The PDF of the article you requested follows this cover page.  

This is an enhanced PDF from The Journal of Bone and Joint Surgery

2008;90:1508-1516.  doi:10.2106/JBJS.G.00201 J Bone Joint Surg Am.Ron El-Hawary, Lori A. Karol, Kelly A. Jeans and B. Stephens Richards    

or the Ponseti Cast TechniqueGait Analysis of Children Treated for Clubfoot with Physical Therapy

This information is current as of July 2, 2008

Supplementary material

http://www.ejbjs.org/cgi/content/full/90/7/1508/DC1accessed at translated abstracts are available for this article. This information can be Commentary and Perspective, data tables, additional images, video clips and/or

Reprints and Permissions

Permissions] link. and click on the [Reprints andjbjs.orgarticle, or locate the article citation on

to use material from thisorder reprints or request permissionClick here to

Publisher Information

www.jbjs.org20 Pickering Street, Needham, MA 02492-3157The Journal of Bone and Joint Surgery

Gait Analysis of Children Treated for Clubfoot withPhysical Therapy or the Ponseti Cast Technique

By Ron El-Hawary, MD, MSc, Lori A. Karol, MD, Kelly A. Jeans, MS, and B. Stephens Richards, MD

Investigation performed at the Movement Science Laboratory, Texas Scottish Rite Hospital for Children, Dallas, Texas

Background: Currently, clubfoot is initially treated with nonoperative methods including the Ponseti cast techniqueand the French functional physical therapy program. Our goal was to evaluate the function of children treated with thesetechniques.

Methods: We reviewed the cases of 182 patients with idiopathic clubfoot (273 feet) who were initially treatednonoperatively. Seventy-seven patients (119 feet) were excluded because they had either received a combination ofnonoperative treatments or had undergone surgery prior to testing. Gait analysis was performed when the children wereapproximately two years of age. Temporal and kinematic data were classified as abnormal if they were more than onestandard deviation from normal.

Results: Gait analysis was performed on 105 patients (fifty-six treated with casts and forty-nine treated with physicaltherapy) with 154 involved feet (seventy-nine treated with casts and seventy-five treated with physical therapy). Thesepatients were an average of two years and three months of age, and their initial Dimeglio scores ranged between 10and 17. No significant differences in cadence parameters were found between the two groups. The rate of normalkinematic ankle motion in the sagittal plane was higher in the group treated with physical therapy (65% of the feet) thanit was in the group treated with the Ponseti cast technique (47%) (p = 0.0317). More children treated with physicaltherapy walked with knee hyperextension (37% of the feet) (p < 0.0001), an equinus gait (15%) (p = 0.0051), andfootdrop (19%) (p = 0.0072); only one patient treated with casts walked with an equinus gait, and only threedemonstrated footdrop. In contrast, more of the patients in the cast-treatment group demonstrated excessive stance-phase dorsiflexion (48% of the feet) (p < 0.0001) and a calcaneus gait (10%). More feet in the physical therapy grouphad an increased internal foot progression angle (44% compared with 24% in the cast-treatment group; p = 0.0144)and increased shank-based foot rotation (73% compared with 57% in the cast-treatment group; p = 0.05).

Conclusions: While the rate of normal kinematic ankle motion in the sagittal plane was 65% in the group treated withphysical therapy, the gait abnormalities that were seen in that group were characterized by mild equinus and/orfootdrop. The rate of normal kinematic ankle motion in the sagittal plane was 47% in the cast-treatment group, but themost common gait abnormality in this group was mildly increased dorsiflexion in the stance phase. The rates ofcalcaneus gait and equinus gait were £15% in each nonoperative group. The differences between the physical therapyand cast-treatment groups may, in part, be the result of the percutaneous Achilles tendon lengthening that is performedas part of the Ponseti cast technique but not as part of the physical therapy program.

Level of Evidence: Therapeutic Level II. See Instructions to Authors for a complete description of levels of evidence.

Historically, the treatment of clubfoot deformity wasprimarily surgical1-3. More recently, it has been reportedthat these operative treatments involving extensive

soft-tissue releases do not always produce satisfactory long-termclinical results4,5. In addition, gait disturbances such as knee

hyperextension, ankle stiffness, and decreased gastrocnemius-soleus power and strength have been documented6. As a con-sequence of these clinical and functional results, there has been arenewed interest in the nonoperative treatment of clubfootdeformity 5.

Disclosure: The authors did not receive any outside funding or grants in support of their research for or preparation of this work. Neither they nor amember of their immediate families received payments or other benefits or a commitment or agreement to provide such benefits from a commercialentity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, division, center, clinical practice,or other charitable or nonprofit organization with which the authors, or a member of their immediate families, are affiliated or associated.

