Introduction: I always aim for neutral mechanical axis alignment. My principles of a successful TKA are proper alignment in all 3 planes, soft tissue balance in extension first, flexion gap balancing by parallel to tibial cut technique, maintenance of joint line, correct sizing of femoral component, and proper cement fixation.

Long-term Survivorship: There is long-term data that supports the efficacy and durability of the neutral position of proximal tibial cut. Over a 20-year follow-up there was a 92.6% success rate in my study. Other authors have found similarly successful survivorship for mechanical failure.

Balance Technique in TKR: My technique to balance the knee is a balance extension gap first, which requires medial soft tissue balancing. Next, I balance the flexion gap parallel to the tibial cut.

Our Results: In one study, I examined the clinical and radiographic data of 68 varus knees. Average post-operative mechanical alignment was 0 ± 3 degrees. There were no outliers which displays the reproducibility of the technique. This is the method of choice in the hands of most surgeons.

In primary total hip replacements there are numerous options available for providing hip stability in difficult situations (i.e. Down's syndrome, Parkinson's disease).

However, in the revision situation in general and in revision for recurrent dislocation specifically, it is important to have all options available including dual mobility constrained liners in order to optimise the potential for hip stability as well as function of the arthroplasty. Even with the newer options, available dislocation rates of higher than 5% have been reported in the first two years following revision surgery at institutions where high volumes of revision surgery are performed. Because of the deficient abductors, other soft tissue laxity and the requirement for large diameter cups, revision cases will always have more potential for dislocation. In these situations in the lower demand patient and where, a complex acetabular reconstruction that requires time for ingrowth before optimal implant bone stability to occur isn't present, dual mobility with constraint has provided excellent success in terms of preventing dislocation and maintaining implant construct fixation to bone at intermediate term follow-up. Hence in these situations dual mobility with constraint remains the option we utilise. We are also confident in using this device in cases with instability or laxity where there is a secure well-positioned acetabular shell. We cement a dual mobility constrained liner in these situations using the technique described below.

Present indication for dual mobility constrained liners: low demand patient, large outer diameter cups, instability with well-fixed shells that are adequately positioned, abductor muscle deficiency or soft tissue laxity, multiple operations for instability

Technique of cementing liner into shell: score acetabular shell if no holes, score liner in spider web configuration, all one or two millimeters of cement mantle

Results: Constrained Dual Mobility Liner – For Dislocation: 56 Hips, 10 year average follow-up, 7% failure of device, 5% femoral loosening, 4% acetabular loosening. For Difficult Revisions: 101 hips, 10 year average follow-up, 6% failure of device, 4% femoral loosening, 4% acetabular loosening. Cementing Liner into Shell: 31 hips, 3.6 year average follow-up (2–10 years), 2 of 31 failures.

The United States is in the midst of an opioid epidemic, with the World Health Organization reporting that American's consume 99% of the world's supply of hydrocodone and 83% of the world's oxycodone. Additionally, pre-operative opioid use has been associated with worse clinical outcomes and higher rates of complications following total knee arthroplasty (TKA). This is especially important in the TKA population given that approximately 15% of patients are either dissatisfied or very dissatisfied at least one year after their TKA procedure. Given the concerning rise in opioid use the American Academy of Orthopaedic Surgeons (AAOS) has recently released an information statement with practice recommendations for combating this excessive and inappropriate opiate use. However, little is known regarding peri-operative opioid use for TKA patients. Therefore, the purpose of this study was to: 1) identify rates of opioid use prior to primary TKA, 2) evaluate post-operative trends in opioid use throughout the year following TKA and 3) identify risk factors for prolonged opioid use following TKA.

Overall, 31% of TKA patients are prescribed opioids within 3-months prior to TKA; this percentage has increased over 9% during the years included in this study. Pre-operative opioid use was most predictive of increased refills of opioids following TKA, however, other intrinsic patient characteristics were also predictive of prolonged opioid use. These characteristics remained predictive after controlling for opioid user status. The increasing rates of opioid prescribing prior to TKA are concerning, especially given literature concluding opioids have minimal effect on pain or function in patients with osteoarthritis and pre-operative opioid use is associated with poor outcomes and more complications following TKA. This data provides an important baseline for opioid use trends following TKA that can be used for future comparison and identifies risk factors for prolonged use that will be helpful to prescribers as the AAOS works to decreased opioid use, misuse and abuse within the United States.

