Skip to main content

Table 1 Characteristics of the included studies according to surgery and intervention types

From: Technology-assisted rehabilitation following total knee or hip replacement for people with osteoarthritis: a systematic review and meta-analysis

Study Sample size Age (y)* Female Condition Intervention Control Length of intervention Outcomes Time points Results PEDro scores
Total knee replacement
 Telephone-based rehabilitation
  Chen et al. 2016 (China) Total: 202
IG: 101
CG: 101
66.6 68.1% Knee OA Standardised rehabilitation programme monitored via telephone support and counselling Standardised in-patient rehabilitation programme 3 calls (5–10 min each) at week, 1, 3 and 6 VAS pain; ROM; SF-36; Beck Depression Inventory scale Post-surgery baseline, 3, 6, 12 months The mean exercise time and total days in the IG group were significantly higher than CG (P < 0.01).
The pain and Beck Depression Scale scores of the IG were significantly lower than those of the CG (P < 0.01) 3 months after TKA. The IG had greater improvement on MCS scores and active ROM (P < 0.01) after TKA.
7/10
  Han et al. 2015 (Australia) Total: 390 IG: 194
CG: 196
64.8 53.0% Knee OA Home exercise programme monitored via telephone support and counselling Usual care includes access to clinic-based outpatient physiotherapy after discharge 1 call/week for 6 weeks WOMAC; ROM; 50-ft walk time; adverse events; hospital readmission Post-surgery baseline, 6 weeks No significant differences between groups were observed, respectively, for WOMAC pain (MD: 0.1; 95% CI: − 0.7, 0.9), physical function (MD: 0.04; 95% CI: − 2.5, 2.6), knee flexion (MD: − 1.1; 95% CI: − 4.1 to 1.9), knee extension (MD: 0.2; 95% CI: − 1.6 to 1.2), or the 50-ft walk time (MD: − 0.04; 95% CI: − 0.8, 0.7) at 6 weeks after surgery.
No statistically significant difference between groups was observed in the number of hospital readmissions.
8/10
  Kramer et al. 2003 (Canada) Total: 160 IG: 80
CG: 80
68.4 59.0% Knee OA Home exercise monitored via telephone support and counselling Common home exercise + out-patient clinic-based rehabilitation At least 2 calls (10~30 min each) between week 2–6 and 7–12 WOMAC; 6MWT; ROM; SF-36; Knee Society Clinical Rating scale; 30-s stair test Post-surgery baseline, 6 weeks, 3, 6, 12 months No statistically significant differences between groups were observed for the pain outcome measures (WOMAC pain scores and Knee Society Clinical Rating scale) and mobility (30-s stair test and 6MWT) at 12- or 52-weeks post-surgery. 6/10
  Park et al. 2017 (South Korea) Total: 40
IG: 21
CG: 19
50–60 years: N = 18;
70–80: N = 22
89.5% Knee OA Telephone support and counselling only SMS texts after discharge 6 calls at week 1, 3, 5, 7, 9 and 11 WOMAC global; Korean-style ADL; life satisfaction index-Z Pre-surgery baseline, 1, 3 months No statistically significant differences between groups were observed for WOMAC, ADL, and life satisfaction. 5/10
  Szöts et al. 2016 (Demark) Total: 117
IG: 59
CG: 58
67.6 66.7% Knee OA Conventional rehabilitation programme monitored via telephone support and counselling Conventional in-patient and out-patient treatment of TKA 2 calls (11–48 min each) at day 4 and 14 WOMAC; SF-36; general self-efficacy scale Post-surgery baseline, 1, 3 months No statistically significant differences between groups were observed on all WOMAC scores. However, significant differences in scores were identified in favour of the IG on general self-efficacy (between-group difference: 2.0; 95% CI: 0, 3.0) and physical function scale of SF-36 (between-group difference: 10.0; 95% CI: 0, 20.0) at 1 month after TKA, but this effect was not seen at 3 months. 8/10
 Video-teleconferencing
  Moffet et al. 2015 (Canada) Total: 205
IG: 104
CG: 101
66.0 45.0% Knee OA Standardised rehabilitation programme via in-home videoconferencing Standardised rehabilitation programme via face-to-face home visits 16 sessions (45–60 min each) over 2 months WOMAC; 6MWT; ROM; KOOS; timed stair test Pre-surgery baseline, 2, 4 months Non-inferiority of the IG compared with CG for all WOMAC scores, 6MWT, KOOS scores, ROM and timed stair tests at 2 months or 4 months after hospital discharge. 8/10
  Tousignant et al. 2011 (Canada) Total: 41
IG: 21
CG: 20
66.0 NR NR Functional rehabilitation via videoconferencing Usual home care services referred by the institute 2 sessions/week (60 min each) for 8 weeks WOMAC; ROM; TUGT; SF-36; Berg balance scale; 30s chair-stand test; Tinetti test; Functional Autonomy Measurement System Post-surgery baseline, end of treatment, 2 months No statistically significant differences between groups were observed for all clinical variables. The CG had greater improvement on WOMAC difficulty (climbing stairs, walking) (P = 0.047), physical functioning (P = 0.019) and less bodily pain (P = 0.014) after 2 months. 5/10
  Russell et al. 2011 (Australia) Total: 65
IG: 31
CG: 34
67.9 41.0% NR Standard rehabilitation programme via internet-based videoconferencing + motion analysis tools Standard out-patient clinical rehabilitation 1 session/week (45 min each) for 6 weeks VAS pain; WOMAC; TUGT; ROM; Patient-Specific Functional Scale; quadriceps lag; limb girth knee; limb girth calf; Gait Assessment Rating Scale, compliance and satisfaction Post-surgery baseline, 6 weeks No statistically significant differences between groups were observed for knee flexion and extension, muscle strength, limb girth, pain, TUGT, QoL, and clinical gait and WOMAC scores at 6 weeks after intervention.
