Effect of Neuromuscular Electrical Stimulation After Total Knee Arthroplasty: A Systematic Review and Meta-Analysis of Randomized Controlled Trials

Background: This systematic review and meta-analysis aimed to evaluate the effect of neuromuscular electrical stimulation (NMES) on quadriceps muscle strength, pain, and function outcomes following total knee arthroplasty (TKA). Methods: PubMed/Medline, Embase, Web of Science, CENTRAL, Scopus, PsycINFO, PEDro, CINAHL, CNKI, and Wanfang were systematically searched for randomized controlled trials (RCTs) from their inception to 18 June 2021. Results: Nine RCTs that involving 691 patients were included in the meta-analysis. Our pooled analysis showed that NMES improved quadriceps muscle strength after TKA within 1 months [standardized mean difference (SMD): 0.81; 95% CI: 0.51–1.11], 1–2 months (SMD: 0.55; 95% CI: 0.13–0.97), 3–4 months (SMD: 0.42; 95% CI: 0.18–0.66), and 12–13 months (SMD: 0.46; 95% CI: 0.18–0.74), pain between 1 and 2 months [mean difference (MD): −0.62; 95% CI: −1.04 to −0.19], pain between 3 and 6 months (MD: −0.44; 95% CI: −0.74 to −0.14) Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) between 3 and 4 months (MD: −0.43; 95% CI: −0.82 to −0.05), timed up and go test (TUG) within 1 month (MD: −2.23; 95% CI: −3.40 to −1.07), 3 minutes walk test between 3 and 6 months (MD: 28.35; 95% CI: 14.55–42.15), and SF-36 MCS between 3 and 6 months after TKA (MD: 4.20, 95% CI: 2.41–5.98). Conclusion: As a supplementary treatment after TKA, postoperative NMES could improve the short-term to long-term quadriceps muscle strength, mid-term pain, and mid-term function following TKA. However, many outcomes failed to achieve statistically meaningful changes and minimal clinically important difference (MCID), thus the clinical benefits remained to be confirmed. Level of Evidence: Therapeutic level I. Systematic Review Registration: https://www.crd.york.ac.uk/PROSPERO/, identifier CRD42021265609.


INTRODUCTION
Total knee arthroplasty (TKA) is one of the most common and cost-effective procedures for patients with end-stage osteoarthritis of the knee, which has been performed with increasing frequency in recent years (1,2). Although TKA provides patients with reduced pain and a functional range of motion (ROM) of the knee joint, quadriceps strength impairment is common following the surgery (3). Besides, almost all patients suffer from postoperative pain with different levels, which affect postoperative satisfaction and outcomes (4). Studies have shown that nearly 20% of primary TKA patients were not satisfied with their outcomes following the surgery (5).
Standardized physical therapy and pharmacologic analgesia improve muscle strength and pain after TKA (6). However, the content of rehabilitation varies worldwide (7,8). Electrical stimulation is effective in accelerating recovery from surgery (9). Neuromuscular electrical stimulation (NMES) has been utilized since the eighteenth century (10). It was regarded as a potential approach to improve muscle contractility and postoperative quadriceps weakness (11). A systematic review involving a total of 933 participants found that NMES may be an effective treatment for muscle weakness and should be regarded as a crucial part of rehabilitation programs (12). The advantages of NMES have been emphasized in many diseases such as anterior cruciate ligament injury, neck pain, stroke, and cerebral palsy (13)(14)(15)(16)(17).
Recently, some studies have examined the effect of NMES following TKA but remain controversial (18)(19)(20)(21)(22)(23). The clinical effectiveness of NMES following TKA on quadriceps muscle Abbreviations: NMES, neuromuscular electrical stimulation; TKA, total knee arthroplasty; PRISMA, preferred reporting items for systematic reviews and metaanalyses; ROM, range of motion; RCT, randomized controlled trial; MCS, mental component score; PCS, physical component score; SF-36, 36-item short form health survey; CENTRAL, Cochrane central register of controlled trials; CNKI, China national knowledge infrastructure; PCI, physiological cost index; VAS, visual analog scale; NPRS, numerical pain rating scale; WOMAC, western ontario and mcmaster universities osteoarthritis index; TUG, timed up and go test; SCT, stair climbing test; 3MWT, 3 minutes walk test; 6MWT, 6 minutes walk test; MD, mean difference; SMD, standardized mean difference; MVIC, maximal voluntary isometric contraction; BMI, body mass index; mNMES, motor-level NMES; sNMES, sensory-level NMES; KOS ADLS, knee outcome survey activities of daily living scale; HRQoL, health-related quality-of-life; MCID, minimal clinically important difference. strength, pain, and function outcomes remains unclear. We conduct this meta-analysis to evaluate the effect of NMES on quadriceps muscle strength, pain, and function outcomes following TKA further.

