WHAT IS ALREADY KNOWN
Cryotherapy is widely used as part of treatment in the rehabilitation of musculoskeletal injuries and postoperative care. However, clinicians and clinical practice guidelines often recommend cryotherapy based solely on theoretical mechanisms of action.
Previous systematic reviews have found unclear evidence regarding the effectiveness of cryotherapy on clinical outcomes in distinct musculoskeletal disorders.
WHAT ARE THE NEW FINDINGS
This systematic review investigated the effectiveness of cryotherapy on pain intensity, range of motion, swelling and function in the postoperative care of people with musculoskeletal disorders.
We found moderate-certainty evidence of a statistically significant difference supporting cryotherapy for pain intensity and range of motion in the immediate to medium term. However, its effects may be too small to be clinically relevant. Evidence for swelling and function was low-certainty, with no considerable effect sizes observed.
Cryotherapy may not be an indispensable intervention in the postoperative management of musculoskeletal disorders, and its use should consider time, cost, patient preference and the availability of evidence-based alternative interventions.
Clinical practice guidelines must be cautious when recommending cryotherapy during postoperative care due to its modest effect size. High-quality randomised controlled trials on this topic are needed.
Introduction
Cryotherapy is a therapeutic modality that uses low-temperature components to remove heat from tissue, frequently used in clinical practice settings to manage musculoskeletal injuries.1 The proposed mechanisms of action include decreasing nerve conduction velocity, which may attenuate muscle spasms mediated by spinal reflexes.2 3 It can also stimulate thermoreceptors that inhibit nociceptive signals in the central nervous system, consequently increasing the pain threshold.2 Cryotherapy may also lower metabolic demand, potentially preventing secondary injury associated with post-traumatic hypoxia and subsequent cell death.4 The local vasoconstriction effect induced by cold exposure may reduce blood flow and vascular permeability, reducing swelling, haematoma formation and tissue necrosis at the injury site.5–7 Thus, based on its potential mechanisms, cryotherapy may improve pain, swelling and functional recovery, especially when applied within the first few hours after trauma.4 8
Although cryotherapy may be popular among clinicians and patients, as it can be a low-cost, accessible and easy-to-administer technique, evidence for its effectiveness remains controversial. Miranda et al
9 investigated the effects of cryotherapy in acute ankle sprains and found a lack of evidence in the literature to support its use. Aggarwal et al
10 found low certainty evidence of minor effects of cryotherapy on pain, swelling and knee range of motion (ROM) after total knee arthroplasty (TKA), which may not justify its use. Nevertheless, clinical practice guidelines recommend cryotherapy.11 12 Therefore, clarifying the evidence of cryotherapy in musculoskeletal disorders is necessary.
A previous systematic review13 evaluated the efficacy of cryotherapy for musculoskeletal disorders, including postoperative rehabilitation. Although the authors concluded that there was low-certainty evidence suggesting beneficial effects of cryotherapy, these findings were limited by methodological issues. Notably, no quantitative analysis or effect size estimations were conducted. Furthermore, the included studies lacked appropriate comparator groups, including randomised controlled trials (RCTs) comparing cryotherapy with distinct active interventions.14–16 Therefore, a new systematic review addressing these limitations is warranted to better inform patients, clinicians and stakeholders about the current evidence regarding the effectiveness of cryotherapy in postoperative musculoskeletal rehabilitation.
Our systematic review aimed to investigate the effectiveness of cryotherapy on pain intensity, ROM, swelling and function in individuals undergoing postoperative care of musculoskeletal disorders, evaluating the certainty of the evidence using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach.17 18
Methods
Equity, diversity, and inclusion statement
The author group consists of one woman and three men, including a junior, a mid-career and two senior researchers. Three authors work in a developing country (Brazil); one works in Australia. Our study included a population composed of men and women from different countries. The findings may be generalisable to environments with fewer resources; however, we acknowledge that we did not examine the effects of race/ethnicity or socioeconomic status.
