Effects of technology-supported exercise programs on the knee pain, physical function, and quality of life of individuals with knee osteoarthritis and/or chronic knee pain: A systematic review and meta-analysis of randomized controlled trials

Abstract

Objective

The study sought to examine the effects of technology-supported exercise programs on the knee pain, physical function, and quality of life of individuals with knee osteoarthritis and/or chronic knee pain by a systematic review and meta-analysis of randomized controlled trials.

Materials and Methods

We searched MEDLINE, EMBASE, CINAHL Plus, and the Cochrane Library from database inception to August 2020. A meta-analysis and subgroup analyses, stratified by technology type and program feature, were conducted.

Results

Twelve randomized controlled trials were reviewed, all of which implemented the programs for 4 weeks to 6 months. Telephone, Web, mobile app, computer, and virtual reality were used to deliver the programs. The meta-analysis showed that these programs were associated with significant improvements in knee pain (standardized mean difference [SMD] = −0.29; 95% confidence interval [CI], −0.48 to −0.10; P = .003) and quality of life (SMD = 0.25; 95% CI, 0.04 to 0.46; P = .02) but not with significant improvement in physical function (SMD = 0.22; 95% CI, 0 to 0.43; P = .053). Subgroup analyses showed that some technology types and program features were suggestive of potential benefits.

Conclusions

Using technology to deliver the exercise programs appears to offer benefits. The technology types and program features that were associated with health values have been identified, based on which suggestions are discussed for the further research and development of such programs.

INTRODUCTION

Knee osteoarthritis (OA) often causes chronic knee pain^1,2 and is a prevalent musculoskeletal disorder that affects many people, especially the elderly.^3–9 Both knee OA and chronic knee pain are leading causes of disability that substantially decrease the quality of life and well-being of afflicted individuals.^5,10,11 With a rapidly aging population, these health issues are becoming more significant.¹² Thus, effective approaches are needed to facilitate treatment and care for those suffering from knee OA and/or chronic knee pain.

Exercise therapy is one of the most common and effective management strategies for these problems.¹³ Several care guidelines suggest exercise as a first-line treatment for knee OA and chronic knee pain.^12–20 However, the conventional delivery of exercise therapy usually requires patients to participate in face-to-face exercise at clinics, which involves a time commitment, motivation, and a monetary cost, or requires patients to perform prescribed exercise in their home with limited supervision and access to resources, which can lead to poor exercise adherence and performance.^21–23

In recent years, technologies have been adopted to surmount these barriers. For example, telephones are used for remote consultations between patients and physical therapists. In addition, multifeatured Web and mobile app integrated with motion- and physical activity–tracking sensors have been developed for the management of exercise progress and health conditions. Some treatment approaches use exercise games, delivered by computer, video game console, or virtual reality, to improve patients’ interactions with and engagement in exercise therapy.^24,25 The support of such technologies is increasing the accessibility, affordability, and attractiveness of exercise programs to treat knee OA and chronic knee pain.

As a result of the growing interest in such technology-supported exercise programs for knee OA and/or chronic knee pain, a number of randomized controlled trials (RCTs) have been conducted to examine their effects. However, the studies reported mixed results, for which a review and meta-analysis study can be useful for obtaining more precise estimates and a better understanding of the effects of these programs. Although there is a previous review that has focused on those with knee OA, its study sample also included patients with hip OA, and thus its findings may not be specific to those with knee problems.²⁶ In addition, the effects of programs may vary by technology type, which means that a comparison of the treatment effects of programs based on different technologies is important. Moreover, several program features have been designed to facilitate and encourage individuals to perform exercise, such as video-based exercise demonstration, sensor-based motion and physical activity tracking, and educational materials,²⁷ but the independent effects of these features have yet to be assessed. Finally, several RCTs have been published recently^28–32 and have yet to be reviewed. We therefore performed a systematic review and meta-analysis to investigate whether technology-supported exercise programs were effective interventions for a large population with knee OA and/or chronic knee pain. Our aim was to determine whether technology-supported exercise programs were associated with the improvements in the following 3 outcomes that are significant to knee OA and/or chronic knee pain: (1) knee pain, (2) physical function, and (3) quality of life.

