaller study had measured this directly: six months after surgery, patients were still well below healthy peers of the same age in strength, range of motion and every functional test – and essentially back at their own pre-surgery level [7]. The authors concluded that standard rehabilitation is not enough for this.

That is the core message of this article. Surgery is a pain treatment. Function is your work – and physiotherapy's.

3.3 And activity in daily life?

A third finding fits the picture: a systematic review of objectively measured activity (step counters and accelerometers) found virtually no increase in physical activity six months after hip or knee replacement. At one year improvements appeared, yet activity levels remained clearly below those of healthy comparison groups [8].

So people do not automatically move more after surgery just because it hurts less. That, too, is something to tackle actively – see section 11.

4. Why the leg is so weak after surgery

The single most important finding in knee rehabilitation research dates from 2005. Twenty people were measured ten days before and 27 days after surgery. One month afterwards, quadriceps strength had collapsed by 62 per cent. Muscle cross-sectional area had fallen by only 10 per cent; voluntary muscle activation by 17 per cent. Together these two factors explained 85 per cent of the loss of strength – and the failure of activation contributed almost twice as much as the wasting [9].

In plain terms: the muscle is still there. It is simply no longer being called up. The technical term is arthrogenic muscle inhibition – the joint itself, through swelling, pain and altered feedback from the tissue, throttles the nerve signal to the muscle. Why this happens and in what order to tackle it is set out in detail in our article on arthrogenic muscle inhibition.

This distinction has direct consequences for treatment:

• Inhibition is not resolved by time but by activation: contracting deliberately, seeing and feeling what happens, getting feedback. That is why electrical stimulation and biofeedback make sense in the first weeks (section 9.2).
• Very high training intensities achieve little in this phase, because the muscle cannot be fully recruited anyway. A trial comparing high-intensity with low-intensity rehabilitation showed exactly that: no difference at 3 or at 12 months – but both groups improved [10]. The high-intensity programme was safe, but its effect was limited by this very inhibition.
• Real strength work pays off increasingly as the inhibition resolves – so over months, not weeks.

Then there is the second, more mundane factor: lack of movement. A few days of rest cost a disproportionate amount of muscle in older age, and rebuilding is markedly slower than losing [11]. Rest after surgery is therefore a drug with a narrow dose range – necessary, but harmful in excess.

5. Before surgery: the weeks people underestimate

5.1 First, the uncomfortable question

Before we get to preparation: is surgery the right step for you? A Danish randomised trial divided 100 people with moderate to severe knee osteoarthritis, all scheduled for a replacement, into two groups. One had surgery followed by twelve weeks of non-surgical treatment, the other received only the twelve weeks of non-surgical treatment – exercise, education, dietary advice, insoles, pain medication [12].

The result at one year: the surgical group did better, and the difference was substantial. But it also had more serious adverse events. And just under a third of the non-surgical group had not had surgery within the first year and managed with conservative treatment.

It does not follow that replacements should be avoided. It follows that a structured exercise and education programme before the decision is not wasted time – not even when surgery happens in the end. That is exactly why exercise-based therapy comes first in every major osteoarthritis guideline [13], [14], and its benefit in knee osteoarthritis is documented in a Cochrane review of 54 trials [15]. Structured programmes such as GLA:D put exactly this into practice [16]. More on this in our article Osteoarthritis.

5.2 Prehabilitation: training before the operation

Once the decision is made, the weeks until the date are valuable. A meta-analysis of 48 randomised trials with 3570 participants in orthopaedic surgery found improvements in pain, function and muscle strength with prehabilitation – most clearly before surgery, with benefit continuing afterwards [17].

The guideline puts it this way: physical therapists should design and implement a preoperative exercise programme that improves strength, flexibility and endurance; education belongs to it as well, including pain neuroscience education to reduce anxiety about the procedure [1]. The early postoperative benefits are well documented but fade over time – and which type of exercise, which setting and how much supervision are optimal remains unresolved.

In practice: four to eight weeks beforehand, strength training two to three times a week for thigh, calf and hip muscles, plus balance and some endurance work. Train both legs – the non-operated one takes on much of the work in the first weeks. And practise beforehand what you will need afterwards: standing up from a chair, stairs with a handrail, walking sticks.

A footnote for the curious: blood flow restriction training – training with light loads under restricted circulation – has been tested in a randomised trial before knee replacement. Eight weeks of it produced more strength at three months after surgery than no preparation, but no advantage in function or symptoms, and by twelve months the strength advantage had gone as well [18]. It is an option for people who cannot tolerate heavy loads – not a must.

6. The first days: get up early

The guideline is unambiguous here: physical therapy including early mobilisation should start within 24 hours of surgery [1]. Those who start earlier have less pain and more range of motion; in several studies the hospital stay was shorter and the rate of thrombosis, pulmonary embolism and chest infection lower.

This is standard in many hospitals today and part of so-called fast-track or ERAS pathways [19]. Where it is not standard, you are entitled to ask.

