Hamstring muscle injury rehabilitation in sports: a scoping review Authors: Aïda Cadellans-Arróniz, Caroline Prince, Spyridon Iatropoulos, Jurdan Mendiguchia, Kenny Guex, Jean-Benoît Morin, Alexandre Rambaud, Pascal Edouard Published: 2026 ABSTRACT AbstractObjectiveTo describe the methodological characteristics and content of published hamstring muscle injury (HMI) rehabilitation programmes in order to inform clinical practice and guide future research.MethodsWe searched the MEDLINE (PubMed), Scopus and Web of Science electronic databases, on October 2024, using key terms “hamstring” and “rehabilitation”, with strategies refined after a preliminary search. Eligible studies included English-language publications involving non-surgically managed acute HMI in humans and reporting rehabilitation protocols.ResultsWe included 137 articles. Primary research (43.1%, n=59) included mostly cohort and randomised controlled trials (RCTs); secondary research articles (54.0%, n=74) were predominantly narrative reviews. Overall, 54.7% of studies had a low level of evidence. Only 44.2% of primary research focused on HMI rehabilitation programmes of which 47.8% were RCTs. Rehabilitation programme details were often missing. In the subacute/functional phases, strengthening (78.3%) and stretching (55.6%) were most frequently reported. Return-to-sport criteria were reported in only 43.5% of primary research, often based on pain resolution. Across both primary and secondary research, 40 different rehabilitation strategies were identified, with the most frequently reported being eccentric-based strengthening, flexibility exercises, progressive running and agility protocols. Key gaps included lack of standardisation, under-reporting of programme parameters and limited inclusion of diverse sporting populations.ConclusionsCurrent scientific literature on HMI rehabilitation programmes shows wide variety, limited high-level evidence primary research and under-reporting of key clinical details. More rigorous and inclusive studies are needed to improve the standardisation and clinical applicability of rehabilitation strategies. WHAT IS ALREADY KNOWN ON THIS TOPIC Hamstring muscle injury (HMI) is the most prevalent injury in sports involving sprinting at maximal acceleration or velocity, such as football, rugby and athletics. Hamstring muscle injury (HMI) is the most prevalent injury in sports involving sprinting at maximal acceleration or velocity, such as football, rugby and athletics. HMI causes short-term, mid-term and/or long-term consequences supporting the need for a well-conducted rehabilitation process to mitigate structural, neuromuscular and functional impairments. HMI causes short-term, mid-term and/or long-term consequences supporting the need for a well-conducted rehabilitation process to mitigate structural, neuromuscular and functional impairments. HMI recurrences support the importance of a well-conducted rehabilitation process after any HMI. HMI recurrences support the importance of a well-conducted rehabilitation process after any HMI. The persistent rates of HMI and reinjury suggest the need to rethink current approaches. The persistent rates of HMI and reinjury suggest the need to rethink current approaches. WHAT ARE THE NEW FINDINGS The level of evidence of articles on HMI rehabilitation was low in 56.2%, and high (level 1) in 9.5%. The level of evidence of articles on HMI rehabilitation was low in 56.2%, and high (level 1) in 9.5%. Over half of the articles (54.0%) on HMI rehabilitation were secondary research, mostly narrative reviews (81.1%). Over half of the articles (54.0%) on HMI rehabilitation were secondary research, mostly narrative reviews (81.1%). Of the 40 different rehabilitation strategies identified across both primary and secondary research, only half were consistently reported in both primary and secondary research. Of the 40 different rehabilitation strategies identified across both primary and secondary research, only half were consistently reported in both primary and secondary research. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY There is a lack of high-level primary research (eg, randomised controlled trial (RCT)) supporting the need to increase the production of well-designed RCTs, which are essential to strengthen the evidence base and support the development of evidence-based clinical guidelines. There is a lack of high-level primary research (eg, randomised controlled trial (RCT)) supporting the need to increase the production of well-designed RCTs, which are essential to strengthen the evidence base and support the development of evidence-based clinical guidelines. The heterogeneity in exercise type, dosage and progression of HMI highlights the need for greater specificity. The heterogeneity in exercise type, dosage and progression of HMI highlights the need for greater specificity. This scoping review offers an overview of the range of strategies currently reported in the literature, providing clinicians with a repertoire of approaches for managing patients with HMI. This scoping review offers an overview of the range of strategies currently reported in the literature, providing clinicians with a repertoire of approaches for managing patients with HMI. Introduction Hamstring muscle injury (HMI) is the most prevalent injury in sports involving sprinting at maximal acceleration or velocity, such as football,1 rugby and athletics.2 HMI causes short-term impairments in hip and knee range of motion, hamstring muscle strength and functionality of daily life activities.3 In the mid-term, persistent deficits in hamstring muscle strength4 and reduced capacity to produce horizontal force at low velocity during sprint acceleration5 6 have been observed when athletes return-to-sport (RTS). Long-term consequences include muscle atrophy,7 8 architectural changes (eg, shorter fascicles and greater pennation angles)9 or neuromuscular inhibition (eg, reduced level of activation and strength).8 10 11 These aspects support the need of a well-conducted rehabilitation process to mitigate structural, neuromuscular and functional impairments following HMI. Recent advances in biomechanics have contributed to optimising athlete rehabilitation and RTS decision-making.12 HMI is associated with a noticeable recurrence rate (~18%–24%).1 4 Evidence from other highly physically demanding populations further supports the value of comprehensive injury profiling to better understand injury patterns and optimise prevention and rehabilitation strategies.13 HMI recurrences can be partially explained by the above-mentioned persistent deficits,7–11 that reduce the hamstrings’ ability to tolerate and cope with the sports demands, increasing the risk of re-injury.14 Additionally, a history of HMI is a known risk factor