More than 1 million total knee arthroplasties (TKAs) are performed every year.1–4 The procedure has grown in use in the aging population and also increased in patients younger than 65 years.5–8 The number of procedures performed annually is expected to double or triple in the next decade as TKAs and revisions have experienced high single- and double-digit growth rates even in bad economic times.8–10
The standard of care after surgery varies depending on patient, physician, and hospital. Traditional TKA recovery regimens have included hospital stays, inpatient facilities (rehabilitation facility or skilled nursing facility), rehabilitation services in the home, and outpatient physical therapy.11 Others have moved away from inpatient rehabilitation facilities in favor of in-home physical therapy.12 Some have even begun promoting outpatient TKA for qualifying patients followed by rigorous physical therapy.13 Amidst the myriad recovery protocols and procedures adopted around the world, the use of continuous passive motion (CPM) machines in recovery remains a point of discussion and debate among top surgeons worldwide.14–16
The concept of CPM began with work by Salter,17–20 Salter and Field,21 and Salter et al22 in the 1960s. Some CPM articles identified positive outcomes, including reduction of postoperative pain, swelling, and adhesions, while improving local blood circulation and return to motion.11,23–31 However, most recent studies and meta-analyses show little or no long-term benefit of CPM, as measured 3 months after surgery.11,14,16,32–37 By 2002, it was estimated that CPM devices were in use in more than 77 countries at more than 17,000 hospitals and continue to be widely used today.16
In the United States, approximately half of all TKAs are performed on Medicare-eligible patients. The changing regulatory environment has shifted toward bundled payment systems.38,39 Although this approach to health care payments is in its infancy in the United States, it has shed light on the process of selecting TKA recovery techniques that prove efficacious for patient outcomes and efficient for overall systemic cost.40,41 Given these economic trends decreasing reimbursement in bundled payment systems,38–41 the emergence of gravity-assisted passive flexion (GAP-FLEX) represents a potential opportunity to improve patient outcomes. A randomized comparative-outcomes study was designed to quantify the effectiveness of GAP-FLEX as compared with CPM, the current standard of care at the investigating hospitals. The objective was to determine whether there was a difference in recovery between the 2 techniques, leading to the following questions: (1) Can GAP-FLEX improve patient recovery of range of motion (ROM) after TKA? (2) Does GAP-FLEX reduce the patient time and effort (eg, set-up, device use, supervision) associated with therapy compared with CPM? (3) Does GAP-FLEX reduce overall episodic care cost? This article describes the methodology of the study, provides detailed results, and discusses the practical implications.
Materials and Methods
The prospective, randomized, 2-center study received institutional review board approval and was Health Insurance Portability and Accountability Act compliant. The primary endpoint was the degree of active flexion ROM, as measured by goniometer at the week 4 follow-up visit. This primary endpoint is the focus of this study, with a null hypothesis of H0: ROMGAP-FLEX>ROMCPM. A power analysis was conducted using retrospective patient physical therapy ROM data. With an anticipated difference of 10° between 2 samples, a sample standard deviation of 10°, a desired alpha of 0.05, and a power of 0.8, the minimum sample size was calculated to be 13 patients in each of the 2 randomized arms of the study. The study was set up to enroll up to twice as many patients, with a midpoint analysis used to decide whether the data were sufficient to assess the primary endpoint. Multiple secondary endpoints were tracked, along with in-process patient physical therapy compliance. For surgeries, a medial parapatellar arthrotomy approach with a cruciate-retaining implant was used. A Vanguard knee joint replacement (Zimmer Biomet, Warsaw, Indiana) and a drain were used on all patients. Patients were closed at 90° of flexion, and post-acute pain management was done with a periarticular injection, an adductor canal block, and standard pain management medications.