Study design
This systematic review followed the Cochrane Handbook recommendations for systematic reviews.19 It was reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses checklist.20 The protocol was prospectively registered in the International Prospective Registry of Systematic Reviews (PROSPERO) (Register number: CRD42023476237).
Search strategy and study selection
Searches were conducted up to 5 June 2025, on MEDLINE, COCHRANE, EMBASE and PEDro databases without date restriction. Search terms were related to “randomised controlled trial”, “musculoskeletal” and “cryotherapy”. A detailed search strategy is presented in online supplemental file S1: Search strategy. We handsearched systematic reviews published in the field for potentially relevant full texts. Following the searches, retrieved references were exported to an Endnote file, and duplicates were removed. Then, two independent researchers (JPdM and RCCF) screened titles and abstracts and assessed potential full texts. Trials that fulfilled our eligibility criteria were included. Discrepancies between researchers were resolved by a third researcher (VCO).
Eligibility criteria
Study design to be included
This systematic review included only RCTs. In the case of crossover RCTs, we extracted data only from the first period (ie, before the washout period).
Participants and health conditions of interest
Studies involving individuals of any sex and age undergoing postoperative rehabilitation for musculoskeletal disorders (ie, rehabilitation after surgery for ligament reconstruction, partial or total joint arthroplasties, arthroscopies, fracture fixation, among others) were considered eligible.
Eligibility was restricted to trials addressing operative care due to clinically diagnosed musculoskeletal disorders. Thus, studies including healthy participants subjected to experimentally induced tissue damage, as well as those investigating perineal pain following childbirth, were not eligible.
Intervention
We considered cryotherapy as any conservative intervention involving the local or general application of low-temperature components (cold), following the National Library of Medicine’s Medical Subject Headings.21 Cryotherapy may be delivered as an isolated cold modality, such as a cold pack (the application of a refrigerated gel or ice wrapped in cloth directly to the affected area),22 continuous cryotherapy (uninterrupted cold application over extended periods, generally maintained by automated devices that circulate cold water through pads or wraps to sustain tissue cooling),23 cold water immersion (the immersion of one or more limbs, or a substantial portion of the body, in cold water to promote regional or systemic cooling)24 or whole-body cryotherapy (brief exposure to cold air in specialised chambers for full-body cooling).25
Cryotherapy can also be delivered simultaneously in combination with other interventions, such as the RICE (rest, ice, compression and elevation), PRICE (protection, rest, ice, compression and elevation of the injured limb)14 26 and protection, optimal loading, ice, compression and elevation protocols,27 or through continuous cold compression devices, which circulate cold water in combination with pneumatic compression via a cuff (eg, the CryoCuff system),28 among others.
Trials investigating topical agents (eg, creams or sprays) designed to mimic the sensation of cold through chemical substances rather than by actual tissue cooling (such as menthol and ethanol),29 as well as invasive procedures, such as cryoneurolysis30 or intraoperative cold solution irrigation,31 were not considered eligible.
Comparator
Our comparisons of interest were control groups consisting of inert or no therapeutics (eg, placebo, no intervention, waiting list or sham) and comparison groups receiving the same components as the experimental group but without cryotherapy, to investigate the additional effects of cryotherapy as an adjuvant to other interventions.
In cryotherapy protocols where the cold component was combined with other therapeutic elements, such as cold compression devices, we only included trials where both the experimental and comparison groups received the same compression protocol, with refrigerated water applied exclusively to the experimental group. Similarly, for trials using the RICE/PRICE approach, both groups required rest, compression and elevation, with ice added solely to the experimental group. This methodological criterion ensured the effect of the cold component could be adequately assessed.
Outcomes
The outcomes of interest were pain intensity, ROM, swelling and function, assessed with any valid instrument such as Visual Analogue Scale (VAS) or Numerical Rating Scales32 for pain intensity; goniometry33 for ROM; for swelling, we considered volumetry or girth measures such as limb circumference and the figure-of-eight technique.25 34 When the girth was measured at various points of the limb, we collected the most central measurement (eg, the midpoint of the patella for knee swelling). For function, we considered self-reported questionnaires, such as the Lower Extremity Functional Scale (LEFS),35 and other valid outcome measures. We did not investigate pain pressure threshold nor consider opioid consumption as a direct and valid measure of pain intensity.