A commentary is available with the electronic versions of this article, on our web site (www.jbjs.org) and on our quarterly CD-ROM (call oursubscription department, at 781-449-9780, to order the CD-ROM).

1508

COPYRIGHT � 2008 BY THE JOURNAL OF BONE AND JOINT SURGERY, INCORPORATED

J Bone Joint Surg Am. 2008;90:1508-16 d doi:10.2106/JBJS.G.00201

Ponseti’s method of using serial casts was first developedin the 1940s but did not become popular until his long-termresults were published in the 1990s7. His techniques are now inwidespread use and include application of a series of correctivelong leg casts, followed by a period of bracing with a footabduction orthosis.

Masse introduced the French functional (physical ther-apy) technique in the 1970s8, and Bensahel et al.9 and Dimeglioet al.10 reported the results of this technique more recently9,10.The French method includes intensive daily stretching, ma-nipulation, and taping of the infant’s foot by a physical therapist.

Both of these nonoperative techniques have been em-ployed at our institution for a number of years. The purpose ofthis study was to evaluate and compare the gait patterns oftwo-year-old children who had been treated since infancy withthese nonoperative techniques.

Materials and Methods

The parents of patients treated for a clubfoot deformity witheither the Ponseti cast technique or the French functional

(physical therapy) method between February 1998 and May2004 at a single center were invited to have their child par-ticipate in this prospective study, which was approved by aninstitutional review board. The inclusion criteria consisted of adiagnosis of idiopathic clubfoot and an initial Dimeglio classi-fication rating of severe or very severe (10 to 17)11. Teratologicclubfeet, patients initially treated outside of our institution, mildand moderate Dimeglio types, and patients who required sur-gical intervention were excluded.

TreatmentBoth methods of treatment were described in detail in an un-biased manner at the infant’s initial visit, and the parents chosewhich treatment their child would receive.

The Ponseti cast treatment included application of a seriesof long leg casts, with the cast changed weekly for five weeks, afterwhich a final long leg cast was applied and worn for an addi-tional three weeks. All of the casts were applied by physicians.Before each new cast was applied, the foot was gradually stretchedinto an improved position with use of the precise sequence ofmaneuvers described by Ponseti12. Percutaneous Achilles tendonlengthening was performed on selected patients in order to fa-cilitate correction of ankle equinus prior to application of thefinal cast. Since an Achilles tendon lengthening is considered aroutine procedure in the Ponseti protocol, the patients who hadundergone that procedure were not considered to have hadsurgery and were therefore included in this study. After theclubfoot deformity was corrected and the cast-treatment phasewas complete, a foot abduction orthosis was applied and wasworn full-time for three months; it was then worn at night andduring naps until the child was at least two years of age.

The French functional (physical therapy) method includedstretching and manipulation by a skilled physical therapist fivedays per week. Between sessions, the foot was temporarily im-mobilized with adhesive tape in an effort to maintain maximalcorrection. Since 2000, some of the infants have been treated with

a continuous passive motion machine designed specifically forthe infant foot (Kinetec, Tournes, France). Daily therapy con-tinued for the first few months of life, until the deformity wascorrected. The parents were then trained to continue the pro-gram at home, with periodic visits with the physical therapist.This home protocol continued until the patients were two tothree years of age.

EvaluationThe parents were invited to have their child participate in agait analysis when the child was approximately two years ofage. Kinematic data were collected with use of a Vicon mo-tion capture system and analyzed with use of Vicon ClinicalManager (Oxford Metrics Group, Oxford, United Kingdom).If it was not possible to complete testing because of the child’suncooperative behavior, a second attempt was made to collectthe data.

Both temporal and gait parameters were identified andcompared with those of fifteen normal two-year-old children.Values were considered abnormal if they were more than onestandard deviation from the average normal data (Table I).

Standing anteroposterior and lateral radiographs of thefeet were made when the patients were two years of age.The tibiocalcaneal angle, defined as the angle formed betweenthe longitudinal axis of the tibia and the plantar surface of thecalcaneus, was compared with published normative data to de-termine whether there was radiographic evidence of an ankleequinus or ankle calcaneus gait13.

Statistical MethodsTwo-way repeated-measures analysis was used to performcomparisons between the treatment groups as well as withnormal data. Chi-square testing was used to compare the dis-tributions of proportions between the groups. Significance wasdefined as p < 0.05.

TABLE I Definitions of Criteria Used for Gait Deviations

from Normal*

Gait Criteria

Equinus gait <3� of dorsiflexion during stance

Calcaneus gait <3� of plantar flexion at toe-off

Footdrop >9� of plantar flexion during thelast 25% of swing

Increased ankle dorsiflexion >15� of dorsiflexion in stance

Internal shank-based footrotation

>0� average internal rotationduring stance

Internal foot progressionangle

>5� average internal rotationduring stance

Knee hyperextension >7� of knee hyperextensionduring midstance

*All values represent one standard deviation from the averagenormal data for two-year-old children tested in our laboratory.