for subsequent HMIs.14 Proper management, particularly of the first injury, seems thus fundamental to reducing recurrence risk and protecting athletes’ health. Such management involves prioritising tissue healing and guiding both the injured muscle and the athlete through rehabilitation, towards a safe return to physical activity and sport.15 These aspects further highlight the importance of a well-conducted rehabilitation process after HMI. In this context, the need for rehabilitation after HMI is evident. Although some evidence exists for HMI rehabilitation programmes, the persistent rates of HMI and reinjury suggest the need to rethink current approaches. A comprehensive overview of current methodological approaches, as well as the procedures and components described in the literature for HMI rehabilitation programmes, is highly relevant for both clinical practice and research. Additionally, beyond the clinical recommendations described in previous studies,16–23 detailing the methodological aspects of articles is key to ensuring the reliability and validity of their findings.24 Previous systematic reviews only focused on evaluating the effectiveness of rehabilitation strategies,25–31 the return to play criteria,32 33 the functional consequences of HMI,34 the efficacy of platelet-rich plasma for HMI35 or reporting quality and exercises component.36 No previous systematic reviews provided a comprehensive picture of the current evidence on HMI rehabilitation programmes in sports. Therefore, we conducted a scoping review with the aim to describe the methodological characteristics (including the study type, design and level of evidence) and the content of HMI rehabilitation programmes (duration, periodisation, dose, components of the programmes and RTS criteria) reported in published articles. The goal was to present the ‘range of the possible’ to support practitioners and guide future research. Methods Protocol and registration We conducted a scoping review because this approach is superior to a systematic review for addressing an exploratory research question. The methodology was guided by The Joanna Briggs Institute Scoping Review Methodology Group.37 The manuscript was reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR),24 and the completed PRISMA-ScR checklist is provided in online supplemental table 1. This scoping review was registered in the Open Science Framework (DOI: 10.17605/OSF.IO/QSVB9) retrospectively, as this was not a usual practice at the start of the project for scoping reviews. Eligible articles We included: (1) articles published in English (based on evidence showing no significant bias when excluding non-English articles in conventional medicine reviews)38 39; (2) articles involving human subjects; (3) with acute, non-surgically managed hamstring injuries and (4) that reported a rehabilitation programme in the aim, methods or results sections of the article. Eligible study design included original research articles, systematic and scoping reviews, meta-analyses, guidelines, research protocols, opinion pieces, case series, clinical commentaries and technical notes. Exclusion criteria were: (1) non-peer-reviewed sources (eg, conference abstracts); (2) articles involving unrelated conditions (eg, other musculoskeletal disorders, neurological or systemic diseases) and articles mentioning rehabilitation only in the discussion section or without reporting it in the main text. Information sources and search In December 2019, a preliminary search was conducted in MEDLINE and Google Scholar using the terms ‘hamstring’ AND ‘rehabilitation’. From the first 200 results in each database, 56 relevant articles were identified. Titles, abstracts, and index terms were analysed to refine the final search strategy. A high number of irrelevant articles related to unrelated conditions prompted the exclusion of these topics in the final search. The databases (MEDLINE [PubMed], Scopus and Web of Science) were searched from inception to 17 October 2024, with the final search conducted on that date. Online supplemental table 2 shows the complete search strategy. All articles were then uploaded to the Covidence reference management platform (http://www.covidence.org), in which duplicates were automatically removed. Article selection Two independent reviewers (PE and CP or AC-A) first screened titles and abstracts for eligibility using Covidence,40 and then screened the full text. Discrepancies were resolved by consensus, with a third reviewer (JM) available if needed. Reference lists of included articles were also searched to identify additional articles. Data extraction One reviewer (AC-A) extracted data based on the following 13 categories: Author(s). Author(s). Year of publication. Year of publication. Location (country where the study was conducted or, when not explicitly reported, country where the study was approved by an ethics committee). Location (country where the study was conducted or, when not explicitly reported, country where the study was approved by an ethics committee). Study type (primary research: studies presenting original empirical data collected by the authors; secondary research: the author of the source did not generate the research data). Study type (primary research: studies presenting original empirical data collected by the authors; secondary research: the author of the source did not generate the research data). Design of the study. Design of the study. Level of evidence according to Melnyk and Fineout-Overholt41 (level 1: Systematic review, Clinical practice guideline; 2: Randomised controlled trial (RCT); 3: Cohort, Quasi-experimental; 4: Cross-sectional; 5: Narrative review, Case report, Clinical commentary; 6: Qualitative research; 7: Delphi consensus, Editorial. Two additional categories were created to classify certain types of publications that did not fit standard evidence levels: Level 8: For articles without an explicit evidence level, such as research protocols and ‘Other’: For scoping reviews, as their level of evidence depends on the specific content. Level of evidence according to Melnyk and Fineout-Overholt41 (level 1: Systematic review, Clinical practice guideline; 2: Randomised controlled trial (RCT); 3: Cohort, Quasi-experimental; 4: Cross-sectional; 5: Narrative review, Case report, Clinical commentary; 6: Qualitative research; 7: Delphi consensus, Editorial. Two additional categories were created to classify certain types of publications that did not fit standard evidence levels: Level 8: For articles without an explicit evidence level, such as research protocols and ‘Other’: For scoping reviews, as their level of evidence depends on the specific content. Aim(s) of the study. Aim(s) of the study. HMI programme: (1) involves the description of an HMI