To identify relevant studies, an electronic database literature search was conducted in the Embase, Medline, PsycINFO via OVID, and Cochrane Central Register of Controlled Trials (CENTRAL) databases using the following key terms: “Knee arthroplasty OR Knee replacement,” “mobile,” “web,” “remote sensor,” “computer,” “telerehabilitation,” and “m-health” (Tables S1 and S2 in Multimedia Appendix 1). The search was executed in October 2021 and updated in August 2023. The search was restricted to 2001 onward. There were no language restrictions during the search. We searched the reference list of included articles to identify potentially eligible studies.

Predefined inclusion criteria were articles reporting the use of mobile or computer apps or any other technologies such as sensor-based devices for delivering or monitoring rehabilitation either scheduled or following knee joint replacement. We also included proof- of-concept papers that described the development process of a mobile or technology-based solution for rehabilitation. The purpose of technology could be for a health care provider to monitor rehabilitation adherence, to aid patient-health care provider communication, to promote self-management, to act as reminders, or to act as a source of education or any other function that is aimed at rehabilitation care following knee replacement. The app or technology could be used by patients, health care providers, researchers, or a family member. Included studies could have been conducted in the community or home for any clinical setting in any geographic region. The studies were required to be original research articles, and we included experimental and observational studies using quantitative or qualitative research methods. Reviews (narrative or systematic reviews), non-English articles, and articles without abstracts or full texts were excluded.

Screening of manuscript titles and abstracts was conducted by 2 independent reviewers using the web app Rayyan [23]. Prior to screening, reviewers discussed inclusion and exclusion criteria to ensure consistency between individuals. Two reviewers assessed the eligibility of the full text, and disagreements were resolved by discussion. Systematic reviews were not included in the review but were used to obtain potentially relevant references. Multiple publications originating from a single technology were grouped and presented as 1 study.

For data charting purposes, the studies were divided into the following 3 categories: (1) studies that had no rehabilitation program but included an app or a technology to assess ROM or gait and were validation studies, (2) studies reporting the use of a mobile or computer app or a telehealth delivery platform for a rehabilitation program with or without sensor-based devices and wearable sensors, (3) studies that reported end users’ perceptions of the technology used for rehabilitation monitoring. Data on the general information for the studies, features of the technology, the proposed function, and perspectives of health care providers and patients were extracted and entered in Microsoft Excel. If only the protocol of a planned study was available, there was no information on clinical evaluation, or the study included <6 individuals, we did not extract data beyond the general information.

The database search, including the ad hoc search, yielded 5960 articles. Of these articles, 300 articles were considered potentially relevant. Of these, 158 articles were included for data extraction, 131 articles were excluded, and 11 articles were not available (Table S3 in Multimedia Appendix 1). Of the 158 articles, 91 articles (64 studies) reported the clinical evaluation of a technology-based rehabilitation program, 29 articles (28 studies) reported the validation or a proof of concept of technology intended to be used for rehabilitation, and 13 articles were protocols of evaluation studies. In addition, 25 articles reported end users’ perceptions on technology (Figure 1) as stand-alone articles or part of clinical evaluation studies (n=13), totaling 38 studies. The 13 studies that reported the perceptions of technology that were also included in rehabilitation program studies were removed from the final list of included full-text articles to avoid double counting.

Of the 38 studies that reported user perspectives, 2 focused on the perspectives of health care providers, 27 focused on the perspectives of patients and caregivers, and 9 focused on the perspectives of both groups (health care providers and patients and caregivers). The approach for data collection was quantitative (n=23), qualitative (n=9), or mixed methods (n=6). The sample size ranged from 2 to 200 health care providers and from 5 to 2292 patients (Tables S7 and S8 in Multimedia Appendix 1).