1509

TH E J O U R N A L O F B O N E & JO I N T SU R G E RY d J B J S . O R G

VO LU M E 90-A d NU M B E R 7 d J U LY 2008GA I T AN A LY S I S O F CH I L D R E N TR E AT E D F O R CLU B F O O T W I T H

PH Y S I C A L TH E R A P Y O R T H E P O N S E T I CA S T TE C H N I Q U E

Results

We reviewed the cases of 182 patients with idiopathic club-foot (273 feet) (Table II). Feet were then excluded from

the analysis if they (1) had undergone a combination of casttreatment and physiotherapy or had undergone botulinumtoxin A (Botox) treatment or (2) had had a failure of nonop-erative treatment and had undergone surgery prior to gaitanalysis. Because the Achilles tendon lengthening was not ini-tially part of the French physical therapy method, the patientswho had been treated with that method and had undergoneAchilles tendon lengthening for management of residual equinusdeformity (nine feet) were considered to have received surgicaltreatment and therefore were excluded from the study. Thirty-nine patients (fifty-seven feet) who had had a percutaneousAchilles tendon lengthening as a part of the Ponseti cast-treatment protocol were included.

One hundred and five patients (fifty-six treated withcasts and forty-nine, with physical therapy) with 154 involvedfeet (seventy-nine treated with casts and seventy-five, with

physical therapy) were included in the gait analysis. Twenty-four (23%) of 103 feet treated with the Ponseti method andforty-eight (36%) of 132 feet treated with physical therapyhad undergone surgery that included more than an Achillestendon release or lengthening. At the time of gait analysis, thepatients were an average age of two years and three months(range, two years to three years and four months). There wasno significant difference in age between the two treatmentgroups.

The average initial Dimeglio scores were 13.1 (range, 10to 17) in the Ponseti cast-treatment group and 13.3 (range, 10to 17) in the physical therapy group, a difference that was notsignificant (p = 0.52).

Cadence parameters, including walking speed, cadence,and stride length, did not differ between the two study groups(p > 0.05). There were several mild deviations in sagittal planeankle kinematics in both groups (Table III).

An equinus gait was defined as midstance-phase (second-rocker) dorsiflexion of <3�. Eleven feet (15%) in the physical

TABLE II Patients Whose Cases Were Reviewed for This Study

Total Cast Treatment Physical Therapy

All patients whose cases were reviewed 182 (273 feet)

Patients included in study 105 (154 feet) 56 (79 feet [57 feet withAchilles tendon lengthening])

49 (75 feet)

Patients excluded from study

Surgical intervention 52 (81 feet) 14 (24 feet) 38 (57 feet [9 feet with Achillestendon lengthening])

Mixture of treatment 25 (38 feet)

TABLE III Demographic and Kinematic Data for the Two Treatment Groups

Cast Treatment Physical Therapy P Value

No. of patients 56 49

No. of feet 79 75

Average Dimeglio score (range) 13.1 (10-17) 13.3 (10-17) 0.52

Equinus gait* 1 (1%) 11 (15%) 0.0051

Footdrop* 3 (4%) 14 (19%) 0.0072

Increased ankle dorsiflexion* 38 (48%) 9 (12%) <0.0001

Calcaneus gait* 8 (10%) 3 (4%) 0.245

Internal foot progression angle* 19 (24%) 33 (44%) 0.0144

Internal shank-based foot rotation* 45 (57%) 55 (73%) 0.05

Knee hyperextension* 6 (8%) 28 (37%) <0.0001

Normal sagittal plane ankle motion* 37 (47%) 49 (65%) 0.0317

Normal gait* 11 (14%) 11 (15%) 0.895

Average tibiocalcaneal angle (range) (deg) 84 (56-110) 90 (72-116) 0.0023

*The values are given as the number of feet with the percentage in parentheses.

1510

TH E J O U R N A L O F B O N E & JO I N T SU R G E RY d J B J S . O R G

VO LU M E 90-A d NU M B E R 7 d J U LY 2008GA I T AN A LY S I S O F CH I L D R E N TR E AT E D F O R CLU B F O O T W I T H

PH Y S I C A L TH E R A P Y O R T H E P O N S E T I CA S T TE C H N I Q U E

therapy group and one foot (1%) in the cast-treatment groupwere associated with an equinus gait. This difference was sig-nificant (p = 0.0051) (Fig. 1).

