700696
research-article2017
CRE0010.1177/0269215517700696Clinical RehabilitationLee et al.
CLINICAL REHABILITATION
Article
Effectiveness of neuromuscular electrical stimulation for management of shoulder subluxation post-stroke: a systematic review with meta-analysis
Clinical Rehabilitation 2017, Vol. 31(11) 1431–1444 © The Author(s) 2017 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav https://doi.org/10.1177/0269215517700696 DOI: 10.1177/0269215517700696 journals.sagepub.com/home/cre
Jae-Hyoung Lee1, Lucinda L Baker2, Robert E Johnson3 and Julie K Tilson2
Abstract Objectives: To examine the effectiveness of neuromuscular electrical stimulation (NMES) for the management of shoulder subluxation after stroke including assessment of short (1 hour or less) and long (more than one hour) daily treatment duration. Data sources: MEDLINE, CENTRAL, CINAHL, WOS, KoreaMed, RISS and reference lists from inception to January 2017 Review methods: We considered randomized controlled trials that reported neuromuscular electrical stimulation for the treatment of shoulder subluxation post-stroke. Two reviewers independently selected trials for inclusion, assessed trial quality, and extracted data. Results: Eleven studies were included (432 participants); seven studies were good quality, four were fair. There was a significant treatment effect of neuromuscular electrical stimulation for reduction of subluxation for persons with acute and subacute stroke (SMD:–1.11; 95% CI:–1.53, –0.68) with either short (SMD:–0.91; 95% CI:–1.43, –0.40) or long (SMD:–1.49; 95% CI:–2.31, –0.67) daily treatment duration. The effect for patients with chronic stroke was not significant (SMD:–1.25; 95% CI:–2.60, 0.11). There was no significant effect of neuromuscular electrical stimulation on arm function or shoulder pain. Conclusion: This meta-analysis suggests a beneficial effect of neuromuscular electrical stimulation, with either short or long daily treatment duration, for reducing shoulder subluxation in persons with acute and subacute stroke. No significant benefits were observed for persons with chronic stroke or for improving arm function or reducing shoulder pain.
1Department
of Physical Therapy, Wonkwang Health Science University, Iksan, Korea 2Division of Biokinesiology and Physical Therapy, University of Southern California, Los Angeles, California, USA 3Norris Medical Library, University of Southern California, Los Angeles, California, USA
Corresponding author: Julie K Tilson, Division of Biokinesiology and Physical Therapy, University of Southern California, Los Angeles, California, USA. Email:
[email protected]
1432
Clinical Rehabilitation 31(11)
Keywords Electrical stimulation, stroke, neuromuscular electrical stimulation, hemiplegia, shoulder subluxation Received: 26 July 2016; accepted: 16 February 2017
Introduction Shoulder subluxation is one of the most common complications for patients with hemiplegia following a stroke.1 In this review, shoulder subluxation post-stroke is defined as partial separation of the humeral head from the glenoid cavity in the inferior direction. For individuals with hemiplegia post-stroke, inferior shoulder subluxation develops as a result of a prolonged downward pull of gravity on the arm against hypotonic muscles, resulting in overstretch of both the glenohumeral capsule and the hypotonic supraspinatus and deltoid muscle.2 Subluxation is most likely to develop during the flaccid stage post-stroke, when individuals experience severe motor deficits associated with hypotonia and/or weakness of the supporting shoulder structures (musculature, ligaments and capsule).3 Shoulder subluxation has been shown to impede recovery of upper extremity function by reducing shoulder range of motion and increasing risk for Complex Regional Pain Syndrome and complex soft tissue shoulder injuries.3,4 These sequelae may lead to extended hospital stays, increased therapy costs and reduced health-related quality of life. Without treatment, subluxation can worsen over time and eventually become uncorrectable.5 Thus, management of shoulder subluxation should be an important part of upper extremity rehabilitation. Various treatment methods have been proposed to reduce shoulder subluxation post-stroke including positioning using lap boards, arm troughs, shoulder slings, shoulder strapping, and electrical stimulation to the shoulder musculature.1–4 Neuromuscular electrical stimulation involves the use of an electrical stimulator that transmits an electrical impulse through the skin. Surface electrodes are used to transmit electrical stimulation to superficial nerve and muscle groups to elicit muscle contraction. Neuromuscular electrical stimulation may reduce shoulder subluxation in hemiplegia and may prevent further joint separation by
strengthening and maintaining muscle mass of the posterior deltoid and supraspinatus muscles. These muscles in particular counteract gravity-induced displacement of the humeral head from the glenoid cavity.6 Five previous systematic reviews7–11 have suggested that electrical stimulation is effective in reducing shoulder subluxation in early stroke. However, previous reviews7–9 included only a small number randomized controlled trials or included non-randomized or controlled studies. The two most recent reviews10,11 were limited to 6 randomized controlled trials.6,12–16 In this systematic review and meta-analysis, 5 new trials were added (4 Korean16,18–21 and 1 Chinese22) for a total of 11 randomized controlled trials. In addition, this review is the first to assess the effectiveness of short versus long daily neuromuscular electrical stimulation treatment duration for reducing shoulder subluxation post-stroke. The primary purpose of this systematic review was to examine the effectiveness of neuromuscular electrical stimulation treatment combined with conventional therapy for the management of shoulder subluxation in patients post-stroke compared to conventional therapy alone. Short versus long daily duration of neuromuscular electrical stimulation treatment was evaluated, each compared to conventional therapy alone. Finally, we examined the effectiveness of shoulder stimulation for improving arm motor function and reducing shoulder pain post-stroke.
