`
`ISSN: 0001-6470 (Print) (Online) journal homepage: http:llwww.tandfonline.com/loi/iort19
`
`Correlations Between Changes in Gait and in
`Clinical Status After Knee Arthroplasty
`
`Gunnar B. J. Andersson, Thomas P. Andriacchi &Jorge 0. Galante
`
`To cite this article: Gunnar B. J. Andersson, Thomas P. Andriacchi & Jorge 0. Galante (1981)
`Correlations Between Changes in Gait and in Clinical Status After Knee Arthroplasty, Acta
`Orthopaedica Scandinavica, 52:5, 569-573, DOI: 10.3109/17453678108992148
`
`To link to this article: http://dx.doi.org/10.3109/17453678108992148
`
`© 1981 lnforma UK Ltd All rights reserved:
`reproduction in whole or part not permitted
`
`Published online: 08 Jul 2009.
`
`~~ Submit your article to this journal 8"
`
`l.illl Article views: 210
`~ ....
`
`.:.I View related articles 8"
`
`~ Citing articles: 24 View citing articles 8"
`
`Full Terms & Conditions of access and use can be found at
`http://www.tandfonllne.com/action/journallnformation?journa1Code=iort19
`
`Download by: [66.201.48.146]
`
`Date: 25 August 2017, At: 13:22
`
`1
`
`
`
`Acta orthop. scand. 52, 569-573, 1981
`
`CORRELATIONS BETWEEN CHANGES IN GAIT AND IN
`CLINICAL STATUS AFfER KNEE ARTHROPLASTY
`
`GUNNAR B. J. ANDERSSON, THOMAS P. ANDRIACCHI & JORGE 0. GALANTE
`
`Department of Orthopaedic Surgery I, Sahlgren Hospital, Goteborg, Sweden, and Department of Orthopaedic
`Surgery, Rush-Presbyterian-St. Luke's Medical Center, Chicago, Illinois, USA
`
`Twenty-six knee arthroplasties in 22 subjects were studied clinically and. in a gait
`labor·atory before and on one or more occasions after surgery. The purpose was to
`identify changes in gait characteristics and relate them to changes in the clinical
`situation of the patients.
`The gait patterns of the patients changed following surgery mainly with respect to
`their time-distance characteristics; stride length, cadence, time of swing, and time of
`support. Positive correlations were noted between these gait changes and the pa(cid:173)
`tient's pain, his limp, and his walking distance. It is concluded that the information
`obtained in a sophisticated gait laboratory is of limited value to the surgeon in his
`clinical assessment.
`
`Key words: biomechanical analysis; clinical assessment; gait; gait analysis; knee
`arthroplasty
`Accepted 2.ii.81
`
`While the main indication for knee arthroplasty is
`pain, a further objective is to improve the pa(cid:173)
`tient's functional ability, and more specifically,
`his gait. This is achieved in part by pain allevia(cid:173)
`tion, in part by correction of deformities, and in
`part by replacement of the articulating surfaces.
`In addition to qualitative observations and
`clinical evaluations, quantitative biomechanical
`measurements have been used
`to describe
`changes in the function of subjects with knee dis(cid:173)
`abilities before and after surgery of the knee
`(Andriacchi et al. 1977, Arborelius et al. 1976,
`Chao 1975, Chao & Stauffer 1975, Collopy et al.
`1977, Kettelkamp et al. 1972, ·Stauffer et al.
`1977, Stauffer et al. 1975).
`Gait can be characterized quantitatively by
`time-distance measurements, such as cadence,
`step-length, time of support and time of swing,
`and by ground-force measurements. As these gait
`parameters are known to change with a change in
`walking speed, both in normal and gait-impaired
`
`subjects, gait information is useful only when
`considered in relation to walking speed (Andriac(cid:173)
`chi et al. 1977, Cavanagh & Gregory 197 5,
`Grieve & Gear 1966, Jacobs et al. 1972,
`Lamoreux 1971, Murray et al. 1966).