programme; (2) mentions a rehabilitation programme without a proper description or (3) does not mention nor describe a rehabilitation programme but presents complementary aspects that can help design rehabilitation programmes (anatomic differences, progression or return-to-play criteria after HMI). HMI programme: (1) involves the description of an HMI programme; (2) mentions a rehabilitation programme without a proper description or (3) does not mention nor describe a rehabilitation programme but presents complementary aspects that can help design rehabilitation programmes (anatomic differences, progression or return-to-play criteria after HMI). Population (number, age, sex, sport and level and sample size, if applicable). Population (number, age, sex, sport and level and sample size, if applicable). Hamstring injury characteristics (location, severity, diagnosis methods), if applicable. Hamstring injury characteristics (location, severity, diagnosis methods), if applicable. Details of rehabilitation programme: (1) total duration, (2) total number of sessions, (3) number of sessions per week, (4) type of intervention in acute phase, if applicable, (5) type of intervention in postacute phase, (6) anatomical location receiving the intervention, (7) analytical (involving a single muscle group or a specific joint) and/or functional exercises (involving several muscles and joints at the same time, simulating movements of daily life activities or sports). Details of rehabilitation programme: (1) total duration, (2) total number of sessions, (3) number of sessions per week, (4) type of intervention in acute phase, if applicable, (5) type of intervention in postacute phase, (6) anatomical location receiving the intervention, (7) analytical (involving a single muscle group or a specific joint) and/or functional exercises (involving several muscles and joints at the same time, simulating movements of daily life activities or sports). RTS criteria, if applicable. RTS criteria, if applicable. Outcomes and details of RTS, if applicable. Outcomes and details of RTS, if applicable. A quality control process was performed by an additional reviewer (PE), who independently extracted data from a random sample of two-thirds of the included articles, allowing data extraction to be partially double-checked. Any discrepancies were resolved during a consensus meeting between AC-A and PE, with a third reviewer (JM) being available if needed. Collating, summarising and reporting the results The results were organised in two parts: A numerical analysis mapping the data in tabular and diagrammatic form, showing the distribution of articles by period of publication, country of location, study type, study design, and level of evidence, presented in absolute (n) and relative (%) values. A numerical analysis mapping the data in tabular and diagrammatic form, showing the distribution of articles by period of publication, country of location, study type, study design, and level of evidence, presented in absolute (n) and relative (%) values. A further analysis was conducted exclusively on articles focused on HMI Rehabilitation Programme, as defined by the authors: those whose main objective involved the description of an HMI rehabilitation programme and that specified the type of intervention implemented. These articles were analysed separately according to whether they were primary or secondary research. For the primary research articles, a descriptive analysis of the characteristics of the population was included. In both primary and secondary articles, we described the methodological characteristics, HMI severity and the details of the rehabilitation programme. For the categorisation of the type of intervention (item that is part of point 11: Detail of rehabilitation programme), categories were added until information saturation was reached due to repetition or redundancy. Data were presented in absolute (n) and relative (%) values. A further analysis was conducted exclusively on articles focused on HMI Rehabilitation Programme, as defined by the authors: those whose main objective involved the description of an HMI rehabilitation programme and that specified the type of intervention implemented. These articles were analysed separately according to whether they were primary or secondary research. For the primary research articles, a descriptive analysis of the characteristics of the population was included. In both primary and secondary articles, we described the methodological characteristics, HMI severity and the details of the rehabilitation programme. For the categorisation of the type of intervention (item that is part of point 11: Detail of rehabilitation programme), categories were added until information saturation was reached due to repetition or redundancy. Data were presented in absolute (n) and relative (%) values. Patient and public involvement There was no public or patient involvement. Equity, diversity and inclusion statement All relevant articles were included regardless of country, sex, age, participant level, ethnicity, socioeconomic status or marginalised group representation. The author team was balanced regarding gender, experiences (junior, mid-career and senior researchers), specialty (sports physiotherapy, medicine, data science, sports science, physical medicine and rehabilitation, sports medicine) and countries (France, Greece, Spain, Switzerland). Results Study selection The systematic search identified 9963 records. After removing duplicates and screening titles and abstracts, 316 full-text articles were assessed for eligibility. No additional articles from systematic or scoping review were included. Following full-text screening, 137 articles were finally included with discrepancies resolved by consensus. The flow chart of articles’ selection is detailed in figure 1. Flow chart of articles selection. *Consider, if feasible to do so, reporting the number of records identified from each database or register searched (rather than the total number across all databases/registers): numbers are provided for each database. **If automation tools were used, indicate how many records were excluded by a human and how many were excluded by automation tools: no automation tools were used. Characteristics of all the included articles Publication period and country Of the 137 included articles, most were published after 2005, with 43.1% between 2015 and 2024 and 37.2% between 2005 and 2014 (online supplemental figure 1). Among the 60 (43.8%) articles reporting the country where the study was conducted (or, when not reported, the country where ethical approval was obtained), 41.7% (n=25) were conducted in Europe, 18.3% (n=11) in Asia, 16.7% (n=10) in North America, 13.3% (n=8) in Oceania, 8.3% (n=5) in South America and only one