What counts in the first days:

• Getting up and walking, several times a day, as far as pain medication allows.
• Straightening the knee. Flexion is not the problem – it almost always returns. It is full extension that is lost and hard to win back. When lying down, put the roll under your heel, not under your knee.
• Switching the thigh on: deliberate contractions with the leg straight, several times a day, few repetitions. That is activation training, not strength training.
• Cooling and elevating – more on this shortly.
• Sitting beats lying, walking beats sitting. What worked yesterday usually works again today.

People who are older, have several conditions or were already short of strength before surgery lose independence fastest in this phase. We have devoted a separate article to this: Leaving hospital weaker.

7. Range of motion: what helps – and what you can skip

The guideline contains four recommendations on range of motion, and two of them are refusals [1].

Continuous passive motion (CPM) should not be used. This is the strongest recommendation of the chapter and rests on four high-quality and twelve moderate-quality studies. The motion machine moves the knee passively and feels like therapy – but it improves neither short-term nor long-term range of motion, function or quality of life. It costs money, can be uncomfortable and takes time away from active practice.

Braces and splints should not be used routinely to improve range of motion. Here too the studies found no meaningful effect, but did find drawbacks: they hinder early walking and are often not worn.

Range-of-motion exercises, by contrast, should be done – passive, active-assisted and active. No single method proved superior to the others; what matters is that the knee is moved regularly at all.

Manual therapy and devices may be added (weak recommendation). A high-quality study showed that joint mobilisation, soft tissue techniques and scar treatment in addition to exercise improved pain, function and satisfaction – though not range of motion at two months compared with exercise alone. Cycle ergometers, slide boards and similar devices are likewise not superior, but they are harmless and many people like them.

And a word on extension. If the knee cannot be fully straightened in the first weeks, that is a reason to address it actively – not to wait. An extension deficit permanently alters the way you walk and loads the other side.

8. Pain and swelling

Here too the guideline separates cleanly between what works and what is merely well meant [1].

Cold should be used – for pain relief as well as for swelling. Eighteen studies support this. No method was clearly superior: a bag of crushed ice performed as well as a compression cooling device in one trial. A meta-analysis found some advantages for continuous cooling systems in pain and blood loss [20] – which does not by itself justify buying an expensive device. How long, how often and from when to cool is unresolved; the studies are too heterogeneous.

The leg should be elevated, with 30 to 90 degrees of knee flexion. This reduces swelling and blood loss and improves early range of motion. An important caveat from the guideline: this position may put extension at risk, so targeted extension exercises should run in parallel.

TENS, kinesio taping, manual therapy and psychologically informed techniques may be used (weak recommendation). TENS improved pain on walking in two studies during the first six weeks; taping helped in the first one to two weeks, with the difference gone at three months. Small effects, then – but hardly any risks either.

What is not recommended: manual lymphatic drainage, compression dressings and continuous passive motion for swelling. The work group found no robust evidence for any of the three [1]. This is a consensus recommendation on weak evidence, not a categorical statement – but anyone paying for these treatments out of pocket should know it.

Psychologically informed techniques deserve a separate mention. A high-quality study tested cognitive behavioural therapy in people with marked fear of movement and found improvements in fear, pain perception and quality of life at one and six months. The effect was tied to the starting point: people with little fear of movement benefited hardly at all [1].

9. Strength: the core of rehabilitation

This is where the strongest recommendation of the whole guideline sits. Physical therapists should design, implement and teach a progressive strength training programme, beginning in the early postacute phase, to improve function, strength and range of motion. Quality of evidence: high. Strength of recommendation: strong. It rests on seven high-quality and seventeen moderate-quality studies [1].

What is notable is how broad the effective programmes were: closed and open chain, concentric and eccentric, machines and bands – all showed benefit. There is no single right exercise. There is the requirement that the load increases.

In addition, a meta-analysis of 18 randomised trials found that people who received physiotherapy with exercise after discharge had better function and less pain at three to four months than people with minimal supervision [21]. Whether the exercises took place in a clinic or at home made no difference.

9.1 How heavy, how often, how long

The guideline sets no numbers, because the studies are too varied. From the underlying literature and general training principles, however, a usable framework can be derived:

• Two to three sessions a week, for at least three months. Plan for a quarter of a year, not a month.
• Choose a resistance that allows 8 to 12 clean repetitions and leaves you feeling at the end of the set that two to four more would have been possible – no more. If you suddenly manage 15 with ease, increase the load.
• Progression is the active ingredient. Two years with the same resistance band is no longer strength training, it is a habit.
• Train both legs, not just the operated one.

How do you know it was too much? The progression criteria used in the trials watch four things: soreness, pain, range of motion and swelling [1]. In practice: a knee that is thicker and markedly stiffer the day after training got too much. A knee that is a bit tired and moves normally after warming up got enough.

And do not train on the principle of “the more pain the better”. An accepted approach in joint problems is to allow pain during exercise up to about 5 out of 10, provided it settles back to baseline within 24 hours. That is a guide, not a law of nature – discuss it with your therapist.

9.2 Electrical stimulation for the thigh

Because the main problem is an activation failure (section 4), one method is of particular interest: one that takes over the activation from outside.

The guideline recommends: neuromuscular electrical stimulation (NMES) at least daily to the quadriceps, starting in the early postoperative period, at the highest tolerable intensity, to improve strength, gait and performance-based outcomes [1]. Six moderate-quality studies support this. Three details from those studies matter: start early (possible from day two), once or twice daily, and keep going for at least three weeks.