Data extraction
Two independent researchers (JPdM and RCCF) extracted characteristics and outcome data from included trials using a standardised data extraction form. Any disagreements between researchers were resolved through consensus or by a third researcher (VCO). The extracted data included information regarding the study design, the participants, details about the interventions, the comparator, outcomes and time points of the groups of interest. For our outcomes of interest, we extracted postintervention means (as the first option) or within-group mean changes over time (change scores), SDs and sample sizes for each of our groups of interest to investigate the immediate effects, as well as the short-term, medium-term and long-term effects. Immediate effects were defined as up to 3 days after allocation. Short-term effects were defined as over 3 days, but up to 14 days after allocation, medium-term effects were defined as a follow-up over 14 days but up to 12 weeks after allocation, and long-term effects were defined as a follow-up over 12 weeks after allocation.
When multiple time points were available within the same follow-up period of the included trial, the one closest to the end of the intervention was considered. When an included study had three or more arms and multiple groups of different cryotherapy modalities, we combined the group pairs to create a single group for pooling using recommended methods.19 When outcome data were not adequately provided, authors were first contacted. If no answer was received, we imputed missing data when possible. When authors did not respond, or imputation was not feasible, we reported the available data and the trial was excluded from the quantitative analysis. All procedures followed the recommended methods.19
Risk of bias assessment
Two independent researchers (JPdM and RCCF) assessed the risk of bias in the included trials using the 0–10 PEDro scale.36 According to this scale, higher scores represent a lower risk of bias. Discrepancies were resolved by consensus or a third investigator (VCO).
Data analysis
When possible, we performed meta-analysis using the random-effects model. The outcome data were converted to a common unit of measurement to report mean differences (MDs) and 95% CIs in forest plots. Data were converted to a 0–10 point scale for pain intensity (higher scores indicate worse pain intensity) and to a 0–100 point scale for function, where higher scores indicate better function. The magnitude of the effect was interpreted by comparing the estimated effect sizes and 95% CIs with the minimum clinically important difference (MCID) for the outcome of interest. An MCID of 2 points for pain intensity37 and 11 points for function38 was used. When we could not present results as MDs due to unmatching units of measurement, estimates were presented as standardised MDs (SMDs), using Hedges’ g to account for small sample sizes.19 In these cases, the magnitude of the effect sizes was interpreted according to Cohen’s d benchmarks of small (SMD <0.50), medium (SMD 0.50–0.79) and large effects (SMD ≥0.80).39 40
Quantitative analysis was conducted using Comprehensive Meta-analysis software, V.2.2.04 (Biostat, Englewood, New Jersey, USA). Statistical heterogeneity was assessed using the I² statistic. Different postoperative musculoskeletal disorders were combined in the same meta-analysis for a more robust estimate. Sensitivity and subgroup analysis were performed to explore the potential impact of specific diagnoses, modalities of cryotherapy and statistical heterogeneity when sufficient trials were included in the pooling. A funnel plot was generated when more than ten trials were included in a meta-analysis to identify possible publication bias. In cases where meta-analysis could not be performed due to insufficient or skewed data, we presented individual trials’ MDs and 95%CIs or median and IQRs as appropriate. These results were reported as supplementary files. All procedures adhered to recommended methods.19 41
Two independent researchers (JPdM and RCCF) assessed the certainty of the evidence using the GRADE system.17 18 Any disagreement was resolved by a third researcher (VCO). According to the four-level GRADE system, evidence may range from high to very low certainty. Low certainty indicates that future high-quality trials are likely to change estimated effects. In the current review, evidence began from high certainty and was downgraded by one point for each of the following issues: imprecision when analysed sample of less than 400 participants42; serious risk of bias when more than 25% of the analysed participants were from trials with a high risk of bias (ie, PEDro score less than 7 out of 10)43; severe inconsistency of results when I² statistics >50%, visual inspection of forest plot, or when pooling was not possible18; publication bias using visual inspection of funnel plot and the Egger’s test adopting an α=0.1 when data from at least ten trials were pooled in the same meta-analysis.44
Results
A total of 6261 records were retrieved from our searches, 1832 duplicates were removed and the remaining 4429 titles and abstracts were screened. Then, 165 potential full texts were assessed for eligibility, and 28 trials were included.45–72 The study selection flow diagram is available in figure 1.
