Content uploaded by Ingo Haase
Author content
All content in this area was uploaded by Ingo Haase on Jun 08, 2016
Content may be subject to copyright.
Authors:
Uwe Eisermann, PhD
Ingo Haase, PhD
Bernd Kladny, MD
Affiliations:
From the University of Applied
Sciences KufsteinTirol, Kufstein,
Austria (UE); Research and Quality
Management, Clinic Group
Enzensberg, Füssen, Germany (IH);
and the Department of Orthopedic
Rehabilitation, Rehabilitation Clinic
Herzogenaurach, Herzogenaurach,
Germany (BK).
Disclosures:
Siemens paid benefits to a research
fund of a health institution
(Rehabilitation Clinic
Herzogenaurach) with which two of
the authors (B. Kladny, I. Haase) are
affiliated or associated.
FIM™ is a trademark of the Uniform
Data System for Medical
Rehabilitation, a division of UB
Foundation Activities, Inc.
Correspondence:
All correspondence and requests for
reprints should be addressed to Uwe
Eisermann, PhD, Ing. A. Gerber-Str.
4, A-6330 Kufstein, Austria.
0894-9115/04/8309-0670/0
American Journal of Physical
Medicine & Rehabilitation
Copyright © 2004 by Lippincott
Williams & Wilkins
DOI: 10.1097/01.PHM.0000137307.44173.5D
Computer-Aided Multimedia
Training in Orthopedic
Rehabilitation
ABSTRACT
Eisermann U, Haase I, Kladny B: Computer-aided multimedia training in or-
thopedic rehabilitation. Am J Phys Med Rehabil 2004;83:670–680.
Objective: To compare a computer-aided training program with a conven-
tional training program in orthopedic rehabilitation.
Design: The study was a randomized, nonblinded, controlled trial in which
follow-up data were obtained at 6 mos. In an inpatient rehabilitation center, a
consecutive sample was taken of patients with first total hip replacements or
first total knee replacements 23– 42 days after surgery. Indication groups were
examined separately. The study population included 189 women and 85 men.
Mean age was 69 yrs (38–86 yrs). Patients received either computer-aided
training (case group) or conventional training (control group) within the frame-
work of their inpatient rehabilitation program. The main outcome measures
were levels of acceptance and effectiveness (Harris Hip Score, Hospital for
Special Surgery Score, FIM™ instrument, and Hanover Functional Ability
Questionnaire).
Results: Both forms of training showed significant improvements until dis-
charge in scores and items used for the result evaluation independent of
patient sex, age, and educational level. The 6-mo follow-up showed that be-
tween the groups, there was no statistically significant difference in the level of
improvement concerning functional capacity. Furthermore, patients displayed
their acceptance of the system by rating it with average values between
“good” and “very good.”
Conclusions: For patients with total hip replacements or total knee replace-
ments, computer-aided training can be regarded as the equivalent to conven-
tional training in relationship to the results of the rehabilitation program. The
system is a new tool in orthopedic rehabilitation. To identify the relative impor-
tance of the system, further research is needed.
Key Words: Rehabilitation, Total Hip Replacement, Total Knee Replace-
ment, Tele-rehabilitation, Computer-Aided Training
670 Am. J. Phys. Med. Rehabil. ●Vol. 83, No. 9
Research Article
Orthopedics
Orthopedic knee and hip impair-
ments can lead to permanent disabil-
ities because of lack of training dur-
ing and after inpatient rehabilitation.
A new concept developed in Germany
supplements this training with re-
mote monitoring and periodic reas-
sessment. This concept, called Uni-
versal Training Assistant, is a tele-
rehabilitation system consisting of
hardware and software components
with training and training analysis
and with information and communi-
cation functions. It enables a patient
at all stages of orthopedic rehabilita-
tion to benefit from remote medical
or therapeutic treatment.
Tele-rehabilitation in general is
the remote delivery of rehabilitative
services such as monitoring and
training of persons with disabilities
using telecommunication technolo-
gy.
1
Reported findings from early ex-
ploratory trials are encouraging.
2–4
However, there have been no ran-
domized, controlled trials demon-
strating clinical effectiveness in reha-
bilitation so far.
The Universal Training Assistant
was tested in a German rehabilitation
clinic with respect to effectiveness
using this new type of training
method. The basic idea was to com-
pare this computer-aided training us-
ing the Training Assistant with train-
ing provided in a conventional
manner. The intention was to raise
the levels of acceptance (measured in
terms of declaration of consent) and
effectiveness (effects on functional
status, activities of daily living, pain,
etc.).
Conceptually, the idea was to de-
velop comprehensive, closed-loop
tele-medical treatment and care for
patients undergoing orthopedic reha-
bilitation. In this context, the word
“comprehensive”not only stands for
defining ideal individual training pro-
grams for the patient by the thera-
pist, but also for analyzing the pa-
tient’s current training performance
and course of therapy, keeping the
patient informed and involved and
communicating and interacting with
him or her. The closed loop is defined
by qualitative and quantitative vari-
ables entered manually by the patient
or automatically calculated for train-
ing and course of therapy, which are
transferred on completion of training
to the therapist, such as details of
intensity and frequency of training.