study was originally from Africa (1.7%) (online supplemental figure 2, online supplemental table 3). Type, design and level of evidence Over half of the articles (54.0%, n=74) were secondary research, mostly narrative reviews (81.1%, n=60). Primary research represented 43.1% (n=59), including observational cohorts (40.7%, n=24), RCTs (32.2%, n=19) and case reports (16.9%, n=10). A small portion (2.9%, n=4) were classified as ‘other’, including editorials, RCT protocols (no published results yet) and one scoping review (table 1). Table 1Number and proportion of articles included according to study type and study designStudy typeAll research articles included (n=137)Articles focus on HMI rehabilitation programmes (n=52)Designn%Designn%Primary studiesCase report1016.9Case report417.4Clinical commentary11.7Delphi14.3Delphi11.7Cohort626.1Cohort2440.7RCT1147.8Cross-sectional23.4Quasi-experimental14.3Case–control11.7Randomised controlled trial (RCT)1932.2Quasi-experimental11.7Total5943.1Total2344.2Secondary studiesScoping review11.4Scoping review13.7Systematic review1114.9Systematic review414.8Narrative review6081.1Narrative review2281.5Clinical practice guideline11.4Appraisal of clinical practice guideline11.4Total7454.0Total2751.9OtherEditorial250.0Editorial150.0Protocol of RCT250.0RCT protocol150.0Total42.9Total23.8HMI, hamstring muscle injury. Number and proportion of articles included according to study type and study design HMI, hamstring muscle injury. The level of evidence was low in 56.2% (n=77) of the included articles (level 7: 2.9%, n=4; level 6: n=0; level 5: 51.8%, n=71) and moderate to high in 43.8% (n=60) (level 4: 1.5%, n=2, level 3: 19%, n=26; level 2: 13.9%, n=19; level 1: 9.5%, n=13) (figure 2). Among primary research (n=59), most had a moderate (45.8%, n=27) to low (22%, n=13) level of evidence and only 32.2% (n=19) had a high level of evidence. For the secondary research (n=74), 82.4% (n=61) were low-level evidence (figure 2). Design and level of evidence of all included articles.(A) Design and level of evidence of all included articles (n=137), (B) design and level of the primary included articles (n=59), (C) design and level of evidence of secondary included articles (n=74). Articles focused on HMI rehabilitation The description of an HMI rehabilitation programme and the type of intervention implemented was mentioned in the aims of 38% articles (n=52), while in 62.0% (n=85) it appeared as a secondary objective or within the methods section (table 1). Primary research articles focusing on HMI rehabilitation programmes Design and level of evidence Of the 23 primary research articles focusing on HMI rehabilitation programme, 47.8% (n=11) had a high level of evidence (level 2, RCT, n=11).16–23 42–44 The remaining 52.2% (n=12) had moderate to low evidence: cohort articles (26.1%, n=6),45–50 case reports (17.4%, n=4),51–54 Delphi study (4.3%, n=1)55 and quasi-experimental design (4.3%, n=1).56 General characteristics Excluding Delphi studies and case reports, the remaining 18 articles included a total of 691 participants, with a median sample size per study of 37.5 participants (IQR: 21–50). Of these participants, 78.4% (n=542) were men and 21.6% (n=149) were women. Most were competitive (30.3%, n=216) or professional athletes (19.8%, n=173), while 17.1% (n=118) trained recreationally. Sports level was unspecified for 24.6% (n=170), and 18.8% (n=130) were involved in sport without further detail; in 23.2% (n=160) this information was unclear. The most common sport was football (31.1%, n=215), followed by athletics (track and field) (24.6%, n=170) and dance (2.3%, n=16). Diagnosis was based on clinical assessment alone (21.7%, n=5), combined with imaging (21.7%, n=5) or imaging alone (13.0%, n=3), using MRI and echography MRI or ultrasound. However, 43.5% (n=10) of primary research focusing on HMI rehabilitation programmes did not report the diagnostic method. The specific hamstring head involved was identified in only 26.1% (n=6) of articles, while 73.9% (n=17) did not specify it. Injury severity was reported in 52.2% (n=12) and classified as grades I to III strain. Population characteristics of the 23 primary articles are detailed in online supplemental table 4. Details of rehabilitation programme The total duration of the rehabilitation programme was reported in 39.1% of articles (mean±SD: 45.5±32.4 days) and the number of sessions per week in 34.8% of articles (mean±SD: 3.7±1.1 days). Acute phase (0–14 days after the injury).15 Only 34.8% (n=8)17 43 46–51 of the primary research articles reported indications on the procedures to be performed in the acute phase of an HMI rehabilitation programme, with RICE (Rest, Ice, Compression and Elevation) or ice application being the most commonly reported practice (62.5%; n=5). Subacute and regeneration/functional phases (0–45 days after the injury).15 78.3% (n=18) of articles included strengthening,16 19–23 43–46 48 50 54–57 55.6% (n=13) stretching,16 17 19–23 43 48 50 51 55 57 43.5% (n=10) trunk/core/pelvis conditioning,16 19 21 22 44 50 51 54 55 57 13.0% (n=3) manual therapy (hands-on manipulation and mobilisation of muscles and joints to improve movement),19 42 47 13.0% (n=3) of massage (deep tissue massage, sports massage),43 50 51 8.7% (n=2) homework,20 23 and 4.3% (n=1) foot/ankle exercises.50 Analytical (56.5%, n=13) and functional (52.2%, n=12) approaches were similarly reported. Running or sprinting was addressed in 39.1% (n=9) of articles. Among the intervention components, strengthening (89.5%, n=17/19) and stretching (76.9%, n=10/13) were the most comprehensively described, whereas the remaining components were reported with less detail. RTS criteria RTS criteria were reported in 43.5% (n=10) of articles,16 19 22 43 44 48 50 55 57 58 with 70.0% (n=7) focusing on pain absence (during testing or during palpation). Of these 85.7% (n=6) combined pain with other criteria such as (1) improvements in strength and function (33.3%, n=2), (2) improvements in strength and function and the ability to perform sports activities without pain (33.3%, n=2) and (3) improvements in mobility and asymmetry between the legs (33.3%, n=2). Some of these criteria were also considered for RTS in isolation (20%, n=2). Only one article (10%) specified clinical test (Askling-H test)16 (table 2). Table 2Details of return to sport criteria according to the study typeStudy typeReturn to sports criterian%DetailPrimaryAbsence of pain770During testing: active knee extension (ATKE) or passive straight leg raise test (ROM ≥90% of contralateral leg), stretching, maximal isometric contraction, single-leg hop test, Askling-H test, resisted isometric test, maximal eccentric contractions in the lengthened state.During palpationAbility to perform sports activities without pain550Drills, rolling