This study used a pretest-posttest quasi-experimental design.

Eligible participants for the study included patients with radiologically confirmed knee osteoarthritis for not less than 3 months, who were 45 years and older, had a Kellgren-Lawrence grade of II and III, and were without any obvious deformities of the trunk or upper and lower extremities. Exclusion criteria were patients with other comorbidities such as acute inflammation, knee surgery, cardiovascular, respiratory, neurologic, and metabolic diseases, and pregnant women. Non–mobile phone owners after initial contact and screening were also excluded.

Sample size estimation for the study was based on the equation c×π1(1–π1)+π2 (1–1π2)/(π1–π2)² [35], where c=7.9 for 80% power, and π1 and π2 are proportion estimates (π1=0.25 and π2=0.65). Therefore, n=7.9 × (0.25 (1–0.25)+0.65 (1–0.65)/(0.25–0.65)=20.49, which is approximately 21. Hence, the calculated N is 42 (21 per group). To account for 10% possible attrition, the approximate minimum sample size was 46.

This study was conducted at the Physiotherapy Departments of the Federal Medical Centre and Benue State University Teaching Hospital, both in Makurdi, Benue State, Nigeria. The study sites were selected purposefully based on the caseload at both centers and, second, one of the centers had fewer physiotherapists. The center with more physiotherapists was chosen as the control CbSE, while the other was the experimental AVbSE. Using 2 sites in this study was intended to minimize the spillover effect (spillover bias refers to an unintended effect that happens to those for whom the experimental intervention was not designed) by geographically separating the CbSE (control) from the AVbSE group. Also, the single-blind approach minimized the Hawthorne effect (that’s a situation where participants in the experimental or control group show exceptionally higher outcome scores or performance just because they are aware they are being studied). Thus, a research assistant who was not involved in the interventions conducted the recruitment and assessments. The study sample was randomly drawn from the list of patients with knee osteoarthritis attending both clinics.

The primary outcomes in this study were pain, range of motion, and muscle strength, while osteoarthritis-specific QoL and physical function were the secondary outcomes. The instruments used in the study were the Quadruple Visual Analog Scale (QVAS), goniometer, tensiometer, and Ibadan Knee and Hip Osteoarthritis Outcome Measure (IKHOAM) questionnaire, and Osteoarthritis Knee and Hip Quality of Life (OAKHQoL) questionnaire.

QVAS

This tool was used to measure pain intensity based on current pain, average pain, least pain, and worst pain, respectively. The QVAS is a segmented numeric alternative to the visual analog scale where a whole number (0-10 integers) that best reflects the intensity of the pain for the different dimensions is chosen by the respondent [32]. The total score of the QVAS is derived by scoring 1+2+4, divided by 3, and multiplied by 10. Pain intensity can then be classified as low intensity (50) or high intensity (50) [36].

Goniometer

A goniometer (12-inch plastic [model 12-1000], Fabrication Enterprises) was used to measure the range of motion at the knee joint. It ranges from 0° to 180°.

Tensiometer

A tensiometer (model and pocket balance) was used to measure quadriceps muscle strength. The scale is from 0 kg to 100 kg.

The IKHOAM Questionnaire

The IKHOAM questionnaire was used to measure the activities of daily living and societal and physical performance. The IKHOAM questionnaire is a 3-domain, 33-item instrument consisting of an activity limitations domain comprising 25 activities of daily living that are performed by individuals with knee and hip osteoarthritis rated on a 5-point (0-4) ordinal scale, a participation restriction domain comprising 3 activities in societal participation rated on a 4 point (0-3) ordinal scale, and a physical performance tests domain including 5 tests rated by a clinician. These tests are (1) 250 m walk test rated on a 6-point (0-5) ordinal scale, (2) single-leg stance test rated on a 6-point (0-5) ordinal scale, (3) stair climbing test rated on a 5-point (0-4) ordinal scale, (4) squat test rated on a 5-point (0-4) ordinal scale, and (5) balance test rated on a 6-point (0-5) ordinal scale. The maximum obtainable score for the tests is 23 (5 scores for each of the 250 m walk test, 1 single-leg stance test, and one balance test, with 4 scores for the stair climbing test and squat test, respectively). The maximum obtainable score on the IKHOAM questionnaire is 232 (ie, 200+9+23). The score for each participant was calculated in percentage as a score divided by the total possible score and multiplied by 100. A low score on the IKHOAM questionnaire implies a low level of physical functioning ability, and a high score means a high level of physical functioning ability [37].

The OAKHQoL Questionnaire

The OAKHQoL questionnaire was used to assess the QoL. The OAKHQoL is the first specific QoL questionnaire for patients with knee and hip osteoarthritis. It is self-administered and comprises 43 items in 5 dimensions: physical activity, mental health, pain, social support, and social activities, plus 3 independent items. Dimension scores are standardized from 0 (worst QoL) to 100 (best QoL) [38].