Based on these results and the infor-
mation originating from the interac-
tion with the patient, the therapist
specifies further individual optimized
training programs. The system also
allows incorporation of new exercises
into the patient’s training schedule to
keep the treatment up to date in ac-
cordance with current guidelines and
the patient’s training progress.
METHODS
Patients received either comput-
er-aided training (case group) or con-
ventional self-training (control
group) within the framework of their
3- or 4-wk inpatient rehabilitation
program. To perform the computer-
aided training, the patient was in-
structed for half an hour by a phys-
iotherapist, who explained the system
and selected and modified the exer-
cises for the patient on the Training
Assistant. In accordance with their
individual capacities, the patients
used the system 3–5 times a week for
30 mins without supervision. The
conventional self-training was per-
formed in a group of 8 –10 patients
under supervision of a physiothera-
pist. The patients received an exercise
program that was either performed
without or with simple tools like balls
or rubber bands. This training was
performed 3–5 times a week for 30
mins. The subjects were assessed dur-
ing a 4-mo interval between Decem-
ber 2000 and March 2001. A compre-
hensive program of inpatient
rehabilitation was provided to each
patient, tailored to meet each pa-
tient’s skills and abilities. The follow-
ing services were provided as part of
the rehabilitation program to both
groups: medical services, nursing,
physical therapy, and occupational
therapy.
Description of the Intervention
Computer-Aided System. The Uni-
versal Training Assistant is used to
perform a selection of computer-
aided, multimedia, real-time training
exercises relevant to rehabilitation of
motor abilities and skills (basic train-
ing and special training). The system
features two different versions espe-
cially designed for the needs of the
user groups, patients, and therapists.
In the therapist’s version, the special-
ist creates the ideal training program
for the individual performance capac-
ity of the patient by defining suitable
exercises (Fig. 1).
The exercises are grouped into
“pools”for specific indications and
correspond exactly with those used in
conventional rehabilitative training.
In addition, the therapist can set
analysis variables for each exercise to
obtain feedback on training perfor-
mance. Depending on the specifica-
tions, devices for describing or re-
cording movement, such as
accelerometers, Web cams, chest sen-
sors, and wristbands, are activated
automatically. By implementing
these devices into the training sched-
ule, the visualization and interpreta-
tion of the training and the assess-
ment of the patient’s compliance is
feasible. Furthermore, the therapist
can specify questions the patient has
to answer at the end of a training unit
(e.g., about the course of training and
acceptance, perceived pain, and
“strenuousness;”these factors are
important for planning the course of
further training). Finally, the thera-
pist can receive a record of the train-
ing unit (date and time, duration,
type and execution of the training,
answers, and patient’s remarks) and
also refer back to the patient record
and send news to the patient.
September 2004 Computer-Aided Training 671
The patient’s version offers the
training programs created by the
therapist (i.e., the Training Assistant
conducts the patient through the
program). Assistance is available for
each exercise, and each set comprises
movement description, training load,
and multimedia assists (video anima-
tions and audio sequences). To en-
hance communication, the patient
has the opportunity to send a mes-
sage to the therapist. With the aim to
offer quick and reliable program op-
eration, especially for computer-illit-
erate users, a simple, clear, and visu-
ally attractive interface has been
designed that features large buttons
and lettering and virtually excludes
the possibility of input errors (Fig. 2).
Structure. All data generated by the
patient using the patient station are
transferred to a server of the service
supplier or provider. The therapist
can review the data using his or her
therapist station, evaluate it, and if
appropriate, select a new training
program adapted to the altered abili-
ties and skills of the patient. This
program will then again be trans-
ferred to the server and will appear
during the next training session. The
basis for this system structure (Fig.
3) is a client-server communication
infrastructure consisting of a central
server and dedicated clients (patient
and therapist workstations). This in-
frastructure makes possible tele-
training, tele-monitoring, and tele-
coaching, and it fulfills all security-
relevant criteria (e.g., data security,
authentication).
Subjects
Selection criteria for the sample
included all persons referred to a re-
habilitation clinic within 5 mos with
total hip replacement or total knee
replacement. Other prerequisites
were that patients had to be able to
train and to fill in a questionnaire.
Patients who fulfilled these inclusion
criteria were randomized to receive
either computer-aided training with
the Training Assistant (case group) or
conventional training (control
group) within the frame of their in-
patient rehabilitation program.
Patients with hip arthroplasty
and patients with knee arthroplasty
were examined separately. The target
per study was at least 2 ⫻60 patients
(cases and controls). This database is
large enough to be able to draw con-
clusions concerning statistically sig-
nificant average values for both de-
pendent and independent samples
with an error probability of
␣
⫽0.05
and 1 ⫺

⫽0.80.
5
Figures 4 and 5 outline the
progress of subjects through the var-
ious phases of the randomized, con-
trolled trial. There were no statisti-
cally significant differences between
Figure 1: Example of patients’time schedule.
672 Eisermann et al. Am. J. Phys. Med. Rehabil. ●Vol. 83, No. 9
the groups for the baseline demo-
graphics of patients.