MATERIALS AND METHODS

This study was registered with PROSPERO (CRD42020202854) and followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines.³³ The literature search, screening of studies, data extraction, and quality assessment were performed independently by the 2 authors (T.C. and J.C.), and disagreements were resolved by discussion to reach a consensus.

Literature search and study selection

On August 5, 2020, we searched MEDLINE via Ovid (from 1946 to present), EMBASE via Ovid (from 1974 to present), CINAHL Plus via EBSCOhost (from 1937 to present), and the Cochrane Library (no date restriction) to identify relevant studies, without language restrictions. The search used both Medical Subject Headings and free-text words related to knee OA, knee pain, exercise therapy, supporting technology, the outcomes of interest to this review, as well as the Cochrane Highly Sensitive Search Strategy to identify RCTs (see Supplementary Appendix 1 for the search details). Duplicate citations were removed using EndNote and by manual screening. We first screened the titles and abstracts, then reviewed the full texts of potentially relevant citations to determine whether they should be included in the analysis. Finally, the reference lists of the identified RCTs and relevant reviews were manually searched to identify any studies that had been missed and warranted inclusion.

Inclusion and exclusion criteria

Studies were included for review if they (1) were RCTs; (2) included adults (≥18 years of age); (3) included individuals who had a diagnosis of knee OA or had chronic knee pain for at least 1 month in the last 12 months prior to the studies; (4) examined the effects of technology-supported exercise programs on knee pain, physical function, or quality of life; (5) were written in English; and (6) were published in peer-reviewed journals. Review articles and case reports were excluded.

Data extraction

Two authors (T.C. and J.C.) used a standardized form to independently extract data on the studies’ characteristics, the details of the technology-supported exercise programs, and the statistical results. The outcomes of interest were changes in knee pain, physical function, and quality of life from baseline to the end of the implementation period of the intervention. We compared the differences in the changes between the intervention and control groups to determine the effects of technology-supported exercise programs.

Assessment of risk of bias

The methodological quality of each included RCT was assessed independently by T.C. and J.C. using the Cochrane risk-of-bias tool.³⁴ Seven aspects were assessed, as follows: sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective outcome reporting, and other sources of bias. Each RCT was assessed as having a low risk of bias, a high risk of bias, or an unclear risk of bias.

Data analysis

A meta-analysis was conducted to examine the main effects. We pooled data across RCTs using random effects models and calculated the standardized mean difference (SMD) for each outcome. The I² statistic was used to measure heterogeneity, where an I² of 0%-50% indicated low to moderate heterogeneity, an I² of 50%-75% indicated substantial heterogeneity, and an I² of 75%-100% indicated considerable heterogeneity.³⁴ Egger’s regression test was used to assess the possibility of publication bias.³⁵ In addition, subgroup analyses were conducted, stratified by (1) the types of technology used to deliver the exercise programs and (2) the program features involved in the programs that were designed to facilitate and encourage participants to perform exercise.

Assessment of quality of evidence

The quality of evidence for each outcome was assessed using the GRADE (Grading of Recommendations Assessment, Development and Evaluation) approach.³⁶ For each of the outcomes, the quality of evidence was downgraded by 1 level from high quality for each serious problem found in the domains of risk of bias, inconsistency, indirectness, imprecision, and publication bias.

RESULTS

Figure 1 illustrates the literature search and selection process. Thirteen studies were identified as being eligible,^{28–32,37–44} but the findings reported in 2 of them^41,42 were obtained from the same RCT. Therefore, the final analysis of this review comprised only 12 independent RCTs.

Figure 1.

The randomized controlled trial (RCT) selection process. OA: osteoarthritis.