METHODS
This study was based on the previous published RCTs. Thus, the ethical approval and consent to participate were not necessary. This systematic review and meta-analysis is performed following the Cochrane Handbook for Systematic Reviews of Interventions (24) and Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (25). The protocol was registered in the PROSPERO (Registration number: CRD42021265609).

Search Strategy
The PubMed/Medline, Embase, Web of Science, the Cochrane Central Register of Controlled Trials (CENTRAL), Scopus, PsycINFO, Physiotherapy Evidence Database (PEDro), CINAHL, China National Knowledge Infrastructure (CNKI), and Wanfang (a Chinese database) were systematically searched for randomized controlled trials (RCTs) from their inception to 18 June 2021 by two independent reviewers (LBP and KXW). The search strategies were shown in Appendix 1.

Eligibility Criteria
The studies included in the meta-analysis were required to meet the following inclusion criteria: (1) Patients: adult patients undergoing primary TKA; (2) Intervention: postoperative NMES. NMES was utilized in the intervention group after the TKA surgery. Patients who received preoperative NMES were excluded; (3) Comparison: conventional rehabilitation or conventional physical therapy. Patients who received any form of electrical stimulation in the control group were excluded; (4) Outcomes: The primary outcome measures, such as quadriceps muscle strength [maximal volitional isometric contraction (MVIC)], physiological cost index (PCI), pain such as visual analog scale (VAS), and numerical pain rating scale (NPRS), Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), Timed Up and Go Test (TUG), Stair-Climbing Test (SCT), 3 Minutes Walk Test (3MWT), 6 Minutes Walk Test (6MWT), range of motion (ROM), and 36-Item Short-Form Health Survey (SF-36); (5) Study design: randomized controlled trials (RCTs); language and published time restrictions were not employed.
Maximal volitional isometric contraction is a classic method to calculate muscle strength for patients with neuromuscular disorders by providing intrinsic factors such as units of kilograms and Newtons of force (26). As a measure of energy cost, PCI was calculated by dividing the heart rate increase (heart rate at the end of the 3MWT minus resting heart rate) by walking speed (m/min) (18,21). A lower PCI score indicated a lower energy cost during walking (27). To assess the TUG score, patients were asked to rise from an armchair, walk 3 m away, then turn and walk back to sit down on the same chair (28,29). The TUG is an excellent representation of essential mobility, strength, balance, and agility (29). The SCT was used to assess the lower extremity strength, power, and balance (29). The 3MWT is a simple, non-incremental, and easy to conduct submaximal strength test (30). As a standard walking test, 6MWT has been widely used to determine the progress following rehabilitation intervention (31). The SF-36 is the most widely used healthrelated quality-of-life (HRQoL) in the USA (32). The SF-36 is consists of eight individual subscales. Scores of those subscales can be combined into two higher-order summary scores: PCS and MCS (33).

Study Selection
Firstly, all the identified studies were imported into the Endnote X9 (Thomson Reuters, CA, USA). After removing the duplicate studies, two reviewers (YGW and HBS) scanned the titles, abstracts, and full texts independently. Any disagreements were resolved by discussion with a senior reviewer (YZ). Commentaries, letters, case reports, trial protocols, reviews, and retrospective studies were excluded from our systematic review and meta-analysis.

Data Extraction
Two authors (LBP and YGW) extracted the following data independently and discussed with a senior reviewer (HBS) if disagreements existed.

Study Quality Assessment
Two authors (KXW and YZ) evaluated the methodological quality of the included studies independently with the Cochrane bias risk assessment tool and discussed with a senior reviewer (BS) if any disagreements existed (34). Each study was documented with low, high, or unclear risk of bias in each domain.

Statistical Analysis
The review manager software (RevMan 5.3, Oxford, United Kingdom) was used to conduct our meta-analysis and produce forest plots. All the continuous variable outcomes were presented as the mean difference (MD) with a 95% CI to calculate the total effect of NMES on patients following TKA. The standardized mean difference (SMD) was used to calculate the total effect if different scales were utilized among the studies (35). I 2 statistics measured heterogeneity among the studies. The random-effects model was used if substantial heterogeneity exists (P < 0.05 or I 2 > 50%). If not, the fixed-effects model was adopted. A P < 0.05 demonstrated a statistically significant difference.