Commonly used quantitative questionnaires to assess satisfaction were the System Usability Scale [129] and the net promoter score [130]. To ratify the experience with telerehabilitation, the Telemedicine Perception Questionnaire was used [131]. Acceptability and usability were assessed using the acceptance of information technology questionnaire [132] and the Telemedicine Usability Questionnaire [133]. Some studies used bespoke questionnaires to report user experience and satisfaction [32, 39, 61, 90, 94, 105, 109, 134-146].

Overall, health care providers perceived telerehabilitation and the use of technology such as biosensors as a way of improving efficiency in providing care [146], patient adherence to exercises [39, 136, 146], patient-physician communication [136], and case management [137, 146]. The main factors associated with user satisfaction with e-consultations were reliable technology, good voice or image synchronization, the refresh rate of images, sound quality, and operability of the peripherals [94, 96, 138, 139]. The key factors they perceived would influence use and uptake of technology were decreased workload (rather than increased) [140], reliability of measurements aided by technology [146], ability to measure functional outcomes objectively [141], clearer criteria when choosing appropriate patients to be enrolled in the program [140], self-efficacy in the use of technology [94, 138, 146], and ease of reporting and tracking of patient data [90]. Patients and health care providers felt e-learning modules, push notifications, and appropriate feedback from sensors and virtual reality improved self-management [138, 142-144] (Figure 4).

Patient satisfaction levels were reported when teleconsultation was provided via a computer, smartphone, or tablet [34, 39, 55, 56, 80, 92, 105, 121, 123, 134, 135, 145, 147-149]; telephone [61], videoconferencing [38, 77, 94, 105, 139, 141, 150-152], a web-based system [32, 50, 90, 96, 140, 153], and an mHealth-enabled integrated care model [46, 88, 138]. Patients were satisfied with telemonitoring due to improved access to services, continued support after discharge from hospital, ability for self-management, reduced need for clinic visits, reduction in cost and travel time, ability of health care providers to provide personalized care [32, 61, 94, 121, 136, 138, 140, 141, 145, 153-155], ease of use [34, 50, 55, 56, 92, 105, 135, 138, 147, 148], motivation to perform exercises [134, 135], sense of security with remote monitoring [134, 155], and empathetic communication by a health care provider [121, 135, 136, 145, 152, 155]. The reasons for dissatisfaction were lack of an in-person examination, shorter appointment times, delay in receiving reports (eg, x-ray), and an inability to transfer pictures from one technology to another [140, 145, 149, 153]. Patients provided suggestions for the app functionalities to improve the ease of use such as minimal clicks, an instructional video for app navigation, and restriction of commercial advertisements [149]. Home modifications [149], emotional well-being, information related to activities of daily living in simple text, dietary advice, frequently asked questions, and use of traditional medicine for postoperative pain management were a few of the suggestions for app content [121].

Patients were generally satisfied with the telerehabilitation program and were ready to recommend it to others [39, 80, 85, 96, 121, 135, 151]. The use of technology for rehabilitation was influenced by computer literacy [141, 150]. However, interruption of virtual physiotherapy sessions due to poor internet issues [139] was not commonly reported (Figure 4).

This scoping review summarized the extent, user perceptions, range, and nature of technologies used to support rehabilitation following knee arthroplasty. All studies reported in this review were from upper and middle-to-upper–income countries, with a steep increase in studies in the last decade. The technologies focused on enabling patients to remember prescribed exercises as well as be able to perform them appropriately by providing synchronous and asynchronous feedback via biosensors or virtual reality. Motivation and support during recovery via technology-enabled 1-way or 2-way communication gave patients access to health care providers. Self-management and monitoring of progress were dependent on active input using e-diaries by patients or passive input through wearables. In the context where these technologies were evaluated, end users were satisfied and found remote monitoring to be acceptable for routine use.

