|Home | About | Journals | Submit | Contact Us | Français|
Rehabilitation of patients following anterior cruciate ligament (ACL) reconstruction has undergone remarkable improvements over the past two decades. During this time, ACL research has been at the forefront of many orthopaedic and sports physical therapy clinics. With over 20 years of ACL rehabilitation experience (senior author) and prior collaboration with accelerated ACL rehabilitation pioneer K. Donald Shelbourne, the authors wish to present a unique perspective on the evolution of ACL rehabilitation.
Prior to the classic article by Paulos et al in 1981,1 literature on ACL rehabilitation was quite sparse. The basis for ACL rehabilitation at this time was founded in basic science studies conducted with animal models. In an effort to protect the graft, emphasis was placed on immobilization, extension limitation, restricted weight bearing, and delayed return to activity. Despite achieving good ligamentous stability, patients often experienced a spectrum of complications.
In 1990, Shelbourne and Nitz2 proposed an accelerated rehabilitation protocol following ACL reconstruction based on clinical experience. Their program emphasized delayed surgery, earlier range of motion and weight bearing, and full extension. As a result, patients experienced better clinical outcomes while maintaining knee stability.
The rehabilitation program presented in this paper is still largely based on the principles of the accelerated protocol. As evidence-based practice and the call for prospective, randomized clinical research continues, the continued progress in treating patients with this injury will be enhanced. Furthermore, clinicians are urged not to lose sight of the clinical reasoning that helped evolve the ACL rehabilitation process where it is today.
Rehabilitation of patients following anterior cruciate ligament (ACL) reconstruction has undergone remarkable improvements over the past two decades. In 1983, one author reported on “The Anterior Cruciate Ligament Problem,” indicating no ideal treatment for a patient with ACL disruption existed.3 Initial surgical treatment of ACL injuries resulted in a high incidence of complications, which led many authors to favor nonoperative treatment and conservative rehabilitation.3 As surgical techniques improved and surgical outcomes became more predictable, postoperative rehabilitation became the key variable in determining successful outcomes.
Prior to the classic article by Paulos et al,1 literature regarding ACL rehabilitation was scarce. Initial reports of rehabilitation of patients following ACL reconstruction consisted of a few general paragraphs at the end of an article regarding surgical procedures. From 1980 to 1985, ACL literature increased dramatically as the period produced more articles than the previous 80 years.4 Since this time, ACL research has been at the forefront of many orthopaedic and sports physical therapy clinics.
With over 20 years of ACL rehabilitation experience (senior author) and prior collaboration with accelerated ACL rehabilitation pioneer, K. D. Shelbourne MD, the authors wish to present a unique perspective on the evolution of ACL rehabilitation. The term accelerated rehabilitation will be used for the context of this paper to refer to the concept of a rehabilitation progression designed to allow early, yet safe return to activities following ACL reconstruction. The term traditional rehabilitaion will be used to describe the more conservative, time-based, protocols that were commonly used in the past. This commentary will provide a brief history of the basic science models that led to the traditional rehabilitation protocols, highlight rehabilitation models proposed by Paulos et al1 and Shelbourne and Nitz,2 provide evidence for the principles behind the accelerated rehabilitation program following ACL reconstruction, and re-emphasize key points to successful rehabilitation outcomes following ACL reconstruction. Although the impact of surgical technique and graft selection on the rehabilitation process is important, such topics are beyond the scope of this paper.
During the 1970's, basic science studies conducted with animal models provided the framework for the traditional rehabilitation model. This data was extrapolated and applied to humans. Application of animal studies should be done with caution due to the differences between the species. Many authors openly stated this limited applicability to clinical human cases.1,5,6 However, clinicians acted on the best available evidence at the time.
Great uncertainty existed as to the role of the ACL in knee joint stability and the long-term effects of knee instability.7 A classic study by Marshall and Olsson8 transected the ACL in 10 dogs and two dogs were used as a control group. The dogs were followed for up to 23 months. Macroscopic, histological, and microroentgenographic examinations revealed osteophytes progressively increasing in size for up to one year as well as proliferative and degenerative changes in the articular tissues. A close relationship existed between instability evaluated by an anterior drawer test and articular changes. The authors concluded that early stabilization was indicated in cases of ACL rupture. A biomechanical study by Butler et al9 sought to determine the importance of knee ligaments in resisting joint translation. They used 14 human cadaver knees secured to a load cell and moving actuator to measure the restraining forces against the anterior and posterior drawer tests. Ligaments were sectioned individually and the test repeated to determine the contribution to restraint. Test result indicated that the ACL provided up to 86% of restraint against anterior translation in the knee. As the function of the ACL and the need for stabilization became clearer, the number and type of surgical procedures increased.4
Following ACL disruption and reconstruction or repair, immobilization for an extended period of time was the standard form of treatment. Noyes and colleagues5,10 reported the functional properties of ligaments in monkeys. Wild primates were immobilized for eight weeks in total body plaster prior to undergoing mechanical testing in tension to failure under high strain-rate conditions. The results of testing on 100 knee specimens showed a significant decrease in ligament strength and stiffness following 8-weeks of immobilization. Two subgroups of monkeys underwent 5 and 12 month reconditioning periods prior to testing. Results showed incomplete recovery of ligament properties 5 months after resumed activity and strength properties required up to 12 months to return to normal. As a result, they suggested delayed return to activity for an extended period of time following immobility, delayed return to strenuous activity for 6 to 12 months, and prescribed protective measures during rehabilitation. The application to humans suggested an extended delay for return to activity rather than shortening the length of immobilization.
