Delivery of Allied Health Interventions Using Telehealth Modalities: A Rapid Systematic Review of Randomized Controlled Trials

Objectives: To determine whether allied health interventions delivered using telehealth provide similar or better outcomes for patients compared with traditional face-to-face delivery modes. Study design: A rapid systematic review using the Cochrane methodology to extract eligible randomized trials. Eligible trials: Trials were eligible for inclusion if they compared a comparable dose of face-to-face to telehealth interventions delivered by a neuropsychologist, occupational therapist, physiotherapist, podiatrist, psychologist, and/or speech pathologist; reported patient-level outcomes; and included adult participants. Data sources: MEDLINE, CENTRAL, CINAHL, and EMBASE databases were first searched from inception for systematic reviews and eligible trials were extracted from these systematic reviews. These databases were then searched for randomized clinical trials published after the date of the most recent systematic review search in each discipline (2017). The reference lists of included trials were also hand-searched to identify potentially missed trials. The risk of bias was assessed using the Cochrane Risk of Bias Tool Version 1. Data Synthesis: Fifty-two trials (62 reports, n = 4470) met the inclusion criteria. Populations included adults with musculoskeletal conditions, stroke, post-traumatic stress disorder, depression, and/or pain. Synchronous and asynchronous telehealth approaches were used with varied modalities that included telephone, videoconferencing, apps, web portals, and remote monitoring, Overall, telehealth delivered similar improvements to face-to-face interventions for knee range, Health-Related Quality of Life, pain, language function, depression, anxiety, and Post-Traumatic Stress Disorder. This meta-analysis was limited for some outcomes and disciplines such as occupational therapy and speech pathology. Telehealth was safe and similar levels of satisfaction and adherence were found across modes of delivery and disciplines compared to face-to-face interventions. Conclusions: Many allied health interventions are equally as effective as face-to-face when delivered via telehealth. Incorporating telehealth into models of care may afford greater access to allied health professionals, however further comparative research is still required. In particular, significant gaps exist in our understanding of the efficacy of telehealth from podiatrists, occupational therapists, speech pathologists, and neuropsychologists. Protocol Registration Number: PROSPERO (CRD42020203128).


Introduction
Clinicians are increasingly providing allied health treatment using telehealth, where healthcare is delivered using information and communication technologies such as telephone or videoconference [1].Delivery of interventions via telehealth, rather than face-toface, is acknowledged to alleviate the barriers of proximity to care, with greater access to clinicians for people with mobility restrictions, chronic health conditions, living remotely, and/or the inability to travel [2].It may also help patients avoid taking time away from work, family, and/or other commitments, and reduce direct and indirect costs incurred by both patients and clinicians [3].The convenience of telehealth may also increase treatment attendance rates and improve adherence to exercise interventions [4].
While telephone-based psychotherapy interventions have been delivered for decades [5][6][7], other allied health professions have been slower to adopt a telehealth mode of delivery.Key barriers include preference for face-to-face delivery reported by both patients and clinicians, technology and internet access issues, and lack of support for reimbursement by organizations and health insurance [4].For older adults, additional challenges that impact the acceptability of telehealth can include digital literacy [2], as well as issues with fatigue, hearing, vision, and/or cognitive impairment [8,9].
Previous systematic reviews that have synthesized the evidence for allied health interventions delivered via telehealth have been condition-specific [10][11][12][13] or included trials that did not compare the efficacy of telehealth relative to traditional face-to-face interventions [14][15][16].Consequently, it is unclear whether patient outcomes following telehealth-delivered allied health interventions are comparable to face-to-face interventions across a range of diagnoses and clinical populations.In 2020, many governments worldwide imposed restrictions on travel and non-emergency face-to-face healthcare as a response to COVID-19 [17].For most healthcare domains, this led to a rapid transition to the use of telehealth models of care, often with little or no prior experience in this delivery modality.Evidence arising throughout that period predominantly failed to compare outcomes with face-to-face interventions, so gaps in the synthesis of evidence remain.
In response to these identified issues, the aim of this rapid review was to evaluate the efficacy of delivering allied health interventions using telehealth modalities for any condition.Comparisons between telehealth delivery and face-to-face delivery were specifically examined to assist clinician decision-making regarding the most effective mode of delivery.The specific research questions were as follows: What outcomes are influenced (impairment, activity limitation, participation)?c.
Are there differences in adherence and safety between telehealth and face-toface delivery?