The participants were recruited consecutively at each study site and followed through until they completed the 8-week treatment program. The participants were informed of the research’s purpose at the onset of the recruitment process. An informed written permission form, which was also translated into the local language by professionals, was given to every participant (N=64) who was evaluated for study eligibility. The number of people invited to participate (n=57), the number of people who rejected (n=3), the number of screened patients who were ineligible (n=4), and their justifications for declining participation were all tracked by a research assistant.

Each individual included in the study had a baseline assessment completed. Measurements included anthropometric variables such as height, weight, and BMI. Each participant’s information, including their age and gender, was recorded. All the questionnaires were applied, and the measurements of the range of motion of the knee joint and muscle strength (quadriceps) were taken at baseline, week 4, and week 8 of the study. Active graded exercises (active range of motion and quadriceps strengthening exercises) were administered to the patients in the 2 groups (CbSE and AVbSE) for 8 weeks.

The participants recruited into the AVbSE group received an asynchronous video containing exercises, instructions, and demonstrations of each exercise’s performance. The AVbSE was developed to be operable on a mobile device with at least an Android OS of 4.1 or an iPhone interface (iOS) for participants who own and could operate a smartphone. The video was also available as a CD drive for those who had or preferred it (n=8). The AVbSE video was an asynchronous video-based intervention intended for usage at home (a safe exercise space). The participants were encouraged to follow the video and complete the tasks thrice weekly. The exercises in the video include a warm-up of active seated knee flexion and extension, quadriceps isometric setting, quadriceps strengthening exercise, hamstring clenches, wall squats, and cool down of active seated knee flexion and extension. The participants in the AVbSE group were telemonitored with voice calls and SMSs to remind them of their exercise schedules and provide symptom reviews and feedback on performance. Also, the telemonitoring was aimed at guiding patients through the exercises and ensuring safety by asking patients to discontinue exercise in case of any adverse event. The CbSE involves the traditional in-person services exercise session led by a physiotherapist trained in the protocols in the AVbSE. The research assistant carried out all baseline treatment outcome assessments (at weeks 4 and 8) for both the CbSE and AVbSE at the clinics. The details of the exercises on the development and feasibility testing of a smartphone video-based exercise program for patients with knee osteoarthritis are already published in an earlier study by Mbada et al [24]. The exercise intervention was developed using a modified Delphi approach, which involved 3 rounds and included a panel of 4 experts and a patient with knee osteoarthritis. After reaching a consensus, the video program for knee osteoarthritis was developed, including 5 main types of exercises: seated knee flexion and extension, quadriceps isometric setting, quadriceps strengthening exercise, hamstring clenches, and wall squats [24]. The highlights of the intervention are (1) warm-up (active range of motion exercises) for 5-10 seconds, and the movement was performed up to 10 times; (2) quadriceps isometric setting repeated up to 10 times; (3) quadriceps strengthening exercise repeated up to 10 times; (4) hamstring clenches held for 10 seconds and then done 10 more times; (5) wall squat held for 10 seconds, then slide back up while keeping their backs firmly against the wall, and the movement was performed up to 10 times; and (6) cool down (active range of motion exercises) for 5-10 seconds, and the movement was performed up to 10 times [24].

Descriptive statistics of frequency, mean, and SD were used to summarize data, and an independent t test was used to compare patients’ general and clinical characteristics at baseline. A 2-way mixed-model ANOVA with repeated measures on the time points was used to determine the main effects of the two treatment programs (between-subject factor) and time points (within-subject factor) on the outcome parameters. Parametric tests were used in this study under the assumption that all the measurements are continuous variables on at least an interval scale, are normally distributed, and have equivalent variances. The α level was set at .05. The analysis was done using IBM SPSS (version 27).

This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving research study participants were approved by the Health Research Ethical Committee, Federal Medical Centre Makurdi, Nigeria (FMH/FMC/HRE/VL 1-24/03/21). Written informed consent was obtained from all participants.

This study examined the effects of CbSE and AVbSE on clinical and psychosocial outcomes in patients with knee osteoarthritis. The average age of the patients was 56.2 (SD 7.5) years, and most were females. The gender pattern observed in this study might be a coincidence. However, more women are reported to seek care for their musculoskeletal conditions [8,39]. From this study, CbSE and AVbSE significantly improved clinical and psychosocial outcomes in patients with knee osteoarthritis.