Increased stance-phase ankle dorsiflexion was defined asdorsiflexion that differed from normal by greater than onestandard deviation (i.e., that was >15�). Thirty-eight feet(48%) in the Ponseti cast-treatment group and nine feet (12%)in the physical therapy group were associated with increasedstance-phase ankle dorsiflexion (p < 0.0001) (Fig. 2). A cal-caneus gait was defined as increased stance-phase dorsiflexionand terminal-stance-phase (third-rocker) plantar flexion of<3�. Eight feet (10%) in the cast-treatment group and threefeet (4%) in the physical therapy group were associated with acalcaneus gait, but this difference did not reach significance(p = 0.245). All eight feet in the cast-treatment group that wereassociated with a calcaneus gait were also associated with in-creased stance-phase ankle dorsiflexion.

The patients who had undergone an Achilles tenotomyas part of the Ponseti cast treatment were compared with thosewho had not. Of the thirty-eight feet associated with increasedankle dorsiflexion in the Ponseti cast-treatment group, thirty-two (84%) had had a tenotomy and six (16%) had not. Thisdifference was significant (p = 0.022). Thirty-two (56%) of thefifty-seven feet for which the treatment had included anAchilles tendon release as part of the Ponseti protocol wereassociated with increased dorsiflexion during stance phase. Alleight feet in the Ponseti treatment group that were associatedwith a calcaneus gait had undergone a tenotomy.

The twenty-two clubfeet treated with the Ponseti methodbut not with an Achilles tenotomy were compared with theseventy-five feet treated with physical therapy. The only sig-nificant kinematic difference between the two groups was in

stance-phase knee hyperextension: the percentage of feet asso-ciated with knee hyperextension was higher in the physicaltherapy group than it was after treatment with the Ponsetimethod without an Achilles tendon lengthening (p = 0.029). Anequinus gait was associated with one foot (5%) treated with thePonseti method without an Achilles tendon lengthening andwith eleven feet (15%) in the physical therapy group. This dif-ference was not significant.

Footdrop is the inability to dorsiflex the ankle during theswing phase. We defined footdrop as increased ankle plantarflexion (>9� of plantar flexion) during the final 25% of theswing phase. Three feet (4%) in the cast-treatment group andfourteen feet (19%) in the physical therapy group exhibitedfootdrop when the child walked (p = 0.0072).

Normal ankle kinematics in the sagittal plane was de-fined as an absence of an equinus gait, of a calcaneus gait, ofincreased stance-phase dorsiflexion, and of footdrop. Thirty-seven feet (47%) in the cast-treatment group and forty-ninefeet (65%) in the physical therapy group were associated withnormal sagittal plane ankle kinematics (p = 0.0317).

Intoeing was defined as an internal foot progressionangle of >5� in the stance phase. Nineteen feet (24%) in thecast-treatment group and thirty-three feet (44%) in thephysical therapy group had an internal foot progressionangle of >5� (p = 0.0144) (Fig. 3). Shank-based foot rotationis a measure of the rotation of the forefoot relative to theposition of the knee. It represents tibial torsion, medial spinof the hindfoot, and metatarsus adductus, but it does notisolate the tibia from the foot. Forty-five feet (57%) in thecast-treatment group and fifty-five feet (73%) in the physicaltherapy group had an internal shank-based rotation of >0�(p = 0.05).

Fig. 1

Sagittal plane ankle kinematic data over one complete gait cycle, obtained from a representative

patient with an equinus gait in the physical therapy group. Positive values represent dorsiflexion

(DF), and negative values represent plantar flexion (PF). The shaded region represents one standard

deviation above and below the average age-matched normal data.

1511

TH E J O U R N A L O F B O N E & JO I N T SU R G E RY d J B J S . O R G

VO LU M E 90-A d NU M B E R 7 d J U LY 2008GA I T AN A LY S I S O F CH I L D R E N TR E AT E D F O R CLU B F O O T W I T H

PH Y S I C A L TH E R A P Y O R T H E P O N S E T I CA S T TE C H N I Q U E

An external foot progression angle was present in elevenlower extremities. Ten of the eleven were in the Ponseti treat-ment group and accounted for 13% of the limbs in that group.One was in the physical therapy group and accounted for 1%

of the limbs in that group. Six of the ten feet treated with thePonseti method were associated with increased stance-phasedorsiflexion, and three of the six were associated with a cal-caneus gait.

Fig. 2

Sagittal plane ankle kinematic data over one complete gait cycle, obtained from a representative

patient with increased stance-phase dorsiflexion in the Ponseti cast-treatment group. As the third-

rocker plantar flexion was within one standard deviation of normal, this patient was not considered

to have a calcaneus gait. Positive values represent dorsiflexion (DF), and negative values represent

plantar flexion (PF). The shaded region represents one standard deviation above and below the

average age-matched normal data.

Fig. 3

Average foot progression data over one complete gait cycle for both treatment groups. The solid line

represents the values for the Ponseti treatment group, and the dashed line represents the values

for the physical therapy group. Positive values represent internal rotation (INT), and negative values

represent external rotation (EXT). The shaded region represents one standard deviation above and

below the average age-matched normal data.