The target audience for inclusion in the study was patients 50 years and older, undergoing TKA (not a revision), agreeing to the proposed physical therapy regimen, and signing an approved informed consent to participate in the randomized study. The postoperative physical therapy regimen was in-home therapy only, and no inpatient rehabilitation facility was used. Exclusion criteria were intended to eliminate other sources of variation in the study. Major exclusion criteria included flexure contracture greater than 10°, flexion less than 60° before or immediately after the procedure (as measured on the operating table), severe deformities, body mass index greater than 40 kg/m2, and a history of deep venous thrombosis (DVT) or pulmonary embolisms. Intraoperative and postoperative exclusion criteria included identification of a fracture, need for additional fixation devices, additional procedures other than TKA, newly identified conditions, admittance to the intensive care unit, and hospital stay longer than 4 days.
Gravity-assisted passive flexion uses gravity to provide a completely passive flexion of the tibiofemoral joint and surrounding tissue followed by an extension. The decompressing effect minimizes pressure on the knee, which reduces irritation, swelling, and pain in the tibiofemoral joint during therapy. The therapy regimen begins postoperatively in the hospital and continues during inpatient rehabilitation in the home. Figure 1 and Figure 2 show GAP-FLEX therapy instructions and an example of the device in use postoperatively, respectively.
Gravity-assisted passive flexion (GAP-FLEX) therapy guidelines for flexion and extension.
Postoperative recovery using gravity-assisted passive flexion (GAP-FLEX) in the hospital setting.
The study included standard-of-care physical therapy for all patients and use of CPM or GAP-FLEX per standard operating instructions. The standard-of-care protocol for CPM both in hospital and at home was 3 times per day for 2 hours each, followed by use of an extension box for 20 minutes each time. The CPM protocol required an initial setting of −10° to 60°, increasing to 120° as tolerated. The GAP-FLEX protocol was 6 times per day for up to 6 minutes each (depending on comfort level), followed by 10 minutes on an extension block. Data were collected per the schedule of assessments presented in Figure 3. Patients were expected to go from hospital therapy to in-home therapy until the primary endpoint of the study (week 4). The same in-home physical therapists were used for all study patients. Outpatient physical therapy was optional post primary endpoint based on patient recovery and physician assessment of recovery. Active flexion was measured using a goniometer, pain was measured using Wong-Baker FACES Pain Rating Scale, and functional mobility was measured using the Tinetti Balance Assessment Tool.
Schedule of assessments for the study, with primary endpoint at visit (V) 2. Abbreviations: DC, discharge; PT, physical therapy; ROM, range of motion; TKR, total knee replacement.
Patients were prospectively randomized to either GAP-FLEX (treatment group) or CPM (control group) in a 1:1 format at each of the 2 clinical sites. A total of 27 patients completed the study to primary endpoint, with 14 assigned to the CPM group and 13 to the GAP-FLEX group. A summary of patient demographic data is presented in Table 1.
Therapy device use began within 24 hours after surgery, and first measurement occurred 48 hours after surgery in hospital (ROM day 2). GAP-FLEX patients achieved greater postoperative ROM within 2 days and sustained an improvement to the endpoint of the study. Eighty-five percent (11 of 13) GAP-FLEX patients achieved or surpassed their baseline ROM by study endpoint (week 4), compared with 50% (7 of 14) of CPM patients (Table 2).
A positive difference of 10.7° ROM day 2 was seen between the GAP-FLEX and CPM patients, with a one-tailed t test significance of P=.003. This difference decreased slightly at the midpoint of the study (ROM day 14) to 5.6° (P=.065). The endpoint results at day 28 showed a positive difference of 8.5° between the groups, with a significance level of P=.009. This result allows rejecting the null hypothesis H0: ROMGAP-FLEX≤ROMCPM and accepting the alternate hypothesis Ha: ROMGAP-FLEX>ROMCPM. Note that 1 data point for the CPM group was imputed from 2 measurements that bounded the missing time point.