Study Selection
A total of 761 studies were identified from the initial search. After removing 344 records for duplicates, 353 studies were excluded by screening the title and abstract. After excluding three studies for not being retrieved, the remaining 61 studies were screened the full-text for eligibility. Fifty-two reports were excluded by screening the full-text: not NMES (n = 19); not postoperative intervention (n = 3); not conventional rehabilitation in the control group (n = 3); not RCT (n = 8); retrospective study (n = 2); protocol (n = 8); review (n = 4); not TKA (n = 2); and outcomes not related (n = 3). The remaining nine RCTs involving 691 patients met the eligibility criteria and were included in the meta-analysis (18,19,21,22,(36)(37)(38)(39)(40). The PRISMA flow diagram was shown in Figure 1.

Interventions
All the involved patients received similar conventional rehabilitation/physical therapy/exercise. Besides, patients in the NMES groups received similar NMES therapy in all the included nine RCTs. The frequency of NMES ranged from 30 to 100 Hz, and the duration ranged from 300 µs to 1 ms in six studies (18,21,22,36,38,39). Klika et al. set the frequency as 50 pps and the pulse width as 5 ms (40). Two studies did not document the frequency and duration data of the NMES protocol (19,37). Three RCTs reported that the NMES therapy was used twice daily (18,21,22 (40). Seven studies utilized the NMES with maximum tolerable intensity (18,19,21,22,(38)(39)(40). The other two RCTs did not report the intensity data of NMES (36,37). The intervention characteristics were shown in Table 1.

Quadriceps Muscle Strength
Four studies reported the quadriceps muscle strength (19,22,39,40). Since the included studies adopt different intrinsic factors to normalize MVIC values, the SMD was used to calculate the total effect of quadriceps muscle strength. Our pooled analysis involving four studies indicated that NMES improved MVIC after TKA within 1 month (SMD: 0.81; 95% CI: 0.51-1.11, P <  . We found no significant heterogeneity (I 2 = 0%; I 2 = 8%; respectively). The forest plot of PCI was shown in Figure 5.

6MWT
Two studies evaluated the effect of NMES on 6MWT following TKA (19,22). Our pooled analysis involving those two