Characteristics of the RCTs analyzed

The characteristics of the 12 RCTs are summarized in Table 1 (see Supplementary Appendix 2 for detailed information). All of the RCTs implemented the technology-supported exercise programs for a short period (4 weeks to 6 months). During the intervention period, the participants in the intervention group were encouraged to perform exercise at least 3 times per week, and 7 RCTs^{29–31,37–39,41,42} also delivered biweekly to monthly consultation with physical therapists for exercise advice. In the control groups, participants in 6 RCTs received conventional care, such as face-to-face physical therapy^{28,37,41,42,44} or educational materials for OA management (eg, presented in brochures and booklets),^32,38 and those in the other 6 RCTs^{29–31,39,40,43} received minimal or no care.

Table 1.

Summary of characteristics of the 12 RCTs

RCT characteristics
Year of publication, n (%)
Before 2010	1 (8.3)40
2010–2014	2 (16.7)41,42,44,a
2015–2019	9 (75.0)28–32,37–39,43
Region where the RCT took place, n (%)
North America	6 (50.0)28,30–32,39,43
Asia	2 (16.7)40,44
Oceania	3 (25.0)29,37,38
Africa	1 (8.3)41,42,a
Sample size
Median (range)	105 (34-282)
Age of the sample, years
Mean (range)	61.30 (46.4-75.3)
Knee problems examined, n (%)
Knee osteoarthritis	6 (50.0)28,37,40–44,a
Knee osteoarthritis and chronic knee pain	1 (8.3)29
Knee osteoarthritis or chronic knee pain	3 (25.0)30,31,39
Chronic knee pain	2 (16.7)32,38
Intervention duration, n (%)
<3 months	5 (41.7)31,39–42,44,a
3-6 months	7 (58.3)28–30,32,37,38,43
Technology used to support the delivery of exercise programs, n (%)
Telephone	6 (50.0)29–31,37,39,41,42,a
Web	2 (16.7)28,38
Mobile app	2 (16.7)32,43
Computer	1 (8.3)40
Virtual reality	1 (8.3)44
Program features involved to facilitate and encourage exercise performance, n (%)
Individualized exercise recommendation	8 (75.0)28–31,37–39,43
Video-based exercise demonstration	4 (33.3)28,29,32,38
Sensor-based motion and physical activity tracking	7 (58.3)30–32,39,40,43,44
Monitoring of exercise progression	9 (75.0)28–32,37–39,41,42,a
Educational materials	9 (75.0)28–32,37–39,43
Reminder to perform exercise	3 (25.0)28,32,38

RCT: randomized controlled trial.

^aTwo of the studies were related to the same RCT.^41,42

Five types of technology were used to deliver exercise programs. Telephone was the most frequently used technology for remote exercise training and monitoring, with 2 RCTs^37,41,42 using telephone only and the other 4^29–31,39 using telephone as a primary tool, complemented by the use of Web-based exercise videos or daily records of physical activity. Web and mobile app were used to deliver self-management programs, as well as for communication with physical therapists via videoconferencing. Interactive exercise games were developed with the support of computer and virtual reality. In addition, we also identified 6 program features that were used to facilitate and encourage participants to perform exercise (see Supplementary Appendix 3 for the details of the program features in each RCT).

Risk of bias assessment

The results of the risk of bias assessment are shown in Figure 2. The details of the random-sequence generation were not reported in 2 (16.7%) RCTs. It was unclear whether allocation concealment was adequate in 5 (41.7%) RCTs. A lack of or the partial blinding of participants and healthcare providers was identified in 11 (91.7%) RCTs, and it was unclear whether the outcome assessors were blinded in 8 (66.7%) RCTs. Attrition bias was detected in 1 (8.3%) RCT, as the dropout rate was > 50%. Reporting and other sources of bias were not identified in any of the 12 RCTs.

Figure 2.

Risk of bias summary for the 12 randomized controlled trials.