The reference trial: 66 people received, from 48 hours after surgery and on top of standard rehabilitation, 15 contractions twice daily at the maximum tolerable intensity. At three and a half weeks the stimulation group was markedly better in strength, functional tests and active extension; at one year the differences were smaller but still detectable for strength and performance [22].

Intensity is the decisive point. A tingle achieves nothing. A visible, forceful contraction has to occur – which is unpleasant, and precisely why the method is often underdosed in practice.

Do not use in people with demand-type pacemakers, active cancer or deep vein thrombosis [1].

10. Walking, balance and movement patterns

The second strong recommendation concerns a field that often gets short shrift in aftercare: motor function training should be part of treatment – dynamic balance training, gait retraining (including computer- or app-assisted) and movement training with feedback. Quality of evidence: high. Strength of recommendation: strong. Nine high-quality and twelve moderate-quality studies [1].

Why this matters: a knee that has hurt for years has learned a protective pattern – less extension at push-off, less load on the affected side, a shorter step. That pattern does not vanish with the pain. It has to be unlearned, and that requires feedback: a mirror, video, a metronome, the therapist's hand in the right place.

What featured in the effective programmes: uneven surfaces, changing bases of support, agility drills, sit-to-stand transfers, stair negotiation. Some studies used apps or virtual reality – but the benefit did not depend on them. The active ingredients are the task and the feedback, not the device.

One concrete reason to take this seriously: falls. In a study of 134 people, 17.2 per cent fell at least once within six months of surgery; the median time of the first fall was 15 weeks – long after hospital discharge. Two thirds of the falls happened while walking, almost half of them at home [23]. Associated factors were poorer joint position sense in the operated knee, impaired sensory orientation for balance and greater knee pain. More on this in our article Fall prevention.

11. Daily life, sport, work – and kneeling

The guideline explicitly recommends encouraging early physical activity and making a plan to increase it step by step – matched to safety, functional tolerance and the person's goals [1]. It draws on a high-quality trial with telephone support, step goals and collaborative goal setting: the intervention group took markedly more steps per day by the end [24].

According to the guideline, stationary cycling, aquatic exercise and Tai Chi are safe and effective in the early phase. One year after surgery, one study even found that hiking improved stair-climbing performance and self-reported function.

Work and sport. A meta-analysis of 44 studies provides figures to expect [25]:

• Return to work: 65 per cent overall, up to 90 per cent at one year. Mean time to return was 12.9 weeks – with a wide range of 5 to 42 weeks, depending on physical demands.
• Return to sport: 82 per cent overall, on average after 20.1 weeks. For low-impact sports it was 76 per cent, for high-impact sports only 35 per cent.

That gives a usable rule of thumb: hiking, cycling, swimming, cross-country skiing, golf – usually yes. Running, tennis, alpine skiing, ball sports – often restricted, and the decision belongs between you, your surgeon and your physiotherapist.

Kneeling. This question is rarely raised openly, but it occupies many people – because of the garden, the grandchildren or prayer. The answer is uncomfortable: a meta-analysis of 36 studies found that at a minimum of one year, 36.8 per cent could kneel, and at a minimum of three years 47.6 per cent [26]. The ability therefore improves over time but stays limited for many.

Two things matter here. First, part of the limitation is not mechanical but a question of confidence – people do not kneel because they fear damaging the implant, or because sensation over the scar has changed. Second, kneeling is not forbidden and does not damage the implant. A cushion, a garden mat or taking weight through the hands often makes the difference. Ask about it – it is discussed far too rarely.

12. Where and how therapy happens

On the organisation of aftercare the guideline makes three recommendations [1]:

Supervised physical therapy should be provided. “Supervised” means: initiated, monitored and individually tailored by a professional – as opposed to a generic exercise sheet that nobody adjusts. Where this supervision takes place depends on safety, mobility and personal circumstances.

Group or individual sessions – both are possible (weak recommendation). The evidence is contradictory. Groups are cheaper and motivating; they reach their limits when someone is much faster or much slower than the rest.

Digital tools should be considered – in addition to in-clinic care or as an alternative to it. Three high-quality and 25 moderate-quality studies are available. Most commonly they found comparable results between telerehabilitation and in-person treatment; an older meta-analysis of four randomised trials with 442 people even found slight advantages for telerehabilitation in range of motion and strength [27]. Apps and remote monitoring as a supplement mostly produced short-term gains.

On the choice of setting after hospital, the guideline is cautious: where possible, outpatient rather than inpatient or home-based – but that is a weak recommendation on weak evidence, and the differences in the studies were statistical rather than noticeable. It requires that someone can manage safely at home and get to the clinic. In Switzerland, insurance questions and regional availability also play a part – this is not a purely clinical decision.

The most important part happens between appointments anyway. Two therapy hours a week are about 1 per cent of your waking time. What happens in the other 99 per cent decides the outcome. Critical voices from fast-track surgery have been asking for years whether standardised therapy packages for everyone are the right approach at all – or whether resources should be directed at those who actually need supervision [28]. For you that means: insist on a home programme that goes beyond “maintain range of motion”.