The vascular anatomy and healing process of the ACL were described in dogs. Arnoczky et al11 utilized microangiography, histology, and tissue-clearing techniques to analyze the normal vascular anatomy in eight dogs. They reported that the central portion of a normal canine ACL had decreased vascularity. Eight weeks following complete ACL transaction, spontaneous healing had not occurred in one animal. Alm et al12 studied the revascularization process following ACL reconstruction in 29 dogs. Through microangiography and histological study, they found that the original vascularization of the distal and middle portionss of the patellar tendon graft were preserved. The proximal and middle parts of the graft where the suture was attached were initially devoid of functioning vessels and had revascularized by 2 months. The structure of the graft resembled a normal ligament at 4 to 5 months.
Clancy and coworkers13 studied the vascularity of the patellar tendon graft in monkeys at 2, 3, 6, 9, and 12 months following ACL reconstruction. Microangiographic and histologic examinations were performed on one animal at each of the follow-up periods. They found that patellar tendon grafts in monkeys were revascularized after 8 weeks and resembled a normal ligament at 9 months and 12 months. Arnoczky et al14 studied the revascularization pattern of patellar tendon grafts in dogs at 2, 4, 6, 8, 10, 16, 20, 26, and 52 weeks postoperatively. Four animals underwent histological and tissue-clearing techniques at each of the follow-up periods. The authors described the graft undergoing phases of ischemic necrosis, revascularization, proliferation, and remodeling. The transplanted graft had intrinsic vessels by 8 weeks, was completely vascularized by 20 weeks, and had the histological appearance of a normal ligament at one year. Amiel et al15 studied the morphology of patellar tendon grafts in rabbits at 2, 3, 4, 6, and 30 weeks following ACL reconstruction. Histological and biochemical examination were performed on five animals at each of the follow-up periods. The grafts demonstrated a gradual assumption of the microscopic properties of the normal ACL. By 30 weeks postoperatively, collagen concentrations of the graft were the same as the normal ACL and cell morphology appeared ligamentous. The authors referred to this gradual process as “ligamentization.” The common theme during this time period was that vascularization of a transplanted graft required 8 weeks and ligamentization required up to one year.
Rougraff et al16 performed arthroscopic and histologic analysis of patellar tendon autografts following ACL reconstruction. The knees of 23 patients underwent arthroscopy and biopsy from 3 weeks to 6.5 years postoperatively. They observed that human patellar tendon autografts were viable as early as 3 weeks with exception of the central biopsy at 3 weeks. They detected increased neovascularity, nuclear morphology, and fibroblastic activity in the human grafts as compared to the necrotic stage observed in animals. Ligamentization required up to 3 years to complete.
To determine the strength of an ACL substitute, researchers studied the mechanical properties of various graft sources. Clancy et al13 studied the tensile strength of patellar tendon grafts in monkeys at 3, 6, 9, and 12 months postoperatively. Three animals at the first three follow-up periods and five animals at the final follow-up period underwent stress to failure testing. The results demonstrated patellar tendon grafts had regained 81% of their original tensile strength prior to transfer at 9 and 12 months following reconstruction and were 52% of the strength of the normal ACL at 12 months following reconstruction. The authors considered these results significant because Butler et al17 had demonstrated that a third of the patellar tendon in humans had 191% of the strength of the ACL. Noyes et al18 compared the mechanical properties of nine human ligament graft tissues obtained from young trauma victims (mean age 26 years). The tissues studied included the ACL, central and medial portions of the bone-patellar tendon-bone (BPTB), semi-tendinosis, fascia lata, gracilis, distal iliotibial tract, and the medial, central, and lateral portions of the quadriceps tendon-patellar retinaculum-patellar tendon. All tissues were subjected to high-strain-rate failure tests to determine strength and elongation properties. The BPTB graft was the strongest with a mean strength of 159% to 168% of that of an ACL. The strength of the substitute graft would theoretically affect the initiation of motion and strengthening activities during the rehabilitation process.
Controversy surrounded the safety of simple motion and other stresses as researchers attempted to identify strain imposed on the ACL during rehabilitation. Grood et al19 studied the biomechanics of knee extension and the effect of cutting the ACL in human cadavers. They reported increased anterior tibial translation during knee extension from 30° of flexion to full extension. Arms and colleagues20 studied ACL strain during knee ROM and simulated quadriceps contractions in human cadavers. Using a strain transducer, they showed that ACL strain decreased as the knee was passively flexed from 0° until 30°−35° where the ACL underwent minimal strain. Further flexion increased the strain to a maximum at 120 degrees. Isometric and eccentric quadriceps contractions significantly increased ACL strain through the first 45° of knee flexion while isometric contraction at flexion angles greater than 60° decreased ACL strain. Quadriceps activity beyond 60° was determined to be safe. The authors speculated that immobilization might not protect the graft if isometric quadriceps contractions occur. Henning et al21 used an in vivo strain gauge to study the load-elongation of the ACL during rehabilitation exercises. Two subjects with acute grade II ACL sprains were utilized and the results were scaled to an 80-pound Lachman test. Cycling produced 7%, single leg half squat produced 21%, normal walking produced 36%, quadriceps contraction against 20 pounds of resistance at 45° produced 50% and at terminal extension produced 121%, and downhill running produced 125% as much elongation as an 80 pound Lachman test. The authors recommended that knee extension not be performed through a full ROM during the first year following ACL reconstruction. Strain data gave clinical insight to the stress produced on the ACL during various rehabilitation activities. Clinicians used this information to avoid certain exercises and thus protect the healing ACL. However, there are no direct methods of knowing the limits of strain that are safe for a healing ligament or graft.