`Only a few studies exist in which quantitative
`changes in gait are related to the patient's clinical
`assessment (Kettelkamp et al. 1972, Stauffer et
`al. 1975). Kettelkamp et al. (1972) studied 41
`rheumatoid knees, and found that cadence was
`significantly related to pain, and that the stride
`length increased when the range of flexion of the
`knee increased. Stauffer et al. (1977) studied 37
`knee patients before and after a Macintosh
`hemiarthroplasty. Overall, the preoperative and
`postoperative gait characteristics were not related
`significantly to clinically evaluated results. In
`none of these studies was walking speed
`documented, so it is difficult to assess the re(cid:173)
`ported results.
`The purpose of this investigation was to study
`
`0001-6470/81/050569-05 $02.50/0
`
`© 1981 Munksgaard, Copenhagen
`
`2
`
`
`
`570
`
`G. B. J. ANDERSSON ET AL.
`
`changes in gait characteristics following surgery,
`and to relate them to changes in the clinical status
`of the patients. The paper does not attempt to
`describe in detail the clinical results of the knee
`arthroplasties.
`
`PATIENTS AND METHODS
`The study included 26 knee arthroplasties in 22 sub(cid:173)
`jects, 14 females and 8 males. The mean age of the
`subjects was 60 years (range 30-79 years). Weights
`ranged from 50 to 101 kg (mean weight 77 kg) and
`heights from 154 cm to 193 cm (mean height 166.5
`cm).
`The clinical diagnosis was primary osteoarthritis in 15
`knees, rheumatoid arthritis in 9 and in one each post
`traumatic osteoarthritis and avascular necrosis. Three
`different types of prostheses were used; 10 Polycentric,
`10 Gunston, and 6 Geomedic, i.e. 20 unconstrained
`prostheses and 6 partially constrained. Simultaneous
`involvement of one or more other joints was present in
`22 subjects; 17 had knee involvement, one hip in(cid:173)
`volvement, and 3 both knee and hip involvement.
`Gait measurements were made on all subjects just
`before surgery and at one or more of the following
`periods postoperatively; 3 to 6 months, 15 knees; 7 to
`12 months, 18 knees; and 13 to 36 months, 10 knees.
`Eighteen knees were seen in the gait laboratory more
`than once postoperatively (Table 1 ).
`A clinical evaluation was made each time gait was
`quantitatively evaluated. The clinical data were noted
`using a 133 item form previously published (Freeman et
`al. 1977), to assess pain and functional activities and to
`record the physical findings.
`The instrumentation used included photocells, foot
`switches, and a multicomponent force plate. The data
`was acquired and processed using a minicomputer.
`Average velocity was derived from the on-off closure
`of a series of five photocells placed at 2.5 meter inter(cid:173)
`vals along the walkway. Velocity was computed at each
`interval.
`Cadence, time of support and time of swing were
`measured using instrumented foot pads placed inside
`the subject's shoes. The foot pad consisted of an insole
`with four pressure sensitive transducers located at the
`heel, the heads of the fifth and first metatarsal bones
`and the big toe. The on-off closure patterns of the
`transducers were transmitted through a cable attached
`to an overhead trolley which trailed the subject while
`walking.
`The information from the foot switches and the photo
`cells was combined into 13-bit digital words and sam(cid:173)
`pled at the rate of 100 Hz. Each bit was decoded in the
`computer to provide times of switch closures and
`openings.
`Time of swing, and time of support were derived for
`each limb independently. Time of swing was defined as
`
`Table I. Number of subjects and arthroplasties studied at
`each time of observation
`
`Time of
`observation
`(months)
`
`Pre-op
`3-6
`7-12
`13+
`
`*
`
`Knees
`
`*
`
`Patients
`
`26
`16
`19
`10
`
`22
`16
`15
`10
`
`the amount of time during which all switches were open
`and time of support as the amount of time any switch
`was closed.
`The Kristal piezoelectric force plate (Kistler, Swit(cid:173)
`zerland, 40 x 60 cm) was mounted inconspicuously in
`the center of the walkway and used to measure dynami(cid:173)
`cally the three components of foot-ground reaction
`force; media-lateral (X), antero-posterior (Y), and
`vertical (Z). The analog signals from the force plate
`were digitized at a sample rate of 200 Hz and proces(cid:173)
`sed.
`Each gait measurement included 20 traverses of a
`ten-meter walkway. The subject was asked to walk a
`prescribed number of times at his normal walking
`speed, at fast walking speed and at slow walking speed.