Method
The review protocol was registered on PROSPERO (CRD42020203128) prior to commencement.The review methodology met the criteria for abbreviated systematic review methods as described in the Cochrane Rapid Review guidelines [18].

Identification and Selection of Trials
A three-phase search strategy was undertaken.First, CENTRAL, CINAHL, EMBASE, and MEDLINE were searched to identify existing systematic reviews from inception to August 2020, and from these, the included RCTs were screened for inclusion in this rapid review using the pre-defined eligibility criteria (Box 1).Second, these same databases were searched for RCTs from the year of the latest systematic reviews in each discipline (2017) to October 2023.The third step was to hand-search reference lists of included trials for published RCTs, cluster-RCTs, or quasi-RCTs published in English that met the pre-defined eligibility criteria.The searches were conducted by one investigator (MJR) and checked by a second investigator (NAL) (Supplementary Table S1).
To expedite this rapid review, a computer-assisted screening of citations was undertaken with Abstrackr (beta version; Rhode Island, USA) [19] to reduce the workload on dual-screening citations through machine learning by predicting the relevancy of citations [20].All citations from phase one were loaded into Abstrackr and screened once (MJR, LJC, CM) with dual screening until no remaining records were predicted to be relevant.Text mining was then used to reduce the assessment of full-text publications for potential eligibility.All PDFs were loaded into WordStat (MJG) (Version 7.1.21;Provalis, Montreal, QC, Canada) via QDA Miner (Version 5.0.21;Provalis, Montreal, QC, Canada), and a dictionary of terms for allied health disciplines was used to identify reviews of the relevant literature that mentioned any allied health discipline in the review methods or results sections (MJG, MJR) (Supplementary Table S2).Manual eligibility screening was then limited to systematic review full-text articles that included one or more allied health terms, telehealth terms, and RCT terms.Full texts were reviewed by one reviewer (MJR, LJC, SK, CM, JF, BG, SA, CH, JF, SK, ZM, AL).A second reviewer (MJR) cross-checked 20% of all excluded records for accuracy.Discrepancies were resolved by consensus.Reasons for exclusion were recorded for all trials that were excluded after a full-text review (Supplementary Table S3).

•
At least one allied health professional (physiotherapy, occupational therapy, speech pathology, psychology, neuropsychology, or podiatry) providing an intervention

•
The intervention could also include allied health professionals other than those specified above • The intervention could include a mix of telehealth and face-to-face with or without additional technology, such as virtual reality, smartphone applications, or personal computer/tablet programs • Location may include the participant's home (community), or another hospital (not co-located) Comparison: face-to-face allied health intervention • Delivered by at least one allied health professional: physiotherapist, occupational therapist, speech pathologist, psychologist, neuropsychologist, and/or podiatrist

Assessment of Quality of Trials
The quality of the included trials was assessed using the Cochrane Risk of Bias Tool Version 1.All trials were reviewed by one reviewer (MJR) and a second reviewer (LJC) cross-checked a random sample of 30% of the studies and discrepancies were resolved by discussion.

