Effect of continuous passive motion following total knee arthroplasty on knee range of motion and function: a systematic review.
Abstract: The aim of this review was to determine the effect of Continuous Passive Motion on knee range of motion and function following total knee arthroplasty. Databases were searched (Medline, PEDro, Cochrane, CINAHL and Google Scholar) to locate randomised controlled trials published in English from 2000 onwards. Forty-two articles were extracted from the search, Inclusion criteria consisted of: 1) participant had a primary total knee arthroplasty 2) continuous passive motion was the rehabilitation intervention of interest 3) outcome measures included knee range of motion and/ or function, Fifteen of the articles satisfied these criteria. Study methodology was rated using the PEDro scoring system. Six articles rated less than four out of ten, so were excluded. No review articles were included. All nine remaining studies had data on range of motion and five of these found no significant difference between groups. Three studies found short term effects which disappeared during follow-up and one found effects that lasted to the end of follow-up. Six of the nine studies investigated knee function: five of these found no significant difference between groups. One study found a significant difference which lasted to the end of follow-up. These studies show that there may be short term benefits with using continuous passive motion particularly for range of motion, but these effects are generally not long term. There is limited evidence on the beneficial effect of continuous passive motion on knee function.

Key Words: Continuous passive motion, total knee arthroplasty, range of motion, physiotherapy.
Article Type: Report
Subject: Osteoarthritis (Care and treatment)
Joints (Range of motion)
Joints (Physiological aspects)
Authors: Viswanathan, Poornima
Kidd, Martin
Pub Date: 03/01/2010
Publication: Name: New Zealand Journal of Physiotherapy Publisher: New Zealand Society of Physiotherapists Audience: Academic Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2010 New Zealand Society of Physiotherapists ISSN: 0303-7193
Issue: Date: March, 2010 Source Volume: 38 Source Issue: 1
Topic: Event Code: 200 Management dynamics Computer Subject: Company business management
Product: Product Code: 3841520 Physical Therapy Equip NAICS Code: 33911 Medical Equipment and Supplies Manufacturing
Accession Number: 226728746

Total knee arthroplasty (TKA) is a common procedure used for patients with osteoarthritis (OA). As time spent in hospital post surgery is rapidly decreasing, it is important to find rehabilitation techniques that quickly re-establish knee range of motion (ROM) (Lenssen et al 2003a). Ninety-five degrees of knee flexion has been shown to allow activities of daily living to be carried out (Lenssen et al 2008), therefore restoration of post-operative ROM is considered a major indicator of the success of a TKA (Beaupre et al 2001).

Continuous passive motion (CPM) is a machine with an external motor which passively moves a joint within a certain ROM (Lenssen et al 2008). It is used widely in the rehabilitation of TKA. Having outlined four stages of stiffness (bleeding, oedema, granulation tissue and fibrosis), O'Driscoll and Giori (2000) proposed that CPM works like a pump, encouraging blood and fluid away from the joint in the bleeding and oedema stages, thus preventing the joint from becoming stiff.

Proposed benefits of CPM include: reduced knee manipulation rates, decreased analgesia requirement, reduced incidence of deep vein thrombosis (DVT), and increased ROM (Lachiewicz 2000). It has also been suggested that patients who have received CPM following TKA have incurred less total cost to the hospital compared with "conventional rehabilitation" (p. 146) due to decreased length of stay and reduced rate of postoperative manipulations (Lachiewicz 2000).

Although Fuchs et al (2005) reported CPM as being protective against deep vein thrombosis, not all authors agree with this (Lynch et al 1988, Ververli et al 1995, Vince et al 1987). Detrimental effects such as increased bleeding and complications of the wound have been reported (Lachiewicz 2000). In addition, CPM is expensive because of the increased time it takes for staff to set up, maintain and adjust the machine as well as the cost of renting and buying additional equipment needed when using CPM (Ritter et al 1989). Since the effects of CPM as an intervention post-TKA remain contentious, this systematic review aims to evaluate the recent literature on the efficacy of CPM after TKA. In drawing conclusions about the clinical value of CPM following TKA, the authors aim to identify areas for further research.


An extensive search was conducted by one author (PV) in March 2009 to locate randomised controlled trials deemed relevant to the review topic. The authors chose to exclude studies published before 2000, since they felt earlier studies were reviewed in an earlier publication on the same topic (Milne et al 2003).

Studies were identified using electronic databases (Medline, PEDro, Cochrane, CINAHL and Google scholar) and reference lists of relevant articles. Keywords used for database searching are documented in Table 1a and the search strategy used for Medline is presented in Table 1b.