1512

TH E J O U R N A L O F B O N E & JO I N T SU R G E RY d J B J S . O R G

VO LU M E 90-A d NU M B E R 7 d J U LY 2008GA I T AN A LY S I S O F CH I L D R E N TR E AT E D F O R CLU B F O O T W I T H

PH Y S I C A L TH E R A P Y O R T H E P O N S E T I CA S T TE C H N I Q U E

We used a strict definition of normal gait that includednormal ankle kinematics in the sagittal plane, a normal footprogression angle, and normal shank-based foot rotation.Eleven feet (14%) in the cast-treatment group and eleven feet(15%) in the physical therapy group were associated with acompletely normal gait (p = 0.895).

Kinematic data and cadence parameters were com-pared between unilateral and bilateral clubfeet in both thephysical therapy and the Ponseti cast-treatment group.There was no significant difference in cadence parameters orin the prevalence of kinematic abnormalities between theunilateral and bilateral clubfeet in the Ponseti treatmentgroup. In the physical therapy group, the unilateral caseswere associated with a lower initial Dimeglio score (12.43compared with 13.57 for the bilateral cases), a greater walkingspeed, and an increased stride length. However, there was nodifference in the likelihood of kinematic abnormalities be-tween the unilateral and bilateral cases in the physical therapygroup.

The tibiocalcaneal angle was measured on a lateral ra-diograph of the foot. As ankle dorsiflexion improved, the ti-biocalcaneal angle decreased. Radiographs of seventy-eight feetin the cast-treatment group and sixty-two feet in the physicaltherapy group were assessed. (Fourteen radiographs were ex-cluded because of a lack of cooperation by the child.) Themean tibiocalcaneal angle in the cast-treatment group was 84�(range, 56� to 110�), which was significantly different from the

mean of 90� (range, 72� to 116�) in the physical therapy group(p = 0.0023).

Discussion

In the past, few investigators have used gait analysis as a toolto evaluate the outcome of treatment of clubfeet6,14-22. Ar-

onson and Puskarich examined patients who had been fol-lowed for more than ten years after treatment of a clubfootdeformity with either cast immobilization or surgery15. Re-gardless of which treatment had been performed, ankle dor-siflexion was reduced by an average of 65%. In addition, therewas an average 24% decrease in normal ankle plantar flexormuscle strength, which correlated with the number of Achillestendon lengthenings that had been performed in the foot.Karol et al. performed gait analysis and muscle strength testingon patients at an average of ten years after surgical release forthe treatment of clubfoot6. Their findings were in agreementwith those of Aronson and Puskarich in that ankle sagittalplane kinematics and plantar flexion power were substantiallyaffected in the majority of the subjects. Furthermore, genuvalgum, knee hyperextension, an internal foot progressionangle, and inappropriate activation of the tibialis anteriormuscle were observed. Davies et al. studied the cases of twenty-five children, with a mean age of twelve years, who had hadprevious posteromedial release for the treatment of clubfoot18.All patients were evaluated at least five years after their lastintervention. The ankle kinematic and kinetic data reported by

Fig. 4

Average sagittal plane ankle kinematic data over one complete gait cycle for both treatment groups.

The solid line represents the values for the Ponseti treatment group, and the dashed line represents

the values for the physical therapy group. Positive values represent dorsiflexion (DF), and negative

values represent plantar flexion (PF). The shaded region represents one standard deviation above

and below the average age-matched normal data. Both groups have, on the average, normal

kinematics. There is slightly less dorsiflexion in midstance in the physical therapy group and mildly

increased dorsiflexion in terminal stance in the Ponseti treatment group. The vertical line divides

the stance from the swing phase.

1513

TH E J O U R N A L O F B O N E & JO I N T SU R G E RY d J B J S . O R G

VO LU M E 90-A d NU M B E R 7 d J U LY 2008GA I T AN A LY S I S O F CH I L D R E N TR E AT E D F O R CLU B F O O T W I T H

PH Y S I C A L TH E R A P Y O R T H E P O N S E T I CA S T TE C H N I Q U E

Davies et al. were similar to ours. They also observed differ-ences in hip and knee kinematics and kinetics, and theythought that these differences were secondary to a lack ofmotion at the ankle joint. Theologis et al. studied the cases ofpatients between the ages of six and sixteen years who wereconsidered to have had a good clinical result of the surgicaltreatment of clubfoot deformity20. These patients were foundto walk with residual intoeing, footdrop, decreased plantarflexor muscle power, increased midfoot dorsiflexion, and ex-ternal hip rotation.