Data Extraction and Analysis
Data were extracted from RCTs by one reviewer with clinical expertise in the respective allied health discipline (LJC, CM, BG, ZM, CH, SA, AL) and a second reviewer (LJC, CH, MJR, NAL) checked 100% of the data for completeness and accuracy.Discrepancies were resolved by discussion between reviewers.Data were extracted using pilot-tested data extraction forms by one reviewer (MJR) and verified by a second author (LJC) using an Excel spreadsheet.Data extracted included study author, publication year, study design (including methods, and geographic location), setting, participant characteristics, intervention characteristics (including clinicians, frequency, duration, intensity), comparator characteristics and study effects on outcomes of interest, adverse events, satisfaction, and adherence.
Outcome data were extracted for pre-treatment, treatment completion, and, where available, follow-up periods for up to 12 months post-treatment.Data were extracted according to the International Classification of Functioning criteria [21].Outcomes were reported post-intervention as well as at secondary time points (three months or longer).
Continuous outcomes were reported as mean differences (MDs) and standardized mean differences (SMDs) with 95% Confidence Intervals.To account for between-study heterogeneity, random-effects meta-analyses were performed where possible, where the postintervention scores were used to obtain the pooled estimate of the effect of telehealth intervention compared with face-to-face intervention using RevMan 5.4 software (Cochrane Collaboration, Oxford, UK).To interpret the treatment effects, the guidelines suggested by Cohen were applied for interpreting the magnitude of the SMD in the social sciences: small SMD = 0.20; medium SMD = 0.50; and large SMD = 0.80 [22].Dichotomous outcomes were reported as a risk ratio (RR) and 95% Confidence Intervals (CI).Given that the RR describes the multiplication of the risk that occurs with the use of telehealth interventions, we interpreted an RR of 1.00 and 95%CI that included 1.00 to mean that the estimated effects are the same for both interventions (telehealth or face-to-face intervention).Heterogeneity was assessed using a Chi-Square test and I 2 statistic.Sensitivity analyses were planned by comparing results with and without quasi-randomized trials and trials with an unclear or high risk of bias for allocation concealment.Estimations of the sample mean and standard deviation for inclusion in meta-analyses were calculated using the methods outlined by Wan et al. [23].Missing summary statistics for within-study means were calculated using the formula for combining groups outlined in the Cochrane Handbook [18] and missing mean differences were calculated using RevMan 5.4 software (Cochrane Collaboration, Oxford, UK).Forest plots were generated for sufficiently homogenous trials within disciplines, with similar patient cohorts, interventions, and outcomes.Where a meta-analysis was not possible, intervention effects were reported as a narrative synthesis.There were a range of populations included in the trials and, as the telehealth model may potentially be more beneficial for certain disciplines, each discipline was analyzed separately.

The Flow of Trials through the Review
The search for systematic reviews yielded 6474 titles.From this search, 545 potentially relevant reviews that compared telehealth with face-to-face intervention were identified.From these two phases, a total of 8465 titles and abstracts were screened (see Figure 1) and 60 publications (52 trials) were identified to have met the inclusion criteria.

The Flow of Trials through the Review
The search for systematic reviews yielded 6474 titles.From this search, 545 potentially relevant reviews that compared telehealth with face-to-face intervention were identified.From these two phases, a total of 8465 titles and abstracts were screened (see Figure 1) and 60 publications (52 trials) were identified to have met the inclusion criteria.

Risk of Bias
Random sequence generation was reported in 32 trials.Four trials utilized location as a means of randomization [31,41,71,80,84].Eighteen trials were at a high risk of bias due to incomplete data.Full details are summarized in Figure 2.

Design and Effects of Neuropsychology Interventions
One small trial (n = 17) included a neuropsychological intervention providing group cognitive rehabilitation for adults with early-stage Alzheimer's Disease [26].Participants received lexical-semantic stimulation either face-to-face or via VC.Both modes of delivery significantly improved global cognitive performance measured by the Mini-Mental State Examination although there were no between-group differences: MD = 1.20, 95%CI: −0.73,

Cognition
In older adults with amnestic mild cognitive impairment, Torpil et al. [35] reported larger improvements were demonstrated with center-based face-to-face cognitive rehabilitation over VC in three of the eight subscales: visual perception (p < 0.001), spatial perception (p < 0.001), and motor praxis (p < 0.001), as well as total scores (p = 0.006); effect estimates were not reported [35].For the other subscales including visuomotor, thinking operation, memory, and attention/concentration, no differences were found (Table 2).