sprints, agility and running tests, completing a series of progressive sprints working up to full speedImprovement in strength and function330Equivalent hamstring function and strength between injured and uninjured legs; scored 5/5 on manual muscle testing of the hamstrings performed on 4 consecutive repetitions in various knee positions; maximal-effort isometric knee flexor contraction at 0°/0° and 90°/90° of hip/knee flexion; demonstrated 5/5 strength when manually resisting knee flexionImprovement in mobility and asymmetry between legs220The ATKE ROM achieved at 95% or greater, length improvements, absence of between-limbClinical test110Askling-H testSecondaryAbsence of pain777.8Full strength pain free, the length of hamstring tenderness, the presence of painful resisted knee flexion, ability to perform functional training activities, at sports-specific movement testing (accelerations, cutting)During palpationStrength improvements666.7Full concentric and eccentric compared with uninjured side, Norbord at preinjury level, isokinetic strength testing less than 5%Clinical test333.3Askling H-test, active straight leg raise testImaging222.2RMI presence of abnormal imaging findings, oedema size or lengthPeak torque222.2<28 during knee flexion and <8 asymmetry between legs,Asymmetry222.2Resolution of between-limbs asymmetry (<10%–15% asymmetry), >20%Other criteriaROM deficits, full ROM pain free, Optojump biomechanical review, horizontal force during running, No fear or kinesiophobia with Askling H-test, High-speed running drills without discomfortRMI, resonance magnetic image; ROM, range of motion. Details of return to sport criteria according to the study type RMI, resonance magnetic image; ROM, range of motion. Secondary research articles focusing on HMI rehabilitation programmes Design and level of evidence and general characteristics Regarding secondary research focused on rehabilitation programmes for HMI (n=27), 81.5% (n=22) were narrative reviews,45 59–79 14.8% (n=4) systematic reviews25–28 and 3.7% (n=1) scoping review assessing primary articles’ adherence to Consensus on Exercise Reporting Template (CERT) guidelines.36 Population characteristics are detailed in online supplemental table 4. Details of rehabilitation programme Only 18.5% articles specified the duration of the rehabilitation programmes (8.5±0.7 days), and 14.8% (n=4) specified the sessions per week (4.8±1.8 days). Acute phase. 44.4% (n=12) mentioned interventions and all included RICE or ice. Other approaches reported were the use of non-steroidal anti-inflammatory drugs (NSAIDs) (25.0%, n=3), ultrasound therapy and laser (8.3%, n=1), compression of the area and transcutaneous electrical nerve stimulation (8.3%, n=1), immobilisation (without specifying the duration) along with sacroiliac joint mobilisation (8.35%, n=1) and one study recommended avoiding the use of NSAIDs (8.3%, n=1). Subacute and regeneration/functional phases. Strengthening was reported in 92.6% (n=25) of the articles (only 33.3% (n=9) provided specific details) and stretching was mentioned in 70.4% (n=19) (only 25.9% (n=7) provided the details). Trunk, core, and pelvis-focused training were proposed in 70.4% (n=19), while manual therapy (22.2%, n=6), massage (11.1%, n=3), and home exercises (3.7%, n=1) were the least suggested. Analytical training (59.3%, n=16) was more frequent than functional training (44.4%, n=12), with limited running (22.2%, n=6) and sprinting (7.4%, n=2) indications. Only 18.5% articles specified the duration of the rehabilitation programmes (8.5±0.7 days), and 14.8% (n=4) specified the number of sessions per week (4.8±1.8 days). RTS criteria RTS criteria were reported in 33.3% (n=9) of articles.45 59 63 67 72 74 76 77 80 The most common criterion was pain-free palpation or in sports movements (77.8%, n=7), followed by full recovery of hamstring and quadriceps strength (66.7%, n=6), and clinical tests like the Askling H-test or active straight leg raise (33.3%, n=3) (table 2). Components of HMI rehabilitation programmes Of the 49 articles describing HMI rehabilitation programmes (both primary and secondary articles), we identified 40 different intervention components. Of these intervention components, 21 (52.5%) were reported in both primary and secondary articles. Meanwhile, 25% (n=10) were only reported in primary articles and 22.5% (n=9) only in secondary articles. Details of these components are provided in table 3. Table 3Description, example, frequency and proportion of the components reported in the studies focused on HMI rehabilitation programmesComponentFrequency of study reporting (n (%))ExampleComponents reported in both primary and secondary articles (n=21)Strength exercises focused on lengthening11 (52.4)Nordic hamstring curlsHamstring flexibility9 (42.9)Straight-leg raise stretchesStrength exercises focused on trunk and pelvis8 (38.1)Single-leg glute bridges or bird-dogsRICE9 (42.9)Ice pack and leg elevation after strainHamstring stretching14 (66.7)Seated hamstring stretchNeuromuscular control exercises7 (33.3)Balance board with leg swingsGeneral strength14 (66.7)Upper body workouts or squats (within pain limits).Strength exercises focusing on the eccentric phase16 (76.2)Slow lowering during Romanian deadlifts.Neurodynamic techniques7 (33.3)Sciatic nerve glides.Mobilisation/manipulation of lumbar segment, sacroiliac, hips, and knee joints9 (42.9)Manual therapy for hip rotation or lumbar spine movement.Core exercises5 (23.8)Planks or dead bugs.Progressive agility and trunk stabilisation16 (76.2)Ladder drills, shuttle runs, cone drillsDry needling on hamstring2 (9.5)Needle inserted into a trigger point in the biceps femoris.Technical skills exercises8 (38.1)Soccer dribbling or sprint starts.Immobilisation2 (9.5)Use of a brace or crutchesCryotherapy10 (47.6)Ice massage after exercise.Low-level laser4 (19.0)Laser applied directly to the hamstring injury siteIsometric exercises10 (47.6)Hamstring push against a fixed surface.Progressive running and eccentric strengthening3 (14.3)Jog-walk progression+Nordic curls.Plyometric exercises6 (28.6)Bounding, box jumps.Isokinetic exercises4 (19.0)Biodex machine hamstring curls.Components reported only in primary articles (not in the secondary) (n=10)Vibration therapy1 (10.0)No detailProgressive running programme2 (20.0)No detailMassage therapy2 (20.0)No detailPercutaneous needle electrolysis1 (10.0)No detailHamstring and gluteus maximus strength2 (20.0)No detailPostural stabilisation for pelvis or lumbopelvic area4 (40.0)No detailKinesiotape2 (20.0)No detailTechnical sprints1 (10.0)No detailRunning technique2 (20.0)No detailAnkle stability exercises1 (10.0)No detailComponents reported only in secondary articles (not in the primary) (n=9)Progressive stretching and strengthening2 (22.2)No