Figure 1
The flow of studies through the literature search and screening. RCT, randomised controlled trial.
Characteristics of included trials and risk of bias assessment
The included trials were published between 1983 and 2024 and conducted in the USA (nine trials), China (five trials), Germany (two trials), Japan (two trials), South Korea (two trials) and Turkey (two trials). The remaining trials were conducted in England, Brazil, Canada, Australia, India and the Netherlands (one trial each).
Distinct postoperative musculoskeletal disorders were examined, including partial or TKA (12 trials), anterior cruciate ligament reconstruction (ACL-R) (6 trials), partial or total hip arthroplasty (3 trials), postoperative shoulder rehabilitation (PO shoulder) (2 trials), knee arthroscopy (2 trials), radius fracture reduction (1 trial), non-specific postoperative knee rehabilitation (PO Knee) (1 trial) and postoperative spine lumbar spinal rehabilitation (PO spine) (1 trial). The modalities of cryotherapy used were cold pack (14 trials), device-delivered continuous cryotherapy (12 trials) and cold air machine (3 trials). One trial did not specify the cryotherapy modality.47
The most frequently investigated outcome was pain intensity (22 trials), followed by ROM (18 trials), swelling (10 trials) and function (6 trials). The most frequently used outcome measures were the 0–10 VAS (15 trials) for pain intensity, goniometer (13 trials) for ROM, girth measurement using circumference (9 trials) for swelling, and the Western Ontario and McMaster Universities Osteoarthritis Index (2 trials) for function. Further information regarding the characteristics of the included trials is presented in online supplemental S2.
The PEDro scores of the included trials ranged from 1 to 7 points out of 10 (median=4 points). One trial was classified as having a low risk of bias (ie, score ≥7 points). The main sources of bias were related to lack of blinding of participants (28 trials, 100%), therapists (28 trials, 100%) and assessors (23 trials, 82.1%); lack of concealed allocation (24 trials, 85.7%) and absence of intention-to-treat analysis (26 trials, 92.9%). A detailed risk of bias assessment is presented in table 1.
Table 1
Risk of bias assessment using the 0–10 PEDro scale (n=28)
The effects of cryotherapy on pain intensity in postoperative rehabilitation for musculoskeletal disorders
We found moderate-certainty evidence of a statistically significant difference favouring the addition of cryotherapy to postoperative care for musculoskeletal disorders in the immediate (MD −0.77, 95% CI −1.23 to −0.31; I²=0%; 15 trials, n=1209), short term (MD −0.84, 95% CI −1.17 to −0.51; I²=0%; 8 trials, n=424) and medium term (MD −0.41, 95% CI −0.65 to −0.17; I²=0%; 6 trials, n=553). However, the effect sizes did not reach clinical relevance based on the MCID of 2 points. There was no statistically significant difference in the long term (MD −0.70, 95% CI −1.52 to 0.13; I²=0%; 2 trials, n=341; low certainty of evidence). The forest plot in figure 2 illustrates these findings. No publication bias was detected in the immediate effects (Egger’s test p value=0.12) (online supplemental file S3: Funnel plot for immediate effects of cryotherapy on pain intensity).
Figure 2
Forest plot of the effects of cryotherapy on pain intensity in postoperative care of musculoskeletal disorders. The types of surgery were partial and/or total knee arthroplasty (seven trials)45 59 63 67–69 71; anterior cruciate ligament reconstruction (three trials)50 51 61; partial and/or total hip arthroplasty (two trials)45 64; knee arthroscopy (one trial)60; shoulder arthroscopy (one trial)56; non-specific postoperative shoulder surgery (including arthroscopy, open surgery, shoulder arthroplasty and shoulder stabilisation) (one trial)66; radius fractures (one trial)49; spinal fusion (one trial).62 The dotted line represents the MCID. MCID, minimum clinically important difference.