DISCUSSION
The purpose of this systematic review and meta-analysis was to explore the effect of NMES on quadriceps muscle strength, pain, and function following TKA. The most important finding of the current study was that postoperative NMES could improve the short-term to long-term quadriceps muscle strength, mid-term pain, and mid-term function following the TKA surgery. Quadriceps muscle weakness is common following TKA (41). It was reported that 50-60% quadriceps muscle strength deficits might occur compared with the preoperative levels (23). Besides, quadriceps weakness has been found to increase joint loading and contribute to the progress of osteoarthritis (42,43). It is a crucial goal to restore quadriceps strength for postoperative rehabilitation. Monaghan et al. systematically reviewed relative studies up to 2008. They did not find direct evidence to prove the advantage of NMES on quadriceps muscle strength recovery following TKA by limited (only two) included RCTs (44). Conley et al. conducted a systematic review involving eight RCTs to assess the effect of NMES on quadriceps strength after knee surgery, such as anterior cruciate ligament reconstruction (n = 5), TKA (n = 2), and meniscectomy (n = 1) (43). They revealed that NMES improved the recovery of quadriceps strength after the knee surgery with grade B evidence (43). Due to limited pieces of relevant literature published, some other reviews also failed to examine the effect of NMES on MVIC following TKA (45,46). After including more RCTs that have been published recently, for the first time, we proved that NMES improved quadriceps muscle strength in terms of MVIC after TKA within 1 month, 1-2 months, 3-4 months, and 12-13 months with high-quality evidence.
After including only one published RCT, one previous review failed to explore the pooled effect of NMES on PCI following TKA (45). We found that NMES could not improve PCI after TKA with high-quality evidence.
There was a high risk of severe acute postoperative pain following TKA, undermining the recovery and delaying the fasttrack rehabilitation programs (47). Dabadghav et al. included 28 bilateral TKA patients following osteoarthritis. One knee received NMES plus exercise therapy randomly, and the other  knee received exercise merely. After immediate postsurgical rehabilitation of 7 days, no significant difference between the two knees in terms of pain was detected (48). A matched comparison trial demonstrated that patients using the home-based NMES in the first 6 weeks relieved the pain (49). There was no previous meta-analysis that has evaluated the effect of NMES on postoperative pain after TKA. We included five RCTs and found that NMES improved postoperative pain at mid-term (1-2 months and 3-6 months) following TKA. However, the improved differences did not reach the minimal clinically important difference (MCID) in pain (50).
WOMAC is one of the most commonly used questionnaires that assess symptoms and physical function in patients with lower limb osteoarthritis (51,52). Additional NMES therapy of 8 weeks to exercise could not improve the WOMAC in patients with knee OA (53). No previous meta-analysis pooled analyzed the effect of NMES on postoperative function after TKA. We found that NMES improved postoperative WOMAC slightly at mid-term (3-4 months) following TKA with high-quality evidence.
After a matched comparison trial of 6 weeks, patients who used the home-based NMES improved TUG compared with patients in the control group following TKA (49). No previous meta-analysis analyzed the effect of NMES on postoperative TUG following TKA. With high-quality evidence, we found that NMES improved the postoperative TUG at short-term (within 1 month) following TKA. Bruce-Brand et al. found that patients who received home-based NMES for 6 weeks improved the SCT results than patients who received standard care in knee osteoarthritis (54). We discovered that NMES could not improve postoperative SCT with low-quality evidence. The effect of NMES on early TUG (within 1 month) has not been explored in previous literature. The SCT comprises different movements, such as stair ascending, descending, and their transition (55), which may be hard for patients to conduct in the short-term following TKA.
We found that NMES improved postoperative 3MWT at midterm with high-quality evidence but not long-term with lowquality evidence. Another two RCTs reported the effect of NMES on 6MWT in our study. We found no significant difference between NMES and control groups in terms of 6MWT at midterm or long-term. The previous studies emphasized an excellent correlation between 3MWT and 6MWT. 3MWT was easier to learn and repeat than 6MWT for patients (30). The advantage of NMES was found in terms of 3MWT but not 6MWT. The discrepancy may be related to the limited RCTs.
We detected no advantage of NMES in ROM in the present study, such as knee flexion and extension. Dabadghav et al. included 28 postoperative bilateral TKA patients and randomly allocated one knee to NMES plus exercise, while the other knee received exercise (48). They demonstrated no additional effect in terms of ROM between the two knees from 28 bilateral TKA patients (48). Results from an RCT also showed that NMES could not improve the ROM of the hemiplegic shoulder in patients after stroke (56).
We included four RCTs (involving 344 participants) and found that NMES improved SF-36 MCS at mid-term (3-6 months). No differences were found in terms of MCS at shortterm or long-term. Besides, NMES could not improve PCS at short-term, mid-term, or long-term. A previous meta-analysis confirmed the advantage of NMES on SF-36 MCS at mid-term (12 weeks) by Bistolfi et al. (45). However, they did not explore the effect of NMES on SF-36 at other periods.
Compared with a previous meta-analysis, Bistolfi et al. only included four RCTs and pooled evaluate the effect of NMES on SF-36 merely (45). The evidence to prove the advantage of NMES in TKA was limited. Given the insufficiency of available data for comparison, some other reviews failed to explore the quantized effect of NMES on TKA (43,44,46,57,58). As far as we knew, this was the first systematic review and meta-analysis to comprehensively explore the effect of NMES on quadriceps muscle strength, pain, and function following TKA. However, the differences did not reach the MCID in pain (50). Given the included studies adopt different scales in many essential outcomes, the SMD was used to calculate the total effect of quadriceps muscle strength and WOMAC, which may generate issues with heterogeneity.
The study has several limitations. First, all the included studies failed to achieve the performance bias, which contributed to the main bias of this systematic review and meta-analysis. Second, the programs of NMES were not standardized among the included RCTs, which contributes to the heterogeneity. Third, the sample size was relatively small.

CONCLUSION
As a supplementary treatment after TKA, postoperative NMES could improve the short-term to long-term quadriceps muscle strength, mid-term pain, and mid-term function following TKA. However, many outcomes failed to achieve statistically meaningful changes and MCID, thus the clinical benefits remained to be confirmed.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

AUTHOR CONTRIBUTIONS
BS and LP conceived and designed the analysis. LP and KW search studies from the databases and analyzed data. YZ, YW, and HS participated in the selection of the studies. LP, YW, and HS extract the data. KW, YZ, and BS participated in the quality assessment. LP drafted the manuscript. BS and YZ ensured the accuracy of the data and analysis. All authors have read and approved the manuscript.