`No specific instructions were given to the subjects to
`step on the force plate to avoid alterations in normal
`walking patterns. A typical test, therefore, did not pro(cid:173)
`duce 20, but only 10 steps on the measuring surface
`suitable for analysis. Ground reaction force measure(cid:173)
`ments were recorded along with the temporal meas(cid:173)
`urements on magnetic tape for further processing.
`
`Data analysis
`
`The times were averaged over one traverse of the
`walkway (usually about 5 strides) while the subject was
`walking at uniform velocity. Thus, 20 traverses of the
`walkway yielded 20 measurements of swing, and sup(cid:173)
`port for the left and right limbs, over a wide range of
`walking speeds. Average cadence and stride length
`were derived from the foot switch and photo cell meas(cid:173)
`urements.
`Temporal and force plate data were interpreted and
`compared by observing each parameter plotted against
`walking speed. A least square regression analysis was
`used to fit linear, quadratic, and cubic polynomials to
`each data set.
`Each traverse of the walkway produced a single data
`point for each of the temporal and ground reaction
`force parameters. All data points obtained in one
`measurement were plotted against velocity. Data for
`each gait parameter obtained in the knee patients were
`then compared to those previously obtained for normal
`subjects of the same stature (Andriacchi et al. 1977) by
`
`3
`
`
`
`1.50
`
`1.00
`
`en
`0:: w
`I-
`w
`~
`
`I
`I-
`C> z
`w
`_J
`w
`0
`0::
`I-
`(/)
`
`0
`
`0.50
`0.4
`
`GAIT ANALYSIS AND KNEE ARTHROPLASTY
`
`571
`
`+
`._Normal Subject
`x (Data collected on two
`+ separate occasions )
`
`6 and9mos.
`Pos -op
`0
`
`> Knee Patient
`
`0.6
`
`1.2
`1.0
`0.8
`VELOCITY (METERS/SEC.)
`
`1.4
`
`1.6
`
`fi"gure /.An illustration of the stride-length-velocity relationships for a normal subject, and a patient with a total knee
`replacement observed 3, 6 and 9 months postoperatively.
`
`plotting corresponding curves on the same graph, as
`illustrated in Figure 1. Then, the entire curve rather
`than each single measurement was considered.
`The clinical data was entered into the computer and
`correlations established with the gait data. Changes in
`gait and in clinical data were compared individually as
`well as in combinations.
`
`RESULTS
`
`Clinical results
`
`When comparing the history of the patients be(cid:173)
`fore and after surgery the important clinical
`changes were a decrease in pain, which occurred
`in all subjects, and an increase in walking dis(cid:173)
`tance, which occurred in all but 3 patients. Other
`functional abilities, such as stair climbing and
`rising from a chair without the use of arm rests,
`improved in 11 and 9 patients, respectively. The
`physical examination revealed a better alignment
`of the leg in all knees with a preoperative varus or
`valgus deformity, and stability improved in all
`unstable knees. A less severe limp was noted in
`
`19 of the patients. The range of motion improved
`by an average of 10 degrees in 9 knees, decreased
`by an average of 10 degrees in 7, and was un(cid:173)
`changed in the remainder.
`
`Gait analysis
`
`Because of the purpose of the present paper ab(cid:173)
`solute values of the gait parameters are not given.
`They can be obtained on request from the au(cid:173)
`thors. The gait parameters of each knee patient
`were compared to the gait measurements obtained
`from a healthy subject of the same stature and
`sex. These subjects have been described in a pre(cid:173)
`vious study (Andriacchi et al. 1977). The meas(cid:173)
`urements were compared over a range of walking
`speeds as illustrated for stride length in Figure 1.
`The measurement from a patient was considered
`normal only if the entire velocity-relationship of a
`particular gait parameter was the same as for that
`of the healthy subject. A more complete descrip(cid:173)
`tion of the approach was given in a previous pub(cid:173)
`lication (Andriacchi et al. 1977).
`
`4
`
`
`
`572
`
`G. B. J. ANDERSSON ET AL.
`
`An increase in the range of voluntary walking
`speed was found in 18 of the 22 subjects. In 3
`patients walking speed range remained un(cid:173)
`changed, and in one it decreased. The increase in
`walking speed range was, as a rule, continuous up
`to 1 year after surgery; only minor changes
`occurred thereafter.