Rougraff and Shelbourne22 suggested that stresses to the healing tissue that remained below failure threshold would be beneficial and that rehabilitation programs designed to limit stresses may negatively affect ultimate outcome. This postulation was supported by Hannafin et al23 who performed an in vitro study on the effects of stress deprivation on canine tendon. Their results showed a significant decrease in tensile strength over 8 weeks. They suggested that stress may be necessary for optimal graft healing and collagen formation.
These basic science studies led to the belief that intra-articular graft healing was a long-term process that included a maturation phase in which the graft was necrotic and weak. In an effort to protect the graft, emphasis was placed on immobilization, extension limitation, restricted weight bearing, and delayed return to activity.
In 1981, Paulos et al1 published the specifics and rationale of their postoperative rehabilitation program for patients following ACL reconstruction (Figure 1). Although they openly stated that their rehabilitation program was based on preliminary findings,opinions and designed to protect all patients, many practicing clinicians quickly adopted this protocol.2,23,24 The rehabilitation program consisted of five phases that included maximum protection (12 weeks), moderate protection (24 weeks), minimum protection (48 weeks), return to activity (60 weeks), and activity and maintenance.
During the maximum protection phase, patients were placed in a cast, nonweight- bearing (NWB) for 6 weeks in 30° to 60° of flexion. Based on animal research, they estimated healing ligament strength at less than 50% by 12 weeks. Full weight-bearing (FWB) was not allowed prior to 16 weeks. Quadriceps activity was limited through the first 24 weeks to minimize risk to the ACL, while emphasis was placed on hamstring strengthening. Running began approximately 9 to 12 months after surgery when the operative leg achieved 75% strength of the normal leg. The authors recommended a minimum of 9 months to return to full activity with most patients requiring at least a year.
A 1980 international survey performed by Paulos et al1 revealed that 53% of responding knee experts initiated knee motion by 3 weeks. The authors were concerned that the early initiation of knee motion could disrupt attachment site fixation. Of those included in the survey, 75% recommended an immobilization position of 30-60 degrees. The mean time for FWB was 7.7 weeks. The authors cautioned progression to early weight-bearing due to animal studies that demonstrated early graft vascularization at 8 weeks. Full range of motion (ROM) was expected at 6 months by 88% of respondents and the mean time for maximum knee motion was 4.3 months. The majority (63% always, 22% sometimes) felt a brace should be used for protection. Most respondents allowed running by 6 months with the mean at 4.7 months. Mean time for return to full activity was 9.4 months.
Many researchers continued to study graft remodeling and revascularization as graft integrity and viability following ACL reconstruction remained a concern. Studies challenged the standard treatment of immobilization following ACL reconstruction and showed the beneficial effects of immediate joint motion.25,26 In turn, authors reported performing motion exercises sooner following reconstruction. Noyes et al27 reported that utilization of early motion avoided knee stiffness and promoted full knee extension following ACL reconstruction. Their early motion program utilized continuous passive motion (CPM) during hospitalization. Upon discharge a knee splint was worn which allowed an immediate arc of 0° to 90° of flexion and the patient used the opposite leg to assist motion for 10 to 15 minutes every hour. In 1987, Noyes et al28 studied the effects of early knee motion following open and arthroscopic ACL reconstruction. Eighteen patients with acute and chronic ACL deficiencies were randomized into two groups prior to surgery. The motion group started knee motion on the second postoperative day while the delayed motion group initiated motion on the seventh postoperative day. All other aspects of the rehabilitation program were the same. Results showed that CPM performed on the second postoperative day did not increase joint effusion or result in stretching of the ligamentous reconstruction as measured by a KT-1000 arthrometer at 6 months postoperatively. Although not significant, the early motion group also achieved increased mean knee extension and flexion values measured at 1, 2, 3, 4, and 12 weeks postoperatively. Despite the apparent benefits of early motion following ACL reconstruction, the authors were still concerned that utilization of a CPM after reconstruction would disrupt or loosen the graft.29
In 1986, Bilko et al30 published the results of a questionnaire taken at the ACL Study Group meeting in 1984. The survey results were compared to the results of the 1980 international survey by Paulos et al.1 Analysis of 44 returned questionnaires indicated that more surgeons immobilized the knee between 30° and 60° of flexion, yet the length of time immobilized decreased. Of those who responded, 48% were immobilized between 1 and 3 weeks compared to 21% who were immobilized between 5 and 8 weeks. Isometric exercises were not prescribed as often during the 1st week postoperatively, while the use of electrical stimulation and isokinetics during rehabilitation occurred more frequently. The earliest time to full weight-bearing ranged from the 3 to 4 week period to 16 weeks. Less than 7% indicated regular use of continuous passive motion. Full ROM was expected at 3 months by 18% and 6 months by another 68% following ACL reconstruction. Only one surgeon responded that the minimum time for return to full activity was 6 months, while 95% reported return within 10 months postoperatively. All respondents allowed return to full activity by one year. Only 25% did not recommend a brace for return to play.