Moderately to highly crosslinked UHMWPEs have functioned for at least a decade with dramatic reduction in wear volumes in THA. This wear reduction has been associated with a markedly reduced incidence of radiographic osteolysis. However, CT studies have demonstrated that osteolysis is not completely eliminated.

There, however, are still questions which include: Is cost for further improvements warranted?; Is 10 years long enough to assure that no clinically relevant osteolysis occurs, especially in younger patients?; Do we have any data demonstrating improvement in revision scenarios?; With high levels of crosslinking (requiring more radiation) some fractures have been demonstrated at the region of the locking mechanism of the liner to shell. Will this prevalence increase? These materials are softer and can cause quicker crack propagation than conventional polyethylene.; Do better locking mechanisms need to be developed to prevent fracture problems that have been demonstrated in the present generation cementless designs?; Do we need more information as to the optimal counterface choice (cobalt chrome, ceramic, oxinium)?; Can hip results be extrapolated to the knee where fatigue failure is a major problem both on the bearing surface and with the locking mechanism?; Is the oxidation we are beginning to see on the surface of retrieved liners (thought to be related to lipids from the synovium and cyclical loading) the tip of the iceberg?

I too am encouraged by the mid-term results of crosslinked polyethylene. Our own data supports it. However, we must keep in mind the questions outlined.

In primary total hip replacements there are numerous options available for providing hip stability in difficult situations i.e. Down's syndrome, Parkinson's disease. However, in the revision situation, in general, and in revision for recurrent dislocation situations specifically, it is important to have all options available including dual mobility constrained liners in order to optimise the potential for hip stability as well as function of the arthroplasty. Even with the newer options available dislocation rates of higher than 5% have been reported in the first two years following revision surgery at institutions where high volumes of revision surgery are performed [Della Valle, Sporer, Paprosky unpublished data]. Because of the deficient abductors, other soft tissue laxity and the requirement for large diameter cups, revision cases will always have more potential for dislocation. In these situations in the lower demand patient and where, a complex acetabular reconstruction that requires time for ingrowth before optimal implant bone stability to occur isn't present, dual mobility with constraint has provided excellent success in terms of preventing dislocation and maintaining implant construct fixation to bone at intermediate term follow-up. Hence in these situations dual mobility with constraint remains the option we utilise. We are also confident in using this device in cases with instability or laxity where there is a secure well-positioned acetabular shell. We cement a dual mobility constrained liner in these situations using the technique described below.

Present indication for dual mobility constrained liners: low demand patient, abductor muscle deficiency or soft tissue laxity, large outer diameter cups, multiple operations for instability, and instability with well-fixed shells that are adequately positioned.

Technique of cementing liner into shell: score acetabular shell if no holes; score liner in spider web configuration; all one or two millimeters of cement mantle.

Reoperation on the acetabular side of the total hip arthroplasty construct because of acetabular liner wear with or without extensive osteolysis is the most common reoperation performed in revision hip surgery today. The options of revision of the component or component retention, liner exchange (cemented or direct reinsertion) and bone grafting represent a classic surgeon dilemma of choices and compromises.

CT scanning is helpful in determining the size and location of osteolytic lesions. My preference is to retain the existing shell when possible especially when there are large osteolytic lesions but where structural support is maintained.

The advantages of complete revision are easy access to lytic lesions, ability to change component position and the ability to use contemporary designs with optimal bearing surfaces (for wear and dislocation prevention).

The disadvantage is bone disruption including pelvic discontinuity with component removal (less so with Explant Systems) and difficult reconstructions due to excessive bone loss from the osteolytic defects (sometimes requiring cup cages).

The advantage of component retention is that structural integrity of the pelvis is maintained and in general, a higher quality polyethylene is utilised. For large lesions I use windows to debride and bone graft the lesions. If the locking mechanism is inadequate, cementing a liner, including a constrained liner in some cases, that has been scored in a spider web configuration provides durable results at 5-year follow-up. The downside to liner exchange is potential instability. We immobilise all liner exchange patients postoperatively.

