Efficacy and safety of continuous passive motion and physical therapy in recovery from knee arthroplasty: a systematic review and meta-analysis

Postoperative knee rehabilitation is paramount to maintaining joint motion function [1‐4]. The incidence of knee stiffness is reported to be as high as 35% with no or inappropriate rehabilitation [2, 4‐6] and significantly affects patient quality of life and satisfaction. Clinically, two intervention strategies can be used to guide patients in postoperative rehabilitation: continuous passive motion (CPM) and traditional physical therapy. CPM was introduced in the 1970s and relies primarily on moving mechanical clips to improve joint mobility and thereby achieve improvement [7‐9].

CPM has a positive biological effect on tissue healing, edema, and hematoma [10‐12]. Vasileiadis et al. [13] confirmed the role of CPM in the maturation of heterotopic ossification by performing CPM rehabilitation in a 46-year-old male patient with right deviation. Stopping the progression and maintenance of heterotopic ossification became a useful aid in increasing joint mobility. Traditional physical therapy mainly includes dynamic floor exercises, suspension, gait training, closed chain exercises, open chain exercises, and pedal exercises, and the basic idea is an active activity. At present, both rehabilitation strategies are used to guide postoperative rehabilitation, but there is high controversy in the industry regarding the clinical application of both. Therefore, it is extremely important to conduct high-quality clinical evidence-based studies to explore reasonable rehabilitation strategies after knee surgery to guide clinical practice.

Previous studies have reported that the use of CPM has advantages over physical therapy, including reduced swelling, faster return of joint mobility, and reduced analgesia [14]. However, there is still a great deal of controversy about whether it is beneficial for patients' postoperative recovery in the past two decades of research [4, 15‐18]. Many researchers support these benefits; on the contrary, many studies show that the advantages of CPM compared to physical therapy are not as clear [1, 2].

Hence, based on previously presented evidence from high-quality randomized controlled trials, our evidence-based study aimed to determine the effectiveness of CPM compared to physical therapy in postoperative orthopedic rehabilitation, comparing key outcomes including knee range of motion (ROM), The Western Ontario and McMaster University Osteoarthritis Index (WOMAC) pain scores, length of stay, satisfaction of patients, postoperative complications, and medical costs.

This systematic review and meta-analysis following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Statement protocol. This study was registered in the International Prospective Register of Systematic Reviews (PROSPERO) (CRD42023410252).

A total of 1025 publications was retrieved according to the search method, and a total of 6 clinical articles were screened for inclusion in the analysis based on the minimum standard. Figure 1 represents the screening process. Six direct comparisons of 557 cases of CPM after TKA as well as other physiotherapy RCT were included in this meta-analysis [8, 22‐26]. The baseline information for these studies is listed in Table 1.

The present study finds that combined with CPM did not significantly improve postoperative functional recovery compared to physical therapy. There was no difference between the two in terms of time to discharge and patient satisfaction. Overall, CPM did not show an advantage in postoperative patient recovery. Rather, it was associated with increased equipment costs and costs of care. Therefore, the current findings are insufficient to support the routine use of CPM to facilitate the recovery process after arthroplasty. In addition, the heterogeneity of included studies was significant. However, we performed a subgroup analysis of WMD to investigate the source of heterogeneity. The association between CPM and ROM is described at different times, i.e., baseline, day 3 or when the maximum value is reached, probably because these times are highly dependent on the time and angle set by the CPM device. Nonetheless, our subgroup analysis showed that regardless of when ROM was measured, the increase in CPM still produced the same results as physical rehabilitation.