12.1 How many sessions – and how often?

This is the most common question after surgery, and the guideline does not answer it: it gives no number [1]. That is not an omission. The studies use such different amounts that no recommendation can be derived from them. What can be said are orders of magnitude – and the factors that decide the individual case.

What actually happens in practice. The largest analysis covers 12,355 people after knee replacement in an American practice network. The mean was around 10.5 therapy sessions [37]. That is the best available reference point for an uncomplicated knee.

The same work contains a second, more important finding. After a uniform care pathway was introduced in those practices, results improved markedly (5.6 points on a daily-function scale) – with no change in the number of sessions. At the same time the differences between individual practices shrank by 63 per cent [37]. In other words: what happens within the session weighs more than the number of sessions.

What training-focused studies use. There the frequency is usually two to three appointments a week over six to twelve weeks. In the large trial comparing high and low intensity it was 26 sessions over eleven weeks [10] – considerably more than is usual in everyday care, and at the same time with no advantage over the gentler option.

More is not automatically better. In a study with 17 prescribed sessions over six weeks, improvement actually decreased as the number of sessions rose [38]. That figure needs care: people who progress poorly get more appointments – the direction of cause and effect runs backwards here. What it does show is still useful: there is no automatic rule that more appointments mean a better knee.

And how much supervision is needed at all? A synthesis of eleven randomised trials with 1884 cases compared supervised therapy sessions with a taught home programme and found no difference in range of motion, strength or quality of life – neither in the short nor the long term [39]. This does not contradict what is written above: physiotherapy with exercise did better than minimal supervision [21]. Taken together the two give a coherent picture – what matters is that the exercise happens, not who is standing next to you.

The Swiss framework. Here the health insurance benefits ordinance sets the structure: one medical prescription covers a maximum of nine sessions, and the first treatment must take place within five weeks. The first four series – 36 sessions in total – can be prescribed freely by the doctor. If therapy is to go beyond that, it requires a report to the insurer's medical officer and their approval [40].

For an uncomplicated knee replacement that means you will usually stay within one to three series. The 36-session limit rarely becomes an issue – and if it does, that is a signal to review the whole course, not merely a formality.

One distinction simplifies everything: therapy appointments are not the same as training sessions. Training happens two to three times a week – regardless of how many appointments you have. The appointments serve to check the programme, readjust the load and correct mistakes. That is why it makes sense to come more closely spaced in the first weeks (about twice weekly, while swelling and activation set the pace) and to thin out later (every one to two weeks) as the home share grows.

What the number depends on in the individual case. Markedly more supervision is usually needed by people for whom several of the following apply:

• A poor starting point before surgery – little strength, short walking distance, long-standing avoidance. This is the strongest single factor [7].
• Thigh activation does not return, the straight leg cannot be lifted (section 4).
• Range of motion stalls – then measurement gets closer, not rarer (section 13.1).
• Accompanying problems: the other knee, the hip, back trouble, dizziness, a neurological condition.
• Marked fear of movement – it measurably lengthens recovery [31].
• The living situation: living alone, steep stairs, no lift, nobody to step in.
• Demanding goals: returning to physically demanding work or to sport takes more than everyday mobility [25].

Markedly less is needed by those who went into surgery with good strength, exercise reliably at home and have access to equipment or a gym. For this group a few review appointments after the introductory phase can be enough.

The practical test is therefore not “how many sessions am I still entitled to” but: what has changed since the last measurement? As long as strength, range of motion and walking distance are measurably improving, the dose is right. If it stalls for several weeks, the programme changes – or we clarify whether the cause lies outside physiotherapy.

13. When things do not go smoothly

13.1 The knee will not become mobile

A knee that stays stiff is a recognised reason for a second operation: in the Swiss registry, joint stiffness accounted for 8.2 per cent of early revisions [2]. If flexion stays markedly below about 90 degrees or the knee cannot be straightened, this belongs on the table early [29].

The best-known treatment for it is manipulation under anaesthesia: the knee is bent past its resistance while you are asleep, so that the adhesions tear. It works – but it is not the only route, and the question of what comes first is less well studied than you would expect.

The honest starting point: there is no randomised trial pitting “manipulation under anaesthesia” directly against “intensified physiotherapy”. The guideline does not address it either – stiffness is one of the situations it explicitly sets aside as an individual case [1]. What follows is therefore a weighing-up, not a deduction.

What argues for trying therapy first. A feasibility study treated ten people with early-stage scarring from week six onwards for four weeks with a combined programme – manual therapy, exercise and a slowly adjusting splint. Seven out of ten reached a functional flexion of at least 110 degrees and avoided anaesthesia; the final result was as good as in a comparison group that had been manipulated (110 versus 109 degrees) [34]. The size belongs to the picture, though: ten people, with the comparison group assembled retrospectively. That is a signal, not proof. And for three out of ten the four weeks achieved nothing – they were manipulated afterwards anyway.

What argues for manipulation. Its effect is well documented. In a study across 15 centres with 124 people whose flexion was below 90 degrees between weeks 4 and 12, the procedure produced 46 degrees more flexion immediately, 28 degrees at six weeks and 37 degrees at one year [33]. A French analysis of 344 cases still found a gain of 36 degrees five years later, a complication rate of 2.3 per cent and satisfaction of 77.5 per cent [35].