Arthrodesis

Today, knee arthrodesis is most commonly performed for cases of chronic sepsis after total knee arthroplasty in patients who are not candidates for reimplantation. This is typically a host in whom the risk of recurrent infection is high, especially when extensor mechanism problems such as patellar tendon rupture is present. Local and systemic host factors place the patient at this high risk for failure of reimplantation. Local factors include chronic lymphedema, major vessel disease, venous stasis, extensive scarring and radiation fibrosis. Systemic problems include malnutrition, malignancy, extremes of age, hepatic or renal failure, diabetes mellitus and alcohol abuse. Also, at least in one study, patients who fail one two-stage reimplantation have at least a 50% change for recurrence the second time.

Methods of knee arthrodesis include external fixation, single or double plate fixation and intramedullary nailing either monolithic or modular. External fixation can be performed as a single procedure. With external ring fixators leg lengthening has been described. Plate and nailing procedures are commonly performed after the infection is eradicated. If infection recurs but fusion has occurred, removal of a modular nail may be difficult although techniques have been described. Fusion rates of 85% to 100% have been reported with the newer techniques and fusion rates are usually correlated with the amount of bone loss after removal of the prosthesis. Optimal position of fusion is slight anatomic valgus and slight flexion. Placing a bowed intramedullary nail with the bow anteromedially can facilitate this alignment. One study comparing arthrodesis after failed knee arthroplasty with primary total knee replacement found nearly identical Short-Form 36 scores. Physical mobility was better with knee arthroplasty but pain control was better with arthrodesis.

Resection Arthroplasty

Indication: Low demand patient with comorbidities

Results: Falahee et al.; 28 knees;15 patients walked independently; Most severely disabled more satisfied. Less disabled least satisfied.

Amputation - Above the Knee Amputations after TKA (Sierra et al)

Prevalence: 0.36% (all causes – most common was peripheral vascular disease), 0.14% (for causes related to TKA: infection, periprosthetic fracture, pain, bone loss, vascular complication)

Results: 25 amputations for causes related to TKA; Avg 8.6 years after TKA

Complications: deep infection 5, superficial infection 1, skin necrosis 1, perioperative death 1. 9 of 25 fitted with above knee prosthesis; Only 5 were walking even to limited degree with prosthesis

Whether to preserve the posterior cruciate ligament in total knee arthroplasty remains a topic of debate. Long term studies out to 20 or more years have shown both posterior-stabilised (PS) as well as cruciate-retaining (CR) total knees to perform well. Studies supporting both sides are abundant, and there are even studies supporting one being superior to the other, and vice versa.

In closer examination of this issue through a recent meta-analysis, the following conclusions can be drawn:

Flexion: mean difference of 2.24, favoring PS designs (p = 0.009)

Patellar crepitance with PS designs can be minimised with design alternations and techniques for patella preparation

Any issues with post impingement can be minimised using PS rotating platform designs

The debate of PS vs CR total knees remains unresolved. There is not a 100% clear advantage. Surgeons should make their choice based on their own clinical outcomes, objectively monitored over the long term. For me personally, the choice is PS.

There are several clinical scenarios to consider cementing an acetabular liner into a secure cementless shell including cases of: 1) inadequate capturing mechanism, 2) damaged locking mechanisms, 3) unavailability of the mating polyethylene liner, 4) instability following debridement for wear, 5) instability at the time of femoral side revision, and 6) recurrent dislocation. The last two situations are common scenarios for cementing a constrained liner into a secure shell.

Technique includes: 1) scoring the shell in cases with no screw holes or polished inner shells, 2) scoring the acetabular liner in a “spider web” pattern, 3) pressurising cement into the shell, and 4) inserting a liner that allows 2mm of cement mantle.

Results of Cementing Constrained Liner Into Secure Cementless Shell: Callaghan et al. JBJS 2004 (31 hips, 2–10 year follow up, 2 of 31 failed, Technical considerations - Do not cement proud, Do not cement into a malpositioned shell); Haft et al. J Arthroplasty 2002 (17 hips, Minimum 1 year follow up, 1 of 17 failed, Technical considerations - Do not cement proud)

Results of Cementing Non-Constrained Liners Into Secure Cementless Shell: Beaule et al. JBJS 2004 (32 hips, Mean 5.1 year avg follow up, 4 components revised for loosening); Callaghan et al. CORR 2012 (31 hips, Mean 5.3 year follow up, No revisions)

Allograft materials have been the mainstay in addressing bone deficiencies in knee and hip replacement and revision surgery for decades because of the associated donor site morbidity of autografts. Bone graft substitutes have been developed to address allograft issues including potential contamination, disease transmission, and availability. Although non-autogenous products have no osteogenic potential, they do have a variable degree of osteoinductive and osteoconductive properties.