`Of the gait characteristics observed, four were
`found to be particularly indicative of changes in
`gait; stride length, time of swing, time of support
`and degree of gait asymmetry.
`The preoperative stride length was below nor(cid:173)
`mal in 19 subjects (22 knees), and normal in 3 ( 4
`knees). Postoperatively, improvement was found
`in 11 patients (14 knees), and deterioration in 3
`(3 knees). The time of swing preoperatively was
`below normal in all but 2 patients (2 knees).
`Postoperatively, changes were found in 17 pa(cid:173)
`tients (17 knees); 12 improved, and 5 deterior(cid:173)
`ated. Time of support postoperatively was below
`normal in 12 patients (14 knees), and normal in
`10 (12 knees). In 8 patients (8 knees) improve(cid:173)
`ment was noted, in 2 (2 knees) deterioration
`occurred, and the remainder were unchanged.
`Asymmetry during stance and swing phase
`changed in 9 subjects; 6 improved, and 3 de(cid:173)
`teriorated.
`The ground-reaction force amplitudes also
`changed, but in general were not as sensitive to
`gait changes as were the temporal measurements.
`Changes in medio-lateral and antero-posterior
`forces occurred in 3 patients only, while changes
`in the vertical forces occurred in 4.
`
`Correlation between clinical data and gait
`observations
`
`The change in gait velocity was closely related to
`changes in pain. In the 4 subjects in whom the
`range of walking speed was unchanged or de(cid:173)
`creased, there was no relief or incomplete relief
`of day pain and walking pain in 3, while the
`fourth was considered to have a normal walking
`speed range even before surgery. A further indi(cid:173)
`cation of the relationship of range of walking
`speed to pain was that these sometimes changed
`between different periods of observations, but
`always in tandem. Similar positive correlations
`were found between walking speed range and
`
`both the degree of limp, evaluated clinically, and
`the maximum walki11g distance.
`The temporal changes in gait related primarily
`to the changes in pain and in walking distance.
`They also related to simultaneous involvement of
`one or more other joints which was present in 22
`subjects; 17 with knee involvement, one with hip
`involvement and 3 with both hip and knee in(cid:173)
`volvement.
`All patients who showed, postoperatively, an
`increase in ground-reaction force amplitudes also
`reported less pain postoperatively.
`
`DISCUSSION
`
`Following knee arthroplasty the gait patterns of
`patients with knee disabilities changed mainly
`with respect to time-distance characteristics;
`stride length, cadence, time of swing, time of sup(cid:173)
`port. Since these characteristics are all velocity
`dependent, it would be possible that the change
`resulted only from the postoperative increase in
`average walking speed. But, the temporal gait
`characteristics changed even when they were
`compared at the same walking speed as before
`surgery. It seems, therefore, that the change was
`due also to other factors. Such factors could be
`the decrease in pain when walking, and the less
`severe limp. Another explanation might be the
`change in stride length. Preoperatively, the sub(cid:173)
`jects used short rapid steps requiring little single
`support time. Postoperatively the step length in(cid:173)
`creased which influenced the other gait paramet(cid:173)
`ers. The velocity dependence of the temporal gait
`characteristics is so marked, however, that gait
`observations which do not take walking speed
`into account can arrive at wrong conclusions.
`In a second study it was shown that the shorter
`than normal stride length found in patients was
`associated with a less than normal knee flexion
`during stance phase and an abnormal use of the
`extensor and flexor muscles of the knee (An(cid:173)
`driacchi et al. 1980).
`It was gratifying to note that gait became more
`normal after surgery. Even so, the gait did not
`become completely normal. This was in part due
`to the fact that all but one subject had multiple
`joint involvement. The findings, however, em-
`
`5
`
`
`
`GAIT ANALYSIS AND KNEE ARTHROPLASTY
`
`573
`
`phasize that normal gait cannot be expected in
`patients following today's knee joint replacement
`surgery. The design of the implant, the operative
`technique, and the difficulties obtaining stability
`can all be factors of importance in this respect.