Despite achievement of good ligamentous stability, patients often experienced a spectrum of complications that included patellofemoral symptoms, quadriceps weakness, and limited ROM.31–33 From 1982 to 1986, Sachs et al34 prospectively followed 126 patients who underwent ACL reconstruction and were immobilized in 30° of flexion for 3 weeks. At one-year follow-up, quadriceps weakness was defined as less than 80% bilaterally and was present in 65% of patients which correlated positively with flexion contracture and patellar irritability. Flexion contractures ≥ 5° were present in 24% of patients and patellofemoral pain occurred in 19% of patients. Sachs et al35 also published results from the San Diego Kaiser review series of 390 patients with ACL surgeries between 1983 and 1988. One year follow-up statistics revealed 3% of patients with postoperative graft impingement, 7% required manipulation, 20% had flexion contractures, 19% experienced patellofemoral pain, 62% demonstrated quadriceps weakness, 12% exhibited an effusion, and 10% required a secondary procedure within 1 year. To decrease the incidence of joint stiffness and flexion contracture, the authors recommended full ROM and no swelling at the time of surgery as well as immobilization of patients at 0° for 10 -14 days postoperatively.
Trends in ACL rehabilitation in the 1980's revealed earlier ROM and weight bearing.36 Disagreeement was predominant in regard to the initiation of weight-bearing, full ROM, rehabilitation exercises, and return to play. The complication rate remained high during this time but decreased with initiation of earlier motion. It is the senior author's opinion that the rehabilitation programs which were published largely emphasized open kinetic chain (OKC) exercises and hamstring strengthening.1,24,37
Rehabilitation of patients following ACL surgery at Methodist Sports Medicine Center initially followed many of the trends begun in the 1980's. In 1982, patients were placed in a cast for 6 weeks following ACL reconstruction. Due to flexion contractures, strict immobilization was replaced in 1983 with the immediate use of a CPM and a 30° removable splint. Like many clinics, the rehabilitation protocol was slightly modified from that used by Paulos et al.1 Patients did not weight bear until 6 weeks, attain full motion until 4 months, or return to activity until 9 to 12 months postoperatively. By 1985, patients were placed in a 0° postoperative splint. Consequently, motion problems decreased while stability remained unchanged. In 1985, the staff studied patient compliance and found that good clinical results, such as full ROM, strength, stability, and return to activity, were not necessarily correlated with subjectively reported patient compliance (unpublished data). In fact, patients who were noncompliant actually had better results than those who complied with the rehabilitation program. Subsequently, a new criterion-based rather than time-based rehabilitation protocol was adopted at Methodist Sports Medicine Center by the end of 1986.
In 1990, Shelbourne and Nitz2 published a clinically based article on accelerated rehabilitation of patients following autogenous bone-patellar tendon-bone (BPTB) ACL reconstruction. The accelerated program called for rapid advancement of goals and emphasized early full knee extension, quadriceps muscle leg control (Figure 1), soft tissue healing, and normalized gait pattern. Patients were not immobilized following surgery. On day 1, CPM was initiated and weight bearing as tolerated was allowed without crutches. Strengthening exercises were predominantly closed kinetic chain (CKC) and OKC quadriceps exercises were minimized. Patients typically returned to light sports activities by 2 months and full activity between 4 and 6 months following reconstruction. Subjective and objective follow-up evaluations were routinely performed, as were isokinetic and KT-1000 evaluations beginning 5 to 6 weeks postoperatively. Shelbourne and Nitz2 reported increased patient compliance, earlier return to normal function, decreased frequency of patellofemoral symptoms, and a significant decrease in the number of procedures required to obtain full knee extension.
Shelbourne and Nitz2 reported a retrospective comparison of follow-up data on 138 patients who performed a traditional rehabilitation program following ACL reconstruction from 1984 to 1985 and 247 patients from 1987-1988 who performed the accelerated rehabilitation program following ACL reconstruction. Subjective knee ratings were similar for both groups from the time of reconstruction to 2-year follow-up. Isokinetic quadriceps strength tests revealed a quicker return of quadriceps strength in the accelerated group at each follow-up period from 4 months to 1 year. Likewise, analysis of KT-1000 scores revealed equal to or better scores than the traditional group at each follow-up comparison from 4 months to 1 year, which indicated no loss in knee stability. Furthermore, 12% of patients who performed the traditional rehabilitation program required surgical intervention to achieve full extension compared to 4% of patients in the accelerated program.
The accelerated program was met with much resistance in the literature. Many authors were concerned that “aggressive” rehabilitation would lead to graft failure,14,38 inappropriate graft strain,39–43 or adversely affect articular cartilage.40 Several authors cited that there was no evidence to support the safety of activities such as early FWB, jogging and agility drills by 5 to 6 weeks, return to sport at 4 to 6 months,43–45 and were alarmed by the lack of long-term follow-up.38,39,43,46 Devita et al45 reported that gait mechanics were abnormal following accelerated rehabilitation while Hardin et al47 suggested that individuals with hyperlaxity had an increased risk for instability following accelerated rehabilitation. Beynnon and Johnson44 questioned the safety of accelerated rehabilitation citing the retrospective nature and possible bias as caution for clinical use.
