Table 4.

Digital health for directly supervised exercise interventions.

Study	Population	Design	Intervention			Comparator or comparators			Primary outcome findings
			Description	Sample size	Description		Sample size
Cuperus et al [73]	Generalized osteoarthritis	RCTa	2 in-person group sessions+telephone monitoring by a nurse	77	Multidisciplinary in-person group intervention led by PTb		81	No difference in daily function on Health Assessment Questionnaire Disability Index between groups
Bennell et al [70]	Inactive adults with knee osteoarthritis	RCT	In-person PT+telephone coaching	84	In-person PT		84	Greater improvements in the NRSc pain score and the WOMACd function in the intervention vs comparator
Kloek et al [59]	Knee or hip osteoarthritis	Cluster RCT	Website+in-person PT	109	Usual in-person PT		99	No difference between groups for KOOSe, timed up and go, and subjective and objective physical activity
De Vries et al [62]	Knee or hip osteoarthritis	Mixed methods study embedded within an RCT [59]	Web-based component of e-exercise used by Kloek et al [59]	Quantitative analysis=90; qualitative analysis=10	N/Af		N/A	Adherence was highest for participants with middle education, 1- to 5-year osteoarthritis duration, and participants who were recruited by physical therapists
Chen et al [58]	Knee osteoarthritis	Quasi-experimental study	Blended intervention: in-person group PT+home exercises, exercise diary, and telephone check-in calls	84	In-person group health education sessions and telephone check-in calls		87	Greater improvements for WOMAC pain and joint stiffness on a Likert scale in the intervention vs comparator
Baker et al [60]	Knee osteoarthritis	Single-blind parallel-arm RCT	BOOST-TLCg (motivational behavior change telephone calls+monthly automated phone reminder messages to exercise)	52	Monthly automated phone reminder messages to exercise		52	No difference between groups in adherence
Doiron-Cadrin et al [63]	Pre-KRh and HRi	RCT	Real-time videoconferencing	12	In-person outpatient PT and usual care		12 (in-person) and 11 (usual care)	High compliance and satisfaction with the teleprehabilitation program
Hinman et al [61]	Knee osteoarthritis	Participant and assessor–blinded RCT	5-10 calls from a physical therapist for exercise advice and prescription+information folder+exercise bands+access to website for exercise videos+≥1 call from a nurse for self-management advice	87	≥1 telephone call from a nurse for self-management advice		88	Improvements in function but not pain in the intervention vs comparator
Lawford et al [71]	Knee osteoarthritis	Exploratory trial using data from the intervention arm of RCT [61]	5-10 calls from a physical therapist for exercise advice and prescription+information folder+exercise bands+access to website for exercise videos+≥1 call from a nurse for self-management advice	87	N/A		N/A	Weak association between therapeutic alliance and improvements in knee pain, self-efficacy, function, quality of life, adherence, and physical activity
Russell et al [72]	Post-KR	RCT	Computer-based system with real-time videoconferencing, measurement tools, and video capture	31	In-person outpatient PT		34	No difference between groups for improvement in WOMAC scores
Tousignant et al [65]	Post-KR	RCT	Custom hardware with videoconferencing and remote-controlled cameras	24	In-person PT		24	No significant difference between groups for knee extension and WOMAC total score
Moffet et al [64]	Post-KR	RCT	Custom hardware with videoconferencing and remote-controlled cameras	104	In-person home-based PT		101	No difference in WOMAC score between groups
Correia et al [69]	Post-KR	RCT	Platform with inertial sensors, phone app, and web portal for PT+2 home visits and telephone support by PT	30	In-person home-based PT		29	Greater improvement in the intervention vs comparator for timed up and go scores at 8 weeks
Correia et al [68]	Post-KR	RCT	Platform with inertial sensors, phone app, and web portal for PT+2 home visits and telephone support by PT	30	In-person home-based PT		29	Greater improvement in the intervention vs comparator for timed up and go scores at 6 months
Bell et al [66]	Post-KR	Pilot RCT	In-person PT+interACTION (monitoring remote rehabilitation platform with portable IMUsj+mobile app with back end clinician portal)	13	In-person PT+unsupervised home exercise program		12	No difference in value (change in activities of daily living scale and total cost) between groups
Chughtai et al [67]	Post-KR	Pre or post	3D motion-tracking cameras, exercise avatar, clinician monitoring, outcome reporting, and communication with a clinician—TKAk and UKAl	18 (TKA) and 139 (UKA)	N/A		N/A	Improvements in Knee Society Scores, WOMAC scores, and Boston University Activity Measure for Post-Acute Care scores
El Ashmawy et al [74]	Post-KR or HR	Retrospective study	Remote joint replacement clinic follow-up at 1-year, 7-years, and every 3-years after in-person consultations at 2 weeks and 6-weeks	1749	N/A		N/A	92% response rate, 87% completed the outcome forms and radiographs, 7% required further in-person appointments, and 89% satisfaction; 1 web-based appointment cost £79 (US $99), with estimated savings of £42,644 (US $53,439.93) per yearm

^aRCT: randomized controlled trial.

^bPT: physical therapy.

^cNRS: Numeric Pain Rating Scale.

^dWOMAC: Western Ontario and McMaster Universities Osteoarthritis Index.

^eKOOS: The Knee Osteoarthritis Outcome Score.

^fN/A: not applicable.

^gBOOST-TLC: Boston Overcoming Osteoarthritis through Strength Training Telephone-linked Communication.

^hKR: knee replacement.

ⁱHR: hip replacement.

^jIMU: inertial motion sensor.

^kTKA: total knee arthroplasty.

^lUKA: unilateral knee arthroplasty.

^mCurrency conversions calculated on May 24, 2022.