Meta-analysis of effects of technology-supported exercise programs

The meta-analysis results are presented in Table 2. Compared with the use of nontechnological or no care services in the control groups, the short-term implementation of technology-supported exercise programs was found to be associated with a significant reduction in knee pain (SMD = −0.29; 95% confidence interval [CI], −0.48 to −0.10; P = .003). Based on the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain subscale, which was most frequently reported in the 10 RCTs that examined knee pain, and a previously reported standard deviation resulted from the typical knee pain population,^45,46 the observed SMD of –0.29 can be translated to a WOMAC pain change of –1.31 points on the 0-20 scale, which can be considered clinically important.⁴⁷ For quality of life, a statistically significant improvement was detected (SMD = 0.25; 95% CI, 0.04 to 0.46; P = .02), and the observed SMD represented an improvement of 4.80 points on the most frequently reported 100-point Knee Injury and Osteoarthritis Outcome Score quality-of-life subscale, which is lower than the suggested minimal clinically important difference.^48,49 However, we did not find evidence for a significant improvement in physical function (SMD = 0.22; 95% CI, 0 to 0.43; P = .053). No publication bias was detected, according to Egger’s regression test. The quality of evidence across all of the RCTs for quality of life was high, while that for knee pain and physical function was moderate due to substantial heterogeneity (I² > 50%). Forest plots for the outcomes are presented in Figure 3.

Table 2.

Meta-analysis results of the effects of technology-supported exercise programs

Outcome	Number of RCTs	Sample Size	SMD (95% CI)	P Value	I 2 (%)	Egger’s Regression Test		Quality of Evidence (GRADE)
						t Value	P Value
Knee paina	10	1335	−0.29 (−0.48 to −0.10)	.003	62	1.20	.26	⊕⊕⊕◯ Moderateb
Physical functionc	11	1390	0.22 (0 to 0.43)	.053	73	0.33	.75	⊕⊕⊕◯ Moderateb
Quality of lifec	8	727	0.25 (0.04 to 0.46)	.02	45	0.72	.50	⊕⊕⊕⊕ High

CI: confidence interval; GRADE: Grading of Recommendations Assessment, Development and Evaluation; RCT: randomized controlled trial; SMD: standardized mean difference.

^aA negative difference favors the intervention group (which had technology-supported exercise programs applied), while a positive difference favors the control group.

^bDowngraded for substantial heterogeneity (I² > 50%).

^cA positive difference favors the intervention group (which had technology-supported exercise programs applied), while a negative difference favors the control group.

Figure 3.

Forest plots for the effects of the 12 randomized controlled trials (RCTs) on (a) knee pain, (b) physical function, and (c) quality of life. CI: confidence interval; SMD: standardized mean difference.

Subgroup analysis for the outcomes by technology type

Table 3 presents the analysis of the effects of exercise programs by technology type. For knee pain, the use of web (SMD = −0.48; 95% CI, −0.90 to −0.05; P = .03) or mobile app (SMD = −0.45; 95% CI, −0.88 to −0.02; P = .04) was associated with a significant reduction in knee pain, which equated to a decrease of more than 2 points on the 20-point WOMAC pain subscale, whereas the use of telephone was not; the subgroup difference was not significant (P = .27). The use of Web to deliver the programs was associated with improved quality of life (SMD = 0.59; 95% CI, 0.11 to 1.07; P = .02), which equated to an improvement of 11.33 points on the 100-point Knee Injury and Osteoarthritis Outcome Score quality-of-life subscale, whereas the use of telephone, mobile app, or computer was not; the subgroup difference was not significant (P = .32). We did not observe any significant improvement in physical function when assessing the individual technology type, and no significant subgroup difference was detected (P = .69).

Table 3.