The last decade has seen an exponential increase in the number of arthroplasties worldwide [156]; however, a corresponding increase in technological solutions to facilitate remote monitoring is nonexistent in resource-limited settings such as LMICs where the need for monitoring and a continuum of care may be higher due to lower literacy levels and lack of access to rehabilitation clinics. Research on this topic that can inform clinical practice is nonexistent in the LMIC context. Despite a high penetration of the smartphone market [157] in LMICs, a higher initial investment to develop the technology, especially in the health care sector [158], or a lack of publication of such efforts could be reasons. In LMICs, there is an increasing trend of lower limb joint replacement procedures [156]. High out-of-pocket expenditures incurred due to home visits by physiotherapists or clinic visits by patients [159] dictate the need for a cost-effective and feasible technology-based strategy to fit the context while using lessons learned from available research.

There is unequivocal evidence that there is a need for physical and psychological support from professionals during the recovery period for pain management, adherence to exercises, and modifications to therapy planning based on one’s progress [3, 160, 161]. The apps were either focused on a single function (such as communication or knowledge transfer) or were multifunctional. They were generally received well by end users; however, the usability and acceptability of these applications or remote monitoring modalities cannot be extrapolated to low health literacy and tech literacy settings. The challenges we expect with using remote monitoring in the LMIC context could be inequitable smartphone access or tech literacy, internet speed, affordability of wearables, the burden to the health system if these needs are provided free of cost, and the need for educational content in multiple languages in countries with a non-native English-speaking, multilingual population such as in India [162].

mHealth interventions have the potential to expand the reach and effectiveness of health support by facilitating behavior change. However, to ensure these “digital behavior health interventions” effectively engage users and are effective, both microengagement (the mHealth interface) and macroengagement (evidence-based behavior change techniques) are essential [163, 164]. However, we found only a handful of studies that reported user involvement during the development stage [28, 32, 51, 58, 64, 65, 75-77, 80, 86, 88, 115, 118, 169]. Studies rarely provided an adequate explanation of the theoretical behavioral framework behind the technology-based interventions [165].

Since the context and technologies are so varied, any new applications that are developed, especially in the LMIC context, should undertake formative research with end users to understand their needs, understand their preferences, and study the local digital regulatory requirements before investing time and effort. Feasibility and pilot testing by a multidisciplinary team should be crucial steps before a full-scale evaluation [69, 166], and embedding end users’ involvement and documenting their experiences at every stage are vital to refining future interventions [164]. Further, the rehabilitation protocols should map the application features with the desired function [167, 168], and this should be confirmed by means of a process evaluation embedded within the clinical evaluation to inform the mechanism of the impact in a real-life setting [147].

This review needs to be interpreted in light of the following limitations. This scoping review focused only on technology interventions for post-knee replacement rehabilitation and hence cannot be extrapolated to other orthopedic procedures. We did not include articles for which the full text was not available. Further, incomplete reporting on the features and functions of the technology is possible and may have affected our qualitative summary and conclusion.

We did not perform a consultation phase as per the guidelines [20], and the research question was formulated upon discussion between the researchers of the scoping review team, physiotherapists, and clinicians. We limited our search from 2001 onward; however, since knee arthroplasty and mHealth came into practice in the last 2 decades, this restriction in the search may not have an implication for our review findings.

Several technologies have been identified to promote adherence, increase self-efficacy, enhance self-management, and support remote monitoring. However, all the available technologies have been developed and used in developed countries. The need for remote monitoring is compelling in resource-limited countries where knee arthroplasty is on the rise. However, irrespective of the context, it is important to involve a multidisciplinary team and include users’ perspectives during the development stage.

Computer and mobile technologies to support rehabilitation following knee arthroplasty are in wide use. Telerehabilitation and remote monitoring are as effective and safe as clinic-based rehabilitation programs. They reduce out-of-pocket expenditure or health cost expenditure by reducing the time to discharge following surgery and the number of clinic visits after discharge.

This study provides a map of the types of technology and the functionality of mobile and computer-based multifunction applications. We summarized end users’ perceptions and reasons for satisfaction or dissatisfaction with available technology. The findings reflect the lack of research and readily available technologies for LMICs.