Agreement among multiple meta-analyses on the effectiveness and safety of virtual reality rehabilitation after stroke: an overview of systematic reviews

Introduction

Worldwide, stroke is the second leading cause of death and a major cause of disability, with over 12 million new strokes reported each year. With advances in health technologies, the range of interventions for stroke survivors is in continuous expansion. Among these, virtual reality (VR) in neurorehabilitation has proved an engaging, interactive, patient-centred, and relatively inexpensive modality to enhance functional recovery. We aim to conduct an overview of systematic reviews exploring the agreement on the effectiveness and the safety of VR technologies for clinical outcomes in stroke survivors to give a comprehensive balance of effects.

Methods

We searched multiple databases up to 17 January 2023 for systematic reviews comparing any kind of VR technology (with or without conventional therapy) versus conventional therapy alone. The primary outcome was upper limb function and activity. The secondary outcomes were gait, mobility and balance, limitation of activities, participation, cognitive and mental function and adverse events. Methodological quality was assessed using the A MeaSurement Tool to Assess systematic Reviews (AMSTAR 2) and the certainty of evidence (CoE) using the Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) approach. Considering reviews assessing the same clinical questions on the same outcome measurement, we examined concordance and discordance of meta-analyses effects sizes (i.e., effective intervention vs no difference) using a conceptual framework based on the Jadad algorithm.

Results

We included 58 reviews of 345 unique primary studies. Overall, 42 (72.4%) had conducted meta-analysis  Many reviews assessed mixed (e.g., both subacute and chronic) (69%) or chronic onset of stroke (17.2%) and were judged critically low in quality by AMSTAR 2 (77.2%). For the primary outcome, meta-analyses reported discordant findings in the direction of effects. Applying the Jadad algorithm, VR with or without conventional therapy seems to be more effective than conventional therapy alone on upper limb function (Fugl-Meyer Assessment for Upper Extremity [ FMA-UE ]), with low to moderate CoE and possible to definite clinical relevance (Figure 1). For secondary outcomes there was uncertainty about the superiority or no difference between groups due to substantial heterogeneity of measurement scales (Figure 2). A few reviews (n=6) reported the occurrence of mild adverse events.

Discussion and Conclusion

Current evidence suggests that multiple meta-analyses agreed on the superiority of VR combined or not with conventional therapy over conventional therapy on FME-UE. These findings support the hypothesis that VR may help to improve the recovery of upper limb motor function and quality of movement. As a safe intervention, clinicians should consider embed VR technologies into their practice and adapt them according to patients’ needs and preferences. Caution in the interpretation of findings is warranted given the poor methodological quality of the reviews.

REFERENCES

Collaborators, G. B. D. Stroke. “Global, Regional, and National Burden of Stroke and Its Risk Factors, 1990-2019: A Systematic Analysis for the Global Burden of Disease Study 2019.” Lancet Neurol 20, no. 10 (Oct 2021): 795-820.

Imbimbo, I., D. Coraci, C. Santilli, C. Loreti, G. Piccinini, D. Ricciardi, L. Castelli, et al. “Parkinson’s Disease and Virtual Reality Rehabilitation: Cognitive Reserve Influences the Walking and Balance Outcome.” Neurol Sci 42, no. 11 (Nov 2021): 4615-21.

Jadad, A. R., D. J. Cook, and G. P. Browman. “A Guide to Interpreting Discordant Systematic Reviews.” [In eng]. Cmaj 156, no. 10 (May 15 1997): 1411-6.

COGNITIVE-MOTOR INTERFERENCE DURING DUAL-TASK IN PATIENTS WITH PARKINSON’S DISEASE

Introduction

Most activities of daily living involve performing several tasks at once: for example, walking while talking or carrying an object requires attention to be divided between competing tasks^1,2.

Patients with Parkinson’s Disease (PD) are characterised by loss of automaticity of gait, along with deficits of executive functions and attention; thus, performing concomitant tasks (i.e., dual-task, DT) may cause Cognitive-Motor Interference (CMI).

According to Plummer³, Dual-Task Effect (DTE) is the percentage of change in performance of a task during dual-task, with negative values reporting a decrease in performance under DT conditions and positive values reporting an improvement.

Interference exerted by cognitive tasks on motor performance received attention in the literature, but little is known about motor over cognitive interference.

The current study investigates the prioritisation strategy during cognitive-motor DT in PD through the theoretical framework proposed by Plummer³.