Given the imperfect clinical and functional results ofsurgical treatment of clubfoot, there has been a resurgence ofinterest in the nonoperative management of this defor-mity5,10,23,24. Both Ponseti’s serial cast technique and the Frenchfunctional (physical therapy) technique have been reported tohave excellent clinical results7,10; however, there have been fewreported analyses of the gait of patients treated with thesenonoperative methods14,25. Alkjaer et al. examined the gait ofnine adult men who had been treated for clubfoot as childrenwith a combination of physical therapy, bracing, and surgery25.As compared with a control group, these patients had smallerankle moments secondary to weak plantar flexors and theycompensated with larger-than-normal moments at the kneeand hip joints26.

In 2005, Karol et al. performed a gait analysis of two-year-old children treated with either physical therapy or surgicalrelease14. Internal rotation was common in both treatmentgroups, but surgical release resulted in a higher likelihood of acalcaneus gait. The most frequent gait disturbances in thephysical therapy group were limited ankle dorsiflexion andfootdrop. Overall, normal kinematic ankle motion in thesagittal plane was observed in 54% of the feet treated withphysical therapy compared with only 39% of the feet treatedwith complete posteromedial releases.

In the present study, a graphic comparison of ankle kine-matics in the sagittal plane showed very little difference be-tween the two treatment groups (Fig. 4), although the Ponseticast-treatment group tended to have greater dorsiflexionthroughout the gait cycle than the physical therapy group. As aresult, the kinematics in the physical therapy group appearedto be superior; however, this may be secondary to the ‘‘aver-aging’’ of the feet associated with an equinus gait (15% of thefeet) with the feet associated with increased dorsiflexion inthe stance phase (12%) in this group. This highlights the im-portance of defining criteria for pathological gait deviations(Table I).

With use of our definitions of abnormal gait parameters,the results identified two clear kinematic patterns. One subsetof children treated with the functional (physical therapy)method walked with knee hyperextension (37% of the feet), anequinus gait (15% of the feet), and footdrop (19% of the feet),while only one foot treated with Ponseti’s method had residualequinus and only three demonstrated footdrop. In contrast,more patients in the Ponseti treatment group demonstratedincreased stance-phase dorsiflexion (48% of the feet) and acalcaneus gait (10% of the feet).

The large number of patients with mild knee hyperex-tension in the functional (physical therapy) group may berelated to the plantar flexion-knee extension couple26, as therewas also a high prevalence of equinus gait in this group. Thislimitation in ankle dorsiflexion may have been a result of thefact that Achilles tendon lengthening was not performed aspart of the physical therapy protocol; however, given this rateof equinus, some patients treated with physical therapy maypotentially benefit from an Achilles tendon lengthening.Dimeglio et al. recently reported incorporation of an earlylengthening of the triceps surae (with use of the Vulpiustechnique) in the French physical therapy regimen for babies27.We have similarly begun performing an Achilles tendon te-notomy in babies being treated with the French method if 25�of dorsiflexion is not achieved; our hope is to reduce thenumber of children with residual equinus and knee hyperex-tension in the future.

The higher rate of increased stance-phase dorsiflexion inthe Ponseti cast-treatment group may be related to the use ofAchilles tendon lengthening prior to the application of thefinal cast in 72% of the feet in this group. This rate of Achillestenotomy is lower than that reported by Morcuende et al.28 butis consistent with the results reported by Ponseti5,12. In ourstudy, 54% of the feet in the Ponseti treatment group that hadhad an Achilles tendon lengthening were associated with eitherincreased stance-phase dorsiflexion or a calcaneus gait. Judi-cious use of Achilles tendon lengthening in patients under-going Ponseti cast treatment may help to improve ankledorsiflexion at the age of two years but also may predispose theankle to excessive dorsiflexion as a result of postlengtheningweakness. Some of these patients with increased dorsiflexion inthe Ponseti treatment group also were unable to plantar flexthe ankle at the end of the stance phase and had a calcaneusgait. We believe that the plantar flexor insufficiency seen inpatients with increased dorsiflexion is less severe than that inpatients with a calcaneus gait. It will be important to followthese children and analyze ankle power at push-off as theygrow to determine the long-term effects on function followinga tenotomy.

The percentage of feet associated with normal kinematicankle motion in the sagittal plane was greater in the physicaltherapy group (65%) than in the cast-treatment group (47%).However, both rates were better than the 39% rate observed inpatients treated with a posteromedial release14.