Accelerated rehabilitation has been previously described in detail.2,48–52 The Methodist Sports Medicine Center rehabilitation program outlined in Figure 1 was largely based on the principles of the accelerated protocol.2 The goal of the Methodist Sports Medicine Center rehabilitation protocol had always been to minimize postoperative complications and return the knee to a normal state as quickly and safely as possible. The protocol continued to be adapted and changed based on clinical experience and the current findings in the literature. With the emergence of evidence-based practice (EBP), much of the accelerated rehabilitation program following autogenous BPTB ACL reconstruction had been well supported in the literature. The term “accelerated” rehabilitation may no longer be appropriate or necessary due to the shift in rehabilitation trends over the past decade.
The Methodist Sports Medicine Center rehabilitation protocol was divided into 5 phases: preoperative, early postoperative, intermediate postoperative, advanced rehabilitation, and return to activity. The time frames presented with each phase were general in nature and based on clinical experience. Progression of patients between phases of rehabilitation were individualized decisions determined by achievement of goals and clinical reasoning.
Phase I rehabilitation began immediately following ACL injury.49 The goals of the preoperative period were to reduce swelling and restore normal motion, gait, and strength prior to surgery. Common exercises for flexion ROM included heel slides (Figure 2), wall slides, and active/assistive flexion. Exercises for extension ROM included heel props (Figure 3), prone hangs, and towel extensions. Once full ROM with minimal swelling was obtained, CKC strengthening was begun with exercises such as leg press, _ squats, step downs (Figure 4), stationary bicycle, and step machines. This time frame also allowed for mental preparation and education of surgery and postoperative rehabilitation. Surgery was scheduled once these goals were attained. The patient underwent preoperative testing for postoperative comparison that included bilateral ROM, KT-1000 ligament arthrometry, isokinetic strength evaluation, and a single leg hop test on the non-involved extremity.50–52
De Carlo et al50 reported a retrospective study of 169 patients who underwent autogenous BPTB ACL reconstruction for acute ACL injury. Patients who had reconstruction within the first week after injury had a significantly increased incidence of arthrofibrosis compared to patients who had reconstruction delayed 21 days or more. Patients who had delayed reconstruction also had better ROM and isokinetic strength scores at 13 weeks following reconstruction. Shelbourne and Foulk53 performed a retrospective review of 143 patients who underwent autogenous BPTB ACL reconstruction within 3 months of injury. Patients were divided into two groups based on when they elected to have surgery. Group 1 delayed surgery a mean of 40 days after injury while group 2 had surgery a mean of 11 days after injury. Results of isokinetic testing determined that patients who delayed ACL reconstruction had significantly better quadriceps strength at 2 and 4 months postoperatively than those who underwent acute surgery. Cosgarea et al54 and Wasilewski et al55 have also confirmed earlier return of motion and strength following delayed ACL reconstruction. Two studies have reported that timing of surgery had no effect on extension loss.56,57 However, both authors defined full extension as 0° rather than the ROM prior to surgery.56,57 Regardless of the time from injury, the senior author believes the condition of the knee prior to reconstruction (minimal swelling, full hyperextension, near normal strength, and normal gait) were directly correlated with the ability to regain early motion and strength postoperatively.
Udry et al58 studied psychological readiness of the patient undergoing ACL reconstruction. They found that adolescents reported higher preoperative mood disturbance levels compared to adults. However, adolescents also reported higher levels of psychological readiness for surgery than adults. Shelbourne and Rask59 reported that patients who had a second ACL procedure for the opposite knee experienced a smoother transition following reconstruction than with the initial procedure. For this reason, a thorough preoperative education was incorporated for all patients. These factors are important to consider because of the effort, motivation, and understanding required of postoperative rehabilitation.
Immediately following surgery, the reconstructed knee was placed in a cold compression cuff with the leg in a CPM machine. Range of motion, quadriceps control, and weight-bearing as tolerated were initiated the day of surgery. The goals for the first postoperative week were to control swelling, obtain full hyperextension, increase passive knee flexion to at least 110°, and establish good quadriceps leg control. The cold compression cuff remained on the knee at all times except when patients performed ROM exercises. The patient remained lying down as much as possible except when exercises were performed or for personal hygiene. Extension ROM exercises, such as heel props and towel extensions, were performed for 10 minutes hourly during the day. Flexion was initiated with the knee rested in the CPM machine set to 110° and held for 10 minutes, four times daily. Early leg control was accomplished with quadriceps setting, straight leg raises, and active knee hyperextension.
By the end of the second postoperative week, the patient should have been able to demonstrate normal gait, full passive extension, 130° of flexion, and good quadriceps leg control. During this week, patients added prone hangs (1-3 lbs could be added if extension was tight) to their daily ROM exercises. Patients were encouraged to stand with their weight over their reconstructed knee with the quadriceps contracted, which locked the knee into full hyperextension. Gait training was necessary if the patient ambulated with a limp or without a normal heel-to-toe pattern. If the patient had full knee hyperex-tension and ambulated normally, strengthening exercises could be initiated which included seated knee extension from 90° to full terminal knee extension and bilateral half squats.