Unfortunately, there are limited reports regarding use of bone graft substitutes for use in total hip and knee arthroplasty. Bone graft substitutes have most frequently been used as an “extender”, in combination with morsellised allograft, to fill cavitary defects. Incorporation of this bone graft substitute and morsellised allograft combination appears to occur incompletely. Stable implant fixation appears to be a prerequisite for incorporation of bone graft substitutes, as these cannot be relied upon for structural support. Although bone graft substitutes appear to perform satisfactorily as “fillers” for contained cavitary bone defects, ultraporous metal augments have become the preferred method of providing structural support for some defects. In view of their substantial cost, high quality clinical, radiographic and retrieval data regarding performance of bone graft substitutes is needed.

Significant hip osteoarthritis has been reported in 8–28% of patients with Down Syndrome. The prolonged life expectancy of these patients has allowed many of them to become disabled by their hip arthritis with the need for hip replacement.

We have been able to perform a multi-center study evaluating total hip replacement in patients with Down Syndrome. Twenty patients (25 hips) with Down Syndrome underwent primary THA at a mean age of 35 years old with a mean 105 month follow-up. Cementless acetabular fixation with screws were used in all cases and all but one femoral component was cementless. Constrained liners were used in 8 cases to enhance stability. Five hips required revision surgery: two femoral components (one for periprosthetic fracture and one for aseptic loosening), two acetabular components (one for recurrent dislocation and one for wear with metallosis), and one hip required a two-stage revision for infection. Other than the hip revised for loosening, no other hip had radiographic evidence of loosening. The mean Harris Hip score improved from 42 points preoperatively to 83 points at final follow-up.

THA is a reliable surgical intervention in patients with Down Syndrome and symptomatic coxarthrosis. These patients and their families have been tremendously satisfied with this procedure.

The designs available today have greatly improved our ability as surgeons to perform successful total knee revision surgery. However, as more and more knee replacements are in service for longer periods of time, the numbers of revisions have increased and have required us as surgeons to address challenging problems including infection, instability and bone loss from wear, osteolysis and loosening.

Understanding the problems needed to be addressed is paramount. Careful preoperative planning is key. Knowing the cause of failure and the aspects of reconstruction that need to be addressed including skin, soft tissues, extensor mechanisms, bone and ligament loss is critical.

Intraoperatively, understanding and applying principles related to establishing joint lines, balancing flexion extension gaps, addressing bone loss and ligament instability and constructing stable knee replacements with the use of stabilising implant articulations, bone deficiency reconstruction with augments and grafts as well as cones and sleeves, and stems for implant stability is also essential.

Postoperatively, rehabilitation and follow-up must be tailored to the individual patient because of the marked nuances of construct in the various revision scenarios.

There are several clinical scenarios to consider cementing an acetabular liner into a secure cementless shell including cases of: 1) inadequate capturing mechanism, 2) damaged locking mechanisms, 3) unavailability of the mating polyethylene liner, 4) instability following debridement for wear, 5) instability at the time of femoral side revision, and 6) recurrent dislocation. The last two situations are common scenarios for cementing a constrained liner into a secure shell.

Technique includes: 1) scoring the shell in cases with no screw holes or polished inner shells, 2) scoring the acetabular liner in a “spider web” pattern, 3) pressurising cement into the shell, and 4) inserting a liner that allows 2 millimeters of cement mantle.

Results of Cementing Constrained Liner Into Secure Cementless Shell A.

Callaghan et al. JBJS 2004

Results of Cementing Non-Constrained Liners Into Secure Cementless Shell A.

Beaule et al. JBJS 2004

In the revision situation, there are times where larger heads are just not enough to obtain and maintain stability. The two most relevant times that this is the case is in patients with very lax tissues, or in patients with insufficient or absent soft tissues, especially abductor mechanisms. In addition, in cases where a revision is being performed for dislocation and components looked well-positioned, constrained liners have been extremely beneficial in our hands.