More feet treated with the French functional method hadan increased internal foot progression angle (44% comparedwith 24% in the cast-treatment group). This difference may bethe result of the routine use of a foot abduction bar in thePonseti treatment group. Because external rotation of the lowerextremity as well as abduction of the foot was maintained withthe orthosis full-time for three months and part-time for two tothree years, it is reasonable to expect that fewer patients in thecast-treatment group would have residual intoeing at the age oftwo years. Intoeing by patients with clubfoot has been found tobe secondary mainly to forefoot adductus and to internal tibialtorsion, and it remains present in a large proportion of children

1514

TH E J O U R N A L O F B O N E & JO I N T SU R G E RY d J B J S . O R G

VO LU M E 90-A d NU M B E R 7 d J U LY 2008GA I T AN A LY S I S O F CH I L D R E N TR E AT E D F O R CLU B F O O T W I T H

PH Y S I C A L TH E R A P Y O R T H E P O N S E T I CA S T TE C H N I Q U E

who have been treated with extensive surgical release6,16,17,19,20.The use of foot abduction bars likely decreased the prevalenceof intoeing in our Ponseti cast-treatment group, but an in-ternal foot progression angle was still observed in 24% of thefeet in this group. As we have no objective data on compliancewith use of the abduction bar, we cannot comment on thepresence or absence of a correlation between use of an orthosisand intoeing. This residual intoeing, observed in some patientstreated nonoperatively with either technique, may requirecorrection with an anterior tibial tendon transfer or a tibialderotation osteotomy in the future.

As seen radiographically, the physical therapy group had asignificantly higher tibiocalcaneal angle than did the Ponsetitreatment group (90� compared with 84�; p = 0.0023). In bothgroups, this measurement varied considerably from the mean of70� for age-matched controls13. The higher tibiocalcaneal angleobserved in the physical therapy group represents radiographicequinus and is in agreement with our kinematic data, whichshowed ankle equinus in association with 15% of the feet inthis group. The radiographic finding of equinus and the lack ofdorsiflexion during stance phase prompted our center to uti-lize early Achilles tendon lengthening in patients being treatedwith physical therapy if dorsiflexion is not achieved. We an-ticipate that this will also decrease the need for more extensivesurgery, such as posterior or posteromedial release, which wasnecessary in 36% of the clubfeet treated with physical therapybut only 23% of the feet treated with the Ponseti method.

In conclusion, approximately half of our two-year-oldpatients who had been successfully treated with either thePonseti or the French functional nonoperative method hadnormal kinematic ankle motion in the sagittal plane. When

deviations from normal ankle kinematics occurred, they gen-erally consisted of a mild limitation in dorsiflexion in thephysical therapy group or a slight increase in dorsiflexion in thePonseti cast-treatment group. These differences may have beenthe result of the frequent performance of a percutaneousAchilles tendon lengthening as part of the Ponseti cast tech-nique but not as part of the French functional program. How-ever, despite these differences, both of these nonoperativemethods resulted in rates of normal ankle kinematics that werehigher than what had been previously reported following thesurgical correction of clubfoot6,14. Therefore, we continue torecommend nonoperative treatment, with either method, forbabies with idiopathic clubfoot. Future studies comparing thegait and function of these two treatment methods are war-ranted as early treatment of ankle equinus becomes a part ofthe physical therapy protocol. n

Note: The authors thank Richard Browne, PhD, for helping them to design and perform thestatistical analyses for this study. They also thank the Movement Science Laboratory staff, whospent hundreds of hours collecting data on two-year-old patients.

Ron El-Hawary, MD, MScDivision of Pediatric Orthopaedics, Isaac Walton Killam Health Centre,5850 University Avenue, P.O. Box 9700, Halifax, NS B3K-6R8, Canada

Lori A. Karol, MDKelly A. Jeans, MSB. Stephens Richards, MDDepartment of Orthopaedics (L.A.K. and B.S.R.) and Movement ScienceLaboratory (K.A.J.), Texas Scottish Rite Hospital for Children,2222 Welborn Street, Dallas, TX 75219. E-mail address for L.A. Karol:lori.karol@tsrh.org

References

1. Turco VJ. Surgical correction of the resistant club foot. One-stage posteromedialrelease with internal fixation: a preliminary report. J Bone Joint Surg Am.1971;53:477-97.

2. McKay DW. New concept of and approach to clubfoot treatment: section II—correction of the clubfoot. J Pediatr Orthop. 1983;3:10-21.

3. Simons GW. Complete subtalar release in club feet. Part I—a preliminary report.J Bone Joint Surg Am. 1985;67:1044-55.

4. Kranicz J, Than P, Kustos T. Long-term results of the operative treatment ofclubfoot: a representative study. Orthopedics. 1998;21:669-74.

5. Noonan KJ, Richards BS. Nonsurgical management of idiopathic clubfoot. J AmAcad Orthop Surg. 2003;11:392-402.

6. Karol LA, Concha MC, Johnston CE 2nd. Gait analysis and muscle strength inchildren with surgically treated clubfeet. J Pediatr Orthop. 1997;17:790-5.

7. Cooper DM, Dietz FR. Treatment of idiopathic clubfoot. A thirty-year follow-upnote. J Bone Joint Surg Am. 1995;77:1477-89.

