Meniscus Tear: Surgery or Rehab? What the Evidence Says About Recovery, Prognosis, and Return to Sport

Key Takeaways: Meniscus Tear

A meniscus tear on MRI is not automatically the cause of knee pain. In a general-population study of 991 adults aged 50–90, 61% of those who had a tear reported no knee pain, aching, or stiffness during the previous month [1]. In practice, the MRI finding needs to be interpreted together with the patient’s age, symptom onset, examination and the overall condition of the knee.
For degenerative tears in middle-aged adults, arthroscopic partial meniscectomy has not outperformed placebo surgery or exercise therapy in randomised trials — including the Finnish FIDELITY trial: placebo surgery at 1 and 5 years [2][3], exercise therapy at 2–10 years [5][6][7], and a 10-year letter reporting no benefit and possible harm [4]. Clinically, degenerative meniscal findings often need to be viewed as part of the condition of the whole joint rather than as an isolated pain generator.
In a degenerative knee, catching and occasional locking have not been shown to justify surgery: meniscectomy gave no added relief over sham surgery, and the authors urged caution in using these symptoms as an indication [3][8][9]. A truly locked knee that cannot straighten is a different clinical situation and warrants individual orthopaedic assessment.
In the OMEX trial cohort, most participants improved whichever treatment they received: in an analysis of all 140 participants, 9 in 10 improved at least gradually within 2 years [10], and at 10 years both groups reported improved pain and knee function [7]. Improvement in symptoms, however, is not the same as reversal of underlying joint degeneration.
In younger adults, early surgery was not superior to a rehab-first strategy in two randomised trials: STARR studied recent traumatic tears and DREAM studied a broader group of young adults with meniscal tears [11][12]. Between about a quarter and two-fifths of the rehabilitation groups underwent surgery during trial follow-up, and a secondary DREAM analysis found surgery relieved self-reported mechanical symptoms better [11][12][13].
An important evidence gap remains for clearly repairable tears. STARR excluded tears that appeared suitable for repair, while DREAM included repair within a broader surgical strategy rather than directly comparing meniscal repair with rehabilitation alone [11][12]. Good randomised evidence for repair versus non-operative rehabilitation is therefore still lacking.
Some tear patterns are treated separately in the literature. ESSKA consensus generally recommends preserving or repairing the meniscus for displaced bucket-handle tears, root tears and repairable traumatic tears, although these recommendations rest largely on low-level evidence [14][15].
In pooled athlete studies, the shortest mean return-to-sport time was reported after meniscectomy (2.1 months) and a longer one after repair (5.8 months), with similar return-to-sport rates [17]. Repair may be associated with a lower long-term rate of consultation for knee osteoarthritis, but that evidence is observational [16][17].
The final decision about surgery remains individual. Trial evidence and consensus recommendations help frame the options, but whether a meniscal finding is clinically important — and whether operative treatment is appropriate — depends on the tear pattern, mechanical symptoms, the condition of the rest of the knee and the treating orthopaedic surgeon’s assessment.

Introduction: Meniscus Tear

Knee problems, and meniscal complaints in particular, are a common reason for patients to come to my practice. What makes these cases challenging is that there are several diagnostic traps along the way. Knee pain can arise for many different reasons, sometimes without a single clearly identifiable cause, and pain around the joint line does not automatically mean that the meniscus is responsible. In older patients, osteoarthritis or more general degenerative changes within the joint can produce a very similar clinical picture. Younger patients can also develop less specific knee pain that happens to localise around the joint line.

MRI can add another layer of uncertainty. Meniscal tears are common MRI findings, particularly with increasing age [1], and a tear seen on imaging is not necessarily the structure causing the patient’s symptoms. In practice, that creates an important diagnostic question: does the history, examination and imaging fit together well enough to make the meniscal finding a plausible explanation for the pain?