Inclusion/exclusion criteria (Table 2) were applied to studies identified from the search strategy by the searching author (PV). Studies satisfying the criteria were assessed using the PEDro scale to determine their methodological quality (score out of ten). The PEDro scale is created specifically to assess the methodological quality of physiotherapy studies. The scale's scoring is approximately two times less variable than the Jadad scale which is another commonly used scale (Bhogal et al 2005). The PEDro scale has also been shown to have fair to good inter-rater reliability with an intraclass correlation coefficient of 0.55-0.56 (Maher et al 2003). Furthermore, the PEDro scale encompasses a much wider range of criteria when assessing quality than the Jadad scale which considers only three criteria and focuses on double blinding as its main indicator of quality (Bhogal et al 2005). For these reasons, the authors believed the PEDro scale to be the most appropriate for this review. The components of the PEDro scale are described in Table 3. The PEDro scale also considers introduction of bias through questions on blinding and intention to treat. Thus the total PEDro score attributed to a study depends partially on whether the study protected itself against bias. One study was not found on the PEDro database (Bruun-Olsen et al 2009), but was evaluated using the PEDro scale by an author of this review (PV) and independently cross checked by the second author (MK).

A cut off score of 4/10 was chosen for this review. Six out of ten is usually used for the PEDro scale (Maher et al 2003). However, it has been shown that whilst a cut off score of 4/10 generally does not change the conclusion of a review it may increase the level of evidence for a given conclusion (Maher 2003).

Data from included studies were extracted by one author (PV), documented into tables then checked by the second author (MK) (Table 4). Results documented included: PEDro score, sample size, length of follow up, outcome measures used, interventions used, participant characteristics and the significance of results (p value). Specifically, this review focused on knee ROM measures and functional status of subjects post-operatively.

This review compared the difference in means of studies using CPM as an intervention against control interventions. Statistically significant results are described using p values (a statistically significant finding for this review was defined as p<0.05).


Study selection

A database search (Medline, PEDro, Cochrane, CINAHL and Google scholar) and reference list searching yielded 42 randomised control trials relating to this review. However, nineteen were published before 2000, four did not use CPM after a TKA or did not use CPM on the knee, and four were not in English. Of the remaining 15 articles, six were excluded: one due to poor quality--2/10 PEDro scale (Leach, et al 2006); four were deemed not relevant once read (Davies et al 2003, Fuchs et al 2005, Lenssen et al 2006, Salter, 2004) and one did not have a non-CPM control (Leonard et al 2007). Figure 1 illustrates this selection process. The nine articles included in this review are summarised in Table 5.


Methodological quality

Table 4 compares the methodological features of each study. Random allocation, between group comparisons and point measures were adequate in all studies whereas no study had subject or therapist blinding. All studies except Lau and Chiu (2001) had assessor blinding. Concealment of allocation and intention to treat were also factors missing from many studies.

Range of motion

All nine studies measured ROM. Five studies (Bruun-Olsen et al 2009, Chen et al 2000, Denis et al 2006, Lenssen et al 2003a, Lenssen et al 2008), compared the effect of CPM as an adjunct to physiotherapy (intervention) with physiotherapy alone (control). Three of the five studies (Bruun Olsen et al 2009, Chen et al 2000, Denis et al 2006) found no statistically significant differences in ROM. Lenssen et al (2008), detected a non-significant improvement of 5[degrees] in the intervention group at day seventeen (p=0.06-0.07). At follow-up (six weeks and three months) no difference was found. Lenssen et al (2003a), found passive extension improved significantly in the intervention group compared to the control group at the end of their study (p=0.029).

Three studies (Bennett et al 2005, Lau and Chiu 2001, McDonald et al 2000), compared the use of CPM with a control group that did not receive CPM as part of post-TKA rehabilitation. McDonald et al (2000), found no significant difference between the groups. Bennett et al (2005), observed a statistically significant difference at day five in the early CPM group for active and passive knee flexion (p=.008 and p<.0001). However, at the three month and one year follow-up, no differences were seen. Lau and Chiu (2001), found active flexion improved in the CPM group; (61.3% of CPM group had 90[degrees] active flexion at day seven versus 44.8% of no-CPM group p<0.05) however, this difference disappeared by day 14. Beaupre et al (2001) divided participants into three intervention groups: one group received standard exercises (SE) and CPM; the second group received SE and sliding board therapy (SB); the third group received only SE. They found no significant differences in recovery of ROM between any of the groups.

Knee Function

Six of the nine studies considered knee function as an outcome measure. Out of the five studies investigating CPM and standard physiotherapy compared to standard physiotherapy alone, three (Denis et al 2006, Lenssen et al 2003a, Lenssen et al 2008) explored knee function. Denis et al (2006) used the Western Ontario and McMaster Universities (WOMAC) osteoarthritis index (a 24 item questionnaire that focuses on lower limb activity and covers pain, stiffness and functional difficulty) while Lenssen et al (2008) used both the WOMAC osteoarthritis index and Knee Society Score (KSS) which considers both knee assessment (pain, stability and ROM) and functional ability to climb stairs and walk. Both studies found no significant differences between intervention groups. Lenssen et al (2003a) used the Hospital for Special Surgery Score, which focuses on pain, function (including strength) and the knee joint itself. They found that the standard physiotherapy group's knee function scores were significantly lower (p= 0.003) than the group receiving CPM after seventeen days.