Shelbourne et al60 performed a prospective trial which compared the effectiveness of different methods of postoperative cryotherapy to decrease pain in 400 patients following autogenous BPTB ACL reconstruction. Patients who used a cold compression device had a significantly shorter hospitalization stay compared to patients who used a thermal blanket or ice bag. They used significantly less oral and injectable narcotics compared to patients who used an ice bag. Noyes et al61 conducted a prospective study of early motion versus delayed motion exercises in 18 patients following ACL reconstruction. Subjects in the early motion group began CPM on the second postoperative day while subjects in the delay motion group were braced in 10° of extension and began CPM on the seventh postoperative day. The results showed no deleterious effects of early motion with regard to knee laxity, joint effusion, hemarthrosis, ROM, use of pain medication, and length of hospital stay. The use of cold compression, CPM, and early active motion allowed for elevation of the leg, patient comfort, and predictable return of motion.
Initially, many authors were hesitant to attain full extension in the early postoperative period.62 These concerns were based on biomechanical studies that showed maximal flexion and extension of the knee caused increased stress on the intact ACL.63 However, many authors had reported that gaining extension immediately postoperatively decreased the frequency of flexion contractures.2,27,28,34,54,64 Rubinstein et al65 reviewed the effects of restoring full knee hyperextension immediately following autogenous BPTB ACL reconstruction. Subjects were grouped according to the degree of hyperextension. Group 1 consisted of 97 patients who hyperextended an average of 10° (8°-15°) and group 2 consisted of 97 patients who hyperextended an average of 2° (0 - 5°). No significant differences in KT-1000 arthrometer manual maximum side-to-side scores between groups were found. The authors determined that restored full knee hyperextension immediately postoperatively did not adversely affect stability of the knee.
Several authors had voiced concern that early weight-bearing may have caused excessive forces that harm the graft or fixation and suggested 4 to 6 weeks of crutches to allow for bone healing.43,62,66,67 However, Arnoczky68 reported that a biologic graft was the strongest the day it was placed inside the knee. A prospective, study by Tyler et al69 sought to determine the effect of immediate weight-bearing after autogenous BPTB ACL reconstruction. Forty-nine subjects were randomized into two groups. Group 1 underwent immediate weight-bearing as tolerated while group 2 was non-weight-bearing for 2 weeks. Results showed that immediate weight-bearing after ACL reconstruction resulted in a lower incidence of anterior knee pain, greater vastus medialis oblique electromyography activity, and no effect on knee stability at a mean follow-up of 7.3 months.
The third and fourth week following reconstruction was the intermediate postoperative phase. During this period, strengthening was initiated cautiously as full ROM was obtained. Strengthening progressed as long as minimal swelling and full ROM were maintained. Exercises were predominantly unilateral, high repetition/low resistance, and CKC exercise during this period and included step downs, leg press, leg extension, and half squats. At the end of 4 weeks, patients underwent passive ROM testing and completed their first postoperative isokinetic strength evaluation and KT-1000 ligament arthrometer tests.
Strain studies indicated that CKC exercises allowed increased muscle activity without subjecting the ACL to increased strain values.70–72 A prospective study by Bynum et al73 compared OKC versus CKC exercises during rehabilitation following authogenous BPTB ACL reconstruction. Ninety-seven patients were randomized to the OKC and CKC protocols. Results at a mean follow-up of 19 months demonstrated that CKC exercise following ACL reconstruction resulted in less patellofemoral pain and better subjective scores than OKC exercise. Subsequently, the authors reported using CKC exercise exclusively following ACL reconstruction. A prospective study by Mikkelsen et al74 compared CKC versus combined CKC and OKC exercise initiated 6 weeks after ACL reconstruction. Forty-four patients were randomized into the two groups. Follow-up at 6 months indicated that the addition of OKC exercise produced a significant improvement in quadriceps strength, earlier return to sport, and no increased KT-1000 measurements. Although caution was used with full arc OKC exercise, the Methodist Sports Medicine Center protocol included integration of both OKC and CKC exercises.
Weeks five through eight comprised the advanced rehabilitation phase. The emphasis of this phase was increased strength and initiation of early sports activities. The patient continued to maintain full ROM and advanced strengthening to low repetition/high resistance as indicated. Once patients demonstrated 70% quadriceps strength via isokinetic testing, they performed light agility drills and proprioceptive activity that included a running progression, lateral slides, crossovers, and single leg hopping. If a joint effusion was present, it was carefully monitored as activity increased. An activity-specific functional progression, such as shooting baskets or dribbling a soccer ball, was initiated near the end of this period. At the end of 8 weeks, the patients were evaluated to assess ROM, tested with the KT-1000 ligament arthrometer, performed an isokinetic strength evaluation, and completed a subjective questionnaire.
In 1993, Barber-Westin and Noyes39 reported serial KT-1000 measurements on 84 patients following BPTB allograft ACL reconstruction and controlled rehabilitation for chronic ACL deficiency. Arthrometer measurements were obtained on each patient for at least 2 years following surgery. Of those patients with abnormal anterior-posterior displacements greater than 2.5 mm, 86% were first detected during the intensive strength training or return to sports phases of rehabilitation. In 1999, Barber-Westin et al40 reported a subsequent observational study of 142 patients following ACL reconstruction that used a rehabilitation program similar to the previous study. However, this group of subjects used a BPTB autograft rather than an allograft. They found no association between abnormal displacements and the phase of rehabilitation.