The treatment question is separate. Arthroscopic partial meniscectomy — trimming away the torn part of the meniscus — is one of the most common orthopaedic procedures [2]. For many years the logic seemed simple: the knee hurts, the scan shows a tear, so remove the torn tissue. But over the past decade, that logic has been tested in randomised trials, including the sham-controlled FIDELITY trial, and the results have informed recommendations against routine arthroscopy for degenerative meniscal tears [9]. One important source of evidence comes from Finland: the FIDELITY trial compared partial meniscectomy with placebo surgery. A 2026 letter reports 10-year outcomes and states no benefit and possible harm from the procedure [4].

This is also how I tend to think about these cases clinically: first, is the meniscus actually the likely source of the symptoms? And only then, what treatment approach makes sense? Rehabilitation may be the more appropriate path in many situations, while operative treatment can become relevant in others. Those are separate decisions, and confusing them is one of the reasons meniscal problems can be more complicated than they first appear.

The questions patients ask are usually very concrete. Do I need surgery? If I don’t have it, how long will the knee hurt? Will it become unstable or arthritic? And for athletes: what does this mean for my season — and for my knee ten years from now?

Degenerative vs Traumatic Meniscus Tear: Why the Distinction Matters

A key step in interpreting the evidence is separating two different conditions that share the same name.

The European Society for Sports Traumatology, Knee Surgery and Arthroscopy (ESSKA) defines a traumatic meniscus tear as one linked to a sufficient knee injury with a sudden onset of pain. Vertical tears — longitudinal tears, including bucket-handle tears, and radial tears — are primarily included in this group [14]. A degenerative meniscus lesion, by contrast, develops through slow tissue degeneration without a clear injury, and ESSKA classifies horizontal tears as degenerative in nature even when they occur in younger patients [14]. The consensus separates the two because they are managed differently [14].

A second useful distinction is between stable and unstable tears. In an unstable tear, the inner fragment of the meniscus can shift into the joint, which can cause locking and sudden pain; a longitudinal tear that evolves into a bucket-handle tear is the classic example [14].4][5][7]

In clinical practice, degenerative meniscal findings often appear in a broader context of joint degeneration. In some older patients, for example, the joint space is already visibly narrowed, suggesting developing or established osteoarthritic change. At that stage, I find it less useful to think about the meniscus as an isolated structure. The meniscus itself may also be degenerative, and it can become difficult to know how much of the patient’s pain is actually coming from the meniscus and how much reflects the condition of the joint as a whole. This is one reason a degenerative meniscal finding on imaging does not automatically identify the main pain generator.

This distinction also matters because the research is split along the same line. The large surgery-versus-placebo and surgery-versus-exercise trials for degenerative tears enrolled middle-aged adults — for example, ages 35–65 in FIDELITY [2] and 45–70 in the Dutch ESCAPE trial [5]. The two trials in younger adults were STARR, which studied recent traumatic tears in patients aged 18–45 [11], and DREAM, which enrolled adults aged 18–40 with MRI-verified meniscal tears [12]. The two age groups and tear types should not be treated as interchangeable when interpreting trial results.

How Often Is a Tear Just an Incidental MRI Finding?

Before deciding whether a meniscus tear needs treatment, it helps to know how often tears exist without causing any symptoms at all.

In a population-based study from Framingham, Massachusetts, researchers performed knee MRI on 991 people aged 50–90 who were recruited without regard to whether they had knee problems [1]. The prevalence of a meniscal tear or meniscal destruction ranged from 19% in women aged 50–59 to 56% in men aged 70–90 [1]. Strikingly, 61% of the participants who had a meniscal tear reported no pain, aching, or stiffness during the previous month [1]. Meniscal findings were common in this general-population sample, and their prevalence rose with age [1].

The FIDELITY investigators make a related point: in a degenerative knee, symptoms do not necessarily come from the meniscus itself and may arise from other tissues or processes, such as bone marrow lesions [3].

This is where I think the patient has to be assessed as a whole rather than the MRI finding in isolation. In practice, I pay attention to the overall appearance of the joint: is there already joint-space narrowing or broader degenerative change, or does the rest of the knee look relatively well preserved? Age, functional status and the way the symptoms started also affect how convincing I find a meniscal finding as the explanation for the pain.