Methods

A cross-sectional analysis of 12 participants (7 women) with mild PD (modified Hoehn and Yahr stages <3.0) was performed. Anamnestic and clinical information was collected. All participants underwent the 3-meter Timed Up and Go test (TUG) and a serial number subtraction cognitive task. Both tasks were performed in single-task and dual-task conditions. Total TUG Duration (TTD, s) and the number of correct subtractions made during the countdown were recorded. The corresponding results were compared between conditions. For the TTD, the Minimal Detectable Change (MDC) in a single subject (a 15% change) was available. Arbitrarily, the same MDC was assumed for the cognitive task.

The cognitive and motor DTE were calculated for each task and then cross-plotted on a Performance Operating Characteristic (POC) type diagram² (Figure 1).

Results

4 patients worsened their cognitive DTE beyond MDC (range -22% to -46%); 4 patients worsened their motor DTE (rage -18% to -53%); 2 patients worsened in both their cognitive (range -17% to -27%) and motor (range -23% to -24%) tasks; in 1 patient the motor DTE worsened (-21%) while the cognitive DTE improved (+67%); in 1 patient the cognitive DTE only, improved (+20%).

Discussion and Conclusion

At least 4 different sub-samples emerge from the POC analysis of the DTE: i) subjects who prioritise the motor task at the expenses of the cognitive one; ii) subjects who worsen in their motor performance with no change in cognitive performance; iii) subjects who improved in their cognitive task as if it was facilitated by the motor one; iv) subjects improving in both cognitive and motor DTE.

While it is recognized that DT exercise may be beneficial in PD⁴, better understanding of the individual mechanisms underlying motor and cognitive performance in DT may lead to the development of more specific treatment approaches in PD.

REFERENCES

1. McIsaac TL, Fritz NE, Quinn L, Muratori LM. Cognitive-Motor Interference in Neurodegenerative Disease: A Narrative Review and Implications for Clinical Management. Front Psychol. 2018;9.
2. Kelly VE, Janke AA, Shumway-Cook A. Effects of instructed focus and task difficulty on concurrent walking and cognitive task performance in healthy young adults. Exp Brain Res. 2010;207:65–73.
3. Plummer P, Eskes G. Measuring treatment effects on dual-task performance: A framework for research and clinical practice. Front Hum Neurosci. 2015;9.
4. Xiao Y, Yang T, Shang H. The Impact of Motor-Cognitive Dual-Task Training on Physical and Cognitive Functions in Parkinson’s Disease. Brain Sci . 2023;13:437.

Comparative assessment of dynamic motor abilities in patients with neurological disorders during straight, curved and blindfolded paths: an instrumented approach

Introduction

Neurological disorders such as stroke, Parkinson’s disease (PD), and traumatic brain injury (TBI) are often associated with balance and gait disorders that lead to an increased risk of falling [ 1-3 ]. In the last decades, wearable sensor-based technologies have been used, to objectively quantify different gait domains (i.e., stability, symmetry, and smoothness) [ 4-8 ]. Several studies reported the usefulness of the sensors-based in characterizing dynamic motor abilities impairments in patients with neurological disorders.  [ 9 ]. Therefore, this study aims to quantify dynamic stability through straight, curved, and blindfolded paths of people particularly exposed to the risk of falling to support the implementation of personalized rehabilitation training and the evaluation of the effectiveness of neurorehabilitation treatments.

Methods

Sixty patients with neurological disorders (20 people with TBI (PwTBI) (7 F; 37.1±14.42; y); 20 stroke survivors (PwS) (6 F; 59.55±12.86 y) in sub-acute phase, 20 patients with PD (PwPD) (8 F; 69.15±7.55 y) and 20 healthy participants (9F; 37.35±13.94 y) were involved in the study. Gait and balance evaluations were performed with both clinical and IMU-based assessments. Participants were equipped with five synchronized inertial measurement units All participants were asked to perform three different motor tasks in a randomized order: the 10-Meter-Walk Test (10MWT) (Figure 3-a), the Figure-of-8-Walk Test (Fo8WT) (Figure 3-b), and the Fukuda-Stepping Test (FST) (Figure 3-c). A set of spatiotemporal and gait quality parameters describing stability, symmetry, and smoothness of gait were computed in all directions.

Results

The IMU-based assessment revealed significant differences during linear, curved, and blindfolded walking among PwTBI, PwPD and PwS as well as in the comparison between the control group. Focusing on the former, PwS showed significantly less symmetry and less smoothness (p < 0.017) than both groups PwTBI and PwPD during all dynamic tasks. 