Methods This systematic review was performed in accordance with the Cochrane Collaboration23 and the Preferred Reporting Items for Systematic Review.24 The protocol was registered with the International Prospective Register of Systematic Reviews on 11 October 2015 (CRD42015026299).
1433
Lee et al.
Eligibility criteria
Search strategy
We considered all randomized controlled trials that assessed the effect of neuromuscular electrical stimulation plus conventional therapy on shoulder subluxation after stroke compared to conventional therapy alone. No language, publication date, or publication status restrictions were imposed. We considered studies with participants with shoulder subluxation after stroke with any stroke type, age, gender, social factors, or clinical setting. We excluded studies with participants with shoulder subluxation associated with other neurological conditions (e.g. brain injury, spinal cord injury, cerebral palsy). We included studies assessing electrical stimulation, whether called neuromuscular electrical stimulation, functional electrical stimulation, or electrical stimulation that was used to reduce shoulder subluxation using surface electrodes. Studies using electrical stimulation combined with other treatment techniques for shoulder subluxation were included. We excluded studies assessing electrical stimulation with invasive electrodes (i.e., percutaneous or intramuscular electrical stimulation) because these invasive techniques are not available for practicing clinicians. Additionally, invasive electrodes recruit deeper muscles, providing a fundamentally different intervention compared with surface stimulation. In addition, we excluded studies which used electrical stimulation protocols exclusively designed for pain management because such protocols would purposely avoiding stimulated muscle contraction. We included studies that assessed glenohumeral joint distance using methods of radiographic or thermoplastic jig measurement.25 Studies using finger width, surface caliper, or measuring tape were excluded due to high risk for poor reliability of measurement. Secondary outcome measures were assessed from studies that met the primary outcome measure criteria. Those included motor function as measured by various motor assessments26–29 and shoulder pain as measured by visual analog scale, verbal rating scale, or pain-free shoulder range of motion.
Studies were identified by searching electronic databases, scanning reference lists of articles and consulting with experts in the field. Literature searches were conducted in the following electronic databases: PubMed (Medline), Cochrane Library, Cumulative Index to Nursing and Allied Health Literature, Web of Science, KoreaMed, and Research Information Sharing Service from inception to January 19, 2017. We used the following search terms: electrical stimulation, neuromuscular electrical stimulation, NMES, functional electrical stimulation, FES, shoulder joint, subluxation, pain, motor function, stroke and hemiplegia. The search strategy used for the Medline database is outlined in the Supplementary Appendix.
Study selection and data collection process Two reviewers separately and independently screened the titles and abstracts of studies identified in the initial searches. A standard screening checklist based on the eligibility criteria was used for each study. Studies that did not meet the eligibility criteria according to the titles or abstracts were excluded. Full text versions of the remaining studies, including those identified as potentially eligible or uncertain, were retrieved for further independent review by the two reviewers to determine eligibility. Disagreements regarding study eligibility were resolved by discussion between the reviewers. When consensus was not reached, a third reviewer arbitrated. The reviewers were not blinded to the authors and institutions of the studies undergoing review. Two reviewers independently assessed the risk of bias of the included studies using the Cochrane Collaboration's tool for assessing risk of bias in randomized trials.30 The two reviewers also assessed the methodological quality of the included studies using the PEDro scale.31 Studies scoring 9-10 on the PEDro scale were considered to be of excellent methodologic quality, 6-8 were considered good, 4-5 were considered fair, and scores below 4 were considered poor quality.32
1434
Clinical Rehabilitation 31(11)
Two reviewers independently extracted data from included studies using a data extraction sheet based on the Cochrane Consumers and Communication Review Group’s data extraction template.33 The following data were extracted from included studies: type of study, number of participants, intervention protocol, comparison groups, and results. We contacted study authors via e-mail at least 2 times to obtain missing data. If authors did not provide needed data, the study in question was excluded from meta-analyses as necessary. For quantitative synthesis, mean change and associated standard deviation were obtained by comparing change from baseline to end of treatment for each group. If standard error was reported, data was calculated to determine standard deviation using the formula Standard Error x √n. If median and minimum/maximum were reported, data was converted to mean and standard deviation using the formulae described by Hozo et al.34 If standard deviations were not reported in an article’s text, they were estimated by measuring the length of lines indicating standard deviation from bar graph figures when available.