Burden of Care
Two trials examined the effect of telehealth versus face-to-face intervention on burden of care [25,33].A carer program including education and problem-solving skills for carers of adults with dementia delivered at home with or without VC revealed no significant differences between groups in carer confidence and perceptions of care with the Caregiving Mastery Index or the Perceived Change Scale [33].Similarly, the Bobath program, proprioceptive neuromuscular facilitation and electromyography-triggered neuromuscular stimulation for people living with dementia (home-based VC or outpatient setting), showed no differences post-intervention or at a 24-week follow-up for caregivers on the Caregiver Strain Index [25] (Table 2).

Motor Outcomes
Three studies investigated motor outcomes in people with stroke delivered at home via VC or in an outpatient setting [25,30,32].There were no between-group differences post-intervention for the modified Barthel Index or upper and lower limb muscle activity after physical therapies and electromyography-triggered neuromuscular stimulation [25], the Fugl-Meyer assessment after motor therapy and stroke education [32], or the Motor Activity Log-Arm Use Subscale and Wolf Motor Function Test (measuring motor arm capacity) after CIMT [30].Sanford and colleagues compared VC with in-person home visits by occupational therapists and physiotherapists for adults with new mobility aids [34].Although there was a significant improvement for the face-to-face group only on mobility self-efficacy scores, there were no differences between groups: MD = 3.10, 95%CI: −6.6, 12.80, p = 0.53.

Design and Effects of Physiotherapy Interventions
Twenty-six trials (29 reports) examined face-to-face versus telehealth physiotherapy intervention; three are described above as they also included occupational therapists.There were a wide variety of interventions which included single or combined interventions; not all were reported in detail.Interventions included aerobic exercise (including cardiac and pulmonary rehabilitation programs), strength training, balance training, flexibility exercises, and functional and/or gait retraining.

Joint Range of Motion
Five trials reported joint range of motion across musculoskeletal populations [29,41,42,53,58].Two trials reported no significant differences between groups for shoulder range abduction [42] and all joints measured [58] (Table 3).The pooled data from two trials revealed no difference between groups for active knee flexion (MD = −0.23 degrees, 95%CI: −1.96, 1.50;I 2 = 0%) and knee extension (MD = 0.17 degrees, 95%CI: −0.80, 1.15;I 2 = 44%) [29,53] whereas one trial observed greater gains in hip range in the telehealth group [41].Three of the four trials that reported longer-term outcomes reported no significant between-group differences for the range of motion at long-term follow-up [29,42,53]; the fourth reported that 100% of burns patients receiving face-to-face intervention achieved a full range at 3 months compared with 70% who received telehealth interventions, p = 0.005 [58].

Strength
Four trials in musculoskeletal populations reported on strength outcomes with varied results [29,36,53,58] (Table 3).Two trials reported comparable improvements in knee extension [53,58] and grip strength [58] between groups, which remained at a long-term followup.Conversely, two trials reported greater improvement in knee extension strength [29,36] in favor of telehealth, in one of which these differences were maintained at a 3-month follow-up [29].

Strength
Four trials in musculoskeletal populations reported on strength outcomes with varied results [29,36,53,58] (Table 3).Two trials reported comparable improvements in knee extension [53,58] and grip strength [58] between groups, which remained at a long-term follow-up.Conversely, two trials reported greater improvement in knee extension strength [29,36] in favor of telehealth, in one of which these differences were maintained at a 3-month follow-up [29].

Gait and Physical Activity
Overall, trials that investigated gait/walking time and physical activity outcomes reported no significant differences between telehealth and face-to-face physiotherapy (Table 3).Two trials reported using the 10-Meter Walk Test time and Self-Paced Walk Test and both reported no significant differences between delivery modes post-intervention [43,48] or at follow-up [48].The pooled data of gait speed from three trials (n = 182) revealed comparable results with telehealth and face-to-face interventions: MD = −0.02m/s, 95%CI −0.31, 0.27; I 2 = 69% [36,44,60].Two trials reported gait speed at a long-term follow-up (n

