Pain has been a problem since the inception of the CPM.  Pain reduces patient compliance, and patients resent being tethered to a painful machine. To counteract this, O’ Driscoll, who advocated only allowing patients out of bed to use the bathroom, argued that patients should receive all the pain medication required to keep them comfortable during more than 20 hours a day being attached to the machine.

The arc of the machine is a major setting that proponents argue is a key to the machine’s reputed efficacy, the patient’s leg does not circumscribe the arc of the machine, the leg often experiences less than 70% of the machine’s arc.

Ironically, no study has ever established optimal CPM usage, and because the CPM does not enhance ROM in the modern world of knee rehabilitation—whatever duration, angle, speed, arc etc. that is prescribed is equally ineffective– whether it is 24 hours of use/day or 30 minutes/day—the machine is equally ineffective at all settings.

Finally, while the leg is being flexed, patients are suffering the ill effects of too much bedrest. Modern norms of treatment have reduced hospital stays, get patients ambulating the day of surgery and argue for a more active recovery, a pattern the military established in WWI. If the decision to use or not use the CPM was based on quality of supporting evidence (e.g., level 1), then it would no longer be used today.

Coutts et al., (1982), the first person to use a CPM following a TKA. He argued that patients did not mind 20 to 24 hours of enforced bedrest, because “patients were introduced to CPM and the mechanical device and its intended effect long before their total knee surgery…by time the patient awakened in the recovery room, the patient was well acquainted with the device”.²¹ However, in 1989 Ritter reports patients resented their mobility being restricted for 20 to 24 hours a day while they used the CPM. Patients were not compliant because they resented the enforced immobility ^82,92,93,102 and there were other problems. Bed rest promotes: contractures and loss of normal motion to joints, atrophy of muscles (15% loss per week of inactivity), calcium is lost from bones, the heart is deconditioned, poor circulation, increased infection in the urinary tract, increased ulcers, and increased depression and anxiety.^{47-5082,77,107,79,94}  (Table 2).

Dr. Coutts and those that followed him compared the CPM to immobilization (Table 4).  Coutts did not include a treatment of allowing motion (active) nor early ambulation without the CPM.  The most direct way to determine if the CPM has any efficacy is to add a third treatment group, “nothing—no CPM”.  This did become the control, decades after Salter & Coutts did their seminal work. In fact, most modern studies do use “nothing” as the control. It is immobilization that has been dropped from recent studies as it is now rightly out of favor (Table 5).

Compared to immobilization the CPM appeared to have a huge effect (Table 2). However, when the control group changed to no CPM, the CPM had little or no effect—this is particularly true of Level I studies (Table 3).  We can now reinterpret the results of earlier studies. Immobilization is harmful. This does not mean that treatment with the CPM was helpful. It merely means that some motion is helpful. In humans, not being tethered to a CPM, allows enough motion that there is no difference between the CPM and no CPM group (Table 4).  In table 8, we have examined the incidence of MUAs. Again, when the control was immobilized legs, the CPM appeared efficacious. However, when the control is no CPM there is little direct evidence that indicates that the CPM has any positive effect. In table 6 we have shown the change that have occurred over time in the length of hospital stay, the amount of immobilization, how many hours a day the CPM was used and the discharge range of motion. We have also reviewed how physical therapy protocols have changed over time and summarized the level of evidence that supports or refutes the use of the CPM (Table 9).

The single, most obvious overall trend is that patients are spending less time in the hospital, less time in bed, they ambulate the day of surgery, hospital physical therapy protocols have less immobilization and more active motion. Yet, despite this consistent movement towards the military’s position of active motion, the CPM remains commonly prescribed and this is an anathema to the modern trends. Why must patients take to their beds for protracted periods of time (Table 7) which achieve nothing and ill serves the patient as established above?

Early CPM studies also suffered from the low quality of design. Many early CPM studies were level 3 due to being retrospective^{66-72, 77, 79-83,98}, and non-randomized^{66-72,76-77,80-81,87-90,103}, nor blinded^78,94. However, prior to the 1970s doctor’s opinion and experience determine not only what interventions would be prescribed but also a patient’s diagnosis. Thus, a patient’s safety and the procedure used depended upon the surgeon’s opinions and beliefs.⁴⁸ (Table 3 & 7). Hence, the popularity of the CPM was not based upon results.

Conclusion

In the end, what are we left with? There is the compelling theory: Moving the joint should reduce the blood and edema in the joint and reduce adhesions. But, when studies that use no CPM as a control reach the same range of motion as those that used the CPM, we have to question Salter’s original assertion, as articulated by O’Driscoll that, “He (Salter) reasoned that because immobilization is obviously unhealthy for joints, and if intermittent movement is healthier for both normal and injured joints, then perhaps continuous motion would be even better.” This assertion is false, because patients without the CPM gain the same ROM as patients with a CPM. The motion need not be continuous; it can be episodic throughout the day. Without continuous motion, there is no logical reason why the motion, must of necessity be passive. Passive motion may initially help with neuromuscular re-education but as that re-education proceeds active motion is not only possible but is, as the Surgeon General’s circular 46 noted… “desirable”.  There is no logical reason to pay to keep patients in bed to their determent.

Millions of patients have had to endure a CPM. Some may have found it comforting, others found it painful, and it robbed them of their freedom. Moreover, it forced them to stay in bed with all the debilitating effects of bedrest, not to mention that it did not reduce the arthrogenic muscle inhabitation (AMI) and so left them with a deep strength deficit (that are often not erased even a decade after their TKA.^135-139 Dr. Rusk was and is correct, “that men did not get ready for full duty by playing black jack or listening to music.” He noted that in recovery time, (which was immobilization), patients were deconditioning and bored”.¹

As we have argued above, immobilization is not the only means by which decondition and boredom are produced. Lying in bed for hours a day appears sufficient. Total Knee Arthroplasties are well on their way to becoming ambulatory surgeries. If this is to be the case, then new ways of restoring range of motion, strengthening muscles and reducing AMI must be found. These new methods must actually work. Insurers and the government should stop paying for CPMs. The machine, however well-intentioned, has been proven unsuccessful, and tax payers and insurers need not pay because we are unwilling to break old habits.