In an older patient with an already degenerative-looking knee, my attention is often less on the meniscus as an isolated structure and more on the condition of the joint as a whole. In a younger patient with an otherwise relatively well-preserved knee, a sudden onset of pain during a specific event — particularly when accompanied by true mechanical locking — can make a structural meniscal injury more clinically plausible. That does not mean the MRI alone establishes the diagnosis; rather, the imaging becomes more meaningful when it fits the history and the rest of the clinical picture.

Ultimately, however, the final assessment is individual. When the diagnosis or treatment choice remains uncertain, an orthopaedic evaluation can become important in deciding how much clinical significance to give the meniscal finding and whether operative treatment needs to be considered.

Degenerative Tears: Does Surgery Help?

The FIDELITY trial: surgery versus placebo surgery

The FIDELITY trial randomised 146 patients aged 35–65 with symptoms of a degenerative medial meniscus tear and no knee osteoarthritis to either arthroscopic partial meniscectomy or a sham procedure in which a meniscectomy was simulated [2]. Participants had had symptoms for more than 3 months that had not responded to conventional conservative care, and patients with an obvious traumatic onset or a recent locked knee were excluded [3].

At 12 months, both groups had improved by a similar amount. The Lysholm knee score improved by 21.7 points after meniscectomy and 23.3 points after sham surgery (between-group difference −1.6; 95% CI −7.2 to 4.0), with similar results for the meniscus-specific WOMET score and for knee pain after exercise [2].

At 5 years, there were still no relevant between-group differences in WOMET, Lysholm, or pain after exercise, and both groups reported sustained improvement [3]. Radiographic knee osteoarthritis, however, progressed slightly more often after meniscectomy: 72% versus 60% of participants progressed by at least one Kellgren–Lawrence grade (adjusted risk difference 13%; 95% CI −2% to 28%), and the OARSI radiographic sum score showed more progression in the meniscectomy group (difference 0.7; 95% CI 0.1 to 1.3) [3].

The 10-year results were published as a letter in 2026. It reports that participant-reported and radiologic outcomes over 10 years of follow-up suggested no benefit and possible harm from partial meniscectomy [4]. The letter’s exact scores, confidence intervals and event counts are not restated here. The original trial enrolled patients with no knee osteoarthritis [2]; my interpretation is that this is the group in which a benefit from surgery should have been easiest to detect.

Surgery versus exercise therapy

Sham-controlled trials answer whether removing the torn tissue does anything specific. Trials against exercise therapy answer the practical question: what should a patient actually do?

In the ESCAPE trial, 321 patients aged 45–70 with a degenerative, MRI-confirmed meniscal tear were randomised to partial meniscectomy or to 16 sessions of exercise-based physical therapy over 8 weeks; patients with a locked knee were excluded [5]. Over 5 years, physical therapy remained non-inferior. Knee function (IKDC score) improved by 29.6 points after surgery and 25.1 points after physical therapy, and the overall crude between-group difference of 3.5 points (95% CI 0.7 to 6.3) stayed within the trial’s non-inferiority threshold of 11 points [5]. During follow-up, 52 of 162 patients (32.1%) in the physical therapy group went on to have a delayed meniscectomy — and in the as-treated analysis, those who crossed over to surgery did not end up with better knee function than those who completed physical therapy alone [5].

The Norwegian–Danish OMEX trial randomised 140 adults (mean age 49.5) with a degenerative medial meniscal tear to 12 weeks of supervised exercise therapy or partial meniscectomy [6]. At 2 years, the difference in KOOS4 was 0.9 points (95% CI −4.3 to 6.1) — not clinically relevant — and thigh muscle strength had improved in the exercise group at 3 months [6]. Nineteen percent of the exercise group crossed over to surgery, without additional benefit [6]. At 10 years, radiographic knee osteoarthritis had developed in 23% after meniscectomy and 20% after exercise therapy (adjusted risk difference 3%; 95% CI −13% to 19%), with no clinically relevant differences in patient-reported outcomes or knee muscle strength, and both groups had improved [7].

For degenerative meniscal tears, the expected benefit from surgery is often limited. This is also reflected in how I approach these cases clinically: when the overall picture is clearly degenerative, I would not assume that removing part of the meniscus will necessarily address the patient’s symptoms.