Gait and Physical Activity
Overall, trials that investigated gait/walking time and physical activity outcomes reported no significant differences between telehealth and face-to-face physiotherapy (Table 3).Two trials reported using the 10-Meter Walk Test time and Self-Paced Walk Test and both reported no significant differences between delivery modes post-intervention [43,48] or at follow-up [48].The pooled data of gait speed from three trials (n = 182) revealed comparable results with telehealth and face-to-face interventions: MD = −0.02m/s, 95%CI −0.31, 0.27; I 2 = 69% [36,44,60].Two trials reported gait speed at a long-term follow-up (n = 117) with pooled data revealing no significant difference between groups: MD = −0.16,95%CI −0.35, 0.04, I 2 = 0%, p = 0.12 [36,60].There were no differences between delivery modes in the freezing of gait in adults with PD post-intervention or long-term [44] or for the Performance Orientated Mobility Assessment-Gait subscale in adults post-stroke [51].One trial reported on physical activity levels post-intervention and long-term and demonstrated no significant difference between telehealth and face-to-face cardiac rehabilitation [49].

Function and Disability
Function and disability were reported using a variety of measures (Table 1).Thirteen musculoskeletal trials reported no difference in function or disability between modes of delivery as measured by the Anterior Knee Pain Scale [54], Constant-Murley and Quick-Dash [42], Hip Disability and Osteoarthritis Outcome Score [41], Ibadan Knee/Hip Osteoarthritis Outcome Measure [55], Knee Injury and Osteoarthritis Outcome Score [38,53], Neck Disability Index, [57], Oswestry Disability Questionnaire [40], and Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) [29,36,43,53].Likewise, there were no differences between face-to-face and telehealth interventions in neurological populations using the Barthel Index [27] and the Functional Independence Measure (FIM) [59].

Function and Disability
Function and disability were reported using a variety of measures (Table 1).Thirteen musculoskeletal trials reported no difference in function or disability between modes of delivery as measured by the Anterior Knee Pain Scale [54], Constant-Murley and Quick-Dash [42], Hip Disability and Osteoarthritis Outcome Score [41], Ibadan Knee/Hip Osteoarthritis Outcome Measure [55], Knee Injury and Osteoarthritis Outcome Score [38,53], Neck Disability Index, [57], Oswestry Disability Questionnaire [40], and Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) [29,36,43,53].Likewise, there were no differences between face-to-face and telehealth interventions in neurological populations using the Barthel Index [27] and the Functional Independence Measure (FIM) [59].

Binge Eating and Purging
A trial of CBT for people with bulimia nervosa found similar reductions in binge eating and purging behavior with no difference between delivery modes, although the authors noted improvements occurred faster with face-to-face therapy [73].
3.8.6.Health-Related Quality of Life Five trials measured HRQoL and the results were inconsistent [31,61,63,64,73].In adults with chronic pain [64] or bulimia nervosa [73], comparable results in HRQoL were reported post-intervention and at a longer-term follow-up for telehealth and face-to-face delivery.Two trials reported no significant differences between delivery modes for HRQoL with the exclusion of the physical domain, which favored telehealth [31,63].These differences were not maintained at a 6-month follow-up in either trial.One trial reported a significantly greater improvement in HRQoL outcomes with face-to-face over telehealth as measured by the Quality of Life Inventory (p < 0.001) [61].

Cognition-Communication
One trial assessed measures of cognitive and linguistic function in adults post-stroke using the Cognitive-Linguistic Quick Test and reported no significant differences between delivery modes for all sub-scores including language and executive function [79] (Table 4).