Patients in the GAP-FLEX group on average achieved a postoperative ROM day 2 that patients in the CPM group achieved a week later (Figure 4). At day 14, GAP-FLEX patients achieved an average ROM equal to that of the CPM patients at the study’s primary endpoint. Ninety-two percent (12 of 13) of GAP-FLEX patients achieved 110° ROM or more at endpoint, compared with 50% (7 of 14) in the CPM group. The ROM trend indicates that CPM patients on average have a lower improvement rate than GAP-FLEX patients. From a patient burden perspective, the absolute time savings between therapy techniques was 151.2 hours with GAP-FLEX. This represents a 90% reduction in scheduled device therapy time. Therapy time for the extension box was equivalent for both groups at 28 hours (a total of 1 hour per day for 28 days). Pain scales showed no statistical difference between groups, whereas functional mobility showed a significant difference of 1.93 (P=.026) at study endpoint (Figure 4).
Active range of motion (ROM) recovery over time. Abbreviations: CPM, continuous passive motion; GFLX, gravity-assisted passive flexion.
Studies continue to be published arguing for or against the value of CPM machines in TKA recovery therapy. There have been some data collected concerning manipulation under anesthesia, yet there has been little discussion about the amount, intensity, and cost of physical therapy required for patients not using CPM.14,16 This study introduces a new type of TKA recovery method using GAP-FLEX and attempted to answer 3 main research questions. Patients using GAP-FLEX were able to achieve a greater active ROM at endpoint. This result is both statistically significant and clearly answers the first question. Furthermore, the GAP-FLEX results at endpoint are significantly higher than those reported in many other meta-analyses regardless of whether they used CPM or physical therapy and manipulation only.24,42,43 As for the second question, the prescribed standard of care using GAP-FLEX requires 90% less patient time than the CPM standard of care (36 minutes vs 360 minutes per day), with equivalent extension box time for both groups (1 hour per day).
Finally, in answer to the third question, the ability of patients to handle the device may reduce the need for admission to a rehabilitation facility or skilled nursing facility. The earlier recovery precludes the need for prolonged outpatient physical therapy, and the faster recovery returns patients to normal life sooner. The number of patients electing optional outpatient physical therapy was greater in the CPM group (12 of 14) compared with the GAP-FLEX group (10 of 13). Patient compliance is a critical success factor in any prescribed therapeutic regimen.44,45 The faster recovery and lowered time burden for patients using GAP-FLEX can improve patient compliance because it is easier to use and requires much less time than CPM. Anecdotal comments from inhospital staff and in-home physical therapists about the general level of happiness with the device also support this thesis. From a systemic cost perspective, GAP-FLEX eliminates the cost of CPM, including the burden associated with maintaining, cleaning, and renting devices, and the potential for hospital-borne infection as a result of device contamination through reuse. By contrast, the GAP-FLEX is a single-patient system that the patient takes home and keeps. Older patients using CPM have much more difficulty getting on and off the machine and require assistance. The ease of use of the GAP-FLEX device compared with CPM has enabled patients, especially individuals who live alone, to forego a rehabilitation facility and elect for in-home physical therapy. This does not fully answer the third question but leads the authors to believe that further study may prove useful in quantifying potential economic benefits.
This study had the limitations of any study with a small sample. However, using a prospective, randomized approach to patient enrollment, the level of statistical significance observed in the primary endpoint provides some comfort. Where possible, adverse events were categorized as device related and/or procedure related. There were 4 adverse events reported during the study, including 3 patients using the CPM device and 1 patient using the GAP-FLEX device. During use of the CPM device, patient 01-020 reported nausea (not related to device use) and patients 01-018 and 02-008 both reported sciatic nerve pain related to device use. During the use of GAP-FLEX, patient 02-011 reported pain and discomfort (not related to device use) and was discovered to have May-Thurner syndrome after TKA. This patient was removed from the study. No device-related adverse events have been reported from the broader population of greater than 500 patients who have been treated with GAP-FLEX at hospitals currently prescribing the therapy. To date, no contraindications have been reported; however, there may exist preexisting conditions or comorbidities that prevent a patient from using GAP-FLEX. In rare and unusual cases where the patient has a preexisting condition that prevents standard mobilization treatment or use of GAP-FLEX, then CPM or intense physical therapy and manipulation may be useful. Further investigation may include the use of GAP-FLEX for recovery from revisions, fractures, anterior cruciate ligament surgery, and meniscus repairs.