Moreover, significant results emerged for the stability domain (p < 0.017) where accelerations were higher in PwS than in PwTBI and PwPD, highlighting the fact that stroke patients suffer of reduced stability. Interestingly, significant differences emerge between PwTBI and PwPD, where PwTBI demonstrates lower stride frequency and higher nRMS (AP)  during the execution of the straight walk task. In addition, from the results it is possible to see significantly lower LDLJv (AP) values (p < 0.017) of PwTBI compared to PwPD, highlighting a less smooth path in both the straight and curvilinear tasks.

Discussion and Conclusion

The use of IMU-based assessments allowed for objective and quantitative measurements of gait symmetry, postural stability, and smoothness during straight, curved, and blindfolded walking tasks. PwS shows the least symmetry and smoothness during all dynamic tasks compared to PwTBI and PwPD, probably due to the typical hemiparetic asymmetrical gait. Furthermore, these gait impairments could support also the instability significantly higher in PwS compared to the other two groups. Our results can inform the development of personalized rehabilitation programs and interventions to improve the dynamic postural stability and gait characteristics of patients with neurological disorders. Further research and clinical implementation of such assessments could lead to better outcomes and enhanced quality of life for individuals affected by these conditions.

REFERENCES

1. Silva-Batista C 2017 Resistance Training Improves Sleep Quality in Subjects With Moderate Parkinson’s Disease
2. Cattaneo 2019 Educational and Exercise Intervention to Prevent Falls and Improve Participation in Subjects With Neurological Conditions: The NEUROFALL Randomized Controlled Trial
3. Dever A 2022 Gait Impairment in Traumatic Brain Injury: A Systematic Review
4. Mazzà C2008. Control of the upper body accelerations in young and elderly women during level walking
5. Pasciuto I Overcoming the limitations of the Harmonic Ratio for the reliable assessment of gait symmetry
6. Kavanagh JJ 2008 Accelerometry: A technique for quantifying movement patterns during walking
7. Balasubramanian S2015. On the analysis of movement smoothness
8. Melendez-Calderon A 2021. Estimating Movement Smoothness From Inertial Measurement Units
9. Hendriks, M.M.S Using Sensor Technology to Measure Gait Capacity and Gait Performance in Rehabilitation Inpatients with Neurological Disorders

EFFECTS OF MANUAL SUBOCCIPITAL MUSCLES “RELEASE” ON PAIN AND DISABILITY IN ADULTS WITH TENSIVE/CERVICOGENIC HEADACHE OR NECK PAIN: SYSTEMATIC LITERATURE REVIEW AND META-ANALYSIS

Introduction

The Suboccipital Muscle Inhibition Technique (MSIT) induces muscle relaxation in the area between occiput and cervical spine. The technique applies a pressure on suboccipital area while the patient lies supine. Suboccipital muscles are involved in the control of posture and head movements[ 1 ]. Several studies proposed the use of this technique to treat pain caused by tension-type/cervicogenic headaches and neck pain[ 2 ]. The clinical effects are thought to be mediated by the autonomic nervous system. Indeed, MSIT seems to release neurotransmitters with both psycho-emotional and general well-being effects[ 3 ]. This systematic review and meta-analysis aims to investigate the effectiveness of MSIT on pain and disability in adults with tensive/cervicogenic headaches or neck pain.

Methods

The review was conducted following the PRISMA statement 2020. Adults with head-neck dysfunction (tension-type/cervicogenic headaches and neck pain) were included. The MSIT had to be performed as follows: subject supine, therapist with hands placed below the patient’s head to create pressure at the level of the suboccipital muscles. Secondary research was excluded. A search was conducted in the PubMed, Scopus, Epistemonikos, PEDro, and Web of Science databases. Two blinded reviewers checked the studies for adherence to inclusion and exclusion criteria. A third reviewer addressed any conflict. The outcomes related to pain and disability were extracted. Table ROB-2, was used to assess risk of bias. The meta-analysis was conducted following standard guidelines using the R statistical environment and the meta and metasens packages.