These findings are consistent with previous reports that exercise interventions have beneficial clinical effects for patients with osteoarthritis in the lower limb [40,41]. The available evidence strongly supports the incorporation of exercise therapy as a fundamental component of nonpharmacological treatment for osteoarthritis of the knee [42]. Numerous studies have demonstrated that muscle strengthening and aerobic exercises can reduce pain and enhance physical function in individuals with mild to moderate knee osteoarthritis [43,44]. A systematic review by Turner et al [45] indicated that resistance training can improve pain and physical function, but the most effective resistance training regimen remains uncertain. Nevertheless, it appears that a total of 24 sessions over 8-12 weeks can yield significant benefits. A meta-analysis of 17 clinical trials involving 1705 patients revealed that resistance exercise contributes to a reduction in pain and stiffness, as well as an improvement in physical function for individuals with knee osteoarthritis [46]. Furthermore, it has been established that strength training exercises carried out over a period of 8 weeks, 3-5 times a week, can be both safe and effective for patients with knee osteoarthritis.

Furthermore, there is evidence for the effectiveness of resistance exercises conducted as a home exercise program for patients with knee osteoarthritis [40,41]. Doi et al [47] conducted a randomized controlled trial involving 142 patients with knee osteoarthritis. They found that a home-based quadriceps strengthening program involving 20 repetitions of a slow knee extension movement implemented 4 times daily for 8 weeks was as effective as nonsteroidal anti-inflammatory drugs for improved pain, stiffness, physical function, and QoL outcomes. However, in a systematic review, Saengpromma et al [48] found that a home exercise program with tracking improves pain and function and increases adherence and changes in behavior in older people compared with the untracked group. Doi et al [47] also corroborate that the effectiveness of home-based exercises in patients with knee osteoarthritis depends on patient adherence. Ettinger et al [49] and Thomas et al [50] submit that adherence to strengthening exercise is a key predictor of response in patients with knee osteoarthritis. However, the best way to deliver in-home strengthening exercises and promote adherence is still unclear.

Few studies emanating from West Africa explored digital rehabilitation for knee osteoarthritis. Odole and Ojo [28] implemented an in-home intervention of standardized exercise programs for patients with knee osteoarthritis, and adherence was monitored using telephone calls. The authors found that in-home telerehabilitation was comparable with clinic-based treatment in terms of improved QoL. Similarly, Ojoawo et al [51] compared clinic-based and telemonitored home-based interventions involving exercise pamphlets and telephone calls. They found that both interventions were effective in managing knee osteoarthritis, but the clinic-based intervention was better than the telemonitored home-based intervention. Both interventions implemented low-tech digital rehabilitation involving preintervention demonstration of exercises and the use of pictorial pamphlets as a guide. This is the first study to implement an asynchronous digital health tool for the rehabilitation of Nigerian patients with knee osteoarthritis. Implementing high-tech digital health interventions in the study setting is hamstrung by many factors, including poverty, lack of needed infrastructure (such as internet availability), and digital illiteracy [52,53].

In-home telerehabilitation creates opportunities to mitigate some of the challenges of home-based interventions (such as lack of support and good communication) using technological platforms [54]. Technology is not efficacious in itself; it depends on innovative ideas and providers’ creative abilities to ensure its use is sustainable [55]. According to Russell [56], telerehabilitation’s primary aim is to help patients receive equal access to rehabilitation services via images, sensors, virtual reality, and environments despite their impairments or barriers in accessibility. In physiotherapy, there has been an increase in the adoption of telecommunication technology, such as telephone or videoconferencing, to provide remote access to physiotherapy services when one-on-one contact is not feasible [28,57]. Telerehabilitation is a convenient way for patients to perform basic physiotherapy exercises on their own within their homes to avoid long-distance travel [58]. This study compared an asynchronous video-based telerehabilitation program with a clinic-based intervention. The treatment outcomes between the groups for both interventions are comparable except for average pain and least pain and OAKHQoL domains, where AVbSE produced significantly higher mean change in average pain and all domains of the OAKHQoL except social activities. Also, interaction effects show that the AVbSE had higher mean change across time points for PA and mental health OAKHQoL domain across all time points. Also, the CbSE score was only higher in the physical performance domain of the IKHOAM in the eighth week. Hence, AVbSE is a viable alternative to delivering physiotherapy for patients with knee osteoarthritis. The findings of this study support some related studies. Odole and Ojo [28] documented the efficacy of telephone-based rehabilitation on pain, physical function outcomes, and QoL in knee osteoarthritis. Chandra and Keerthi [58] also reported the efficacy of telerehabilitation using telephonic consultation and videoconferencing for patients with knee osteoarthritis, using exercise as a home-based treatment. Bennell et al [59] investigated the cost and clinical effectiveness of combining a physiotherapist-delivered intervention with a nurse-delivered telephone coaching for people with knee osteoarthritis. Better outcomes were achieved in the group of patients who received physiotherapy treatment and the nurse-delivered telephone coaching than those who received only physiotherapy treatment. According to Adhikari et al [60], telephone-based treatments used in telerehabilitation were both feasible and efficient, significantly reducing pain caused by various musculoskeletal problems in resource-limited settings. Another study by Multani et al [61] found significant improvement in pain, muscle strength, and functional ability within 4 weeks of telerehabilitation in patients with knee osteoarthritis.