The effects of cryotherapy on ROM in postoperative rehabilitation for musculoskeletal disorders
Given the heterogeneity of measurement units, estimates were reported as SMDs. Moderate certainty evidence showed a statistically significant difference in favour of the addition of cryotherapy to postoperative care for musculoskeletal disorders in the immediate (SMD 0.37, 95% CI 0.09 to 0.66; I²=17.7%; 7 trials, n=533; small effect size), short term (SMD 0.51, 95% CI 0.25 to 0.77; I²=2.83%; 8 trials, n=356; medium effect size) and medium term (SMD 0.61, 95% CI 0.04 to 1.18; I²=6.3%; 3 trials, n=140; medium effect size). The detailed forest plot is in figure 3.
Figure 3
Forest plot of the effects of cryotherapy on flexion ROM in postoperative musculoskeletal disorders. The types of surgery were partial and/or total knee arthroplasty (seven trials)45 55 59 60 65 67 72; anterior cruciate ligament reconstruction (four trials)50 58 61 70; hip arthroplasty (one trial)45; knee arthroscopy (one trial)60; radius fractures (one trial)49; non-specific postoperative knee surgery (PO knee) (including meniscus injuries, ligament repairs, fractures, chondromalacia and prior surgical interventions such as arthroscopy, osteotomy and partial knee arthroplasty) (one trial).57 ROM, range of motion.
The effects of cryotherapy on swelling in postoperative rehabilitation for musculoskeletal disorders
Given the heterogeneity of measurement units, SMDs were used, with the direction of the effects set as negative for better interpretation. We found a statistically significant difference favouring the addition of cryotherapy to postoperative care for musculoskeletal disorders in the short term (SMD −0.35, 95% CI −0.64 to −0.05; I²=5.7%; 7 trials, n=387; small effect size). No statistically significant difference was found in the immediate (SMD −0.37, 95% CI −0.90 to 0.17; I²=12.71%; 4 trials, n=256) or medium term (SMD −0.36, 95% CI −0.75 to 0.03; I²=14.2%; 4 trials, n=189). The certainty of evidence was low for all estimates. Forest plots are available in figure 4.
Figure 4
Forest plot of the effects of cryotherapy on swelling in postoperative musculoskeletal disorders. The types of surgery were total knee arthroplasty (six trials)52 54 59 67 68 72; shoulder arthroscopy (one trial)56; radius fractures (one trial)49; knee arthroscopy (one trial).60
The effects of cryotherapy on function in postoperative rehabilitation for musculoskeletal disorders
All studies included in the pooling investigated the additional effects of cryotherapy in postoperative TKA rehabilitation. We found a statistically significant difference favouring cryotherapy in the short term (MD 3.45, 95% CI 0.69 to 6.20; I²=28.36%; three trials, n=158). However, the effect size did not reach clinical relevance based on the MCID of 11 points. There was no statistically significant difference in the medium (MD 0.41, 95% CI −1.15 to 1.96; I²=0%; 3 trials, n=158) and long term (MD 0.49, 95% CI −0.75 to 1.72; I²=0%; 2 trials, n=300). The certainty of evidence was low for all estimates. The forest plot is in figure 5.
Figure 5
Forest plot of the effects of cryotherapy on function in postoperative musculoskeletal disorders. The type of surgery was total knee arthroplasty (four trials).55 68 71 72 The dotted line represents the MCID. MCID, minimum clinically important difference.