Results

From a total of 3844 records, 13 randomized controlled clinical trials (RCTs) involving 745 subjects were included. MSIT was applied with varying frequency and duration: from 1 to 5 times weekly, for 1 to 8 weeks with sessions lasting from 4 to 20 minutes. Specifically, 8 out of 13 RCTs proposed 1/2 sessions per week for 4 weeks of treatment while 10 out of 13 RCTs applied MSIT lasting 5/10 minutes. In addition, 8 out of 13 RCTs investigated the effects of MSIT I addition to exercises or cervical manipulation. Most of the studies showed significant effect of MSIT on pain and disability (reduction in headache frequency and intensity). These improvements increased when MSIT was associated with other treatment. The studies showed overall uncertain risk of bias. The meta-analysis involved 9 RCTs, showing moderate significance (P=0.04, RR=0.59, 95% CI 0.53-0.67; substantial heterogeneity I2=51%) in favor of using the treatment over control groups on pain.

Discussion and Conclusion

The results suggest that MSIT can reduce pain and disability in subjects with head-neck dysfunction. However, the study has some limitations including the lack of uniformity of measurement scales used in the studies to assess different outcomes such as pain and disability. Future studies with long-term follow-up are needed to identify the optimal effects of manual therapy approaches in terms of number of sessions, duration of treatment and interaction with other interventions. The application of this technique in combination with other manipulative techniques and cervical exercise is recommended.

REFERENCES

1. Cho SH, Kim SH, Park DJ. The comparison of the immediate effects of application of the suboccipital muscle inhibition and self-myofascial release techniques in the suboccipital region on short hamstring. J Phys Ther Sci, 2015 Jan;27(1):195-7.
2. Kuchera ML. Applying osteopathic principles to formulate treatment for patients with chronic pain. J Am Osteopath Assoc 2007;107(10 Suppl 6):Es28–38.
3. Santos G.J.B., Severiano M.I.R. A importância do toque terapêutico. FIEP Bulletin, 2011, 81: 1-7.

Effect of botulinum toxin injection on clinical and instrumental measures of walking ability in post-stroke patients with equinus foot deviation. A prospective cohort study.

Introduction

Equinus foot deviation (EFD) is the most frequent lower limb acquired deformity in stroke survivors. It affects ankle stability during the stance phase of gait and hinders foot clearance during swing, increasing the risk of falling and reducing both participation and quality of life. EFD may result from several factors, including the presence of triceps surae spasticity. Botulinum toxin (BoNT-A) is the first-line treatment for spasticity and is typically associated with adjuvant treatments, inclusive of physiotherapy, to potentiate its effect [ 1 ]. This study aims to describe the effects of BoNT-A injection alone at the triceps surae of post-stroke patients with EFD on ankle ROM and spasticity, loading and propulsive abilities during gait, and on the patient’s overall walking ability.

Methods

Prospective cohort study. Inclusion criteria: hemiparesis consequent to a first stroke, >1 y from the lesion, age <80 y, ability to walk for at least 10 m without help, Modified Tardieu Scale (MTS) ≥ 1 at the calf muscles, treatment by BoNT-A at the triceps surae with no physiotherapy thereafter. Exclusion criteria: cognitive barriers, orthopaedic pathologies at the lower limbs, ongoing antispastic therapy. Patients were assessed 1 week before and 4-6 weeks after BoNT-A injection. Clinical assessment included: ankle maximum passive dorsiflexion with the knee extended and flexed (pDF_KE, pDF_KF), MTS score and spasticity angle (SA), walking speed, FAC, WHS, and RMI. Dynamic loading ability (DLA) and dynamic propulsive ability (DPA) were computed from ground reaction force (GRF) data [ 2 ]. DLA is the mean value of the vertical component of the GRF. DPA is the mean value of the positive part of the fore-aft component [ 2 ]. The Wilcoxon test was used to compare paired variables.

Results

20 adult patients with chronic stroke and EFD, 4F/16M, age 42 (15) years were included. In baseline, pDF_KE was -4 (7)°, pDF_KF was 4 (8)°, median MTS score was 2 in both conditions (KE, KF), spasticity angle was 9 (5)° at the gastro-soleus complex (KE) and 9 (7)° at the soleus (KF). FAC ranged between 3 and 4, WHS between 3 and 6 and RMI between 5 and 15. On average, pDF_KE and pDF_KF did not vary after treatment (p=0.15, p=0.54). MTS score and SA did not vary at the soleus (p=0.23, p=0.18), while a nearly significant improvement was found at the gastro-soleus complex for both MTS score, reduced by 1 point (p=0.065), and SA, reduced by 3° (p=0.053). Walking speed was 33 (12) %height/s before treatment and 36 (14) %height/s after treatment (p=0.173). DLA minimally increased from 66 (8) to 68 (9) %BW (p=0.053). DPA remained stable at 3 (2) %BW (p=0.68). FAC, WHS, and RMI did not vary (p>0.78). Walking speed improved in 6 subjects, was stable in 11, and worsened in 3 cases.