Two (Bennett et al 2005, McDonald et al 2000) of the three studies comparing a CPM group with a non-CPM group considered knee function. Both studies used the KSS and found no significant difference between the groups. In addition, the Beaupre et al (2001) study investigated knee function using the WOMAC osteoarthritis index and found no significant differences between the three participant groups.


This systematic review investigated the efficacy of CPM after TKA with results from nine randomised control trials. These ranged in quality from 4/10 to 8/10 on the PEDro scale.

Overall, three types of post-TKA intervention were used: CPM and physiotherapy compared to physiotherapy alone (five studies); CPM compared to no CPM (three studies); and a mixture of standard exercises, sliding board exercises, and CPM (one study).

CPM intervention parameters were different in each study. The initiation of intervention was also found to be varied. Six studies began their intervention day one post-operatively. One study started day two after surgery, while another started day two or three depending on the group the participant was in. Finally, one study gave all patients standard physiotherapy including CPM for four days before initiating the intervention. These variations mean reliable comparison of the results is difficult.

Among the participants in all the studies, there were an unequal proportion of males and females (generally the number of females outweighed males). As gender characteristics may influence results, the results are thus likely to be more generalizable to females. Another noticeable feature of six of the nine studies was the inclusion of only participants with osteoarthritis of the knee and these tended to be older patients. Thus results may not apply to patients who received a TKA for reasons other than osteoarthritis (for example rheumatoid arthritis).

This review focused specifically on ROM and function, which are important outcomes when gauging the success of a TKA.

Range of Motion

In all nine studies, a goniometer was used to measure ROM. However, what was measured differed from study to study. Three studies measured active flexion and extension, two measured passive flexion and extension, while a further two measured active and passive flexion and extension. Quadriceps lag, passive extension and active & passive flexion were measured in Bennett et al (2005). McDonald et al (2000) did not define whether flexion and extension was active or passive.

Notwithstanding variation in measurement, five (Beaupre et al 2001, Bruun-Olsen et al 2009, Denis et al 2006, Chen et al 2000, McDonald et al 2000) of the nine studies found no statistically significant change in ROM. Only one study found a significant difference (p=0.029) in passive extension that lasted to post-operative day 17, the end of follow-up (Lenssen et al 2003a). Follow-up did not extend for long enough after cessation of the intervention for conclusions to be made on long term effects of CPM on ROM; nonetheless, the findings seem to indicate that if CPM has any benefit for ROM, it may be short term.

Knee Function

Of the six studies which investigated the effect of CPM on knee function, five found no significant differences between intervention groups. It should be noted, however, that different tools were used to measure knee function (WOMAC osteoarthritis index, Knee Society Score and Hospital for Special Surgery Score). Each of these tools differs slightly in focus, which may affect comparability of results. However in general it appears CPM has no superior effect over standard physiotherapy in terms of improving knee function.

General Discussion

Statistical power illustrates the strength of a study. Four studies did not define their power. Denis et al (2006) had a power of 86% while Lenssen et al (2008), Bennett et al (2005), Beaupre et al (2001), and Bruun-Olsen et al (2009) had powers of 80%. Chen et al (2000) required 32 participants per group for a power of 80% but the study failed to meet this requirement.

The lack of blinding in the studies (subject and therapist) could introduce performance bias. However, considering the intervention, blinding would be difficult to achieve. Nonetheless, to reduce bias as much as possible, assessor blinding should be present in all studies. This was the case in all studies except Lau and Chiu (2001).

The possibility of publication bias must also be considered when interpreting the findings of a study. Positive or desirable results may be more likely to be published. This can cause results to inaccurately represent true research findings. Furthermore, limiting the language of included studies to English means there could be a language bias on the types of results identified. Therefore a limitation of the review is that only published English material was included. Unpublished material was not considered.

Adequate follow-up which ensures results are not influenced by loss of potentially significant data was not present in two studies. Intention-to-treat was lacking in five of the nine trials: studies which exclude data from participants who do not complete the study may alter the results and introduce exclusion bias. Finally, only four of the studies exhibited concealed allocation.

As the length of follow up between included studies was varied, it is difficult to draw conclusions on the long term effects of CPM after a TKA. Whilst the majority of studies had a follow up period of at least three months and found no statistically significant differences between intervention

groups, Denis et al (2006), Lenssen et al (2003a) and Chen et al (2000) had much shorter follow up periods. Therefore whether CPM use had a long-term effect on subjects of those studies could not be determined. Lenssen et al (2003a) reported significant improvement in extension in the CPM group as well as decreased functional status in their control group at day 17. However, as follow up did not extend beyond day 17 it is difficult to draw conclusions on the maintenance of the short-term ROM improvement due to CPM. Though further research would be required to determine the exact time frame involved, the studies reviewed show that any benefit to ROM and functional recovery in the post-TKA knee due to CPM disappears over a relatively short period of time, and there is no evidence to indicate that CPM groups recover at a different rate from control groups in the long-term.

Whilst many of the studies had poor methodology, and varied interventions (which made robust comparison difficult), the results appear to illustrate that any benefit of CPM over physiotherapy alone is short lived. In the long term it seems CPM does not improve ROM or knee function any more than physiotherapy.