For patients undergoing TKA, GAP-FLEX therapy is highly effective, achieving greater ROM and faster recovery than CPM. Ease of use makes implementation in hospital and home health care easy. A single-patient system reduces chances of hospital-borne infections from device reuse. Further evaluation of GAP-FLEX can address the ability of patients to elect home recovery as opposed to an inpatient rehabilitation facility and track potential reductions in need for outpatient physical therapy. A follow-up study is being launched with a larger sample size evaluating total episodic cost using physical therapy only as the control. Together, these studies could represent a potential change to the standard of care that returns patients to normal life faster while potentially reducing total episodic care cost.
- Koh IJ, Kim TK, Chang CB, Cho HJ, In Y. Trends in use of total knee arthroplasty in Korea from 2001 to 2010. Clin Orthop Relat Res. 2013;471(5):1441–1450. doi:10.1007/s11999-012-2622-y [CrossRef] PMID:23054516
- Kurtz SM, Ong KL, Lau E, et al. International survey of primary and revision total knee replacement. Int Orthop. 2011;35(12): 1783–1789. doi:10.1007/s00264-011-1235-5 [CrossRef] PMID:21404023
- Memtsoudis SG, Della Valle AG, Besculides MC, Gaber L, Laskin R. Trends in demographics, comorbidity profiles, in-hospital complications and mortality associated with primary knee arthroplasty. J Arthroplasty. 2009;24(4):518–527. doi:10.1016/j.arth.2008.01.307 [CrossRef] PMID:18534410
- Singh JA, Vessely MB, Harmsen WS, et al. A population-based study of trends in the use of total hip and total knee arthroplasty, 1969–2008. Mayo Clin Proc. 2010;85(10):898–904. doi:10.4065/mcp.2010.0115 [CrossRef] PMID:20823375
- Cram P, Lu X, Kates SL, Singh JA, Li Y, Wolf BR. Total knee arthroplasty volume, utilization, and outcomes among Medicare beneficiaries, 1991–2010. JAMA. 2012;308(12):1227–1236. doi:10.1001/2012.jama.11153 [CrossRef] PMID:23011713
- Crowninshield RD, Rosenberg AG, Sporer SM. Changing demographics of patients with total joint replacement. Clin Orthop Relat Res. 2006;443(443):266–272. doi:10.1097/01.blo.0000188066.01833.4f [CrossRef] PMID:16462450
- Kerkhoffs GM, Servien E, Dunn W, Dahm D, Bramer JA, Haverkamp D. The influence of obesity on the complication rate and outcome of total knee arthroplasty. J Bone Joint Surg Am. 2012;94(20):1839–1844. doi:10.2106/JBJS.K.00820 [CrossRef] PMID:23079875
- Kurtz S, Ong K, Lau E, Mowat F, Halpern M. Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030. J Bone Joint Surg Am. 2007;89(4):780–785. doi:10.2106/00004623-200704000-00012 [CrossRef] PMID:17403800
- Bozic KJ, Kurtz SM, Lau E, et al. The epidemiology of revision total knee arthroplasty in the United States. Clin Orthop Relat Res. 2010;468(1):45–51. doi:10.1007/s11999-009-0945-0 [CrossRef] PMID:19554385
- Kurtz SM, Ong KL, Lau E, Bozic KJ. Impact of the economic downturn on total joint replacement demand in the United States: updated projections to 2021. J Bone Joint Surg Am. 2014;96(8):624–630. doi:10.2106/JBJS.M.00285 [CrossRef] PMID:24740658
- Brosseau L, Milne S, Wells G, et al. Efficacy of continuous passive motion following total knee arthroplasty: a meta-analysis. J Rheumatol. 2004;31(11):2251–2264. PMID:15517640
- Kramer JF, Speechley M, Bourne R, Rorabeck C, Vaz M. Comparison of clinic- and home-based rehabilitation programs after total knee arthroplasty. Clin Orthop Relat Res. 2003;410:225–234. doi:10.1097/01.blo.0000063600.67412.11 [CrossRef] PMID:12771834