Better outcomes were found in the IG for the Patient-Specific Functional Scale (between-group difference: −1.08; 95% CI: − 1.86, − 0.30) and the WOMAC stiffness (between-group difference: 1.46; 95% CI: 0.24, 2.68) at 6 weeks. The intervention was well received by participants, who reported a high level of satisfaction with this novel technology.
8/10
  Piqueras et al. 2013 (Spain) Total: 142
IG: 72
CG: 70
73.3 ± 6.5 83.0% Knee OA Weight-bearing functional exercise via a videoconference software with a 3D avatar + wireless sensors (accelerometer and gyroscopes) + web portal for therapist to evaluate patient data Standardised rehabilitation programme 1 session/day (60 min each) for 10 days (supervised sessions for 5 days followed by home self-care sessions for 5 days) VAS pain; WOMAC; TUGT; ROM; quadriceps muscle strength; hamstring muscle strength Post-surgery baseline, 2 weeks; 3 months Active extension ROM had a greater increase at 5 days post-surgery (P = 0.045), but the increase became equal at 3 months. IG achieved a greater increase in muscle strength (P = 0.011) and it was maintained after 3 months (P = 0.018). CG had a lower baseline level for TUGT, therefore had a greater increase at 3 months (P = 0.008). 6/10
 Game-based therapy/Visual biofeedback
  Christiansen et al. 2015 (U.S.) Total: 26
IG: 13
CG: 13
67.4 46.2% Knee OA In-patient post-operative physical therapy + home exercise programme + weight-bearing biofeedback training with a Nintendo Wii Fit balance board In-patient post-operative physical therapy + home exercise program IG: 1 session/day for 6 weeks
CG: 2 sessions/day for 6 weeks
Weight-Bearing Ratio; hip, knee and ankle moment Post-surgery baseline, 6 weeks, 26 weeks No statistically significant differences were found between groups for weight-bearing ratios, knee extension moment. FTSST improved in the IG compared with the CG at 6 (between-group difference: −2.3; 95% CI: − 4.2, − 0.4) and 26 weeks (between-group difference: − 1.3; 95% CI: − 2.3, − 0.2). 7/10
  Ficklscherer et al. 2016 (Germany) Total: 30
IG: 17
CG: 13
53.0 38.5% TKR and ACL Standard physiotherapy + exercise training with the Nintendo Wii (two Wii controllers were placed at the knee and ankle) + a motion analysis software Standard physiotherapy alone 1 session/day (10 min or until fatigue of the participant) after surgery until discharge (average 3.2 sessions) IKDC; Modified Cincinnati Rating System; Tegner Lysholm Knee Score Pre-surgery baseline, before discharge, 4 weeks after surgery No statistically significant differences were observed between groups for IKDC scores, the Cincinnati Rating scores, and the Tegner Lysholm Knee Score at 4 weeks. 4/10
  Fung et al. 2012 (Canada) Total: 50
IG: 27
CG: 23
68.1 42.0% NR Physiotherapy + exercise training with a Nintendo Wii Fit balance board Physiotherapy + lower extremity exercise includes balance, posture, weight lifting and strengthening) 1 session (15 min each) in total NPRS; ROM; 2-min walk test; Lower Extremity Functional Scale; Activity-specific Balance Confidence Scale; length of rehabilitation; satisfaction Post-surgery baseline, at discharge (~ 50 days after surgery) No significant differences were observed between groups for pain, knee ROM, walking speed, timed standing tasks, Lower Extremity Functional Scale, Activity-specific Balance Confidence Scale or patient satisfaction with therapy services between the groups. 5/10
  Jin et al. 2018 (China) Total: 66
IG: 33
CG:33
66.5 ± 3.5 57.6% Knee OA Conventional rehabilitation + rowing exercises with a VR Conventional rehabilitation including quadriceps muscle strengthening + ROM exercises + psychological intervention + pain management education 3 sessions (30 min each)/day WOMAC index; HSS score; VAS pain; ROM Pre-surgery baseline, 1, 3, 6 months (WOMAC, HSS);
Post-surgery baseline, 1, 3, 5, 7 days (VAS pain);
Pre-surgery baseline, 3, 7, 14 days (ROM)