Although CPM seems to be a commonly accepted and current rehabilitation technique after TKA, it is a contentious and potentially costly and time-consuming intervention post-TKA. Based on the results of this review, the use of CPM does not seem justified as an adjunct to standard physiotherapy. An earlier systematic review by Milne et al (2003) concluded that CPM with physiotherapy may be beneficial over physiotherapy alone only in the short term. Similarly, Lenssen et al (2003b) found CPM provided no long term effects on ROM. They recommended CPM be used if ROM needed to be regained quickly. The results of this review are in agreement with both of these studies.


Recent evidence for the use of CPM after TKA is varied and conflicting. The majority of studies found either no significant benefit or a small benefit in ROM that disappeared during medium-term follow-up. There is no evidence to support the use of CPM in achieving long-term benefits in either knee ROM or function. Consequently, regardless of the possible costs and effort associated with application of CPM (itself a contentious issue), it seems physiotherapy is the more valuable rehabilitation technique. It is suggested that physiotherapy is primarily used after TKA but if cost and effort are not limiting factors, CPM may offer faster rehabilitation than physiotherapy alone. This review has focused primarily on the effects of CPM on ROM and knee function. However, other proposed CPM benefits such as reduction in need for analgesia and reduced DVT incidence were not reviewed, and may form the basis of further research. Furthermore, as many different doses of both physiotherapy and CPM were used, this review cannot favour one particular dosage as being most beneficial. Research on the optimal dosage of CPM is needed so that clinicians can have confidence in CPM protocols, and a gold standard in CPM application post-TKA may be found.

Key Points

Following a total knee arthroplasty:

* There is conflicting evidence regarding for the effectiveness of CPM for short term increase in knee ROM

* There is evidence of no long term benefit of CPM to restoration of either knee ROM or function

* There is currently no evidence regarding optimal dosage of CPM


Article used as template for review

Dodd KJ, Taylor NF and Damiano DL (2002): A systematic review of the effectiveness of strength-training programs for people with cerebral palsy. Archives of Physical Medicine and Rehabilitation 83: 1157-1164.

Articles used in review

Beaupre LA, Davies DM, Jones CA, and Cinats JG (2001): Exercise combined with continuous passive motion or slider board therapy compared with exercise only: A randomised controlled trial of patients following total knee arthroplasty. Physical Therapy 81: 1029-1037.

Bennett LA, Brearley SC, Hart JAL and Bailey MJ (2005): A comparison of 2 continuous passive motion protocols after total knee arthroplasty. Journal of Arthroplasty 20: 225-233.

Bruun-Olsen V, Heiberg KE and Mengshoel AM (2009): Continuous passive motion as an adjunct to active exercises in early rehabilitation following total knee arthroplasty--a randomised controlled trial. Disability and Rehabilitation 31: 277-283.

Chen B, Zimmerman JR, Soulen L and DeLisa JA (2000): Continuous passive motion after total knee arthroplasty: A prospective study. American Journal of Physical Medicine and Rehabilitation 79: 421-426.

Denis M, Moffet H, Caron F, Ouellet D, Paquet J and Nolet L (2006): Effectiveness of continuous passive motion and conventional physical therapy after total knee arthroplasty: A randomised clinical trial. Physical Therapy 86: 174-185.

Lau S.KK and Chiu KY (2001): Use of continuous passive motion after total knee arthroplasty. Journal of Arthroplasty 16: 336-339.

Lenssen AF, De Bie RA, Bulstra SK and van Steyn MJA (2003a): Continuous passive motion (CPM) in rehabilitation following total knee arthroplasty: A randomised controlled trial. Physical Therapy Reviews 8: 123-129.

Lenssen TAF, van Steyn MJA, Crijns YHF, Waltje EMH, Roox GM, Geesink RJT et al (2008): Effectiveness of prolonged use of continuous passive motion (CPM), as an adjunct to physiotherapy, after total knee arthroplasty. BMC Musculoskeletal Disorders 9: 60.

McDonald SJ, Bourne RB, Rorabeck CH, McCalden RW, Kramer J and Vaz M (2000): Prospective randomized clinical trial of continuous passive motion after total knee arthroplasty. Clinical Orthopaedics and Related Research 380: 30-35.

Other References

Bhogal SK, Teasell RW, Foley NC and Speechley MR (2005):

The PEDro scale provides a more comprehensive measure of methodological quality than the Jadad scale in stroke rehabilitation literature. Journal of Clinical Epidemiology 58: 668-673.

Brosseau L, Milne S, Wells G, Tugwell P, Robinson V, Casimiro L et al (2004): Efficacy of continuous passive motion following total knee arthroplasty: A meta-analysis. Journal of Rheumatology 31: 2251-2264.

Davies DM, Johnston WC, Beaupre LA and Lier DA (2003): Effect of adjunctive range-of-motion therapy after primary total knee arthroplasty on the use of health services after hospital discharge. Canadian Journal of Surgery 46: 30-36.