- Berger RA, Sanders S, Gerlinger T, Della Valle C, Jacobs JJ, Rosenberg AG. Outpatient total knee arthroplasty with a minimally invasive technique. J Arthroplasty. 2005;20(7) (suppl 3):33–38. doi:10.1016/j.arth.2005.05.021 [CrossRef] PMID:16214000
- Harvey LA, Brosseau L, Herbert RD. Continuous passive motion following total knee arthroplasty in people with arthritis. Cochrane Database Syst Rev. 2014;(2):CD004260. doi:10.1002/14651858.CD004260.pub3 [CrossRef] PMID:24500904
- He ML, Xiao ZM, Lei M, Li TS, Wu H, Liao J. Continuous passive motion for preventing venous thromboembolism after total knee arthroplasty. Cochrane Database of Systematic Reviews. 2014; (1). doi:10.1002/14651858.CD008207.pub3 [CrossRef]
- Joshi RN, White PB, Murray-Weir M, Alexiades MM, Sculco TP, Ranawat AS. Prospective randomized trial of the efficacy of continuous passive motion post total knee arthroplasty: experience of the hospital for special surgery. J Arthroplasty. 2015;30(12):2364–2369. doi:10.1016/j.arth.2015.06.006 [CrossRef] PMID:26165955
- Salter RB. Motion vs rest: why immobilize joints?J Bone Joint Surg. 1982;64-B:251–254.
- Salter RB. The physiologic basis of continuous passive motion for articular cartilage healing and regeneration. Hand Clin. 1994;10(2):211–219. PMID:8040199
- Salter RB. History of rest and motion and the scientific basis for early continuous passive motion. Hand Clin. 1996;12(1):1–11. PMID:8655611
- Salter RB. Continuous passive motion: from origination to research to clinical applications. J Rheumatol. 2004;31(11):2104–2105. PMID:15517619
- Salter RB, Field P. The effects of continuous compression on living articular cartilage: an experimental investigation. J Bone Joint Surg. 1960;42(1):31–90. doi:10.2106/00004623-196042010-00004 [CrossRef]
- Salter RB, Hamilton HW, Wedge JH, et al. Clinical application of basic research on continuous passive motion for disorders and injuries of synovial joints: a preliminary report of a feasibility study. J Orthop Res. 1984;1(3):325–342. doi:10.1002/jor.1100010313 [CrossRef] PMID:6481515
- Bennett LA, Brearley SC, Hart JA, Bailey MJ. A comparison of 2 continuous passive motion protocols after total knee arthroplasty: a controlled and randomized study. J Arthroplasty. 2005;20(2):225–233. doi:10.1016/j.arth.2004.08.009 [CrossRef] PMID:15902862
- Johnson DP. The effect of continuous passive motion on wound-healing and joint mobility after knee arthroplasty. J Bone Joint Surg Am. 1990;72(3):421–426. doi:10.2106/00004623-199072030-00016 [CrossRef] PMID:2179218
- Johnson DP, Eastwood DM. Beneficial effects of continuous passive motion after total condylar knee arthroplasty. Ann R Coll Surg Engl. 1992;74(6):412–416. PMID:1471839
- Maloney WJ, Schurman DJ, Hangen D, Goodman SB, Edworthy S, Bloch DA. The influence of continuous passive motion on outcome in total knee arthroplasty. Clin Orthop Relat Res. 1990;(256):162–168. doi:10.1097/00003086-199007000-00023 [CrossRef] PMID:2364605
- Montgomery F, Eliasson M. Continuous passive motion compared to active physical therapy after knee arthroplasty: similar hospitalization times in a randomized study of 68 patients. Acta Orthop Scand. 1996;67(1):7–9. doi:10.3109/17453679608995599 [CrossRef] PMID:8615108
- O’DriscollShawn W, Nicholas J. Giori. Continuous passive motion (CPM): theory and principles of clinical application. Journal of Rehabilitation Research and Development. 2000;37.2:179.