Constrained acetabular liners have been available for close to two decades. Two basic types of liners are available. The type first developed by Joint Medical Products was the SROM constrained liner which captured the femoral head with a locking ring in the polyethylene. These liners may have better results with larger head sizes because the hip can be taken through a larger range of motion (with larger head sizes) before the locking ring is stressed. The second type of constraining liner was developed by Osteonics (Stryker). It consisted of a tripolar replacement which is constrained by a locking ring in the outer polyethylene of the device. Indications for constrained liners include patients undergoing primary arthroplasty who are low demand and have dementia or hip muscle weakness or spasticity. Indications for constrained liners in the revision situation include cases with previously failed operations for instability, elderly low demand patients with instability, cases with poor or absent hip musculature, and cases with well positioned acetabular and femoral components and with hip instability. In this last scenario we cement the liners into fixed shells.

Our results at average 10-year follow-up in 101 hips, demonstrate a 6% failure of the device. Four hips were revised for acetabular loosening and four hips for femoral loosening. One additional hip was revised for acetabular osteolysis. Considering the difficulty of the cases we consider these results to be quite encouraging. At average 3.9 year follow-up of 31 cases where the liner was cemented into the secure shell only one case failed by dislodgement of the liner and one case by fracture of the locking mechanism.

Our experience has led to the following technical recommendation: (1) if cementing the component score the liner and make sure it is contained within the shell (2) avoid inserting the liner into a grossly malpositioned shell (3) avoid positioning the elevated rim of the liner into a position where impingement might occur and (4) avoid placing the shell and constrained liner in cases with massive acetabular allografts unless additional fixation, i.e. cages, are utilized. Especially in the elderly, these liners are our components of choice for many pre-operative and intra-operative cases of instability.

The first 101 posterior cruciate retaining modular tibial components of a single design performed by a single surgeon in 75 patients were evaluated at a minimum 20-year follow-up. All components were fixed with cement. These patients had been prospectively followed at five-year intervals and evaluated clinically using Knee Society ratings and documenting any need for reoperation. Serial radiographs were evaluated for radiolucencies, osteolysis or component migration until the time of patient death or at minimum 20-year follow-up.

At minimum 20-year follow-up, five knees (5%) had required a revision operation. All revisions occurred greater than 10 years following the index procedures. Benefits of modularity (i.e. retention of the tibial tray) were utilized in three of five cases in this closely followed cohort. Survivorship from any revision was 90.8% at 20 years. For the 16 living patients with 22 knees, the average Knee Society Clinical and Functional scores were 91 and 59, respectively, and the average range of motion was 115 degrees.

When considering gamma irradiated in air polyethylene and a first generation locking mechanism were utilized, these results encourage the authors to continue to use modular tibial trays.

Purpose: Femoroacetabular impingement (FAI) is a precursor to osteoarthritis (OA) of the hip. To investigate the fate of impingement abnormalities we analyzed the contralateral hip in patients undergoing THA for advanced FAI. Our purpose was to determine the bilaterality of FAI abnormalities, and to describe the prognosis of these deformities.

Method: We reviewed 508 patients ≤50 years of age treated with THA. Radiographic review identified 70 hips that had OA secondary to FAI (71% cam, 5% pincer, 24% combined). Bilaterality was determined radiographically, and the fate of the contralateral hip was analyzed by determining radiographic presence and progression of OA, and the need for subsequent THA.

Results: 71% of the patients were male and the average age was 43.2 years. The contralateral hip was analyzed on radiographs over an average 9 year period (range, 4–30 years). 100% of the contralateral hips had radiographic features of FAI. 49 (70%) of the contralateral hips demonstrated degenerative disease. 14 had advanced OA at presentation, 41 had progressive joint space narrowing, 25 had progression of Tonnis OA grade and 26 underwent subsequent THA. Statistical analysis showed that alpha angle, LCEA, joint space width, and head-neck ratio have strong predictive value for subsequent THA (p< 0.05).

Conclusion: This study demonstrates that FAI abnormalities are commonly bilateral and are associated with OA progression in the majority of hips. Patients diagnosed with FAI should have both hips monitored, and joint preservation surgery or THA considered when appropriate.