Fuchs S, Heyse T, Rudofsky G, Gosheger G and Chylarecki C (2005): Continuous passive motion in the prevention of deep-vein thrombosis: A randomised comparison in trauma patients. Journal of Bone and Joint Surgery 87: 1117-1122.

Lachiewicz PF (2000): The role of continuous passive motion after total knee arthroplasty. Clinical Orthopaedics and Related Research 380: 144-150.

Leach W, Reid J and Murphy F (2006): Continuous passive motion following total knee replacement: a prospective randomized trial with follow-up to one year. Knee Surgery, Sports Traumatology, Arthroscopy 14: 922-926.

Lenssen AF, Crijns YH, Waltje EM, Roox GM, van Steyn MJ, Geesink RJ et al (2006): Effectiveness of prolonged use of continuous passive motion (CPM) as an adjunct to physiotherapy following total knee arthroplasty: design of a randomised controlled trial. BMC Musculoskeletal Disorders 7: 15.

Lenssen AF, Koke AJA, De Bie RA and Geesink RGT (2003b): Continuous passive motion following primary total knee arthroplasty: Short- and long-term effects on range of motion. Physical Therapy Reviews 8: 113-121.

Leonard GM, Tremblay LE, Chabot M, Lariviere J and Papadopoulos P (2007): The effects of early continuous passive motion after total knee arthroplasty. Physiotherapy Canada 59: 111-117.

Lynch AF, Bourne RB, Rorabeck CH, Rankin RN and Donald A (1988): Deep-vein thrombosis and continuous passive motion after total knee arthroplasty. Journal of Bone and Joint Surgery 70: 11-14.

Maher CG (2000): A systematic review of workplace interventions to prevent low back pain. Australian Journal of Physiotherapy 46: 259-269.

Maher CG, Sherrington C, Herbert RD, Moseley AM and Elkins M (2003). Reliability of the PEDro scale for rating quality of randomized controlled trials. Physical Therapy 83: 713-721.

Milne S, Brosseau L, Welch V, Noel MJ, Davis J, Drouin H et al (2003): Continuous passive motion following total knee arthroplasty. Cochrane Database of Systematic Reviews, 2 Art. No.: CD004260. DOI:10.1002/14651858.CD004260.

Moher D, Liberati A, Tetzlaff J and Altman DG (2009). Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. British Medical Journal 339: 332-339.

Moseley AM, Herbert RD, Sherrington C and Maher CG (2002). Evidence for physiotherapy practice: A survey of the physiotherapy evidence database (PEDro). Australian Journal of Physiotherapy 48: 43-49.

Naylor J, Harmer A, Fransen M, Crosbie J and Innes L (2006): Status of physiotherapy rehabilitation after total knee replacement in Australia. Physiotherapy Research International 11: 35-47.

O'Driscoll SW and Giori NJ (2000): Continuous passive motion (CPM): Theory and principles of clinical application. Journal of Rehabilitation Research and Development 37: 179-188.

PEDro: The Physiotherapy Evidence Database (PEDro) frequently asked questions: how are trials rated? http://PEDro.org.au/ scale_item.html [Accessed on April 14, 2009].

Ritter MA, Gandolf VS and Holston KS (1989): Continuous passive motion versus physical therapy in total knee arthroplasty. Clinical Orthopaedics and Related Research 244: 239-243.

Salter RB (2004): Continuous passive motion: From origination to research to clinical applications. Journal of Rheumatology 31: 2104-2105.

Ververli PA, Sutton DC, Hearn SL, Booth RE, Hozack WJ and Rothman RR (1995): Continuous passive motion after total knee arthroplasty analysis of cost and benefits. Clinical orthopaedics and related research 321: 208-215.

Vince KG, Kelly MA, Beck J and Insall JN (1987): Continuous passive motion after total knee arthroplasty. Journal of Arthroplasty 2: 281-284.

Poornima Viswanathan

School of Physiotherapy, University of Otago, Dunedin, New Zealand

Martin Kidd, BSc, DipPT, PGDipTertTchg, MHealSci

Clinical Coordinator, Professional Practice Fellow, School of Physiotherapy, University of Otago, Dunedin, New Zealand


Martin Kidd, PO Box 56, School of Physiotherapy, University of Otago, Dunedin, New Zealand, Email: martin.kidd@otago.ac.nz, Phone: +64 03 4798436, Fax: +64 03 4795161
Table 1a: Keywords used in search strategy

"OR"   Arthroplasty, Replacement knee        AND
       Knee Joint

"OR"   Motion therapy, continuous passive
       Physical therapy

Other keywords used: Rehabilitation;
Range of motion (ROM)

Table 1b: Medline search strategy

1. Arthroplasty, Replacement, Knee/
2. TKA.mp.
3. Knee Joint/
4. 1 or 2 or 3
5. Motion Therapy, Continuous Passive/
6. 4 and 5
7. CPM.mp.
8. 5 or 7
9. 4 and 8
10. Physical Therapy Modalities/
11. 9 and 10
12. Rehabilitation/
13. 8 and 12
14. 4 and 12
15. 4 and 10
16. 12 and 15
17. Osteoarthritis/ or Osteoarthritis, Knee/
18. 8 and 17
19. 10 and 12 and 17

Table 2: Inclusion/exclusion criteria for studies
for this review.