- Postel J-M, Thoumie P, Missaoui B, et al. French Physical Medicine and Rehabilitation Society. Continuous passive motion compared with intermittent mobilization after total knee arthroplasty: elaboration of French clinical practice guidelines. Ann Readapt Med Phys. 2007;50(4):244–257.
- Ververeli PA, Sutton DC, Hearn SL, Booth RE Jr, Hozack WJ, Rothman RR. Continuous passive motion after total knee arthroplasty: analysis of cost and benefits. Clin Orthop Relat Res. 1995;(321):208–215. PMID:7497671
- Vince KG, Kelly MA, Beck J, Insall JN. Continuous passive motion after total knee arthroplasty. J Arthroplasty. 1987;2(4):281–284. doi:10.1016/S0883-5403(87)80060-8 [CrossRef] PMID:3430154
- Alkire MR, Swank ML. Use of inpatient continuous passive motion versus no CPM in computer-assisted total knee arthroplasty. Orthop Nurs. 2010;29(1):36–40. doi:10.1097/NOR.0b013e3181c8ce23 [CrossRef] PMID:20142693
- Chen LH, Chen CH, Lin SY, et al. Aggressive continuous passive motion exercise does not improve knee range of motion after total knee arthroplasty. J Clin Nurs. 2013;22(3–4):389–394. doi:10.1111/j.1365-2702.2012.04106.x [CrossRef] PMID:23020840
- Denis M, Moffet H, Caron F, Ouellet D, Paquet J, Nolet L. Effectiveness of continuous passive motion and conventional physical therapy after total knee arthroplasty: a randomized clinical trial. Phys Ther. 2006;86(2):174–185. doi:10.1093/ptj/86.2.174 [CrossRef] PMID:16445331
- Leach W, Reid J, Murphy F. Continuous passive motion following total knee replacement: a prospective randomized trial with follow-up to 1 year. Knee Surg Sports Traumatol Arthrosc. 2006;14(10):922–926. doi:10.1007/s00167-006-0042-9 [CrossRef] PMID:16489477
- MacDonald SJ, Bourne RB, Rorabeck CH, McCalden RW, Kramer J, Vaz M. Prospective randomized clinical trial of continuous passive motion after total knee arthroplasty. Clin Orthop Relat Res. 2000;380:30–35. doi:10.1097/00003086-200011000-00005 [CrossRef] PMID:11064970
- Maniar RN, Baviskar JV, Singhi T, Rathi SS. To use or not to use continuous passive motion post-total knee arthroplasty presenting functional assessment results in early recovery. J Arthroplasty. 2012;27(2):193–200.e1. doi:10.1016/j.arth.2011.04.009 [CrossRef] PMID:21752575
- Froimson MI, Rana A, White RE Jr, et al. Bundled payments for care improvement initiative: the next evolution of payment formulations: AAHKS Bundled Payment Task Force. J Arthroplasty. 2013;28(8)(suppl):157–165. doi:10.1016/j.arth.2013.07.012 [CrossRef] PMID:24034511
- Miller DC, Gust C, Dimick JB, Birkmeyer N, Skinner J, Birkmeyer JD. Large variations in Medicare payments for surgery highlight savings potential from bundled payment programs. Health Aff (Millwood). 2011;30(11):2107–2115. doi:10.1377/hlthaff.2011.0783 [CrossRef] PMID:22068403
- Bozic KJ, Ward L, Vail TP, Maze M. Bundled payments in total joint arthroplasty: targeting opportunities for quality improvement and cost reduction. Clin Orthop Relat Res. 2014;472(1):188–193. doi:10.1007/s11999-013-3034-3 [CrossRef] PMID:23649225
- Porter ME, Kaplan RS. How to pay for health care. Harvard Business Review. 2016;94.7–8:88.