No significant between-group differences were found in preoperative WOMAC, HSS score and knee ROM (P > 0.05). WOMAC indexes were significantly lower and HSS scores were significantly higher in IG than in CG at 1, 3, and 6 months after TKA, respectively (P < 0.05). VAS pain was significantly lower in IG than CG at 3, 5, and 7 days after TKA (P < 0.05). Knee ROMs were significantly higher in IG than CG at 3, 7, and 14 days after TKA (P < 0.05). 6/10
  Li et al. 2013 (China) Total: 60
IG: 30
CG: 30
65 ± 12 68.3% Knee OA Robot-assisted walking training + VR + knee joint CPM training + peri-knee neuromuscular electrical stimulation + exercise Knee joint CPM training + peri-knee neuromuscular electrical stimulation + exercise + walker-assisted in-door ambulation training 2 sessions/day (30 min each), 5 days/week for 2 weeks 6MWT; HSS score; knee kinesthesia grade; knee proprioception grade; FAC; Berg balance score; 10-m sitting-standing time Post-surgery baseline, 1, 2 weeks, 1, 3, 6, 12 months The HSS scores were significantly higher in IG compared with CG from 1 month and the difference reached a peak at 12 months. The Berg scores were significantly higher in IG from 1 week and the difference reached a peak at 3 months, lasting until the end of the study. The 10-m sitting–standing time was significantly higher in IG from 2 weeks which lasted until the end of the study. The 6MWT was higher in the IG from 1 week and the most significant difference appeared at 3 months, which lasted until the end of the study.
The knee kinesthesia grade, knee proprioception grade, and FAC score were better in the IG but not statistically significant.
2/10
 Web-based therapy
  Bini et al. 2016 (U.S.) Total: 29 IG: 14
CG: 15
63.3 40.0% NR Standard rehabilitation programme + asynchronous educational video application on a mobile device Standard in-person out-patient physical therapy 3 months no limit use VAS pain; SF-36; VR-12 item health survey PCS, MCS; KOOS-PS; satisfaction Pre-surgery baseline, 3 months No statistically significant differences were found between groups in any of the clinical outcomes (VAS, KOOS, SF-36 PCS and MCS). There was no difference in the percentage of people that had improved more than the MCSI for both the VAS and VR-12.
The overall utilization of hospital-based resources was 60% less in the IG compared with the CG.
6/10
  Culliton et al. 2018 (Canada) Total: 416
IG: 209
CG: 207
63 64% Knee OA Online e-learning tool during their preadmission clinic visit in addition
to the 31-page guide
Standard patient education; a 31-page hard copy of “My Guide to Total Knee Joint Replacement” 12 months no limit use Patient expectation, satisfaction, Knee Society Scoring System, KOOS, SF-12, Hospital Anxiety and Depression Scale; PCS; UCLA Activity Score; Social Role Participation Questionnaire Pre-surgery baseline, 12 months One year postoperatively, the risk that expectations of patients were not met was 21.8% in the CG and 21.4% in the IG for an adjusted risk difference of 1.3% (P = 0.78). The proportion of patients satisfied with their TKA at 12 months postoperative was 78.6% in the IG and 78.2% in CG.
There are significant between-group differences in favour of the CG for the new Knee Society Knee Scoring System symptoms score (P = 0.04) and the functional activities score (P = 0.04) at 12 months. We also found that CG had less anxiety (P = 0.02) and lower scores for rumination (P = 0.02), magnification (P = 0.02), and helplessness (P = 0.02) than IG on the PCS.
7/10
  Eisermann et al. 2004a (Germany) Total: 149
IG: 75
CG:72
70 79.4% NR Exercise training with a computer-aided multimedia, real-time educational software Self-training under supervision 3–5 sessions/week (30 min each) for 3–4 weeks Staffelstein Score for TKR; Hospital for Special Surgery; FIM instrument; Hanover Functional Ability Questionnaire; patient acceptance Post-surgery baseline, 6 months The average functional capacity of IG has significantly improved from 46.4 ± 14.4 to 76.9 ± 16.8 (P < 0.001) at 6 months. The CG also increased from 48.3 ± 16.7 to 70.6 ± 20.6. Differences between follow-up and admission scores showed a small effect on the credit of the IG (effect size = 0.38). However, there was no statistically significant improvement for the IG (P = 0.153).