Inclusion Criteria:      Exclusion Criteria:

1. Patients had          1. Review Articles
   undergone a
   primary TKA

2. CPM was used          2. Studies with PEDro
   as a rehabilitation   scores below 4/10

3. Outcome measures
   included knee
   ROM and/or function

Table 3: PEDro scale components used to score randomized
controlled trials out of ten (component 1 is not
considered part of the score).

Components of PEDro Scale *

1) Eligibility criteria defined

2) Random allocation

3) Concealed allocation

4) Baseline comparison

5) Subject Blinding

6) Therapist blinding

7) Assessor blinding

8) Adequate Follow up (>85% of participants)

9) Intention to treat

10) Between group comparison

11) Point measures (treatment effect)

* PEDro http://PEDro.org.au/scale_item.html; Maher et al 2003;
Bhogal et al 2005; Moseley et al 2002.

Table 4: A full summary of the nine studies
included in this review

Study Name      PEDro    Sample Size    Outcome
                Score    and Follow     variables

Denis et al     8/10     81 Subjects    - Active knee flexion
(2006),                                 + extension ROM
                         Follow up      (goniometer)
                         7-8 days       - Timed up and go
                                        - Hospital length of

Lenssen et al   8/10     60 Subjects    - Functional status
(2008),                                 (WOMAC and Knee
                         Follow up      Society score)
                         Day 4, 17,     - Active + passive
                         6 weeks, 3     lexion and
                         months         extension knee ROM
                                        - Perceived effect
                                        - Post-operative
                                        - Satisfaction and

Beaupre et      8/10     120 Subjects   - Active lexion +
al (2001),                              extension knee ROM
                         Follow up      (goniometer)
                         5-7days,       - Joint pain, stiffness
                         3 and 6        and function
                         months         (WOMAC)
                                        -Health status

Bruun-Olsen     7/10*    63 Subjects    - Active + passive
et al (2009),                           lexion and
                         Follow up      extension knee ROM
                         1 week, 3      (goniometer)
                         months         - Pain (VAS)
                                        - Walking ability
                                        (Timed up and go)
                                        - Swelling

Lenssen et al   7/10     40 Subjects    - Passive lexion +
(2003a),                                extension knee ROM
                         Follow up      (goniometer)
                         Day 4 and      - Function -Hospital
                         17             for special surgery
                                        score (HSS)
                                        - Length of hospital
                                        - Pain (VAS)
                                        - Muscle strength
                                        (Hand held
                                        - Medication
                                        - Satisfaction

Bennett et al   6/10     147 Subjects   - Quadriceps lag,
(2005),                                 passive extension,
                         Follow up      active + passive
                         Day 5,3        lexion knee ROM
                         months, 1      (goniometer)
                         year           - Hospital stay length
                                        - Wound healing
                                        - Knee function (Knee
                                        Society Score)
                                        - Perceived health
                                        - Pain (VAS)

McDonald et     5/10     120 Subjects   - Flexion + extension
al (2000),                              knee ROM
                         Follow up      - Analgesic use
                         Discharge,     - Length of stay
                         6, 12, 26      - Knee function (Knee
                         weeks, 1       Society Score)

Chen et al      5/10     51 Subjects    - Passive lexion +
(2000),                                 extension knee ROM
                         Follow up      (goniometer)
                         Day 3, 7,      - Knee effusion
                         discharge      (circumference)

Lau & Chiu      4/10     43 Subjects    Active flexion t
(2001).                                 extension knee ROM
                         Follow up      (goniometer)
                         Day 3, 5,      - Total drain output
                         7, 14, 28,     - Total hospital stay
                         42, 3 and 6
                         months, 1

Study Name      Intervention

Denis et al     - Control (n=27);
(2006),         Physiotherapy only
                - Intervention group 1
                (n=26); 35 minutes daily
                CPM + Physiotherapy
                - Intervention group 2
                (n=28); 2 hours daily CPM
                + Physiotherapy

Lenssen et al   All patients had 20
(2008),         minutes of Physiotherapy
                and 4 hours of CPM (twice
                daily for 2 hours) for 4
                days, After this:
                - Control (n=30);
                Physiotherapy (30 minutes
                - Intervention group
                (n=30); Physiotherapy
                and additional 2 weeks of
                CPM (4 hours daily),

Beaupre et      - SE with CPM (n=40)
al (2001),      - SE and SB (n=40)
                - SE alone (n=40)
                Exercise started day three
                post surgery; (3 sets of 10-
                15 repetitions daily for 30
                CPM; (3x 2 hour sessions
                daily) 0-30[degrees] increased as
                SB; at least 2x 10 minute
                sessions daily,