Discussion and Conclusion

A subset of patients only had an improvement after treatment, while the remaining subjects did not vary or even worsened. This explains the lack of statistical significance in the results. In our study, walking speed increased in only 1/3 of the patients after treatment, with limited or no effect on functional scales. On the one hand, this may depend on the lack of adjunctive physiotherapy following BoNT-A, which is instead recommended. On the other hand, a preliminary assessment of calf muscles by sEMG during walking might have modified the treatment selection, as in [ 3 ]. Finally, GRF-based indices can be a valid compromise to obtain an instrumental evaluation over time of the effects of BoNT-A with extremely low evaluation times and costs. Patient recruitment is ongoing to increase the sample size and the consequent statistical power.

REFERENCES

[ 1 ] Picelli A et al. Ann Phys Rehabil Med 2019;62(4):291-296

[ 2 ] Campanini I et al. Gait Posture 2009;30(2):127-31

[ 3 ] Ferrarin M et al. Eur J Phys Rehabil Med 2015;51(2):171-84

Walking and fatigue in People with Multiple Sclerosis: Gait compensatory strategies to control clearance during the mid-swing phase. A qualitative study.

Introduction

People with Multiple Sclerosis (PwMS) commonly experience falls or near falls, of which one-third seem to be associated with fatigue or tripping. Lately, different studies have inquired about changes in gait parameters related to fatigue, but none have depicted the clinical compensations in the swinging limb that PwMS implement when they get fatigued, to reduce the risk of tripping. The present study tries to describe the strategies that PwMS carry out to control the clearance of the swinging limb when they get fatigued.

Methods

Thirty-two PwMS (EDSS 3.0±1.5) and 8 healthy subjects (HS) were recruited. Kinematic data were collected using a SMART-D motion capture system (BTS, Milano, Italy) with LAMB protocol. Subjects were asked to walk continuously at a steady cadence (spontaneous + 15%) suggested by a metronome. Every minute physical exertion was recorded on the Borg scale (RPE); the test ended as the subject reached a score of 17 (very hard).

We estimated the trend over time for clearance, foot drop, and lower limb length (LL), calculated as the distance between the ipsilateral anterior superior iliac spine and lateral malleolus. For each variable, we derived the slope trends and considered the slope coefficient (k) to describe our findings.

Finite mixture models were used to provide a cluster analysis: a) Univariate model of k clearance; b) Multivariate model of k-LL and k-footdrop, considering the subjects with a negative clearance according to the previous univariate analysis.

Results

PwMS walked less time (13.9±10.22 vs 30.0±1 min) HS reaching an RPE of 17, while HS walked 30 minutes reaching an RPE≤11. The cluster analysis of the k clearance showed 2 different patterns (Figure 1): 1a) showing a minimal clearance variability -0.11(0.03)mm/min (light-blue dots); 1b) a higher clearance variability -0.62(1.1)mm/min (red dots).

The multivariate model (Figure 2), considering subjects in 1b) and with a negative k-clearance, showed 3 different patterns related to k-LL and k-footdrop: the first group (green triangles) had an increased LL over time (k-LL=4.8(0.5)mm/min) and a reduction in the foot drop (k-footdrop=-2.0 (0.5)mm/min). The second group (red squares) showed a minimal variation in both parameters (k-footdrop=0.2(1.0)mm/min, k-LL = 0.3(1.0)mm/min). The third group (blue dots) had an increased k-footdrop (5.6(2.8)mm/min) associated with a reduction in the LL (k-LL=-4.8 (2.8)mm/min).

Discussion and Conclusion

The present findings seem to have the potential to better guide gait rehabilitation. In subjects with a stable clearance, the fatiguability seems associated more to deconditioning and general stability. For subjects who increase clearance (over-compensating), the treatment could be aimed toward more energy-conservative strategies. While, for subjects more at risk of tripping we found three different patterns: a group had a progressive deficit in ankle dorsiflexion partially compensated by a shortening of the limb in flight; another group had a slight change in both; finally, a third group had a deficit related to limb length in flight partially compensated by an increase in ankle dorsiflexion. Thus, rehabilitation intervention could be directed to proximal or, distal muscle function or both, and ankle-orthosis prescription could be suggested to subjects with a real need. Other factors (eg. trunk, pelvis) should be explored in future studies.