Effects of technology-supported exercise programs on the outcomes by technology type

Type of Technology	Number of RCTs	SMD (95% CI)	P Value
Knee pain
Telephone	6	−0.12 (−0.41 to 0.17)	.42
Web	2	−0.48 (−0.90 to −0.05)	.03
Mobile app	2	−0.45 (−0.88 to −0.02)	.04
Subgroup difference			.27
Physical function
Telephone	6	0.08 (−0.29 to 0.44)	.68
Web	2	0.50 (−0.08 to 1.07)	.09
Mobile app	2	0.23 (−0.35 to 0.81)	.43
Computer	1	0.24 (−0.69 to 1.16)	.62
Subgroup difference			.69
Quality of life
Telephone	6	0.18 (−0.05 to 0.41)	.13
Web	1	0.59 (0.11 to 1.07)	.02
Virtual reality	1	0.24 (−0.47 to 0.96)	.51
Subgroup difference			.32

CI: confidence interval; RCT: randomized controlled trial; SMD: standardized mean difference.

Subgroup analysis of the outcomes by program feature

Table 4 shows the results of subgroup analysis by program feature. We found that each of the 6 features led to the intervention groups experiencing a greater reduction in knee pain than the control groups of nontechnological or no care, although there was no evidence that the effects significantly differ between subgroups (ie, programs that involved or did not involve the feature).

Table 4.

Effects of technology-supported exercise programs on the outcomes by program feature

Features Examined for Each Outcome	Programs That Involved the Feature			Programs That Did Not Involve the Feature			Subgroup Difference (P Value)
	Number of RCTs	SMD (95% CI)	P Value	Number of RCTs	SMD (95% CI)	P Value
Knee pain
Individualized exercise recommendation	8	−0.30 (−0.53 to −0.08)	.01	2	−0.20 (−0.67 to 0.27)	.40	.70
Video-based exercise demonstration	4	−0.46 (−0.73 to −0.19)	.001	6	−0.13 (−0.39 to 0.14)	.34	.09
Sensor-based motion and physical activity tracking	5	−0.32 (−0.59 to −0.05)	.02	5	−0.24 (−0.54 to 0.07)	.13	.69
Monitoring of exercise progression	9	−0.27 (−0.49 to −0.05)	.02	1	−0.39 (−0.97 to 0.20)	.19	.71
Educational materials	9	−0.34 (−0.52 to −0.15)	.001	1	0.28 (−0.41 to 0.97)	.42	.09
Reminder to perform exercise	3	−0.48 (−0.81 to −0.16)	.004	7	−0.18 (−0.42 to 0.07)	.16	.14
Physical function
Individualized exercise recommendation	8	0.22 (−0.04 to 0.49)	.10	3	0.19 (−0.27 to 0.64)	.43	.89
Video-based exercise demonstration	4	0.48 (0.19 to 0.77)	.001	7	0.02 (−0.24 to 0.27)	.90	.02
Sensor-based motion and physical activity tracking	6	0.14 (−0.19 to 0.46)	.40	5	0.29 (−0.04 to 0.61)	.08	.52
Monitoring of exercise progression	9	0.24 (−0.01 to 0.49)	.06	2	0.10 (−0.43 to 0.63)	.71	.64
Educational materials	9	0.25 (0.01 to 0.49)	.04	2	−0.01 (−0.58 to 0.57)	.98	.41
Reminder to perform exercise	3	0.48 (0.09 to 0.87)	.02	8	0.09 (−0.17 to 0.36)	.50	.10
Quality of life
Individualized exercise recommendation	6	0.31 (0.08 to 0.54)	.01	2	0.01 (−0.48 to 0.50)	.97	.29
Video-based exercise demonstration	2	0.29 (−0.12 to 0.69)	.16	6	0.23 (−0.06 to 0.52)	.12	.83
Sensor-based motion and physical activity tracking	4	0.25 (−0.12 to 0.62)	.19	4	0.25 (−0.04 to 0.54)	.09	.99
Monitoring of exercise progression	7	0.25 (0.02 to 0.49)	.04	1	0.24 (−0.52 to 1.00)	.53	.98
Educational materials	6	0.31 (0.08 to 0.54)	.01	2	0.01 (−0.48 to 0.50)	.97	.29
Reminder to perform exercise	1	0.59 (0.16 to 1.02)	.01	7	0.19 (−0.02 to 0.39)	.07	.10

CI: confidence interval; RCT: randomized controlled trial; SMD: standardized mean difference.