Patient postoperative outcome can be accurately predicted by the patient’s preoperative HHS or WOMAC score. Prospective, cohort studies of one hundred and seventy-five THAs. SF-36, WOMAC and Harris Hip Score (HHS)questionnaires were used to determine pre-operative and two year final outcome. Student’s t-test, 95% confidence intervals, receiver operator characteristic curves, simple regression analysis and probability were measured. Patients with a HHS = sixty-five pre-operatively had a 100% probability of having an excellent result postoperatively. A preoperative HHS value of thirty-four, and preoperative WOMAC (physical function) value of fifty were the best cutoff points to attain a significantly better postoperative functional outcome.

Total hip arthroplasty (THA) has been well documented to enhance patient function, but patient outcome is dependant on preoperative statuts. The exact timing of surgery to optimize patient outcome after THA remains unknown. This study determines the ideal timing for surgery to obtain the best possible functional outcome.

Prospective, multicenter, cohort studies of one hundred and seventy-five identical, cemented THAs. General health (SF-36) and disease specific (WOMAC and Harris Hip Score(HHS)) questionnaires were used to determine preoperative and two year final outcome. Student’s t-test, 95% confidence intervals, receiver operator characteristic curves, simple regression analysis and probabilty were measured.

All functional scores were improved significantly postoperatively (p< 0.001). Patients with a HHS = sixtyfive preoperatively had a 100% probability of having an excellent result postoperatively. A preoperative HHS value of thirty-four, and preoperative WOMAC (physical function) value of fifty were the best cutoff points to attain a significantly better postoperative functional outcome.

Patient postoperative outcome can be accurately predicted by the patient’s preoperative HHS or WOMAC score. Optimization of surgical timing, by prioritizing wait lists or deciding to treat the arthritis operatively, based on these preoperative score guidelines will help ensure an excellent outcome post THA.

With dislocation, prevention is more optimal than treatment. Causes of dislocation include patient compliance, soft tissue tensioning, and component malpositioning problems. Dislocation can occur from bone impingement, component impingement, and spontaneously.

Most work on dislocation has evaluated the differences in the propensity for dislocation based on head size. The larger the head to neck ratio the more motion that can be obtained before cup neck impingement occurs. Skirted modular femoral heads decrease the head to neck ratio and in addition the modular necks must be wider because of strength issues. Because of modularity propensity for dislocation has been on the rise.

As wear has been associated with larger head sizes and thinner polyethylene when conventional polyethylene is used, use of larger head sizes was not considered an option for dislocation prevention and treatment. Constrained liners and bipolar components were used instead. If the newer hyper crosslinked polyethylene components do markedly reduce wear it will be possible to use larger head sizes to prevent dislocation. However, there may be problems with bony impingement associated with larger head sizes.

Historically we know that all femoral, cemented stems have not performed the same. The Muller stem with its sharp comers did not perform as well as the Charnley femoral component. Titanium femoral components have not performed as well in the cemented situation as have stiffer chrome cobalt components.

Today we have come to recognise that the durability of a cemented femoral component is dependent on a number of variables to include stem geometry and surface finish as well as the cement technique.

Since several designs including the Trapezoidal-28, the Exeter, and the Iowa have incorporated various surface finishes over time, the issue of surface finish is one that some investigators think is relatively important. Components with rougher surface finishes adhere better to cement, decrease cement strains, and when they become loose are more likely to abrade cement. Components with smooth surface finishes do not adhere to cement, place cement under compression and when they become loose they are less likely to abrade cement. When these three stems were evaluated with both smooth and matte finishes the smooth surface finish stems have always demonstrated better results with less loosening and less osteolysis. However some matte finish designs have performed well including the CAD and the HD-2. Whatever prosthesis is utilised all investigators agree that it is optimal to have the prosthesis surrounded by cement, hence the use of centralisers.

Hence today all would agree that the surgeon should use a torsionally sound stem and place an adequate cement mantle around that stem at the time of surgery .If that can be achieved all stems may be equal, however if cement mantle defects are inevitable a smooth surface finish probably has better durability.

One hundred hips in patients who had had primary uncemented replacements were followed up for one or two years, and assessed by five different methods. All produced different results. The Hospital for Special Surgery rating produced the most optimistic assessment and the Merle d'Aubigne rating the most pessimistic. The functional class of the patients, as defined by Charnley in 1979, significantly affected the ratings, and these should clearly be included in all rating systems. Moreover, if systems are to be compared, they should all use descriptive words, such as limp or pain, in precisely the same way.