detailNSAIDs (non-steroidal anti-inflammatory drugs)2 (22.2)No detailEarly motion2 (22.2)No detailPlatelet-rich plasma2 (22.2)No detailElectrotherapy5 (55.6)No detailUltrasound therapy4 (44.4)No detailImmobilisation1 (11.1)No detailNSAIDs (avoid)6 (66.7)No detailInstrument-assisted soft tissue mobilisation/soft tissue mobilisations1 (11.1)No detailHMI, hamstring muscle injury; RICE, Rest, Ice, Compression, Elevation. Description, example, frequency and proportion of the components reported in the studies focused on HMI rehabilitation programmes HMI, hamstring muscle injury; RICE, Rest, Ice, Compression, Elevation. Discussion The main findings of the present scoping review were as follows. First, the available evidence was predominantly based on secondary research articles, while most primary research articles had a moderate to low level of evidence and comprised few RCTs. Second, the contents of HMI rehabilitation programmes showed considerable heterogeneity, with limited information on duration, frequency or progression, partial consensus on key components and unclear consensus on RTS criteria. These findings present the ‘range of the possible’ regarding HMI rehabilitation programmes. These suggest that clinicians should follow these programmes with caution and that future research should emphasise well-reported RCTs with standardised protocols to support more precise and evidence-based clinical decision-making. Methodological variability of HMI rehabilitation programmes research Our scoping review showed a predominance of secondary research (54.0%, n=74) over primary research (43.1%, n=59). This pattern is common in emerging fields becoming more established.81 Temporal analysis (online supplemental figure 1) shows that secondary research began to increase in 2005, when there was a greater number of primary research, and surpassed primary research in the following decade. A predominance of secondary research might suggest a limited generation of new primary data, which can limit real progress in addressing knowledge gaps or controversies identified in reviews.81 Regarding articles focusing on HMI rehabilitation programmes, a greater proportion of secondary research was also identified compared with primary research. Moreover, among primary research, 52.2% were classified as having a moderate to low level of evidence. Only 19 of 137 were articles with a high level of evidence (RCTs). Considering the 40-year review period and the high HMI incidence, this limited number of RCTs reveals a significant gap in high-quality evidence. As the gold standard in clinical research, the scarcity of RCTs weakens the overall evidence base and reduces confidence in clinical recommendations. This shortage may be a consequence of various challenges inherent to rehabilitation research, such as difficulties in standardising rehabilitation protocols, implementing blinded and controlled study designs, and conducting long-term follow-ups. Consequently, the predominance of lower-level evidence, like narrative reviews (81.1% of secondary articles), suggests that current rehabilitation strategies rely more on clinical experience or theory than on scientific validation. We only identified one systematic review,36 which highlights a significant gap in critical, structured evidence synthesis. This scarcity aligns with the low number of RCTs identified and suggests a lack of consolidated knowledge, hindering the standardisation of effective interventions. Existing recommendations should therefore be interpreted with caution. Our results highlight a significant gap in the scientific evidence supporting HMI rehabilitation programmes, emphasising the need for continued knowledge generation through primary research. Future research should prioritise well-designed RCTs that evaluate both the efficacy and safety of HMI rehabilitation programmes. Using standardised reporting guidelines such as Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) (for trial protocols),82 CERT (for exercise-based interventions)83 or Consolidated Standards of Reporting Trials (CONSORT) (for randomised trial results) can enhance the transparency and reproducibility of future RCTs.84 Population characteristics of research on HMI rehabilitation programmes The analysis of the population reveals important gaps in participant representation and injury characterisation, hindering the applicability and generalisability of findings. We observed a gender disparity, with nearly 80% of participants being men. This is consistent with the existing literature. A systematic review of studies published between 2017 and 2021 in six sports medicine journals showed that 70.7% of studies focused exclusively on men, vs only 8.8% on women. Additionally, mixed-sex studies also overrepresented men.85 We believe that several factors may contribute to this imbalance. First, HMI are more prevalent in men, especially in sports involving sprinting and sudden direction changes, with male athletes being 2–4 times more likely to be affected than females.2 Although these sports are not only male-dominated in terms of participation, higher injury rates may result in a sampling bias favouring men. Second, nearly 60% of the studies included competitive or professional athletes. Despite growing female participation, women remain under-represented at elite levels. For example, FIFA reports 5.4 million male vs 0.8 million female players in Germany, 1.3 million vs 98 000 in England, and 1.7 million vs 49 000 in France.15 This underscores the need for inclusive research considering sex differences in physiology and biomechanics for HMI rehabilitation. In addition, we also highlighted disparities in the HMI diagnosis, with different approaches to make the diagnosis (clinical alone, imaging alone, or combination of clinical and imaging) and almost half of the primary research articles, which did not mention the diagnostic method. Consensus-based diagnostic strategies can represent an important perspective for clinical and research activities. Only 26.1% of the studies reported the specific location of the HMI. Since injury location affects severity, recovery time and treatment response, rehabilitation protocols should be tailored accordingly.86 87 Future studies should include injury characterisation to develop rehabilitation strategies tailored to athletes’ clinical profiles. Rehabilitation strategies during the acute phase of HMI Only 34.8% of primary research reported specific guidelines for interventions during the acute phase. The most commonly reported strategy was the RICE protocol or cryotherapy. The acute phase is critical for minimising secondary damage, modulating inflammation and establishing the conditions for optimal functional recovery.15 88 The lack of clear guidelines in early clinical management