- Lenssen TA, van Steyn MJ, Crijns YH, et al. Effectiveness of prolonged use of continuous passive motion (CPM), as an adjunct to physiotherapy, after total knee arthroplasty. BMC Musculoskelet Disord. 2008;9(1):60. doi:10.1186/1471-2474-9-60 [CrossRef] PMID:18442423
- Bruun-Olsen V, Heiberg KE, Mengshoel AM. Continuous passive motion as an adjunct to active exercises in early rehabilitation following total knee arthroplasty: a randomized controlled trial. Disabil Rehabil. 2009;31(4):277–283. doi:10.1080/09638280801931204 [CrossRef] PMID:18608367
- Pisters MF, Veenhof C, de Bakker DH, Schellevis FG, Dekker J. Exercise adherence and physical activity in patients with osteoarthritis of hip or knee: a randomized clinical trial comparing behavioral graded activity and usual exercise therapy. J Physiother. 2010;56:41–47. doi:10.1016/S1836-9553(10)70053-9 [CrossRef] PMID:20500136
- Sluijs EM, Kok GJ, van der Zee J. Correlates of exercise compliance in physical therapy. Phys Ther. 1993;73(11):771–782. doi:10.1093/ptj/73.11.771 [CrossRef] PMID:8234458
|Characteristic||CPM (n=14)||GAP-FLEX (n=13)||Total (N=27)|
| Median (range)||66.5 (53–77)||66.0 (50–77)||66 (50–77)|
| Female||7 (50%)||6 (46.2%)||13 (48.1%)|
| Male||7 (50%)||7 (53.8%)||14 (51.9%)|
| White||13 (92.9%)||13 (100%)||26 (96.3%)|
| Black||1 (7.1%)||0 (0%)||1 (3.7%)|
| American Indian or Alaska Native||0||0||0|
| Native Hawaiian or other Pacific Islander||0||0||0|
| Hispanic or Latino||3 (21.4%)||2 (15.4%)||5 (18.5%)|
| Not Hispanic or Latino||11 (78.6%)||11 (84.6%)||22 (81.5%)|
|Smoking history, No.|
| Current smoker, less than 1 pack per day||1 (7.1%)||1 (7.7%)||2 (7.4%)|
| Never smoked or has not smoked within the past 10 years||13 (92.9%)||12 (92.3%)||25 (92.6%)|
|Preoperative baseline active flexion ROM, mean±SD||108.7°±17.8°||110.5°±9.8°||109.6°±14.3°|
|Postoperative baseline maximum passive flexion on operating table (ROM), mean±SD||131.4°±6.0°||132.7°±4.8°||132.0°±5.4°|
|Outcome||CPM (n=14)||GAP-FLEX (n=13)||Difference||P|
| Baseline active ROM||108.7±17.8||110.5±9.8||+1.8||.374|
| Day 2||82.9±8.8||93.6±9.0||+10.7||.003|
| Day 14||103.4±10.0||109.0±8.4||+5.6||.065|
| Day 28 (endpoint)||109.8±7.8||118.3±9.5||+8.5||.009|
| Day 28 (endpoint)||1.21±2.01||1.15±1.96||−0.06||.459|
|Functional mobility, mean±SD|
| Day 28 (endpoint)||24.1±3.1||26.1±1.5||+1.9||.026|
|Optional therapy, No.|
| Yes||12 (85.7%)||10 (76.9%)||22 (81.5%)|
| No||2 (14.3%)||3 (23.1%)||5 (18.5%)|