And what argues against it. Those 2.3 per cent are few, but they are serious: among them were a rupture of the patellar tendon and an avulsion of the tibial tuberosity – the piece of bone that tendon attaches to [35]. There is also a finding that gets quoted often: in a large database analysis, the rate of further surgery two years after a manipulation was higher than without one (7.3 versus 4.9 per cent) [36]. That figure needs reading with care – anyone who needs a manipulation has the more difficult knee to begin with. The comparison therefore says more about how those knees fare than about what the manipulation does.

The real conflict is time pressure. A meta-analysis of 14 studies covering 13,445 knees compared early manipulation (mostly within three months) with delayed manipulation: the mean gain in flexion was 32 degrees early and only 19 degrees late [32]. So the longer a therapy attempt runs, the less well manipulation works if it turns out to be needed after all. That, and not which method is “better”, is where the difficulty lies.

What can practically be drawn from this:

• Do not wait – concentrate. If range of motion stalls, the right answer is not “let us look again in two weeks” but to intensify therapy and measure more often.
• An attempt with a deadline. In the study cited, the window was four weeks – anyone showing no progress within it was manipulated [34]. Setting a clear deadline is better than carrying on open-endedly.
• The decision falls within the first quarter of a year, not after it [32]. The reference point used in the studies is flexion below 90 degrees between weeks 4 and 12 [33].
• Anti-inflammatories as an add-on achieve nothing. In the only randomised trial on this, a combination of a steroid and a painkiller improved the result after manipulation neither at six weeks nor at one year [33].
• The decision belongs to the operating clinic, not to physiotherapy. Our job is to measure the course and report in good time.

This is one of the few points where waiting causes harm. If you notice at week six that range of motion has stalled, get in touch – not at week twelve.

13.2 The pain persists

Around a fifth of people still have relevant pain at three months [4]. For a long time there was no structured offer for them. A large British trial tested a care pathway: people with pain three months after surgery received a detailed assessment by a specialist practitioner, targeted onward referrals depending on the cause found, and telephone follow-up over twelve months. The pathway improved pain severity and pain interference compared with usual care [30].

The message lies less in the size of the effect than in the attitude: persistent pain after a knee replacement is a finding to be investigated – not a state to be accepted. Possible causes range from muscle weakness through instability or a kneecap problem to sensitisation of the pain system, and they are treated in entirely different ways.

13.3 Fear of movement

A systematic review shows that marked fear of movement (the technical term is kinesiophobia) impairs function for up to a year after a knee replacement, and active range of motion for up to six months [31]. The sequence is familiar: someone afraid of damaging the joint loads it less. Loading less means getting weaker. Getting weaker means more symptoms – and thus more reason for caution.

This is not a character flaw but a treatable factor. The guideline explicitly recommends psychologically informed techniques for people with high fear of movement [1]. In practice it usually starts more simply: with explanation. Knowing that an implant carries several times your body weight changes a surprising amount.

14. A realistic timeline

The following figures are orientation values drawn from the cited literature and from practice, not prescriptions. Individual courses vary – in both directions.

• Day 0 to 3: get up, walk, cool, elevate, secure extension, contract the thigh deliberately. First walking distances with a walking aid.
• Week 1 to 3: build range of motion, electrical stimulation if activation is a problem, increase walking distance, practise stairs. Swelling sets the pace.
• Week 4 to 12: the actual training phase. Progressive strength training two to three times a week, balance, gait retraining. This is where most of it is decided.
• Month 3 to 6: the biggest improvements fall in this window [6]. Return to work on average around week 13 [25], to light sport around week 20.
• Month 6 to 12: slower progress, but progress nonetheless – provided training continues. This is exactly where most people stop.
• After one year: whatever is still missing is rarely a question of the implant and usually a question of strength.

And a reassuring figure for the beginning: the worst measurement is often at week four. Being worse in the first weeks than before surgery is the normal course, not an alarm signal [7].

15. Six common misunderstandings

“The implant makes the knee healthy again.”
It makes it largely pain-free. Measured function, by contrast, barely improves on its own in the studies [6], and six months afterwards it is still well below that of healthy peers [7]. Whatever is meant to come back has to be trained.

“The motion machine helps the knee become mobile.”
The guideline explicitly advises against it, on the basis of four high-quality and twelve moderate-quality studies [1]. Being moved passively is no substitute for active practice.

“I have to spare the knee so the implant lasts.”
Four out of five total replacements last 25 years [3], and the most common early reasons for revision – kneecap problems, infection, instability – have nothing to do with training volume [2]. More dangerous than loading is the loss of strength that leads to falls [23].

“If I have less pain I will automatically move more.”
Measured objectively, hardly anyone does. Six months after surgery activity was essentially unchanged, and at one year it was still clearly below that of healthy comparison groups [8].

“After six weeks I should be back to normal.”
The biggest gains fall in the first three to six months [6], the mean return to work is at just under 13 weeks [25], to sport at 20 weeks. Six weeks is the beginning, not the goal.