which values above 25% and 50% were considered indicative of moderate and high heterogeneity, respectively. Fixed effects models were used for nonsignificant findings of heterogeneity; random effects models were used in the case of significant findings of heterogeneity. We also performed sensitivity analyses, based on the PEDro scale (6 or more points vs less than 6 points), method of subluxation measurement (radiographic vs thermoplastic jig measurement), and source of standard deviation data (reported directly in study vs measured from standard deviation lines in figures).36 All analyses were performed using Review Manager (RevMan) software, version 5.3.37
Data analysis
Description of studies
We qualitatively assessed the articles for clinical homogeneity and performed meta-analysis when studies were sufficiently similar to combine. To perform the meta-analysis, mean and standard deviation change scores of each outcome measure were pooled. We divided study participants by acuity for meta-analysis: acute/subacute or chronic. The cut-off point for acute/subacute and chronic stage was defined as 6 months poststroke.35 In addition, we divided studies by daily treatment duration: short (⩽1 hour/day) and long (>1 hour/day). For continuous outcomes, if the unit of measurement was consistent across trials, the results were presented as the weighted mean difference with 95% confidence intervals (95% CIs); if the unit of measurement was inconsistent, results were expressed as the standard mean difference with 95% CI. Statistical heterogeneity was assessed using heterogeneity (x2) and inconsistency (I2) tests, in
The characteristics of each study (e.g. study design, number of participants, intervention, outcomes, and results) are described in Table 1. The eleven studies included 432 participants; 216 in neuromuscular electrical stimulation plus conventional therapy and 216 in conventional therapy only groups. One study14,15 reported subcategories of participants based on acuity and were analyzed accordingly. Participants with acute stroke (9 studies, n=325) were 2.0±2.2 months and 1.6±1.7 months post-stroke in intervention groups (n=163) and control groups (n=162), respectively. Participants with chronic stroke (4 studies, n=107) were 9.4±4.1 months and 9.1±3.9 months poststroke in intervention groups (n=53) and control groups (n=54), respectively. All studies measured subluxation in millimeters using antero-posterior radiograph of the shoulder except one study21 that measured subluxation on the surface of the skin in centimeters using a thermoplastic jig.
Results We initially identified 1480 articles after deleting duplicates from 1822 articles identified in six databases (Figure 1). Two articles14,15 reported different outcome measures from the same study and were therefore considered one study. Ultimately, eleven randomized controlled trials met the inclusion criteria and were included in the systematic review.
1435
Lee et al.
Figure 1. Flow diagram of included studies.24
Details of each study’s neuromuscular electrical stimulation parameters (e.g. pulse rate, pulse duration) are described in Supplementary Table 1. Neuromuscular electrical stimulation frequency ranged from 10 to 36 pulses per second. Mean treatment program parameters were 1.4±0.9 sessions per day (range 1-4), 136.0±133.8 minutes per day (range 20-335), and 5.2±1.0 sessions per week
(range 3-7). Total mean neuromuscular electrical stimulation treatment duration per week was 12.6±12.7 hours (range 1-31.5).
Risk of bias in included studies Supplementary Figure 1 summarizes the risk of bias evaluations for the included studies. PEDro
Design/ PEDro score
2- arm RCT 5/10
2- arm RCT 4/10
2- arm RCT 6/10
Study
Baker and Parker, 1986.6
Faghri et al., 199412
Linn et al., 199913 N=40 Experimental group (n=20): Age: 71 years Onset: 2 days Control group (n=20): Age: 73 years Onset: 2 days
N=63 Experimental group (n=31): Age: 56±13 years Onset: 49±32 days Control group (n=32): Age: 55±12 years Onset: 46±51 days N=26 Experimental group (n=13): Age: 65±13 years Onset: 16±5 days Control group (n=13): Age: 69±12 years Onset: 17±4 days
Participants
Table 1. Characteristics of included studies. Results • Subluxation: - Reduced significantly (p