REFERENCES

Comber L, Galvin R, Coote S. Gait deficits in people with multiple sclerosis: A systematic review and meta-analysis. Gait Posture. 2017 Jan;51:25-35. doi: 10.1016/j.gaitpost.2016.09.026. Epub 2016 Sep 26. PMID: 27693958.

Broscheid KC, Behrens M, Bilgin-Egner P, Peters A, Dettmers C, Jöbges M, Schega L. Instrumented Assessment of Motor Performance Fatigability During the 6-Min Walk Test in Mildly Affected People With Multiple Sclerosis. Front Neurol. 2022 May 9;13:802516. doi: 10.3389/fneur.2022.802516. PMID: 35614920; PMCID: PMC9125148.

Fritz NE, Eloyan A, Baynes M, Newsome SD, Calabresi PA, Zackowski KM. Distinguishing among multiple sclerosis fallers, near-fallers and non-fallers. Mult Scler Relat Disord. 2018 Jan;19:99-104. doi: 10.1016/j.msard.2017.11.019. Epub 2017 Nov 22. PMID: 29182996; PMCID: PMC5803437.

Efficacy of telemedicine for musculoskeletal disorders: an umbrella review

Introduction

Telemedicine is a broad term encompassing many applications, such as diagnostic asynchronous evaluation, continuous monitoring using biosensors and synchronous video consultations, including multiple variations on each theme. This definition includes “Telerehabilitation”, “Health Technologies”, “Digital Medicine” and other similar keywords (1, 2). In addition, in recent years, an increasing number of studies use patient-reported outcomes measurements (PROMs) and patient-reported experience measurement (PREMs) to evaluate telemedicine services (3). Several systematic reviews (SRs) assessing the use of telemedicine for musculoskeletal conditions have been published in last years. However, the landscape of evidence on multiple clinical outcomes  remaines unclear. The aim of this overview is to explore the efficacy of telemedicine and rehabilitation in the treatment of musculoskeletal conditions in terms of PROMs, PREMs and objective outcomes.

Methods

We conducted an overview of SRs (PROSPERO n:CRD42022347366) searching PubMed and EMBASE up to July 25, 2022 for SRs of randomized controlled trials assessing patients with any musculoskeletal or orthopedic condition, undergoing any kind of interventions based on advanced technology systems named as “Telemedicine”, “Telerehabilitation”, “Health Technologies” and “Digital Medicine”, delivered both in synchronous and asynchronous modalities, compared to in-person treatment or usual care/no treatment. We collected PROMs regarding pain, HRQoL, physical function, social function, emotional function, cognitive function, health literacy, side effects, adherence; PREMs, categorized into treatment and technology; and objective measures, including direct and indirect costs. We assessed the methodological quality by A Measurement Tool to Assess Reviews 2 (AMSTAR 2). Findings were reported qualitatively.

Results

Overall, 35 SRs published between 2015 and 2022 were included (Figure 1). The majority of reviews assessed “telerehabilitation” (n=29) in patients with osteoarthritis (n=13) using PROMs (n=142 outcomes mapped with 60 meta-analyses). Table 1 shows SRs’ general characteristics. Proportion of PROMs and PREMs by number of review is displayed in figure 2. Most reviews (68.6%) were rated as critically low by AMSTAR 2. A substantive body of evidence meta-analyzed found telemedicine to benefit or being equal in terms of PROMs compared to conventional care (n=57 meta-analyses).  Meta-analyses showed no differences between groups in PREMs (n=4), while objectives measure (i.e. ‘physical function’) were mainly in favour of telemedicine or showing no differences (9 out of 13). Figure 3 shows directions of SRs’ effects and AMSTAR II by outcomes and by type of population. All SRs showed significant lower costs for telemedicine compared to in-person visit.

Discussion and Conclusion

To our knowledge, this is the first overview of reviews encompassing any kind of telemedicine for different musculoskeletal disorders. Telemedicine can provide more accessible tailored health care with non-inferior results in various clinical outcomes in comparison with conventional care. The assessment of telemedicine is largely represented by PROMs, reflecting how relevant is patient-centered care. Clinicians and stakeholders should consider the adoption of the best available telemedicine technologies to meet patients’ acute and chronic conditions; evidence-based exercise and education can be tailored and delivered remotely, for instance, to increase patient’s compliance to treatment. In a cost-effectiveness point of view, future studies should put efforts in investigating PREMs, objective measures and costs filling the gaps on this promising area.