“If it still hurts at three months, the operation failed.”
At three months around 22 per cent still have relevant pain [4]. There is a tested assessment and treatment pathway for it [30] – but you have to raise the subject.

16. When to get in touch

Seek medical assessment immediately – these are not matters for physiotherapy:

• Fever, chills, increasing redness or warmth, wound discharge or suddenly sharply increasing pain – suspected infection.
• New calf pain, one-sided swelling of the lower leg, shortness of breath or chest pain – suspected thrombosis or pulmonary embolism.
• A knee that suddenly gives way or locks.

Raise in physiotherapy:

• Range of motion has stalled for two to three weeks, or full extension is still missing.
• The knee is markedly thicker after every session and moves worse the following day.
• You cannot contract the thigh deliberately or lift the straight leg.
• You have fallen or feel unsteady when walking.
• You avoid movements for fear of damaging the joint.
• Three months after surgery you still have relevant pain.
• You have a goal that has never been discussed – kneeling, cycling, back up the mountain, back to work.

What to expect in our practice: a baseline assessment (range of motion, circumference, thigh strength and activation, sit-to-stand test, walking speed, balance), from which a training programme with clear progression steps is built – and, this is the decisive part, a home programme for the time between appointments. After four to six weeks we measure again. If nothing improves although you have trained regularly, the question goes back to the operating clinic.

17. In summary

A knee replacement is a good operation with a clear, narrowly defined promise: less pain. It keeps that promise for most people, and four out of five joints last a quarter of a century.

Everything else – strength, walking, stairs, steadiness on your feet, the return to work and sport – does not come with the implant. The evidence on this is unusually clear: progressive strength training and the training of walking, balance and movement patterns carry the two strongest recommendations of the guideline [1]. Motion machines, splints, lymphatic drainage and compression dressings carry none.

What you need for it is manageable: use the weeks before surgery, get up on day one, secure extension early, make the thigh recruitable again in the first weeks – with electricity if necessary – then two to three demanding strength sessions a week for at least a quarter of a year, with balance and gait retraining alongside, and the willingness not to settle after six weeks for whatever came back by itself.

The knee is adaptable at 75 too. More slowly than at 40 – but it answers.

If you would like to practise going down and coming back up while learning something about the knee: our game Kniewerk takes you from the valley to the summit in fifteen clean squats. It does not replace therapy – but it makes tempo and depth tangible.

References

All Digital Object Identifiers (DOIs) were individually verified against the Crossref register. The links in the reference list lead via the DOI service to the publishers' pages. Some of these are located outside Switzerland and the EU. When you click them, your IP address is transmitted to the provider concerned – on our own site that does not happen.

[1] Bove AM, Carroll LA, Cone S, et al. Clinical practice guideline for physical therapist management of total knee arthroplasty: revision 2026. Physical Therapy. 2026;106(7):pzag058. https://doi.org/10.1093/ptj/pzag058

[2] Stiftung für Qualitätssicherung in der Implantationsmedizin, ANQ. Implantatregister SIRIS Hüfte und Knie. Kurzfassung – SIRIS Report 2024: Jahresbericht Hüft- und Knieimplantate 2012–2023. Bern; Dezember 2024. anq.ch (PDF)

[3] Evans JT, Walker RW, Evans JP, Blom AW, Sayers A, Whitehouse MR. How long does a knee replacement last? A systematic review and meta-analysis of case series and national registry reports with more than 15 years of follow-up. The Lancet. 2019;393(10172):655–663. https://doi.org/10.1016/S0140-6736(18)32531-5

[4] Cheng HY, Beswick AD, Bertram W, et al. What proportion of people have long-term pain after total hip or knee replacement? An update of a systematic review and meta-analysis. BMJ Open. 2025;15(5):e088975. https://doi.org/10.1136/bmjopen-2024-088975

[5] DeFrance MJ, Scuderi GR. Are 20% of Patients Actually Dissatisfied Following Total Knee Arthroplasty? A Systematic Review of the Literature. The Journal of Arthroplasty. 2023;38(3):594–599. https://doi.org/10.1016/j.arth.2022.10.011

[6] Orange GM, Hince DA, Travers MJ, et al. Physical Function Following Total Knee Arthroplasty for Osteoarthritis: A Longitudinal Systematic Review With Meta-analysis. Journal of Orthopaedic & Sports Physical Therapy. 2025;55(1):1–11. https://doi.org/10.2519/jospt.2024.12570

[7] Bade MJ, Kohrt WM, Stevens-Lapsley JE. Outcomes Before and After Total Knee Arthroplasty Compared to Healthy Adults. Journal of Orthopaedic & Sports Physical Therapy. 2010;40(9):559–567. https://doi.org/10.2519/jospt.2010.3317

[8] Arnold JB, Walters JL, Ferrar KE. Does Physical Activity Increase After Total Hip or Knee Arthroplasty for Osteoarthritis? A Systematic Review. Journal of Orthopaedic & Sports Physical Therapy. 2016;46(6):431–442. https://doi.org/10.2519/jospt.2016.6449