`The information obtained in a sophisticated
`gait laboratory appears to be of limited value to
`the surgeon in this clinical assessment. Such
`measurements are time-consuming and expensive
`and can perhaps be replaced by questions about
`the walking distance, observations of the patient's
`limp, and measurements of his maximum walking
`speed in a simple laboratory environment since
`these three factors relate directly to gait. The
`usefulness of gait observations in the laboratory
`lies more in the classification of particular gait
`abnormalities associated with joint pathologies
`and neuromuscular disorders and in the study of
`forces and moment acting at the joints for pur(cid:173)
`poses of the design of artificial joints.
`
`CONCLUSIONS
`
`The study shows that measurements of temporal
`and ground reaction forces can be used to indi(cid:173)
`cate gait changes following knee arthroplasties.
`Changes were found primarily in the temporal
`measurements which were clearly velocity de(cid:173)
`pendent. Positive correlations were noted be(cid:173)
`tween these temporal gait changes and the pa(cid:173)
`tient's pain, his limp and his walking distance.
`
`ACKNOWLEDGEMENT
`This work was supported by grants from The Scholl
`Foundation, The Arthritis Foundation, NIH AM*
`00498 and AM* 20702. The study was made in the
`John L. and Beatrice Keeshin Human Motion Research
`Laboratory.
`
`REFERENCES
`
`Andriacchi, T. P., Ogle, J. A. & Galante, J. 0. (1977)
`Walking speed as a basis for normal and abnormal
`gait measurements. J. Biomech. 10, 261.
`Andriacchi, T. P., Galante, J. 0. & Fermier, R. W.
`(1980) Gait evaluation following total knee replace(cid:173)
`ment: A comparative study. Proc. 26th Orthop. Res.
`Soc., p. 309.
`Arborelius, M. M., Carlsson, A. S. & Nilsson, B. E.
`(1976) Oxygen intake and walking speed before and
`after total hip replacement. Clin. Orthop. 121, 113.
`Cavanagh, P. R. & Gregory, R. J. (1975) Knee joint
`torque during the swing phase of normal treadmill
`walking. J. Biomech. 8, 337.
`Chao, E. Y. (1975) Functional evaluation of total knee
`replacement patients through gait analysis. ASME
`Publication, No. 75-APMB-5.
`Chao, E. Y. & Stauffer, R. M. (1975) Biomechanical
`evaluation of geometric and polycentric knee ar(cid:173)
`throplasty. In: Total knee replacement, pp. 52---00,
`Instm. Mech. Engrs., London.
`Collopy, M. C., Murray, M. P., Gardner, G. M., Di
`Ulio, R. A. & Gore, D. R. (1977) Kinesiologic
`measurements of functional performance before and
`after geometric total knee replacement. One-year
`follow-up of twenty cases. Clin. Orthop. 126, 196.
`Freeman, M. A. R., Todd, R. C. & Cundy, A. D.
`(1977) A technique for recording the results of knee
`surgery. Clin. Orthop. 128, 216.
`Grieve, D. W. & Gear, R. J. (1966) The relationship
`between the length of stride, step frequency, time of
`swing and speed of walking for children and adults.
`Ergonomics 9, 379.
`Jacobs, N. A., Skorecki, J. & Charnley, J. (1972)
`Analysis of the vertical component of force in normal
`and pathological gait. J. Biomech. S, 11.
`Kettelkamp, D. B., Leaverton, P. E. & Misol, S. (1972)
`Gait characteristics of the rheumatoid knee. Arch.
`Surg. 104, 30.
`Lamoreux, L. (1971) Kinematic measurements in the
`study of human walking. Bull. Pros. Res., pp. 3-8,
`Biomechanics Lab, University of California.
`Murray, M. P., Kory, R. C., Clarkson, B. H. & Sepic, S.
`B. (1966) Comparison of free and fast speed walking
`patterns of normal men. Am. J. Phys. Med. 45, 8.
`Stauffer, R. M., Chao, E. Y. S. & Gyory, A. N. (1977)
`Biomechanical gait analysis of the diseased knee
`joint. Clin. Orthop. 126, 246.
`Stauffer, R., Kettelkamp, D. B., Thompson, C. &
`Wenger, D. (1975) The Macintosh prosthesis. Pros(cid:173)
`pective clinical and gait evaluation. Arch. Surg. 110,
`717.
`
`Correspondence to: Gunnar Andersson, M.D., Department of Orthopaedic Surgery I, Sahlgren Hospital,
`S-413 45 Goteborg, Sweden.
`
`6
`
`