8. Masse P. [Le traitement du pied bot par la methode ‘‘fonctionnelle’’]. In:Cahier d’enseignement de la SOFCOT. Paris: Expansion Scientific; 1977. p 51-6.French.

9. Bensahel H, Guillaume A, Czukonyi Z, Desgrippes Y. Results of physicaltherapy for idiopathic clubfoot: a long-term follow-up study. J Pediatr Orthop.1990;10:189-92.

10. Dimeglio A, Bonnet F, Mazeau P, De Rosa V. Orthopaedic treatment andpassive motion machine: consequences for the surgical treatment of clubfoot.J Pediatr Orthop B. 1996;5:173-80.

11. Dimeglio A, Bensahel H, Souchet P, Mazeau P, Bonnet F. Classification ofclubfoot. J Pediatr Orthop B. 1995;4:129-36.

12. Ponseti IV. Congenital clubfoot: fundamentals of treatment. Oxford: OxfordUniversity Press; 1996.

13. Vanderwilde R, Staheli LT, Chew DE, Malagon V. Measurements on radio-graphs of the foot in normal infants and children. J Bone Joint Surg Am.1988;70:407-15.

14. Karol LA, O’Brien SE, Wilson H, Johnston CE, Richards BS. Gait analysis inchildren with severe clubfeet: early results of physiotherapy versus surgical re-lease. J Pediatr Orthop. 2005;25:236-40.

15. Aronson J, Puskarich CL. Deformity and disability from treated clubfoot.J Pediatr Orthop. 1990;10:109-19.

16. Yngve DA. Foot-progression angle in clubfeet. J Pediatr Orthop. 1990;10:467-72.

17. Yamamoto H, Muneta T, Furuya K. Cause of toe-in gait after posteromedialrelease for congenital clubfoot. J Pediatr Orthop. 1994;14:369-71.

18. Davies TC, Kiefer G, Zernicke RF. Ankle and first metatarsophalangealjoint dorsiflexion in children with clubfoot. J Pediatr Orthop. 2001;21:727-30.

19. Beyaert C, Haumont T, Paysant J, Lascombes P, Andre JM. The effect ofinturning of the foot on knee kinematics and kinetics in children with treatedidiopathic clubfoot. Clin Biomech (Bristol, Avon). 2003;18:670-6.

20. Theologis TN, Harrington ME, Thompson N, Benson MK. Dynamic foot move-ment in children treated for congenital talipes equinovarus. J Bone Joint Surg Br.2003;85:572-7.

21. O’Brien SE, Karol LA, Johnston CE 2nd. Calcaneus gait following treatmentfor clubfoot: preliminary results of surgical correction. J Pediatr Orthop B.2004;13:43-7.

1515

TH E J O U R N A L O F B O N E & JO I N T SU R G E RY d J B J S . O R G

VO LU M E 90-A d NU M B E R 7 d J U LY 2008GA I T AN A LY S I S O F CH I L D R E N TR E AT E D F O R CLU B F O O T W I T H

PH Y S I C A L TH E R A P Y O R T H E P O N S E T I CA S T TE C H N I Q U E

22. Muratli HH, Dagli C, Yavuzer G, Celebi L, Bicximoglu A. Gait characteristicsof patients with bilateral club feet following posteromedial release procedure.J Pediatr Orthop B. 2005;14:206-11.

23. Ponseti IV. Clubfoot management. J Pediatr Orthop. 2000;20:699-700.

24. Herzenberg JE, Radler C, Bor N. Ponseti versus traditional methodsof casting for idiopathic clubfoot. J Pediatr Orthop. 2002;22:517-21.

25. Alkjaer T, Pedersen EN, Simonsen EB. Evaluation of the walking pattern inclubfoot patients who received early intensive treatment. J Pediatr Orthop.2000;20:642-7.

26. Gage JR. An overview of normal walking. Instr Course Lect. 1990;39:291-303.

27. Dimeglio A, Canavese F, Bonnet F, Bentahar T. Clubfoot: functionaltreatment and passive motion machine. A fifteen years experience.Paper No 30. POSNA Annual Meeting; 2006 May 5; San Diego,California.

28. Morcuende JA, Abbasi D, Dolan LA, Ponseti IV. Results of anaccelerated Ponseti protocol for clubfoot. J Pediatr Orthop. 2005;25:623-6.

1516

TH E J O U R N A L O F B O N E & JO I N T SU R G E RY d J B J S . O R G

VO LU M E 90-A d NU M B E R 7 d J U LY 2008GA I T AN A LY S I S O F CH I L D R E N TR E AT E D F O R CLU B F O O T W I T H

PH Y S I C A L TH E R A P Y O R T H E P O N S E T I CA S T TE C H N I Q U E