Shelbourne and Davis75 followed 603 patients who underwent autogenous BPTB ACL reconstruction and participated in a sports agility program at a mean of 5.1 weeks. These patients were evaluated to determine if program effected knee stability. Patients were required to have full hyperextension, knee flexion to 120°, and at least 60% quadriceps strength compared to the normal leg. The KT-1000 manual maximum arthrometer scores revealed that 92.7% of patients at a mean of 5 weeks and 93.2% of patients at a mean follow-up of 24 weeks had displacement differences of 3 mm or less. The results showed that early return to sports agility activities did not compromise graft integrity measured 24 weeks following ACL reconstruction.
Return to activity was very individualized and was designed to match the patient's goals. The patient continued to increase strength and increase the intensity and duration of athletic activities. A functional progression (Figure 5) that followed a half to three-quarter to full speed progression of sport-specific activities was incorporated in this phase. The patient achieved 85% quadriceps strength and completed a functional progression program prior to return to full athletic activity. While some patients returned to activity as early as 2 months, typically patients returned to full activity between 4 and 6 months after ACL reconstruction.
Many authors continued to base return to activity guidelines on histological studies that reported full maturation and required 12 months to complete.76 However, Rougraff et al16 reported that ligamentization could require up to 3 years to complete. Glasgow et al77 studied the effects of early (5 months) versus late (9 months) return to vigorous cutting activities on outcome in 64 patients following patellar tendon autograft ACL reconstruction. Return to vigorous activity was based on a minimum of 8 weeks postoperation, negative Lachman test, absence of effusion, and patient desire to return. At a mean follow-up of 46 months, no differences were found in KT-1000 scores, subjective evaluations, or isokinetic strength. Interestingly, in a review of 1288 patients who underwent autogenous BPTB ACL and accelerated rehabilitation, Shelbourne and Davis75 reported that more patients tore their normal, contralateral ACL (4.4%) than their reconstructed ACL (2.4%). They proposed that graft failure was not the result of a weakened graft, but rather the consequence of normal return to sport.
In 1995, Shelbourne et al78 reported KT-1000 manual maximum difference scores in a 2 to 6 year follow-up of 209 patients after autogenous BPTB ACL reconstruction and accelerated rehabilitation. The mean KT-1000 score was 2.06 mm at full ROM and 2.10 mm at a mean 2.7 year follow-up. In 1997, Shelbourne and Gray79 reported objective data on 806 patients and subjective data on 948 patients in a 2 to 9 year follow-up after autogenous BPTB ACL reconstruction and accelerated rehabilitation. Of those patients who underwent acute reconstruction, the mean manual maximum KT-1000 arthrometer difference was 2.0 mm with 90% of patients less than or equal to 3 mm and 98% of patients less than 5 mm of laxity. No joint space narrowing was seen in 94% of patients, isokinetic quadriceps evaluation revealed 94% strength, mean motion was 5° of hyperextension and 140° of flexion, and mean subjective modified Noyes questionnaire87 score was 93.2 out of 100 possible. In 2000, Shelbourne and Gray79 reported on the effects of meniscus and articular cartilage status on autogenous BPTB ACL reconstruction and accelerated rehabilitation in a 5 to 15 year follow-up. Of those patients with both menisci present and normal articular cartilage at the time of surgery, 97% had normal or near normal radiographs. Based on these findings, evidence supported that accelerated rehabilitation following autogenous BPTB ACL reconstruction produced excellent long-term results without affecting long-term stability.
Trends in ACL rehabilitation in the 1990's revealed remarkable changes compared to the 1980's. Many authors began adopting protocols similar to the accelerated program.80–85 The rehabilitation programs were characterized by preoperative rehabilitation, immediate ROM and weight bearing, full passive knee extension, and functional exercise. Meanwhile, some authors continued to share concerns regarding the wide scale use of these new protocols, particularly in specific patient groups or with specific graft sources.40,45,46
Recently, the buzzword in the physical therapy profession has been “evidence-based practice.” Evidence based practice is a very positive trend that may ultimately result in improved quality and effectiveness of patient care. Sackett et al86 defined EBP as “the integration of best research evidence with clinical expertise and patient values.” Evidence based practice could be the trend that defines ACL rehabilitation in the 2000's.
Several authors41,44,70,87 have published well-conducted research on the strain behavior of the ACL during common rehabilitation activities. A comprehensive database has been compiled based on peak strain values in which the authors used to design rehabilitation programs to be compared in a prospective, randomized, double-blind trial. The results of these studies will help delineate rehabilitation programs that are safe for a healing ACL graft. A need exists for prospective, randomized, blinded clinical trials to compare accelerated rehabilitation with more conservative rehabilitation before accelerated rehabilitation can be considered safe and appropriate.88
Recently, Beynnon et al89 reported the results of a prospective, randomized, double-blind study comparing accelerated versus nonaccelerated rehabilitation in 22 patients following BPTB ACL reconstruction. The rehabilitation programs were based on their previous work of ACL strain data during rehabilitation activities. Exercises that had been shown to produce significant strain to the ACL were initiated earlier in the accelerated program and delayed in the nonaccelerated program. Exercises that did not produce significant ACL strain were initiated in both rehabilitation programs during the same time frame. The accelerated program, characterized by early unrestricted weight-bearing and early use of quadriceps-dominated exercises, lasted 19 weeks and return to sports was possible by 24 weeks while the nonaccelerated program lasted 32 weeks and return to sports was possible also at 32 weeks. At 2-year follow-up, their results demonstrated no difference in anterior knee laxity between accelerated and nonaccelerated rehabilitation. The authors also found that both programs produced the same outcomes in clinical assessment, patient satisfaction, functional performance, and articular cartilage metabolism. Furthermore, total compliance measured at the end of each program was significantly less in the nonaccelerated group.