Communication
One trial investigated conversation techniques delivered center-based, in-person, or home-based via VC to adults with aphasia and cognitive-linguistic communication disorders.There were greater improvements in participant-reported communication confidence in favor of face-to-face delivery in both adults with aphasia and cognitive-linguistic communication disorders: MD = 2.61, 95%CI: 0.48, 4.74, p = 0.02 and MD = 2.90, 95%CI: 0.37, 5.43, p = 0.02, respectively [79].Patient-reported communication ability was investigated in one trial of TBIconneCT training via home visits or home-based VC with no significant differences between delivery modes [80].
Communication partner ratings were investigated in three trials.All found no significant difference between delivery modes for PD-associated dysarthria [83], aphasia [79], cognitive-linguistic communication disorders [79] and TBI [80], except for the transaction subscale in Measure of Participation in Conversation where there were greater improvements in face-to-face participants over telehealth participants (mean(SD) face-to-face: 2.31(0.66)versus telehealth: 19(0.68),p = 0.03) [80].This latter trial reported no differences between groups on these outcomes at the 3-month follow-up [80].
Aphasia was assessed in two trials using the Western Aphasia Battery-Revised, Part A [79] and word retrieval in conversation and spoken picture naming [84].One trial reported no significant difference between face-to-face and telehealth intervention in Western Aphasia Battery-Revised scores [79].The other trial investigated telehealth delivered by clinicians based at a university lab versus an outpatient setting versus face-toface face.There were differences between the two telehealth groups: outpatient-delivered telehealth resulted in better word retrieval compared to university-delivered telehealth, and face-to-face therapy scored more highly than university-delivered telehealth.However, further analysis of reported data revealed that there were no differences between telehealth overall (combined groups) and face-to-face intervention: MD = −2.80(−16.84, 11.24), p = 0.70.(p = 0.05) [84].
Two trials investigated telehealth and face-to-face speech and language therapy on fluency in people with chronic stuttering [78] and voice function outcomes in people with PD-associated dysarthria [83].There were no differences between groups as measured by stuttering frequency, self-reported fluency, and stuttering severity rating scale postintervention, with no difference between groups 9 months after randomization [78].For voice functions, there were similar significant improvements in monologue and sustained phonation, and reading sound pressure levels (dB), with no difference between groups [83].

Health-Related Quality of Life
One trial investigated HRQoL in adults with PD and reported no significant difference between delivery modes as measured by the Dysarthria Impact Profile and the Parkinson's Disease Questionnaire [83].

Design and Effects of Podiatry Interventions
Although there were trials evaluating wound care and diabetic foot ulcer management via telehealth, the care teams did not include podiatrists or other allied health professionals and were therefore not included [85,86].

Study-Related and Possibly Study-Related Adverse Events
There were no significant differences in rates or the severity of adverse events between face-to-face and telehealth occupational therapy in the three trials that reported on this out-come [25,32,33].Similarly, ten physiotherapy trials also reported no significant difference between delivery modes for the occurrence or severity of adverse events [37,39,41,42,44,47,49,53,54,60].The pooled data of study-related and possibly study-related events from four trials revealed that RR = 1.01, 95%CI: 0.36, 2.89; I 2 = 29% [37,41,48,53].No major events were reported.Two trials reported the occurrences of diagnosis-related hospitalizations during the trial and follow-up period [39,56].One reported that four telehealth participants and two face-to-face participants experienced a respiratory-related hospitalization (4% of the participant group) [39] and the other reported that two telehealth participants and eight face-to-face participants were hospitalized for cardiac reasons [56].Four psychology trials reported on the presence of adverse events [28,62,69,75]; one reported that one of the 42 telehealth participants required further evaluation for distress during the trial [69].No adverse events were reported in the remaining trials.Adverse events were not reported in neuropsychology nor speech pathology trials.