This study supports other studies, including reviewed literature that suggest that muscle strengthening interventions improve pain and physical functioning in persons with mild to moderate osteoarthritis of the knee. Also, the study reveals that asynchronous video-based exercises effectively manage patients with knee osteoarthritis and produce comparable effects in physiologic and psychosocial outcomes to conventional clinic-based physiotherapy. Furthermore, the findings of this study confirm an earlier submission by Winters [62] that telerehabilitation brings the hope of enabling access for all, as well as helping address societal challenges in the delivery of rehabilitative services once barriers such as distance and reimbursement are overcome. Also, according to Lamplot et al [63], as technology improves in remote and rural locations, initial knee examination and follow-up visits may also be conducted virtually [63]. The findings of this study support earlier advocacy on the necessity for the development and implementation of telerehabilitation services to support persons with potentially disabling conditions who may have limited access to rehabilitation services [64,65].

The study’s relevance lies in its provision of evidence on digital rehabilitation to expand access to rehabilitation for individuals with knee osteoarthritis in Nigeria and similar medically underserved contexts. The exercises used in this study were developed based on available evidence and expert opinions following a Delphi study. The interventions were applied in accordance with evidence from a systematic review, which showed that exercise programs lasting 8-12 weeks, with 3-5 sessions per week, are safe and effective for patients with knee osteoarthritis. This study aligns with recent recommendations for self-management in patients with chronic conditions [66], particularly for those with osteoarthritis [67]. Digital interventions have been beneficial in promoting self-management among patients with various conditions [68,69], including osteoarthritis [70,71]. Using asynchronous video-based exercises as digital rehabilitation for patients with knee osteoarthritis would improve access to physiotherapy services, eliminate geographical and distance barriers, and reduce the burden of clinic visits and waiting times.

This study has some potential limitations that may affect the generalizability of findings. As such, some characteristics of the sample must be considered. First, the significant difference in the baseline parameter between the intervention and control group regarding the OAKHQoL domains, physical performance domain of IKHOAM, and least pain domain of QVAS may be a source of bias on the outcome of the interventions. Second, most participants were primarily middle-aged adults who were largely uneducated; this may affect the ease of use of technological tools. Third, this study used a dual clinical approach to recruiting participants, which could have potential inherent selection bias. Selecting multiple sites for this study may introduce selection bias related to patient demographics or treatment practices, which could independently influence the outcomes of the intervention. However, the findings of the study suggest that the general characteristics of patients at baseline were comparable across groups, indicating that the potential for selection bias is limited. Although both clinics are classified as tertiary health care institutions in Nigeria, the distinctive attributes of teaching hospitals often make them the preferred option for patients seeking specialized and comprehensive health care. As such, the control site, which is the teaching hospital in this case, may have more standardized treatment practices, potentially leading to performance bias and affecting treatment outcomes. Fourth, it is challenging to guarantee the adherence of the participants in the AVbSE group, as it is commonly reported in asynchronous tele-exercise programs, despite the telemonitoring used in this study. Also, asynchronous tele-exercises offer scheduling flexibility to patients but do not provide for real-time interface and response from the provider.

Furthermore, excluding patients without mobile phones from data collection may be a limitation of this study. While mobile phone ownership is relatively high worldwide, it is still a problem in specific populations, such as adults in rural or developing areas, and children [72]. In Nigeria, smartphone penetration is increasing and is expected to reach around 60% by 2025 [73]. However, mobile phone ownership in poorer areas is still limited, with only 58% of Nigerians in urban areas and 32% in rural areas owning smartphones in 2022 [74]. Therefore, the pervasive digital divide or digital poverty in many African settings hinders the continent from fully benefiting from digital technology [75]. For instance, the lack of mobile phone ownership could impede the implementation of mobile health interventions in Africa [76]. Researchers conducting self-report research using mobile phones should be aware that providing a mobile phone to standardize participant response platforms can impact response behavior [72].

Future research investigating the short- and long-term effects of telerehabilitation among a larger population of patients with knee osteoarthritis from multiple sites is recommended to validate the findings of this study.

Clinic-based strengthening exercise circuit training and its asynchronous video-based variant have been found to effectively reduce pain, impairment, and disability and increase patients’ QoL. The video-based variant was associated with higher improvement in most health-related QoL domains. In comparison, clinic-based strengthening exercise circuit training led to higher improvement in pain intensity.