REFERENCES

1. Cottrell MA, Russell TG. Telehealth for musculoskeletal physiotherapy. Musculoskelet Sci Pract. 2020;48:102193.
2. Russell TG. Physical rehabilitation using telemedicine. J Telemed Telecare. 2007;13(5):217-20.
3. Knapp A, Harst L, Hager S, Schmitt J, Scheibe M. Use of Patient-Reported Outcome Measures and Patient-Reported Experience Measures Within Evaluation Studies of Telemedicine Applications: Systematic Review. J Med Internet Res. 2021;23(11):e30042.

 

Are there sex and gender differences in low back pain interventions of randomized controlled trials? A meta-research study

Introduction

Low back pain (LBP) is the leading cause of Years Lived with Disability worldwide. The global prevalence of LBP is higher among females compared with males across all age groups (1). To improve LBP management, various rehabilitation interventions recommended by high quality clinical practice guidelines are effective (2). However, treatment effects can be different in male and female. This can also depend on the recruitments of participants in the randomized controlled trials (RCTs). Thus, we investigated the prevalence of different sex and gender participants in LBP trials to improve knowledge in sex and gender differences, enhancing tailored healthcare and external validity of randomized controlled trials.

Methods

Results

Discussion and Conclusion

REFERENCES

Performance of ChatGPT compared to clinical practice guidelines in making informed decisions for low back pain and sciatica: A cross-sectional study

Introduction

ChatGPT is a language model developed by OpenAI that is trained to generate human-like text based on large amounts of data and has the potential for role-playing during informed decisions. We aim to assess internal consistency, reliability, and accuracy of ChatGPT compared to recommendations from international clinical practice guidelines (CPGs) in providing answers to a complex clinical question on low back pain and sciatica.

Methods

This cross-sectional study compares ChatGPT answers to CPGs recommendations in diagnosis and treatment of low back pain and sciatica. All eligible recommendations were classified into ‘should do’, ‘could do’, ‘do not do’, or ‘uncertain’ categories by consensus recommendations across CPGs. Using existing CPGs’ recommendations, relative clinical questions were developed and queried to ChatGPT. We assessed (i) internal consistency of text ChatGPT answers when a clinical question was posed three times, (ii) reliability between two independent reviewers in grading ChatGPT answers into the following categories ‘should do’, ‘could do’, ‘do not do’, or ‘uncertain’, and (iii) accuracy of ChatGPT answers compared to CPGs recommendations in classifying the correct categories. Reliability was calculated using Fleiss’ kappa (κ) coefficients, whereas accuracy was measured by inter-observer agreement (IOA) as frequency of the agreements among all judgements.

Results

We found modest internal consistency of text ChatGPT answers across all three trials in all clinical questions (mean percentage of 49%, standard deviation of 15). Intra (reviewer 1: κ=0·90 standard error (se)=0·09; reviewer 2: κ=0·90 se=0·10) and inter-reliability (κ=0·85 se=0·15) between the two reviewers was “almost perfect”. Accuracy between ChatGPT answers and CPGs recommendations was slight, showing agreement in only 33% of recommendations.

Discussion and Conclusion

ChatGPT showed internal consistency in their text answers but their indications were inappropriate compared to the CPGs’ recommendations in diagnosing and treating low back pain and sciatica. Clinicians and patients should use this AI model cautiously because the system provides misleading indications on average.

REFERENCES

Collaborators GBDLBP. Global, regional, and national burden of low back pain, 1990-2020, its attributable risk factors, and projections to 2050: a systematic analysis of the Global Burden of Disease Study 2021. Lancet Rheumatol 2023; 5(6): e316-e29

Dave T, Athaluri SA, Singh S. ChatGPT in medicine: an overview of its applications, advantages, limitations, future prospects, and ethical considerations. Front Artif Intell 2023; 6: 1169595.

Khorami AK, Oliveira CB, Maher CG, et al. Recommendations for Diagnosis and Treatment of Lumbosacral Radicular Pain: A Systematic Review of Clinical Practice Guidelines. J Clin Med 2021; 10(11).

Sallam M. ChatGPT Utility in Healthcare Education, Research, and Practice: Systematic Review on the Promising Perspectives and Valid Concerns. Healthcare (Basel) 2023; 11(6).

The Berg Balance Scale is a proper tool to measure balance in persons with Multiple Sclerosis and an advanced walking disability: evidence from Rasch analysis

Introduction

Persons with Multiple Sclerosis (PwMS) are at high risk of falling, and falls are proven to be consistently associated with balance impairment.