In Finland, this also has practical consequences outside the consultation room. In my experience, insurance companies may be reluctant to approve payment commitments for surgery when the indication is a degenerative meniscal lesion rather than a clearly traumatic tear. That does not determine whether surgery is appropriate for an individual patient, but it reflects the broader shift away from treating degenerative meniscal findings as an automatic surgical problem.

What the guidelines conclude

Based on this body of evidence, the BMJ Rapid Recommendations panel issued a strong recommendation against arthroscopic surgery for nearly all patients with degenerative knee disease — explicitly including those with meniscal tears, mechanical symptoms, or a sudden onset of symptoms [9]. The exception the panel left open was a persistent, objectively locked knee [9].

Catching and Locking: Do Mechanical Symptoms Mean You Need Surgery?

Mechanical symptoms — a sensation of the knee catching, or briefly locking — are often taken as a reason to operate on a meniscus tear. The logic is intuitive: a loose flap gets caught, so remove it. In degenerative tears, the evidence does not support this.

In a secondary analysis of FIDELITY, 46% of the meniscectomy group and 49% of the sham group reported catching or locking before surgery; at any follow-up, the corresponding figures were 49% and 43%, a risk difference of 0.03 (95% CI −0.06 to 0.12) [8]. The authors concluded that resecting the torn meniscus did not relieve catching or occasional locking any better than sham surgery, and questioned whether these symptoms are caused by the degenerative tear at all [8]. At 5 years, mechanical symptoms were reported more often after meniscectomy (adjusted risk difference 18%; 95% CI 5% to 31%) [3]. I go through the broader evidence on knee noises — and why catching is so often over-read — in Is Joint Clicking Normal?

The picture may be different in younger adults with meniscal tears. In a secondary analysis of the DREAM trial, 63 of 121 young adults reported mechanical symptoms at baseline [13]. At 12 months, 9 of 26 (35%) in the surgery group and 20 of 29 (69%) in the exercise group still reported mechanical symptoms; the estimated risk difference across time points was 28.7% (95% CI 8.6% to 48.8%) [13]. Surgery appeared more effective for relieving self-reported mechanical symptoms — but not for improving pain, function, or quality of life [13]. This is a secondary analysis, so it should be treated as hypothesis-generating.

A truly locked knee is a separate category. The FIDELITY authors note that arthroscopic surgery is indicated when a patient cannot fully extend the knee because of certain tear types, such as a bucket-handle tear — while pointing out that such patients are only a small subset of those currently having meniscal surgery [3].

A truly locked knee is different from intermittent catching or occasional locking. Clinically, these cases can be difficult because the patient is often clearly symptomatic and functionally limited at the time of assessment. In that situation, it can be much harder to take a passive approach, even when the broader evidence around meniscal surgery is cautious.

This is one of the situations where I find treatment decisions less straightforward than they may appear on paper. A young patient with a knee that cannot move normally because of a mechanical block may be in considerable distress, and that clinical picture naturally increases the pressure to reach a definitive treatment plan. The decision is therefore not based only on whether surgery has shown benefit in broader meniscal-tear populations, but also on the severity and mechanical nature of the individual presentation.

In my experience in Finland, patients with a genuinely locked knee often end up being assessed for operative treatment. I would still frame this as an individual orthopaedic decision rather than an automatic indication for surgery, because the exact cause of the locking and the overall clinical picture still matter.

Meniscus Tears in Young, Active Adults: What Do the Trials Show?

Surgery has traditionally been widely used for meniscal tears in young adults. Two randomised trials have now compared early surgery with a rehab-first strategy.

The Dutch STARR trial enrolled 100 patients aged 18–45 (mean age 35.1, 76% male) with a recent, traumatic, isolated meniscal tear and no osteoarthritis; 34 of them were competitive or elite athletes [11]. Patients were excluded if they had a locked knee, a tear that looked suitable for repair on MRI, or a cruciate ligament rupture [11]. They were randomised to a strategy of early arthroscopic partial meniscectomy or physical therapy, with the option of delayed meniscectomy after 3 months. At 24 months, IKDC scores were 78 out of 100 in both groups (difference 0.1; 95% CI −7.6 to 7.7) [11]. In the physical therapy group, 21 patients (41%) eventually had a delayed meniscectomy — a median of 5.5 months after randomisation, ranging from 3 to 21 months — while the remaining 59% (calculated from the reported 41%) did not undergo delayed meniscectomy [11].