might hinder recovery or increase the risk of recurrence.89 90 Although the RICE protocol has remained the standard for acute muscle injuries, its effectiveness has been questioned.89 90 Outside of the articles included in this review, some studies suggested that it may disrupt the natural healing process by suppressing the initial inflammatory response, necessary for tissue regeneration.89 90 Additionally, ice application (a key component of RICE) is widely used for early-pain management.91 However, most supporting evidence for this procedure is derived from animal studies with limited relevance to human muscle tissue, and some even show delayed tissue healing.92 While early postinjury use may be justified for pain relief, continued application beyond 12 hours should be used cautiously until stronger human evidence emerges.92 Rehabilitation strategies during the subacute and functional phases of HMI The subacute phase was the focus in 82.6% of studies, emphasising strengthening with a similar proportion of analytical and functional exercises. This reflects a more active and functional approach, tailored on muscle recovery, mobility restoration and neuromuscular re-education for a gradual RTS activity.15 Indeed, 39.1% of the articles included guidelines for running or sprinting. However, less than half provided information on sessions frequency or the overall programme duration (39.1%). These parameters are fundamental in any rehabilitation process, as they directly influence the dose of stimulus received by the injured tissue and, consequently, effectiveness and speed of recovery as well as safe RTSs.83 93 This omission may contribute to heterogeneity in reported outcomes, complicating comparisons between studies.94 Future research should provide detailed information not only on the exercises but also on these parameters. Components reported in HMI rehabilitation programmes As most of the information on HMI rehabilitation programmes issued from secondary research, caution should be taken in the interpretation and use of these programmes. 25% of the identified components were exclusively reported in primary research, and are thus based on empirical opinion. Relevant examples include vibration therapy,49 percutaneous electrical stimulation,47 progressive high-speed running programmes,22 or specific exercises for ankle stability.16 These components may indicate a trend to combine more specific and personalised rehabilitation programmes that consider not only the directly affected musculature but also biomechanical and neurofunctional factors. The absence of these approaches from narrative reviews may result from their novelty or scarce recent primary research. In contrast, 22.5% of reported components were found exclusively in secondary articles, most of them narrative reviews with a low level of evidence. Examples include NSAIDs, ultrasound therapy, transcutaneous electrical nerve stimulation, immobilisation or laser therapy, and many of these approaches were insufficiently described for practical application. Although secondary research provided a wider range of interventions, the effectiveness of these approaches was not always supported by specific primary evidence. For example, NSAID use remains controversial due to inconclusive evidence and possible interference with muscle regeneration.95 96 The clinical benefit of electrotherapy and ultrasound in musculoskeletal conditions remains unproven in RCTs. Thus, their inclusion in rehabilitation protocols should be applied with caution, until more robust evidence is available. We believe that these recommendations may be based more on expert opinion, previous clinical guidelines or indirect extrapolations that extend beyond what scientific data can support. Unfortunately, this lack of alignment between both levels of research contributes to greater variability in therapeutic practice. Only 52.5% of the components were shared by both primary and secondary articles, indicating a partial consensus between original research and clinical experience shared in narrative reviews. Key components included eccentric-based strengthening, flexibility, neuromuscular control, lumbar stability training, progressive running and agility protocols. RTS criteria after HMI Less than half of primary research and only one-third of secondary research reported specific RTS criteria, underscoring the limited guidance available to support safe reintegration into sports. Absence of pain (on palpation or functional testing) was the most reported factor, in 70% of primary and 77.8% of secondary articles. Pain is widely used in clinical assessment, as it seems to be a clear indicator of active injury or incomplete recovery. This reliance on pain as a clinical marker is also supported by recent literature. For example, Edouard et al15 emphasises that pain is widely used to guide rehabilitation progression in muscle injury management, aiming to maintain a pain-free or limited-pain status during all phases of recovery. This strategy seeks to balance mechanical loading and tissue protection, even though scientifically validated indicators to optimise this balance remain lacking. Importantly, this approach is most effective when analgesics do not mask symptoms and when patients are properly educated to interpret pain as a feedback mechanism throughout rehabilitation.33 In this line, Hickey et al22 conducted a RCT comparing pain-free versus pain-threshold rehabilitation following acute HMI, reporting no significant differences in time to RTS or recurrence rates between groups.22 However, absence of pain alone may be an insufficient criterion to ensure full rehabilitation and prevent recurrence.22 In fact, most of primary research (85.7%) combine pain absence with functional indicators (restored muscle strength and function, the ability to perform sports activities pain-free and the correction of asymmetries and mobility limitations between both legs). This approach is consistent with current evidence from other fields, highlighting the need for a multifactorial assessment that includes both clinical and functional aspects to make informed and safe RTS decisions.97 However, the overall heterogeneity of objective assessments observed may increase subjectivity in clinical decision-making, potentially leading to premature or excessively delayed RTS. Secondary research offered greater specificity in certain areas, specifically related to strength recovery, by identifying the muscle groups involved and the tools used (eg, isokinetic testing). 