Sensitivity and subgroup analysis
Sensitivity analysis assessing the impact of missing data imputations showed no meaningful changes in the direction or magnitude of the estimates. An additional sensitivity analysis was conducted, pooling trials with homogeneous measurement units for ROM (flexion ROM goniometry in degrees) and swelling (limb girth measurement in centimetres) using MDs and 95% CI to confirm the robustness of the SMD analysis and provide a more clinically interpretable effect. Statistically significant improvements in knee flexion ROM were observed in the immediate (MD 10.43°, 95% CI 0.31 to 20.55; I²=0%; 4 trials; n=263), short term (MD 6.64°, 95% CI 0.78 to 12.50; I²=0%; 5 trials; n=207) and medium term (MD 4.82°, 95% CI 1.30 to 8.34; I²=4.6%; 3 trials; n=140). Statistically significant reductions in swelling were also found in the immediate (MD −0.69 cm, 95% CI −0.99 to −0.39; I²=0%; 3 trials; n=173), short term (MD −0.70 cm, 95% CI −0.98 to −0.42; I²=0%; 5 trials; n=237) and medium term (MD −0.73 cm, 95% CI −1.26 to −0.20; I²=0%; 3 trials; n=122). The certainty of evidence for both outcomes was rated as low. Forest plots are presented in online supplemental file S4: Sensitivity analysis.
Subgroup analysis was conducted to explore specific diagnoses based on body segments, including TKA, ACL-R, hip arthroplasty and postoperative shoulder rehabilitation, for the immediate effects of cryotherapy on pain intensity and ROM (online supplemental file S5: Subgroup analysis by specific musculoskeletal condition). Additionally, we explored the immediate effects of different cryotherapy modalities, including device-delivered continuous cryotherapy and cold packs, on pain intensity and ROM. Device-delivered continuous cryotherapy demonstrated a statistically significant difference in pain intensity (MD −1.03, 95% CI −1.70 to −0.36; I²=11.8%; n=8 trials; n=606) and ROM (MD 11.54, 95% CI 0.85 to 22.23; I²=0%; n=3 trials; n=342). Cold pack demonstrated a statistically significant difference in ROM (MD 10.22, 95% CI 2.02 to 18.41; I²=0%; n=3 trials; n=85), but not in pain intensity (MD −0.61, 95% CI −1.27 to 0.05; I²=0%; n=7 trials; n=603). The forest plots for these findings are presented in online supplemental file S6: Subgroup analysis by cryotherapy modalities.
Summary of findings, including estimates and the certainty of the evidence by disorder, outcome and time-point, is presented in online supplemental table 2: Summary of the findings with GRADE assessment.
Discussion
This systematic review with meta-analysis and GRADE assessment investigated the effects of cryotherapy on pain intensity, ROM, swelling and function in the postoperative care of musculoskeletal disorders. The main findings showed moderate-certainty evidence of a statistically significant difference in favour of the addition of cryotherapy to postoperative rehabilitation for musculoskeletal disorders for pain intensity and ROM. However, the effect sizes were predominantly small and may not reach clinical relevance. The certainty of evidence was low for swelling and function, with no clinically meaningful effect sizes observed. The primary reason for downgrading the certainty of evidence was the serious risk of bias across the included trials. Therefore, more high-quality RCTs are needed to strengthen the evidence and draw more robust conclusions.
Our findings align with previous systematic reviews. Martimbianco et al
73 found limited evidence suggesting cryotherapy may reduce pain intensity following arthroscopic ACL-R. Similarly, Aggarwal et al
10 reported low certainty evidence of statistically significant differences for pain intensity, ROM and swelling but not for function after TKA. The authors concluded that the potential benefits of cryotherapy may be too small to justify its use. Further reviews investigated cryotherapy in different musculoskeletal disorders but found insufficient evidence for more robust syntheses.9 13 74 Considering this, our systematic review provides the most updated evidence of cryotherapy in the postoperative management of musculoskeletal disorders.