Evidence based practice has not been limited to prospective, randomized, blinded clinical trials.90 While the authors agree that prospective, randomized, blinded clinical trials are the gold standard and large studies of this nature are required for best evidence practice, the difficulty most clinicians face in performance of such studies must be acknowledged. Prospective long-term outcome studies may also be conducted to gain insight into the effectiveness of clinical intervention.
To date no studies have been published that have determined conservative rehabilitation following ACL reconstruction to have produced better outcomes or long-term stability than those reported with accelerated rehabilitation. Therefore, the current evidence supports the use of the more physiologic progression following BPTB ACL reconstruction.
While research evidence has been a very important part of EBP, it is the senior author's opinion that clinical expertise and patient values are equally important components of EBP for quality patient care. Salter91 reported that the biological concept of CPM for synovial joints was based on clinical observation and deduction. In 1970, the concept of CPM was introduced which was contrary to the initial thought process of joint immobilization for disease or injury.91 The evolution of the Methodist Sports Medicine Center rehabilitation protocol following ACL reconstruction was based on clinical experience and listening to patients.2,49–51 In 1990, accelerated rehabilitation was the antithesis of traditional rehabilitation following ACL reconstruction.2
The clinician must utilize clinical reasoning skills and individualize the care of each patient. No specific exercises or parameters exist for exercise intensity or duration that have been proven to lead to successful outcomes. Guidelines for early application of strain to the healing ACL have not been published. The Methodist Sports Medicine Center rehabilitation protocol the authors have presented has adhered to the basic principles of rehabilitation. Patients increased activity if they had attained full ROM, exhibited minimal effusion and pain, had a normal gait, and demonstrated good leg strength measured isokinetically (quadriceps deficit of ≤30%). The condition of the knee dictated rehabilitation. Patients were not forced to return to activity. Only when the patient was physically and mentally ready was return to activity considered. In addition, the use of a functional progression program allowed the patient and the physical therapist, athletic trainer and, in rare instances, the coach to determine if an athlete was ready to advance.
Many researchers have attempted to replicate the accelerated Methodist Sports Medicine Center rehabilitation protocol45,46,91 or currently utilize a similar protocol. For this reason, the authors felt that it was important to clear some misconceptions and re-emphasize some key points of the Methodist Sports Medicine Center rehabilitation protocol.
The preoperative period was vitally important for successful outcome following ACL reconstruction.49,50,52,93 Patients were required to have full ROM including hyperextension, minimal effusion, good quadriceps strength via isokinetic testing, and normal gait prior to reconstruction. Once these goals were met, surgery was scheduled at a time that was convenient for the patient to allow restricted activity during the first postoperative week.
The emphasis of the first postoperative week was the minimization of swelling.49,59,60 If swelling could be prevented, motion and strengthening would not be inhibited. Although immediate full weight bearing as tolerated with crutches was allowed, activity was not unrestricted. The patients were instructed to remain supine with the leg in the CPM machine except when performing exercises or personal hygiene.50 The ability to prevent swelling during the first postoperative week greatly impacted the progression of return to activity.
Restoration of motion should be aimed to achieve motion equal to the opposite extremity. Normal motion was often thought of as 0° to 135°.46,56,57,66,89 However, a study by De Carlo and Sell94 of 889 preseason athletes found that 96% of individuals demonstrated some degree of hyperextension. The mean ROM was 5° of hyperextension and 140° of flexion for males and 6° of hyperextension and 143° of flexion for females. If the patient had not achieved full ROM, especially hyperextension, equal to the opposite side, the return of normal gait, function, and knee biomechanics would have been inhibited. The importance of obtaining full hyperextension postoperatively has been well documented. 2,16,27,28,34,54,64
The Methodist Sports Medicine Center rehabilitation protocol was criterion-based.49 The time frames given were used as guidelines and were not absolute. Advancement to the next phase depended on the condition of the knee and completion of the goals of the previous phase. The initial phases of the program were very similar for all patients in an attempt to restore normal motion, gait, and strength. The latter phases of the program were much more individualized in an effort to return patients to their previous level of function.
Over the past two decades, rehabilitation of a patient after an ACL injury has made a dramatic shift toward better patient outcomes and quicker return to activity. In their respective times, the traditional and accelerated rehabilitation models have both given clinicians a sound framework for treating patients as well as stimulated further research. A solid base of evidence exists in the literature to support accelerated rehabilitation as both safe and effective. As EBP and the call for prospective, randomized clinical research continues, the continued progress in treating this injury is exciting. Furthermore, clinicians are urged not to lose sight of the clinical reasoning and deduction that assisted in the evolution of the current science of ACL rehabilitation.
The senior author would like to acknowledge the nearly two decade long collaboration with K. Donald Shelbourne, MD. This association was invaluable in gaining clinical reasoning skills and research experience relating to effective management of patients with various knee pathologies.