The Danish DREAM trial randomised 121 young adults aged 18–40 (mean age 29.7) with an MRI-verified tear who were eligible for surgery, either to early meniscal surgery (not limited to meniscectomy) or to 12 weeks of supervised exercise therapy and education, with the option of later surgery [12][13]. At 12 months, KOOS4 had improved by 19.2 points after surgery and 16.4 points after exercise (adjusted difference 5.4; 95% CI −0.7 to 11.4), and 26% of the exercise group had crossed over to surgery [12].

There is also an important gap in the trial evidence. STARR was designed to compare arthroscopic partial meniscectomy with physical therapy and excluded tears that looked repairable on MRI [11]. DREAM included meniscal repair within its broader surgical strategy, but it did not provide a direct head-to-head comparison of meniscal repair versus rehabilitation alone [12]. Good randomised evidence is therefore still lacking for the specific question of how a clearly repairable meniscal tear should be managed compared with a non-operative rehabilitation strategy. ESSKA’s consensus statements on repairable tears are summarised in the next section [14].

Meniscus Tear Prognosis: How Long Will It Hurt, and What Happens to the Knee?

How long does it take to improve?

There is no single number, but the trials give a fairly consistent picture. In a secondary analysis of the OMEX trial, researchers analysed all 140 participants (both treatment groups) and identified three recovery trajectories in patients with degenerative tears: low with minimal improvement (10–12% of participants), moderate with gradual improvement (20–36%), and high with early improvement (53–70%) [10]. Nine in 10 improved at least gradually over 2 years [10]. In ESCAPE, the improvement achieved during the first 2 years was maintained at 5 years [5].

The exercise programmes used in these trials ran for 8–12 weeks [5][6][12][13], and STARR used 3 months of physical therapy as the point at which surgery could be reconsidered [11]. My synthesis of these trial designs: a fair trial of rehabilitation means a structured programme of roughly 2–3 months, with further gradual improvement often continuing after that — not a few weeks of rest followed by a decision.

What does the knee look like years later?

The placebo group of FIDELITY provides long-term randomised data on patients whose torn meniscal tissue was not resected, although they still underwent diagnostic arthroscopy. The sham group still underwent diagnostic arthroscopy, so it is not an untreated natural history. The 10-year letter summarises that comparison as no benefit and possible harm from removing the torn tissue; absolute sham-group scores are not restated here [4]. In OMEX, 20% of the exercise group developed radiographic knee osteoarthritis by 10 years [7], and in ESCAPE about half of participants in both groups had at least a 1-point increase in the OARSI radiographic sum score by 5 years [5]. My interpretation: a degenerative tear is often one feature of a knee that is already ageing, and removing the tear has not been shown to slow that process — if anything, the FIDELITY data hint the opposite [3][4].

Will the knee become unstable?

This question deserves an honest answer: the trials I reviewed provide no direct evidence on recurrent giving way. What they do show is that the evidence for conservative care comes largely from ligament-stable knees — both ESCAPE and STARR excluded cruciate ligament ruptures [5][11]. ESSKA likewise separates functional instability (a symptom) from laxity (a measurable sign) and defines a stable knee as one with intact ligaments [14]. In practice, a knee that genuinely gives way should prompt an assessment of the ligaments rather than being attributed automatically to the meniscus — that is clinical synthesis rather than a trial finding.

Does a meniscus tear heal on its own?

ESSKA states that no results are available on the self-healing potential of isolated meniscal tears [14]. The improvement seen in the conservative arms of the trials above should therefore be understood as symptoms and function improving — not as proof that the tear itself has healed.

One reassuring message for many patients with a meniscal problem is that symptoms often improve over time, regardless of which treatment path is chosen. I think this can be an important part of the discussion, because patients may otherwise interpret the MRI finding as something that will remain permanently symptomatic unless it is fixed.