The Berg Balance Scale (BBS) is one of the most widely used tools to assess balance in PwMS, also within RCTs. Reliability and validity of the BBS in PwMS were evaluated through the Classical Theory Test (concurrent validity with Dynamic Gait Index (r=0.780) and the Timed-Up-and-Go test (r=0.620). It discriminated with a low sensitivity between fallers and non-fallers. Inter-rater and intra-rater reliability were excellent (ICC=0.960).

Unfortunately, these traditional psychometric procedures cannot assess some crucial requirements underlying the use of rating scales such as the BBS. Indeed, Rasch analysis has emerged as a powerful tool to evaluate the measurement quality of a scale.

Hence, this study aims to evaluate the BBS measurement properties in a multicenter sample of PwMS through Rasch analysis.

Methods

Data were collected retrospectively within the outpatient Neuro-rehabilitation services of three Italian centers for 814 PwMS, adhering to these inclusion criteria: clinically or laboratory-definite multiples sclerosis; ability to stand independently for more than 3 seconds. For each participant, we collected the BBS, the Expanded Disability Status Scale (EDSS), the Activity-specific Balance Confidence (ABC) scale, and the number of falls (previous two months).

Using the Confirmatory Factor Analysis and Mokken Analysis, a preliminary unidimensional analysis of the BBS total sample (1220 observations) was performed. The sample was splitted into one validating (B1) and three confirmatory subsamples (Figure 1). Following the Rasch analysis performed on B1, item estimates were exported from B1 and anchored to the other subsamples.

Then, we studied the convergent and discriminant validity of the scale (BBS-MS) with the three external indicators.

Results

CFA and MA showed sufficient preliminary unidimensionality. The Rasch analysis on B1 failed monotonicity, local independence, and unidimensionality, and did not fit the Rasch model. After grouping locally dependent items, the BBS-MS fitted the model (χ²_df=23.8₈; p=.003) and satisfied all requirements for adequate internal construct validity (ICV) (Table 1). However, it was mistargeted to the sample (targeting index=1.922), with a distribution-independent Person Separation Index equal to 0.962, sufficient for individual measurements (Figure 2). The B1 final solution was replicated on A1, A2, and B2 subsamples, and the B1 item estimates were anchored to the confirmatory subsamples, satisfying the fit to the model (χ²=[19.0, 22.8], p-value=[.015, .004]) and all ICV requirements (Table 2).
BBS-MS directly correlated with the ABC scale (rho=.523) and inversely with EDSS (rho=-.573). It significantly differed across groups based on the EDSS, the ABC scale, and the number of falls.

Discussion and Conclusion

To our knowledge, this is the first study reporting on the Rasch analysis of the BBS for PwMS. It supports the ICV, reliability, and targeting of the BBS-MS as a measurement tool in an Italian multicentre sample of PwMS. Using one validation and three confirmation subsamples, we demonstrated the BBS-MS fitting to the Rasch model and the satisfaction of all requirements for adequate ICV.

On the other hand, the scale was slightly mistargeted to our convenience sample as its items were, on average, less difficult than the mean ability of the sample, uncovering significant targeting issues for a precise balance measurement in still ambulatory PwMS.

Indeed, our study suggested that the BBS-MS may be a precise and responsive measurement scale to assess balance in RCTs targeted to more disabled PwMS with an advanced walking disability. Thanks to this validation, we provided interval-level measures of balance ability, allowing parametric statistics to be used.

REFERENCES

1. Nilsagard, C. Lundholm, E. Denison, and L.G. Gunnarsson, Predicting accidental falls in people with multiple sclerosis — a longitudinal study. Clin Rehabil 23 (2009) 259-69
2. V. Jacobs, and S.L. Kasser, Balance impairment in people with multiple sclerosis: preliminary evidence for the Balance Evaluation Systems Test. Gait & posture 36 (2012) 414-8.
3. Cattaneo, J. Jonsdottir, and S. Repetti, Reliability of four scales on balance disorders in persons with multiple sclerosis. Disability and rehabilitation 29 (2007) 1920-5.
4. Hobart J, Cano S. Improving the evaluation of therapeutic interventions in multiple sclerosis: the role of new psychometric methods. Health Technol Assess 13 (2009) (12): iii, ix-x, 1-177.
5. Tennant, and P.G. Conaghan, The Rasch measurement model in rheumatology: what is it and why use it? When should it be applied, and what should one look for in a Rasch paper? Arthritis Rheum 57 (2007) 1358-62.