Bruun-Olsen     - Active exercises and
et al (2009),   CPM 70-100[degrees] (n=30) 2
                sessions of 2 hour daily
                CPM for 1 week
                - Active exercises (n=33);
                Exercises started first
                day after surgery, Done
                daily for 30 minutes,
                (Active movements,
                strengthening and
                walking for 1 week),

Lenssen et al   - CPM with Physiotherapy
(2003a),        (n=20)
                - Physiotherapy (n=20)
                Both groups started day 1
                after surgery for 5 days,
                strengthening + mobilising,
                20 minutes twice daily for
                4 days,
                CPM; 4 hours daily,
                All patients received pain
                After discharge,
                all patients had
                Physiotherapy 3 times
                weekly at home,

Bennett et al   - Standard CPM (n=47);
(2005),         twice daily for 3 hours
                0-40[degrees] increasing 10[degrees] daily
                - Early CPM (n=48); twice
                daily for 3 hours 90-50[degrees]
                initially, progressed to full
                extension by day 3,
                - No CPM (n=52)
                All groups had post-
                operative Physiotherapy,
                CPM occurred over five

McDonald et     - CPM from 0-50[degrees] (n=40);
al (2000),      increasing to tolerance
                - CPM from 70-110[degrees] (n=40)
                - NO CPM (n=40)
                CPM started in recovery
                room for 18-24 hours,
                From day 2 all patients
                had twice daily
                Physiotherapy (ROM
                exercises + mobilisations)
Chen et al      - Group 1 (n=23); CPM (
(2000),         starting at 0[degrees] extension)
                5 hours daily and
                Physiotherapy (2 hours
                - Group 2 (n=28)
                Physiotherapy (2 hours
                Intervention for 8 days

Lau & Chiu      - CPM group; started
(2001).         day 1 after surgery
                0-60[degrees] , increasing to pain
                tolerance. 23 hours daily
                for 6 days.
                - Immobilised group;
                no exercises (ROM or
                walking) were given
                until after day 7. (After
                this, both groups
                received Physiotherapy;
                mobilisations t full weight
                bearing walking)

Study Name      Limitations                 Results

Denis et al     - No blinding of subjects   - No significant
(2006),         or therapists,              differences were
                - All participants had OA   found between
                -WOMAC scale was            the groups for any
                modified                    outcome,
                - Findings apply to the
                dosage of CPM used,
                - 38 male:43 female

Lenssen et al   - Used participants with    - ROM; was 5 [degrees]
(2008),         decreased ROM 4 days        more in the intervention
                after surgery               group than control group
                - No blinding of subjects   at day 17, This
                or therapists,              difference disappeared
                - All participants had OA   at follow up (6 weeks
                - Before surgery,           and 3 months),
                average lexion was
                higher in intervention      - No significant
                group,                      difference was
                - Patients managed          noted for any other
                CPM themselves              outcome,
                - 21 male:39 female

Beaupre et      - Subjects and therapists   - ROM; no significant
al (2001),      not blinded                 differences in flexion
                - All patients received SB  or extension between
                exercises (15 min daily)    the groups,
                as part of SE               - No significant
                - CPM group adherence       differences were
                was an average of 1,8       noted for any of
                sessions per day not        the other outcomes
                3 while the SE group        between groups,
                averaged 1,7 sessions
                per day,
                - 67 male:53 female

Bruun-Olsen     - Subjects and therapist    - There were no
et al (2009),   not blinded                 statistically significant
                - No intention to treat     differences for any
                - All participants had OA   of the outcomes
                - 19 male:44 female         between the groups
                                            at one week or three

Lenssen et al   - All patients had OA       - Passive ROM;
(2003a),        - Allocation not            extension in CPM
                concealed                   group improved
                - Subjects and therapists   significantly (p=0,029),
                not blinded                 - Function (HSS)
                - CPM compliance            declined more
                was not 4 hours daily       in control group;
                as expected (average        (p=0,003 after 17
                3,6hrs daily)               days)
                - Differences in pain       - Strength; CPM
                perception between          group was twice that
                patients,                   of control at day 4,
                - 13 male:27 female         Difference gone after
                                            day 17,
                                            - Pain; less in CPM
                                            group at day 4 and 17
                                            (p=0,005 & p=0,009),
                                            - No differences for
                                            other outcomes,

Bennett et al   - Allocation not            - Active and passive
(2005),         concealed                   knee flexion;
                - Subjects and therapists   improvement at day
                not blinded                 5 was statistically
                - No intention to treat     significant (p=,008
                - All participants had OA   and p<,0001) in early
                - Fixed lexion deformity    CPM group,
                may respond differently
                from CPM (not               At three months and
                investigated fully as       1 year, there was no
                sample size too small)      significant difference
                - 48 male:99 female         between groups for
                                            any variable,

McDonald et     - No concealed              No statistically
al (2000),      allocation                  significant results were
                - Subjects and therapists   found for any of the
                not blinded                 outcomes,
                - Follow up was not
                - No intention to treat
                - All patients had OA