Considering our findings, clinical practice guidelines should approach cryotherapy recommendations cautiously due to insufficient evidence of clinically relevant effects in these conditions. Interventions supported by robust evidence of clinical benefit ought to be prioritised to strengthen the evidence-based practice framework. There are suggestions from animal models that cryotherapy may delay tissue regeneration by disrupting key inflammatory and cellular responses essential for healing.75 76 In human postexercise recovery studies, cryotherapy has been shown to reduce circulating inflammatory and oxidative stress markers, leading to a transient delay in muscle regeneration due to incomplete degradation of damaged organelles.77–79 Moreover, hot water immersion has demonstrated greater benefits on muscle regeneration markers than cold water immersion.80 However, further research in humans with musculoskeletal injuries is needed to clarify the translational relevance of these findings and avoid overstatements. A paradigm shift in clinical decision-making regarding cryotherapy is needed, grounded in high-quality clinical studies with patient-centred outcomes rather than extrapolations from laboratory findings.
Most included trials in this systematic review had a high risk of bias and failed to score in critical methodological domains. The sensory nature of cryotherapy, where the cold sensation is easily perceived, inherently makes blinding the participant and therapist difficult, potentially introducing a risk of performance and detection bias.81 82 Although one trial59 reported blinding participants, no feasibility assessment to confirm their success was provided. Many studies frequently failed to score on the assessor blinding domain because they employed patient-reported outcome measures, in which the participant acts as their own assessor. Additional methodological weaknesses included inadequate allocation concealment and the absence of intention-to-treat analysis, which may increase the risk of selection bias by allowing the foreknowledge of group assignment and overestimating treatment effects by excluding non-compliant withdrawn participants.83 84 None of the included trials investigated potential sex-based differences in cryotherapy’s efficacy, leaving a gap in our understanding.
There is a need for more robust and representative MCIDs in the literature for clinical outcomes in the postoperative care of musculoskeletal disorders and across different cryotherapy modalities. We could not interpret the clinical relevance of cryotherapy’s effects in ROM and swelling subgroup syntheses due to the lack of MCID values in the literature. Although the MCIDs used in this systematic review for pain intensity and function provide a suitable parameter for effect size, it may be valuable for clinicians to consider the specific costs, dosage, facilities, expectations and patients’ goals if they opt to use cryotherapy as part of rehabilitation. Further investigations considering costs and benefits can more accurately guide the patient-oriented clinical decision-making process and define more robust MCIDs.
This systematic review was conducted with rigorous methodological standards, adhering to established guidelines19 and provided an updated synthesis of the evidence on the topic. While clinicians often prescribe cryotherapy based on its theoretical mechanisms, clinical decisions in a biopsychosocial and patient-centred context should focus on meaningful patient outcomes. Accordingly, this review offered evidence on relevant clinical outcomes, evaluated the certainty of the findings and presented interpretations on effect sizes, providing information for clinicians and patients for evidence-based decision-making. Moreover, we included only RCTs that investigated the separate effect of the cold component in cryotherapy modalities. This methodological decision ensured that observed effects could be specifically attributed to the cold component of cryotherapy, without confounding from additional therapeutic elements.
Nevertheless, this review has potential limitations. First, pooling diverse musculoskeletal diagnoses and surgical methods may have introduced heterogeneity. Second, we received no response to data requests from six trial authors.46 52 63 65 67 To address these issues, we performed subgroup analysis by clinical diagnoses and imputations from a previous meta-analysis5 to handle missing information, with sensitivity analysis to confirm that these approaches did not impact the direction or magnitude of the estimates. Finally, planned subgroup analysis of cryotherapy dosage was not feasible due to high variability and insufficient reporting.
Conclusions
Our systematic review and meta-analysis showed that cryotherapy is not supported by high-quality evidence. Moderate-certainty evidence suggested statistically significant effects favouring cryotherapy on pain intensity and ROM in the immediate, short-term and medium-term postoperative care of musculoskeletal disorders; however, the magnitude of the effects was small and may not be clinically relevant. Low-certainty evidence suggested trivial or no significant effects for swelling and function. Therefore, the overall clinical impact of cryotherapy in this context may be limited. Future high-quality RCTs addressing methodological issues are needed to improve the certainty of the evidence.
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information. Not applicable. All relevant data extracted for this systematic review are included within the manuscript and its supplementary information files.
Ethics statements
Patient consent for publication
Ethics approval
Not applicable.