33.3% included standardised clinical tests, including the Askling H-test and the Active Straight Leg Raise. Surprisingly, only one primary article mentioned a specific clinical test (Askling H-test).16 This test has proven effective in evaluating muscle elongation tolerance and in safely replicating high-speed running mechanical loads and extreme stretch conditions.98 Clinical studies also support its effectiveness in identifying residual fear or dysfunctions and predicting re-injury risk.98 99 Therefore, it seems to be an appropriate test to include before making decisions about RTS. Finally, other assessment tools (MRI findings, biomechanical analysis, peak torque output and psychological readiness) were reported only in secondary articles. Although they have also shown diagnostic validity for detecting the presence of muscle abnormalities,100 primary research might focus more on easily measurable clinical criteria, neglecting tools that require specialised equipment or more complex assessments. Recent perspectives have called for the standardisation of sensor technologies and artificial intelligence–based assessment tools to enhance objective, reproducible and clinically meaningful RTS decision-making.101 Limitations This scoping review has some limitations. This review was not prospectively registered which may limit transparency and increase the potential for reporting bias. However, no methodological deviations occurred between the registered protocol and the final conduct of the review. And to ensure reproducibility the methodology was reported following established methodological guidance for scoping reviews (PRISMA-ScR). Data extraction was independently verified for approximately two-thirds of the included studies rather than all records, so the possibility of data extraction errors cannot be completely excluded. Consistent with scoping reviews design,24 no formal risk of bias assessment or critical appraisal of the methodology was conducted. Substantial heterogeneity in programmes descriptions, particularly in session dosage, frequency and duration, hindered comparability and identification of best practices. The CERT checklist was not applied to the included studies, which limits the ability to objectively quantify the level of under-reporting, but the latter did not represent our study aim. Additionally, relevant articles may have been missed due to language restrictions or database limitation. Finally, this review focused on mapping and describing existing literature, rather than assessing the clinical effectiveness of rehabilitation programmes, which future systematic reviews or meta-analyses should address. Perspectives for research The findings of this scoping review highlight several critical areas that warrant further investigation to advance the field of HMI rehabilitation. There is a lack of high-level primary research, particularly RCTs, considered the gold standard in clinical research. Increasing the production of well-designed RCTs is essential to strengthen the evidence base and support the development of evidence-based clinical guidelines. The heterogeneity in exercise type, dosage and progression, also highlights the need for greater specificity. Future studies should provide detailed information on these parameters to support replicability and enable dose-response analyses. Authors should use reporting guidelines such as SPIRIT, CONSORT and CERT to describe the methods and results of their studies. This would enhance transparency, reproducibility and comparability across studies, thereby improving the overall quality of evidence in the field. We also found an under-representation of diverse populations, with most studies focusing on male, professional athletes, limiting generalisability. Few studies report precise injury characteristics such as location or severity, despite their known impact on recovery. Addressing these gaps is crucial for designing individualised rehabilitation programmes. Standardising RTS criteria is another priority. Future research should validate multifactorial models that integrate clinical assessments, strength and functional tests, imaging and psychological readiness to ensure safe and effective reintegration into sport. In addition, emerging artificial intelligence approaches may support personalised rehabilitation by predicting re-injury risk and guiding individualised rehabilitation strategies, although further validation is required.102 Finally, the observed misalignment between primary and secondary literature reflects that narrative reviews sometimes recommend interventions (eg, electrotherapy, cryotherapy) that have not been tested in the primary studies, while certain interventions studied in primary research are not consistently included in secondary literature. This discrepancy underscores the need for more high-level primary research and systematic reviews and meta-analyses to consolidate knowledge, guide clinical practice and inform future research priorities in this field. Clinical implications Based on the findings, proposing a single, optimal rehabilitation programme remains challenging. However, this scoping review offers an overview of the range of strategies currently reported in the literature, offering clinicians a descriptive map of existing approaches for managing patients with HMI. The components most frequently reported in levels 1 and 2 studies included: eccentric-based strengthening exercises, flexibility, neuromuscular control, lumbar stability training and progressive running and agility protocols. These elements may serve as a practical reference when designing individualised rehabilitation plans. However, their level of supporting evidence should be interpreted cautiously, as no formal assessment of risk of bias was conducted in this review. The lack of standardised RTS criteria reinforces the importance of using multifactorial assessments considering not only pain resolution, but also functional, psychological and biomechanical factors. Encouraging clinicians to engage directly with primary research and high-quality evidence is essential. Given the discrepancies observed between primary and secondary research, relying solely on expert opinion or secondary research may limit the precision and effectiveness of clinical practice. Conclusions By mapping the methodological characteristics and content of published articles on HMI rehabilitation, we identified substantial variety and several gaps in both the level of evidence and the reporting of key clinical details. The predominance of secondary research, the limited number of high-level primary evidence and the under-representation of female and recreational populations highlight the need for more rigorous, inclusive and standardised research. These findings offer valuable insights for clinicians aiming to implement evidence-informed rehabilitation programmes and underscore clear directions for future investigations to enhance the effectiveness and applicability of HMI rehabilitation strategies in athletes. Supplementary material