[9] Mizner RL, Petterson SC, Stevens JE, Vandenborne K, Snyder-Mackler L. Early Quadriceps Strength Loss After Total Knee Arthroplasty: The Contributions of Muscle Atrophy and Failure of Voluntary Muscle Activation. The Journal of Bone & Joint Surgery. 2005;87(5):1047–1053. https://doi.org/10.2106/JBJS.D.01992

[10] Bade MJ, Struessel T, Dayton M, et al. Early High-Intensity Versus Low-Intensity Rehabilitation After Total Knee Arthroplasty: A Randomized Controlled Trial. Arthritis Care & Research. 2017;69(9):1360–1368. https://doi.org/10.1002/acr.23139

[11] Nunes EA, Stokes T, McKendry J, Currier BS, Phillips SM. Disuse-induced skeletal muscle atrophy in disease and nondisease states in humans: mechanisms, prevention, and recovery strategies. American Journal of Physiology – Cell Physiology. 2022;322(6):C1068–C1084. https://doi.org/10.1152/ajpcell.00425.2021

[12] Skou ST, Roos EM, Laursen MB, et al. A Randomized, Controlled Trial of Total Knee Replacement. New England Journal of Medicine. 2015;373(17):1597–1606. https://doi.org/10.1056/NEJMoa1505467

[13] Bannuru RR, Osani MC, Vaysbrot EE, et al. OARSI guidelines for the non-surgical management of knee, hip, and polyarticular osteoarthritis. Osteoarthritis and Cartilage. 2019;27(11):1578–1589. https://doi.org/10.1016/j.joca.2019.06.011

[14] Kolasinski SL, Neogi T, Hochberg MC, et al. 2019 American College of Rheumatology/Arthritis Foundation Guideline for the Management of Osteoarthritis of the Hand, Hip, and Knee. Arthritis & Rheumatology. 2020;72(2):220–233. https://doi.org/10.1002/art.41142

[15] Fransen M, McConnell S, Harmer AR, Van der Esch M, Simic M, Bennell KL. Exercise for osteoarthritis of the knee. Cochrane Database of Systematic Reviews. 2015;(1):CD004376. https://doi.org/10.1002/14651858.CD004376.pub3

[16] Skou ST, Roos EM. Good Life with osteoArthritis in Denmark (GLA:D™): evidence-based education and supervised neuromuscular exercise delivered by certified physiotherapists nationwide. BMC Musculoskeletal Disorders. 2017;18(1):72. https://doi.org/10.1186/s12891-017-1439-y

[17] Punnoose A, Claydon-Mueller LS, Weiss O, Zhang J, Rushton A, Khanduja V. Prehabilitation for Patients Undergoing Orthopedic Surgery: A Systematic Review and Meta-analysis. JAMA Network Open. 2023;6(4):e238050. https://doi.org/10.1001/jamanetworkopen.2023.8050

[18] Jørgensen SL, Aagaard P, Bohn MB, et al. The Effect of Blood Flow Restriction Exercise Prior to Total Knee Arthroplasty on Postoperative Physical Function, Lower Limb Strength and Patient-Reported Outcomes: A Randomized Controlled Trial. Scandinavian Journal of Medicine & Science in Sports. 2024;34(11):e14750. https://doi.org/10.1111/sms.14750

[19] Wainwright TW, Gill M, McDonald DA, et al. Consensus statement for perioperative care in total hip replacement and total knee replacement surgery: Enhanced Recovery After Surgery (ERAS®) Society recommendations. Acta Orthopaedica. 2020;91(1):3–19. https://doi.org/10.1080/17453674.2019.1683790

[20] Liu MM, Tian M, Luo C, Wang S, Cheng Y. Continuous cryotherapy vs. traditional cryotherapy after total knee arthroplasty: A systematic review and meta-analysis of randomized controlled trials. Frontiers in Surgery. 2023;9:1073288. https://doi.org/10.3389/fsurg.2022.1073288

[21] Artz N, Elvers KT, Lowe CM, Sackley C, Jepson P, Beswick AD. Effectiveness of physiotherapy exercise following total knee replacement: systematic review and meta-analysis. BMC Musculoskeletal Disorders. 2015;16:15. https://doi.org/10.1186/s12891-015-0469-6

[22] Stevens-Lapsley JE, Balter JE, Wolfe P, Eckhoff DG, Kohrt WM. Early Neuromuscular Electrical Stimulation to Improve Quadriceps Muscle Strength After Total Knee Arthroplasty: A Randomized Controlled Trial. Physical Therapy. 2012;92(2):210–226. https://doi.org/10.2522/ptj.20110124

[23] Chan ACM, Jehu DA, Pang MYC. Falls After Total Knee Arthroplasty: Frequency, Circumstances, and Associated Factors – A Prospective Cohort Study. Physical Therapy. 2018;98(9):767–778. https://doi.org/10.1093/ptj/pzy071

[24] Christiansen CL, Kline PW, Anderson CB, et al. Optimizing Total Knee Arthroplasty Rehabilitation With Telehealth Physical Activity Behavior Change Intervention: A Randomized Clinical Trial. Physical Therapy. 2024;104(10):pzae088. https://doi.org/10.1093/ptj/pzae088