Adherence
For occupational therapy trials, one trial reported similarly high levels of adherence to sessions with no difference between delivery modes [32] and one trial reported that adherence was significantly higher in the telehealth group [34].Meta-analyses of available data from nine physiotherapy trials (n = 520) revealed no significant differences in attendance between telehealth and face-to-face delivery: SMD = −0.15,95%CI: −0.51, 0.20; I 2 = 74% (Figure 10) [36,37,39,43,44,48,50,54,58].Three physiotherapy trials had insufficient data to be included in the meta-analysis [42,47,60].One trial had significantly poorer adherence to telehealth compared with face-to-face [60], whereas two trials reported significantly higher adherence to telehealth than face-to-face [42,47].For psychology interventions, five trials reported no significant difference between groups for adherence [61,62,65,75,76].One trial reported greater adherence to telehealth over face-to-face whereas one trial found greater attrition with telehealth compared with face-to-face psychology [64].Adherence to speech therapy was reported in one trial with similar high attendance rates (≥84%) across both groups; around 94% of those who attended completed all training sessions [82].
Healthcare 2024, 12, x FOR PEER REVIEW 30 of 39 two telehealth participants and eight face-to-face participants were hospitalized for cardiac reasons [56].Four psychology trials reported on the presence of adverse events [28,62,69,75]; one reported that one of the 42 telehealth participants required further evaluation for distress during the trial [69].No adverse events were reported in the remaining trials.Adverse events were not reported in neuropsychology nor speech pathology trials.
The therapist time was reported in three trials.For synchronous telehealth, the time taken to deliver therapy was lower for telehealth participants (10 min per telehealth participant versus 98 min per face-to-face participant) [44].Another trial reported that travel time was reduced for telehealth participants (M = 255.9min versus M = 77.2min, p < 0.0001) [33].Total therapist time was also significantly reduced in asynchronous telehealth (M = 6.5 h, IQR 1.2 versus M = 32.1 h, IQR = 5.2, p < 0.01) [41].

Discussion
The findings of this rapid review demonstrate that telehealth can be successfully implemented as an alternate delivery mode to face-to-face allied health interventions for occupational therapists, physiotherapists, psychologists, and speech pathologists for some conditions in adults.Overall, a wide range of allied health interventions delivered via telehealth resulted in similar outcomes as comparable interventions delivered via face-to-face.These included outcomes such as balance after stroke [27,51,60], walking distance [47,48,53,60], HRQoL [37,38,47,48,53,54,56,61,64,71], and communication [78,79,84].Most trials were comparable in terms of patient satisfaction, adherence, and attendance.Allied health delivered via telehealth demonstrated similar rates of adverse events as interventions provided face-to-face.
Effectiveness studies of podiatry interventions delivered via telehealth have been reported to be acceptable to residents of aged care facilities [87] and show positive outcomes for wound and diabetic foot ulcer management [85,86].However, no RCTs that included podiatrists in the care team were identified for inclusion in the present review.Therefore, further robust research using RCT methods to evaluate the benefits and risks of telehealth use by podiatrists is still required.
Imposed travel and service restrictions in response to COVID-19 led to a forced rapid uptake of telehealth as an acceptable and viable modality for the delivery of health services [88,89], with good levels of satisfaction reported by patients and a range of healthcare providers during COVID-19 restrictions [90,91].The continued delivery of treatment using telehealth-based models of care may be particularly beneficial for people with mobility limitations, chronic health conditions, living remotely, or with limited ability to travel.Only a small number of trials specifically relied on a partner/carer or assistant presence for support with the intervention; therefore, the presence of a carer should not be a barrier to offering allied health intervention via telehealth.Telehealth affords greater access for patients to clinicians with the potential for improving health outcomes in patients who may otherwise miss out.For example, in patients with chronic pain, the intensity and nature of treatment in an outpatient pain clinic differed in relation to the distance that patients had traveled; telehealth may afford similar intensity and nature of treatment regardless of location [20,92].When considering convenience, flexibility, and outcomes for patients, allied health professionals may incorporate telehealth as a single or mixed model of care [93].
Although it has been suggested that telehealth may be limited to those with access to a high-speed internet connection [43], the range of technology options reported in this review demonstrates the potential to cater to an individual's needs and their existing technology while maintaining healthcare data security [69].