We additionally found that programs that involved the feature of video-based exercise demonstration were associated with a significantly greater improvement in physical function, compared with those that did not involve this feature (subgroup difference: P = .02). Moreover, programs that involved the features of educational materials and a reminder to perform exercise led to a significant improvement in physical function, although the effects did not show heterogeneity between subgroups. For other features, we did not identify a significant improvement in physical function when programs involved each of them, and similar changes were observed between subgroups.

Furthermore, we identified that participants experienced a significant improvement in quality of life when they took part in programs that involved individualized exercise recommendation, monitoring of exercise progression, educational materials, or a reminder to perform exercise, although we did not identify the evidence that these effects significantly vary between subgroups. For programs that involved the features of video-based exercise demonstration and sensor-based motion and physical activity tracking, they did not show a significant improvement in quality of life, and similar changes were identified between subgroups.

DISCUSSION

Main findings

In this systematic review, we meta-analyzed 12 RCTs that examined the short-term effects of technology-supported exercise programs on knee pain, physical function, and quality of life among individuals with knee OA and/or chronic knee pain. We found that these programs were associated with significant and clinically important improvements in knee pain and quality of life but not with significant improvement in physical function. Positive effects were observed may be due to technological tools providing more access to healthcare resources (eg, information, knowledge, data) that raise participants’ awareness about the importance of health and exercise management,^50,51 promote daily physical activity,⁵² increase exercise adherence,^53,54 and lead to greater improvements in health-related outcomes.⁵⁵

Among the 5 types of technology that were used to support the delivery of the exercise programs, Web-supported programs were suggestive of potential improvements in knee pain and quality of life, demonstrating small but significant benefits for chronic pain management.^56–58 This may be due to the fact that as part of such Web-supported programs, participants receive plenty of self-management guidance, as well as professional advice from physical therapists via videoconferencing or instant messaging. Consequently, participants would be motivated to regulate their exercise behaviors in a safe, feasible, and well-accepted way.^52,59,60 Similar to Web-supported programs, mobile apps also provided various exercise resources and demonstrate potential benefits for knee pain reduction. In modern society, such mobile health technology is more accessible and flexible than other modes of healthcare provision, enabling individuals to perform management of their health conditions anywhere and at any time.^61–66 However, our review did not identify the evidence of heterogeneity of effects when comparing different types of technology and several technology types produced similar beneficial impacts. Healthcare providers may therefore wish to consider the specific requirements of individuals and select the most suitable technology in each instance. For example, we found that telephone was the most used technology, as it has the advantages of being easy to use and inexpensive, which is especially beneficial for people who are relatively less familiar with modern technologies (eg, older adults or individuals with low information technology literacy). In contrast, programs supported by Web and mobile app usually comprise multiple features, enabling participants to perform self-management with additional support from physical therapists. Moreover, virtual reality technology provides interactive feedback on exercise movements in a 3-dimensional environment,^67–69 enabling participants to develop a clear understanding of their exercise performance. Supported by computer, virtual reality, and robotic technology, game-based exercise therapy could become a highly useful tool.^70–73 However, such systems are often expensive and require a greater amount of technical assistance from healthcare professionals than other technology, and thus may not be widely accepted.^74,75

Various program features were used to facilitate and encourage participants to perform exercise. Programs that involved the feature of video-based exercise demonstration appeared to have some benefits for physical function compared with those that did not. This may be attributable to the fact that exercise videos, together with audio instructions and subtitles, may help participants to develop a correct sense of gestures, speed, and range of motion for performing exercise.⁷⁶ Thus, compared with text-based materials, exercise videos instructed by physical therapists will provide clearer demonstrations of movements. In addition, programs that involved some other features may also have potential benefits, although they did not show evidence of heterogeneity of effects between subgroups. For example, programs that included a reminder to perform exercise were associated with improvements in knee pain, physical function, and quality of life. This is supported by the findings that older adults may adhere poorly to prescribed home exercises due to forgetfulness or limited supervision from physical therapists, and reminders can help resolve this problem by regularly encouraging them to perform exercise and increasing participants’ adherence to exercise therapy.^77–80 Another example is that programs that incorporated educational materials were also associated with improvements in all 3 outcomes, which may be related to increased knowledge and awareness of performing exercise and knee health, leading to improvements in health conditions.