[25] Pasqualini I, Emara AK, Rullan PJ, et al. Return to Sports and Return to Work After Total Knee Arthroplasty: A Systematic Review and Meta-Analysis. JBJS Reviews. 2023;11(7):e22.00250. https://doi.org/10.2106/JBJS.RVW.22.00250

[26] Nadeem S, Mundi R, Chaudhry H. Surgery-related predictors of kneeling ability following total knee arthroplasty: a systematic review and meta-analysis. Knee Surgery & Related Research. 2021;33(1):36. https://doi.org/10.1186/s43019-021-00117-z

[27] Jiang S, Xiang J, Gao X, Guo K, Liu B. The comparison of telerehabilitation and face-to-face rehabilitation after total knee arthroplasty: A systematic review and meta-analysis. Journal of Telemedicine and Telecare. 2018;24(4):257–262. https://doi.org/10.1177/1357633X16686748

[28] Bandholm T, Kehlet H. Physiotherapy Exercise After Fast-Track Total Hip and Knee Arthroplasty: Time for Reconsideration? Archives of Physical Medicine and Rehabilitation. 2012;93(7):1292–1294. https://doi.org/10.1016/j.apmr.2012.02.014

[29] Zachwieja E, Perez J, Hardaker WM, Levine B, Sheth N. Manipulation Under Anesthesia and Stiffness After Total Knee Arthroplasty. JBJS Reviews. 2018;6(4):e2. https://doi.org/10.2106/JBJS.RVW.17.00113

[30] Wylde V, Bertram W, Sanderson E, et al. The STAR care pathway for patients with pain at 3 months after total knee replacement: a multicentre, pragmatic, randomised, controlled trial. The Lancet Rheumatology. 2022;4(3):e188–e197. https://doi.org/10.1016/S2665-9913(21)00371-4

[31] Brown OS, Hu L, Demetriou C, Smith TO, Hing CB. The effects of kinesiophobia on outcome following total knee replacement: a systematic review. Archives of Orthopaedic and Trauma Surgery. 2020;140(12):2057–2070. https://doi.org/10.1007/s00402-020-03582-5

[32] Akhtar M, Razick D, Seibel A, et al. Outcomes of Early Versus Delayed Manipulation Under Anesthesia for Stiffness Following Total Knee Arthroplasty: A Systematic Review and Meta-Analysis. The Journal of Arthroplasty. 2024;39(11):2872–2879. https://doi.org/10.1016/j.arth.2024.05.059

[33] Abdel MP, Salmons HI, Larson DR, et al. The Chitranjan S. Ranawat Award: Manipulation Under Anesthesia to Treat Postoperative Stiffness After Total Knee Arthroplasty: A Multicenter Randomized Clinical Trial. The Journal of Arthroplasty. 2024;39(8S1):S9–S14.e1. https://doi.org/10.1016/j.arth.2024.02.034

[34] Rauzi MR, Foran JRH, Bade MJ. Multimodal conservative management of arthrofibrosis after total knee arthroplasty compared to manipulation under anesthesia: a feasibility study with retrospective cohort comparison. Pilot and Feasibility Studies. 2022;8(1):71. https://doi.org/10.1186/s40814-022-01026-y

[35] Roullet CA, Descamps S, Monadjemi S, et al. Manipulation under anesthesia for stiffness after total knee arthroplasty: A French multicenter study with 5 years' follow-up including 344 cases. Orthopaedics & Traumatology: Surgery & Research. 2026;112:104432. https://doi.org/10.1016/j.otsr.2025.104432

[36] DeFrance MJ, Cheesman QT, Hameed D, DiCiurcio WT, Harrer MF. Manipulation Under Anesthesia Is Associated With an Increased Rate of Early Total Knee Arthroplasty Revision. Orthopedics. 2022;45(5):270–275. https://doi.org/10.3928/01477447-20220608-01

[37] Capin JJ, Minick KI, Stevens-Lapsley JE, et al. Variation in Outcomes and Number of Visits Following Care Guideline Implementation: Part 2 of an Analysis of 12 355 Patients After Total Knee Arthroplasty. Journal of Orthopaedic & Sports Physical Therapy. 2023;53(3):151–158. https://doi.org/10.2519/jospt.2022.11370

[38] Benes G, Adams Z, Dubic M, David J, Leonardi C, Bronstone A, Dasa V. Optimal Duration of Physical Therapy Following Total Knee Arthroplasty. Geriatric Orthopaedic Surgery & Rehabilitation. 2024;15:21514593241250149. https://doi.org/10.1177/21514593241250149

[39] Chaudhry YP, Hayes H, Wells Z, Papadelis E, Khanuja HS, Deirmengian C. Not All Patients Need Supervised Physical Therapy After Primary Total Knee Arthroplasty: A Systematic Review and Meta-Analysis. Cureus. 2023;15(2):e35232. https://doi.org/10.7759/cureus.35232

[40] Verordnung des EDI über Leistungen in der obligatorischen Krankenpflegeversicherung (KLV), SR 832.112.31, Art. 5 (Physiotherapie). Zusammengefasst in: FMH. Verordnung von Physiotherapie im KVG-Bereich – Frequently Asked Questions. Stand 31.10.2022. fmh.ch (PDF)