The rating for acceptance of the system was 1.26 ± 0.81 in the IG compared with a rating of 1.28 ± 0.73 in the CG, which both indicated as “good” to “very good”. There was no statistically significant difference between groups.
3/10
Total hip replacement
 Telephone-based rehabilitation
  Hordam et al. 2010 (Demark) Total: 161
IG: 68
CG: 93
74.9 62.4% Hip OA Conventional rehabilitation monitored via telephone support and counselling Standard postoperative procedure 2 calls (5~15 min each) at week 2 and 10 SF-36 8 subscales Post-surgery baseline, 12 weeks, 9 months Physical function (P = 0.03), general health (P = 0.023) and mental health (P = 0.05) were significantly higher in IG compared with CG after 3 months, but all became non-significance at 9-month follow up. 6/10
 Videoconferencing
  Vesterby et al. 2016 (Demark) Total: 73
IG: 36
CG: 37
IG: 63 (43–80)
CG: 64 (45–84)
47.2% NR Home education and medical records via a TV set + videoconferencing via the internet or mobile In-patient and out-patient standard fast-track plan 2 videoconferences at day 2 and 6 after surgery.
Total intervention for 90 days
TUGT; length of stay; HRQoL; Oxford hip score; VAS anxiety Pre-surgery baseline, 3, 6, 12 months HRQoL increased in both groups, but there were no statistically significant differences between groups (P = 0.4). There were also no statistically significant differences between groups for TUGT at 3 months and the Oxford Hip score at 3 months, 6 months or 12 months. Both groups had a statistically significant gain from baseline to 12-month follow-up (both P < 0.001). At the 12-month follow-up, the rates of complications and readmissions were similar between the groups, but the number of postoperative hospital contacts was lower in the IG.
Length of stay was reduced from 2.1 days (95% CI: 2.0 to 2.3) to 1.1 days (95% CI: 0.9 to 1.4; P < 0.001) in the IG. Post-operative hospital contacts (phone calls) were lower in IG compared with CG at 12-month follow up (P = 0.04)
7/10
 Web-based therapy
  Eisermann et al. 2004b (Germany) Total: 149
IG: 79
CG: 70
68.6 70.3% NR Exercise training with computer-aided multimedia, real-time educational software Self-training under supervision 3–5 sessions/week (30 min each) for 3–4 weeks Staffelstein Score for THR; Harris Hip Score; FIM instrument; Hanover Functional Ability Questionnaire Post-surgery baseline, 6 months The average functional capacity of IG has significantly improved from 37.4 ± 16.8 to 72.7 ± 22.8 (P = 0.001) at 6 months. The CG increased in a very similar way from 38.3 ± 19.2 to 74.8 ± 23.0. There was no effect and no statistically significant difference in improvement between groups.
Patients displayed their acceptance of the system by rating it with average values between “good” and “very good.” The average IG rating was 1.26 ± 0.59 compared with a rating of 1.21 ± 0.73 in the CG. There was no statistically significant difference between the two groups.
3/10
  Wang et al. 2018 (China) Total: 400
IG: 200
CG: 200
55.7 ± 13.8 53.1% Hip OA (25%) Interactive internet platform + videoconference Routine rehabilitation + telephone follow-up by nurses At least 1 chat/week for the 1st month after discharge; at least 1chat/fortnight within 2 to 4 months; at least 1 chat/month within 5 to 6 months. Harris Hip Score; ADL; SF-36 Scale Post-surgery baseline (admission), 3, 6 months after discharge A significant between-group main effect was also found in favouring IG on the Harris hip scores (P < 0.001), ADL scores (P = 0.041) and SF-36 (P = 0.048). 5/10
  1. Abbreviation ADL Activities of Daily Living, CG Control group, CPM Continuous passive motion, FAC Functional ambulation, FIM Functional Independence Measure, HRQoL Health-related quality of life, HSS Hospital for Specific Surgery, KOOS Knee injury and Osteoarthritis Outcome Score, IKDC International Knee Documentation Committee score, IG Intervention group, LBP Low back pain, MCSI Minimal Clinically Significant Improvement, NPRS Numerical pain rating scale, NR Not reported, ODI Oswestry Disability Index, PCS Pain Catastrophizing Scale, ROM Range of motion, SF-12 12 item Short Form Survey, SF-36 Short Form-36, SMS Short messaging service, 6MWT Six-minute walk test, TKR Total knee replacement, THR Total hip replacement, TR Telerehabilitation, TUGT Timed up and go test, UCLA University of California at Los Angeles, VAS Visual analogue scale, VR Virtual reality, WOMAC The Western Ontario and McMaster Universities Osteoarthritis Index
\