Chen et al      - Low power and             - ROM; differences
(2000),         number of participants      between groups not
                - No concealed              significant for knee
                allocation                  lexion and extension
                - Subjects and therapists   (day 3, 7 and
                not blinded                 discharge)
                - Inadequate follow-up      - Knee swelling was
                - No intention to treat     not significantly
                - 15 male:36 female         different,

Lau & Chiu      - Allocation not            -Active flexion; at
(2001).         concealed                   day 7, 44.8% of
                - No baseline               immobilised group
                comparability               had 90[degrees] active
                - No blinding               lexion versus 61.3%
                - No intention to treat     of CPM group.
                - 7 male:36 female          These differences
                                            disappeared by day

                                            - No statistically
                                            significant results were
                                            found for any other

Study Name      Clinical implications

Denis et al     - Implies CPM at the dosage used is
(2006),         no better than Physiotherapy alone in
                rehabilitation after TKA, Therefore cost
                and effort should be considered,

Lenssen et al   - CPM may have a short term effect
(2008),         on ROM; however effect is not long
                term, Thus CPM does not appear
                to be justified over Physiotherapy,
                especially taking cost and effort into
                consideration unless ROM needs to be
                gained quickly,
                - May be limited to patients with limited
                ROM after the in-hospital period,

Beaupre et      - As no significant differences were
al (2001),      noted between the groups, it can
                be concluded that it does not
                make a difference whether SE, SB
                or CPM is used, The end effect is the
                same, Therefore cost and time and
                availability should be considered,

Bruun-Olsen     There was no difference between
et al (2009),   using CPM with exercise and just
                exercise, Therefore it seems more cost
                effective to employ the use of exercise
                after TKA,

Lenssen et al   - This study found CPM with
(2003a),        Physiotherapy to be more beneficial
                than just Physiotherapy in the short

                - These findings differ quite a lot to
                those of many other studies especially
                as many of the effects last to end of
                study (day 17), However is would be
                important to know whether the effects
                are long term,

Bennett et al   - Early CPM improved ROM in the short
(2005),         term, but the effect did not last in the
                long term, Thus the study did not show
                clinically significant results to justify
                CPM over Physiotherapy,
                - CPM may be beneficial if rapid gains
                in ROM are required,

McDonald et     - This study did not show CPM to have
al (2000),      any benefit over Physiotherapy, in the
                rehabilitation of patients after TKA for
                any of the outcomes measured,

Chen et al      - As the differences between the
(2000),         groups for the outcomes were not
                significant, CPM use cannot be justified
                over Physiotherapy as it was just as

Lau & Chiu      - This study shows that CPM has short
(2001).         term advantages that may be useful if
                early discharge is desired but no long
                term benefits.

                - It also showed; immobilisation for one
                week after surgery did not affect the
                final ROM.

Abbreviations: CPM--Continuous passive motion, ROM--Range of
motion, OA--Osteoarthritis, SE--Standard exercise, SB--Slider
board, TKA--Total knee arthroplasty.

* Not scored by PEDro--scored by the authors (PV, MK) using
PEDro scoring criteria

Table 5: Comparison of the methodological quality of the studies
included in this review as outlined by the PEDro scale*. Refer to
table 3 for description of the components numbered 1-11.

Study                         1         2         3         4

Denis et al (2006)         [check]   [check]   [check]   [check]
Bennett et al (2005)       [check]   [check]             [check]
Beaupre et al (2001)       [check]   [check]   [check]   [check]
Lenssen et al (2003a)      [check]   [check]             [check]
Chen et al (2000)          [check]   [check]             [check]
McDonald et al (2000)      [check]   [check]             [check]
Lau & Chiu (2001)          [check]   [check]
Lenssen et al (2008)       [check]   [check]   [check]   [check]
Bruun-Olsen et al (2009)   [check]   [check]   [check]   [check]

Study                         5         6         7         8

Denis et al (2006)                             [check]   [check]
Bennett et al (2005)                           [check]   [check]
Beaupre et al (2001)                           [check]   [check]
Lenssen et al (2003a)                          [check]   [check]
Chen et al (2000)                              [check]
McDonald et al (2000)                          [check]
Lau & Chiu (2001)                                        [check]
Lenssen et al (2008)                           [check]   [check]
Bruun-Olsen et al (2009)                       [check]   [check]

Study                         9        10        11      Score

Denis et al (2006)         [check]   [check]   [check]     8
Bennett et al (2005)                 [check]   [check]     6
Beaupre et al (2001)       [check]   [check]   [check]     8
Lenssen et al (2003a)      [check]   [check]   [check]     7
Chen et al (2000)                    [check]   [check]     5
McDonald et al (2000)                [check]   [check]     5
Lau & Chiu (2001)                    [check]   [check]     4
Lenssen et al (2008)       [check]   [check]   [check]     8
Bruun-Olsen et al (2009)             [check]   [check]     7

* Not scored by PEDro--scored by the authors
(PV, MK) using PEDro scoring criteria
Gale Copyright: Copyright 2010 Gale, Cengage Learning. All rights reserved.