I am more cautious with that reassurance when the knee is already clearly degenerative. If there is established osteoarthritis, I would not usually describe the joint itself as “healing” in the same way. In practice, I tend to explain that symptoms may fluctuate — with better and worse periods — while the underlying degenerative process can continue over time. For a middle-aged patient with a degenerative knee, that distinction matters: the current pain episode may settle, but the joint may still gradually wear over the years.

From a clinical perspective, I therefore try to separate improvement in symptoms from reversal of the underlying structural change. Those are not the same thing, and making that distinction can help set more realistic expectations.

Locked Knees, Root Tears, and Repairable Tears: What the Literature Reports

The trials above mostly concern degenerative tears or isolated, non-locked tears. The literature describes several other tear patterns separately. For most of them the evidence is consensus-level or observational rather than randomised.

A truly locked knee or displaced bucket-handle tear. A knee that cannot fully straighten because of a displaced fragment is the situation in which the FIDELITY authors say arthroscopic surgery is indicated [3], and even the BMJ guideline that recommends against arthroscopy for degenerative knee disease leaves an exception for a persistent, objectively locked knee [9]. For a non-reducible bucket-handle tear combined with an ACL injury, ESSKA recommends prompt repair of the meniscus [14].

Root tears. A meniscal root tear detaches the meniscus at its anchor point, and the ESSKA consensus treats root tears as a distinct category [14]. In a retrospective cohort of 52 patients (mean age 58) with symptomatic medial meniscus posterior root tears treated non-operatively, 31% underwent total knee arthroplasty at a mean of 30 ± 32 months after diagnosis, and 87% overall failed non-operative treatment (failure defined as conversion to arthroplasty or a severely abnormal IKDC score) [15]. ESSKA notes that medial root tears are generally degenerative in nature, yet root repair is typically recommended because it gives better clinical outcomes than conservative treatment or partial meniscectomy; posterolateral root tears, which often occur with ACL tears, should also be repaired, particularly when the meniscofemoral ligament is absent or injured (both graded as low-level evidence) [14]. In the ESSKA consensus, root tears are therefore a distinct category in which repair is typically recommended in suitable cases; this recommendation rests on low-level evidence.

Repairable traumatic tears. ESSKA’s 2019 consensus states that preserving the meniscus should be the first-line treatment for traumatic tears, because long-term clinical and radiological outcomes are worse after partial meniscectomy than after repair [14]. It reserves partial meniscectomy for situations where repair or leaving the tear alone is not possible, such as complex, highly degenerative, or flap tears [14]. ESSKA recommends that when repair is required, it be done as early as possible, although the supporting evidence is low-level [14]. It is worth noting how thin the evidence base is: of the consensus’s 27 questions, only one received the highest evidence grade [14].

Ultimately, the decision about surgery is always individual and belongs to the treating orthopaedic surgeon. Trial results and consensus recommendations can guide that decision, but they cannot replace an assessment of the actual patient, the tear pattern, the symptoms and the overall condition of the knee.

From my perspective, this is also an important point when discussing treatment in Finland. I have a great deal of confidence in the standard of orthopaedic care here, and particularly in the public healthcare system I often see treatment decisions being made with close attention to current evidence and guideline-based practice. That does not mean every case is managed in exactly the same way, but it does mean that a recommendation for or against surgery is usually made in a broader clinical context rather than from the MRI finding alone.

Meniscus Tear in Athletes: What Studies Report on Meniscectomy, Repair, and Rehab

Studies in athletes report on return-to-sport timing, later knee osteoarthritis, and reoperation. Most of this evidence is observational.

Return to sport. A 2026 systematic review and meta-analysis of athletes — pooling studies that were largely observational rather than randomised head-to-head comparisons — found a mean time to return to sport of 2.1 months after meniscectomy, 5.8 months after meniscal repair, and 8.9 months after meniscal allograft transplantation [17]. Return-to-sport rates were high and similar — 98.2% after meniscectomy and 96.9% after repair — and 84.4% versus 81.3% returned to their pre-injury level, without statistically significant differences between groups [17]. The time gap was larger for lateral tears (2.4 months after meniscectomy vs 6.5 months after repair) than for medial tears (2.0 vs 4.8 months) [17]. ESSKA adds that people typically resume normal daily activities within 2–4 weeks after meniscectomy, whereas repair requires considerably longer recovery [14].