Limitations
Limitations to this review include the use of a rapid review process; however, we followed the Cochrane [18] criteria for rapid reviews and integrated technological tools to expedite several processes.The authors acknowledge that this rapid review process may have resulted in missing relevant RCTs published prior to 2017 that were not included in any systematic review, hence impacting the results.The use of machine learning technology to support citation screening, and text mining for the first stage of the full-text review of systematic review eligibility, may have resulted in the omission of relevant trials.Trials and reviews in languages other than English were excluded, which may have influenced the results and likely made our findings most relevant to service provision in English-speaking countries.Future trials could consider including trials in languages other than English.Trial authors were not contacted for additional data which may have limited the scope of the evidence synthesis, particularly for meta-analyses.We limited dual review for trial selection which may have led to missing relevant trials.In addition to screening systematic reviews and reference lists of included reports, several databases were searched from 2017 forward.This approach was taken to reduce duplication of previous work undertaken by other systematic review authors, and a complete search of RCTs from database inception was not undertaken.However, given that 361 systematic reviews were identified that examined telehealth treatments, with the most recent searches conducted in 2017, it would be unlikely that these reviews all missed relevant trials conducted prior to 2017.Nonetheless, the possibility of missing published RCTs not included in a systematic review prior to 2017 is acknowledged which may have influenced the results.There were insufficient studies to allow for a comparison of the differences in the effectiveness of telehealth by modality.In addition, there was a wide variety of interventions tested, and potential variations in efficacy across different types of interventions were not explored.This meta-analysis was limited to a subset of included studies, which limits the generalizability of findings and highlights the need for the provision of all trial results in telehealth research.Allied health disciplines were limited to physiotherapy, occupational therapy, psychology, speech pathology, podiatry, and neuropsychology; provision of telehealth by other allied health disciplines was not investigated.Future research should consider the inclusion of other allied health disciplines, differences between telehealth modalities, longer-term outcomes, and outcomes such as cost-effectiveness and clinician satisfaction.

Conclusions
This rapid systematic review provides evidence that telehealth interventions provided by occupational therapists, physiotherapists, psychologists, and speech pathologists result in similar increases in walking, balance, HRQoL, depression and anxiety symptoms, and communication ability as face-to-face interventions across a range of clinical presentations.Clinicians should be confident in using telehealth interventions for clients who have a preference for this modality, or who may not otherwise be able to access treatment.Few trials were identified that evaluated the efficacy of telehealth for interventions commonly provided by occupational therapists, podiatrists, and neuropsychologists in areas such as self-care training, cognitive rehabilitation, behavioral management, and podiatry interventions.Further RCTs are needed to address these gaps in knowledge, ideally conducted in

1 .
How effective are allied health interventions delivered using telehealth modalities compared with interventions delivered face-to-face of a comparable dose?a.What are the telehealth modalities and interventions used by allied health clinicians?b.
Trials where more than 20% of the participants were aged younger than 18 years, or the sample was receiving obstetric or peri-natal treatment, or interventions addressing drug, tobacco, or alcohol use (to limit the scope of rapid review)•Trials comparing telehealth methods without a face-to-face intervention comparison group•Trials published in any language other than English as the translation was not available • Trials comparing two different types of intervention (e.g., education face-to-face versus active therapy via telehealth) •Trials evaluating allied health assessment or monitoring only with no intervention • Trials where the allied health professional providing the telehealth intervention was co-located • Online-only/computer-based programs without the involvement of allied health professionals (e.g., computer-based independent exercise program) • Interventions that include medical, nursing, or non-listed allied health/pharmacy professionals where the effects of the intervention component delivered by the allied health professional cannot be isolated from the interventions provided by the other professional group/s • Interventions delivered by an allied health or rehabilitation assistant without clear supervision or delegation from an allied health professional

Table 1 .
Summary of included trials.

Table 1 .
Summary of included trials.
stretching, balance, coordination, and ambulation FTF = Center-based exercise, 3 × 60 min supervised sessions per week (non-consecutive days) Telehealth = Home-based exercise, 3 × 60 min sessions per week, with VC calls to monitor adherence and progression, plus center-based session 1 per month to check/revise exercises Duration = 12 weeks, 36 sessions Fatigue (FSS) Function (FIM) Health-Related Quality of Life (NHP, QoLS)

Table 2 .
Results for telehealth versus face-to-face occupational therapy and combined occupational therapy and physiotherapy by outcome.

Table 3 .
Results for immediate effects of telehealth versus face-to-face physiotherapy by outcome for outcomes not included in pooled analysis.
* In favor of telehealth.

Table 4 .
Results for telehealth versus face-to-face speech pathology.