Although several previous studies have confirmed that CPM improves ROM only in the initial postoperative period and does not have much effect on long-term postoperative recovery, this is consistent with the results of our present meta-analysis. The association between CPM and ROM was described at different times, i.e., baseline, day 3, or at the time of maximal value, possibly because these times were highly dependent on the timing and angle of the CPM device settings. However, our subgroup analyses showed that regardless of when ROM was measured, an increase in CPM still produced the same results as physical recovery. Yang et al [27] found that CPM use was not frequently associated with improved knee ROM and functional outcomes from hospital discharge to a final follow-up. In our study, the analysis of patient satisfaction was added, as well as the conclusion that CPM generates more inpatient spending and longer hospital stays. In actual clinical practice, however, the use of CPM devices remains the standard of care in many institutions for rehabilitation [28], although the provision of CPM to patients has now been shown to be associated with insignificant long-term benefits and the short-term therapeutic role of the procedure remains controversial [16, 29]. The primary goal of using a CPM device is to increase short-term postoperative knee ROM, as several studies have reported short-term efficacy of CPM in improving CPM [30], Although most studies have shown nonsignificant results for CPM, CPM is also heavily used, which is related to subjective patient factors as well as recovery expectations, and should be validated by including a larger sample of patients for follow-up. Lee et al studied new CPM machines compared to previous conventional CPM machines to form a clinical assessment of the usefulness and effectiveness of seated CPM machines in patients undergoing total knee arthroplasty, using more objective tools such as digital inclinometers and handheld dynamometers to measure ROM [29].

We clarified that the difference in the effect of CPM and PT on patients' motor function recovery was not significant. On this basis, the patient's satisfaction is important [14]. Several previous studies have shown no statistical difference between the two in terms of patient satisfaction. In Gatewood et al. [31] by analyzing the efficacy of the means of rehabilitation after knee surgery, it was noted that CPM did not improve in terms of patient satisfaction. Wirries et al. [32] prospectively randomized the analysis of patient satisfaction with CPM after TKA through 40 patients, using the WOMAC and the Knee Social Score (KSS), to assess patient satisfaction and knee function, ultimately concluding that there was no significant difference between the both. Our findings also show that CPM does not improve patient satisfaction, possibly because CPM does not show benefit in any of the outcome indicators assessed, provides additional costs, and requires additional training for implementation [33].

In the study conducted by Joshi et al [32], two patients in the CPM group had postoperative complications. One patient was discharged with an acute quadriceps tendon tear and the other had a deep hematoma. One patient in the no-CPM group had a very deep wound dehiscence after a fall. Mau-Moeller et al. [26] systematically evaluated the effectiveness of TKA's new active sling inpatient ROM exercise program; this physical therapy was easy to perform during hospitalization and was less expensive than CPM treatment. Musa Eymir et al.  [34] held that AHSE (active heel gliding exercise) therapy provides more practical rehabilitation and leads to beneficial outcomes for patients with TKA. Therefore, their active exercise approach that encourages patients to participate in rehabilitation should be the first choice for acute postoperative rehabilitation after TKA rather than CPM.

Postoperative knee rehabilitation is essential to maintain joint motor function and significantly affects the quality of life and satisfaction of patients. This study describes in detail the clinical applicability of CPM and PT through meta-analysis, which is of great significance for the selection of rehabilitation exercises and the development of the next rehabilitation program for patients in clinical practice. Meanwhile, our meta-analysis also has some limitations. Firstly, CPM protocols and follow-up periods were inconsistent across all studies, which may lead to the possibility of bias. The long-term impact of CPM should be further assessed. Furthermore, due to the nature of the CPM equipment, it was not possible to blind the subjects to CPM grouping. In addition, some patients had received TKA before this study and therefore knew that the use of CPM devices as standard, could lead to effects that could have uncontrolled patient implications. Therefore, an assessment of the risk of bias revealed a generally high risk of bias in allocation concealment (selection bias) and participant blindness (performance bias). In the case of CPM application, however, these situations are unavoidable. These inconsistent results may be due to inappropriate matching of the CPM machine to the patient as well as measurement errors in ROM between studies.

Combined with CPM did not significantly improve postoperative functional recovery relative to physical therapy. There was no difference between the two in terms of time to hospital discharge and patient satisfaction. Overall, CPM did not show superior benefits for postoperative patient recovery. On the contrary, it was associated with increased equipment costs and care expenses. Therefore, the results of the current study are insufficient to support the routine use of CPM to facilitate the recovery process after arthroplasty. We believe that as CPM is used more in orthopedics, further optimization of measurement structures and device innovations are needed for additional evaluation.