Long-term knee osteoarthritis consultations. In a Swedish registry study of 2,487 patients aged 16–45 with a traumatic meniscus tear, followed for a median of 10 years, 17% consulted a doctor for knee osteoarthritis after meniscectomy, 10.0% after repair, and 2.3% in the general population [16]. The adjusted hazard ratio for repair versus meniscectomy was 0.74 (95% CI 0.48 to 1.15; the interval includes 1), and 0.51 (95% CI 0.27 to 0.96) when cases diagnosed within the first 2 years were excluded [16]. The authors interpreted the point estimates as suggesting roughly a 25–50% lower risk of consultation for knee osteoarthritis after repair, while noting that the consultation rate after repair remained at least twice that of the general population [16]. This is observational data, and the authors acknowledge likely residual confounding — healthier knees may be selected for repair [16].

Reoperations. In the same registry, 32% of repaired patients had a subsequent meniscectomy during follow-up, compared with 14% after an initial meniscectomy [16].

Rehab-first strategy in trials. STARR, which included competitive and elite athletes (34 of 100 participants), randomised patients with isolated, non-locked traumatic tears not judged suitable for repair to early meniscectomy or physical therapy [11]. DREAM studied a broader group of young adults with meniscal tears, in which 26% of the exercise group crossed over to surgery within 12 months [12]. In STARR, the 41% of the physical therapy group who underwent delayed surgery had it a median of 5.5 months after randomisation [11]. That figure is not directly comparable with the return-to-sport times reported in surgical cohorts, which are counted from surgery [17].

Strength. In the OMEX trial, thigh muscle strength had improved in the exercise group at 3 months, and the authors concluded that exercise had positive effects over surgery on thigh strength, at least in the short term [6], but there were no clinically relevant strength differences at 10 years [7].

Conclusion: Meniscus Tear

A meniscus tear is not one condition, and that is probably the most important point to take away. In degenerative tears, particularly in middle-aged adults, several randomised trials now provide long-term evidence that partial meniscectomy does not outperform placebo surgery or structured exercise therapy in terms of patient-reported function, and it does not reliably relieve catching or occasional locking [4][5][7][8]. Most trial participants improved over time regardless of treatment [7][10]. Clinically, however, I would separate improvement in symptoms from reversal of the underlying joint degeneration: a painful episode may settle, while an osteoarthritic or otherwise degenerative knee can still continue to change structurally over the years [4][5][7].

In younger adults, the picture is more nuanced. STARR studied isolated, non-locked traumatic tears that were not considered repairable, while DREAM included a broader group of young adults with meniscal tears. Neither trial showed superiority of early surgery over a rehab-first strategy, although a meaningful proportion of patients in the rehabilitation groups underwent surgery during trial follow-up [11][12]. At the same time, these trials do not settle the specific question of how a clearly repairable tear compares with non-operative rehabilitation. STARR excluded repairable tears, and DREAM included repair within a broader surgical strategy rather than testing repair directly against rehabilitation [11][12]. For locked knees, root tears, and repairable traumatic tears, the available literature relies much more heavily on consensus recommendations and observational evidence [14][15].

For athletes, the trade-off is also less simple than “surgery versus no surgery.” In pooled observational data, meniscectomy was associated with the shortest return-to-sport time, while repair took longer and was associated with a lower rate of consultation for knee osteoarthritis [16][17]. These studies do not establish which strategy is best for an individual athlete, and tear type, mechanical symptoms, the condition of the rest of the knee and sporting demands all affect how relevant the evidence is.

In practice, I therefore try not to make the MRI finding the decision. The more important questions are whether the meniscal lesion actually fits the symptoms, whether the knee is otherwise relatively well preserved or already degenerative, and whether there are features such as true mechanical locking or a clearly repairable tear that change the clinical picture.

The final decision about surgery remains individual and belongs to the treating orthopaedic surgeon. Trial evidence and consensus recommendations are there to guide that decision, not replace it.