Implications for research

This review has the following implications for research. First, some pilot RCTs had a small sample size (eg, 20 or fewer participants in each group); therefore, studies with larger sample sizes are required to reduce the risk of publication bias and the failure to detect significant differences.^81,82 Second, just like the implementation of many other healthcare technology and protocols,^65,83–85 further testing of the usability and acceptance of technology-supported exercise programs is necessary, as the success of these interventions depends primarily on them being well designed and accepted by end users.^{74,75,85–88}

Implications for practice

A majority of the reviewed RCTs were published in 2015 or later, reflecting rapidly increasing interest in the implementation of technology-supported exercise programs.²⁴ The use of advanced technologies, such as Web and mobile app, may overcome the barriers of limited healthcare resources, leading to these technologies supplementing or surrogating the conventional delivery of care.^62,63,89,90 In addition, the feature of video-based exercise demonstration could be considered for future development of exercise programs because the programs that involved this feature led to a significantly greater improvement than did those that did not, while the potential benefits of other features should not be neglected, including reminder to perform exercise, educational materials, individualized exercise recommendation, sensor-based motion and physical activity tracking, and monitoring of exercise progression.

Thus, there are opportunities and challenges to both the development and implementation of technology-based healthcare programs for OA and/or chronic knee pain.⁹⁰ One approach to improve program effects could be to provide personalized information for individuals,⁹¹ given the fact that the same exercise guidelines may not be suitable for people with different knee-related health conditions. It has also been reported that personalized reminders and feedback are more effective than their generic equivalents in healthcare programs.^92–96 Another approach could be to improve the usability of a program. For example, a complex feature may be easily usable by a young adult but very difficult for an older individual to use because of aging or reduced physical function. Program developers should therefore pay more attention to all performance aspects of users, including cognitive, physical, and behavioral performance,⁹⁷ and develop exercise programs that support and fit the users in the target population.^88,98

Strengths and limitations

The review has the following strengths. First, with the use of a systematic and effective methodology, our review provides more supporting evidence for the effects of technology-supported exercise programs for the treatment and management of knee OA and/or chronic knee pain. Second, our review has identified technology types and program features that have demonstrated potential benefits for knee health management, based on which practical suggestions are provided. However, this review has the following limitations. First, although individuals with knee OA and chronic knee pain have overlapping clinical symptoms, they may still have potential clinical diversity, which may cause statistical heterogeneity in meta-analysis. Second, the RCTs included in the review examined the short-term effects of the exercise programs, and the long-term effects are still less clear. Third, a limited number of RCTs used computer and virtual reality to deliver exercise programs, which meant that it was difficult to identify significant changes in outcomes and examine the effects of these types of technology. Finally, participants in most RCTs were aware of the interventions that were used, and thus placebo effects could have existed.⁹⁹

CONCLUSIONS

For individuals with knee OA and/or chronic knee pain, the use of technology-supported exercise programs appears to be an effective approach, at least for short-term improvements in knee pain and quality of life. Healthcare organizations may consider the use of mobile information technology or technological platforms that can facilitate the provision and accessibility of care aids and resources to deliver exercise programs. Further efforts are required to consider the specific needs of individuals and develop usable programs to better facilitate and encourage them to perform exercise. Also, the long-term effects of the exercise programs should be further examined by RCTs with large sample sizes and long implementation durations.

FUNDING

This study was supported by the Seed Fund for Basic Research of the University of Hong Kong (Project No.: 201910159250; Principal Investigator: CKO).