Instruments
For ascertaining aspects of func-
tional status, activities of daily living,
pain, and range of motion, established
assessment instruments were used:
Staffelstein-Score for total hip and total
knee replacements,
6
Harris Hip Score,
7
Hospital for Special Surgery Score,
8
and the FIM™instrument.
9–11
The
data were collected by a number of
rehabilitation physicians who had been
introduced to the instruments and
scores beforehand. Staffelstein-Score is
similar to Harris Hip Score and Hospi-
tal for Special Surgery Score. The
Hanover Functional Ability Question-
naire (FFbH)
12,13
was filled in by pa-
tients. The FFbH for patients with os-
teoarthritis is part of a series of short,
self-administered questionnaires for
the assessment of functional limita-
tions in activities of daily living among
patients with musculoskeletal disor-
ders. The resulting FFbH scores can
range between 0 (minimal functional
capacity) and 100 (maximal functional
capacity). A score of ⬎70 points has to
be viewed as normal functional capac-
ity and an improvement of ⱖ12% as
clinically meaningful.
Procedure
Data were collected during the
clinical inpatient period at two refer-
ence points: the time of admission
and the time of discharge. A 6-mo
follow-up was done to assess the
long-term benefit of the treatment on
patients using a questionnaire, which
included the FFbH.
Descriptive statistics were com-
puted for all relevant variables. Non-
parametric tests were performed to
examine differences regarding scores
of outcome measures. Change from
pretreatment to posttreatment was
analyzed using the Wilcoxon’s test.
Group averages were compared using
the Kolmogorov-Smirnov test. Data
were analyzed with the SPSS 10
package (SPSS, Chicago, IL).
In addition, computer-aided sur-
veys were done. These surveys were
used to establish the level of accep-
tance. They were carried out in the
following way: after a specified num-
ber of training units with the com-
puter, the patients of the case groups
were given a questionnaire by the
computer (i.e., they were questioned
on-line regarding the criteria manag-
ing the system, arrangement of the
training, training times, effectiveness
Figure 2: To offer quick and reliable program operation, especially for computer-illiterate users, a simple, clear, and
visually attractive interface has been designed that features large buttons and lettering and virtually excludes the
possibility of input errors.
September 2004 Computer-Aided Training 673
of the training, recommendations for
training, and enthusiasm for the
training). The patients assessed most
of the criteria on the basis of 5-point
scales ranging from ⫹2to⫺2. For
managing the system in general and
multimedia arrangement of the com-
puter-aided training there were mean
values ranging between “good”and
“very good.”
Supplementing the acceptance as-
pect, physician and therapist focus
group discussions
14
were held. Specific
kinds of data, surveyed using qualita-
tive methods, are only comparable with
a limited extent, but they offer a much
better picture of how the subject actu-
ally thinks and feels. The results ob-
tained from qualitative procedures
were combined with the results ob-
tained from quantitative procedures
and interpreted collectively.
RESULTS
Acceptance and Practicability
Dropouts. In total, 77 of 373 patients
(21%) meeting the inclusion criteria
declined to take part in the study. The
main reason for this was that these
patients (68 of 77, 88%) rejected
computer-aided training; apparently,
these patients had been put off by
warnings related to certain health
problems they had. For example, pa-
tients with poor eyesight were in-
formed that overexertion might affect
their vision. Three patients dropped
out of the study early because they
did not feel comfortable operating
the computer (Figs. 4 and 5).
Acceptance of Computer-Aided
Training from the Patients’ View-
point (Case Groups). To establish the
level of acceptance, all 142 patients of
the case groups were questioned on-
line (see above). Duration (of an in-
dividual training unit) and frequency
(i.e., number of training units per
week) were predominantly consid-
ered to be “reasonable.”Effectiveness
of the training, on average, was rated
1.4 (SD ⫽0.6) regarding physical fit-
ness and 1.3 (SD ⫽0.8) regarding
general strain, in each case on a
5-point scale ranging from ⫹2 (very
helpful) to ⫺2 (useless). A total of
131 of 142 cases (92.3%) would rec-
ommend this training to other pa-
tients with similar complaints, and
only one patient (0.7%) would not be
able to recommend it to others. As-
sessment of enthusiasm for training
was similarly positive.
Acceptance of Computer-Aided
Training from the Viewpoint of the
Figure 3: Client-server communication infrastructure consisting of a central server and dedicated clients (patient and
therapist workstations). This infrastructure makes possible tele-training, tele-monitoring, and tele-coaching, and it fulfills
all security-relevant criteria (e.g., data security, authentication).
674 Eisermann et al. Am. J. Phys. Med. Rehabil. ●Vol. 83, No. 9
Physicians and Therapists. To estab-
lish the level of acceptance, the phy-
sicians and therapists were invited to
give their assessments in a focus
group discussion on the criteria:
course of training, arrangement of
training, its effectiveness, supervi-
sion, and enthusiasm for training.
They were asked to observe the crite-
ria from their own viewpoint and to
describe the patients in this light (ac-
ceptance of the training by the pa-
Figure 4: Recruitment of patients with total hip replacements.
September 2004 Computer-Aided Training 675
tients as evaluated by the physicians
and therapists).
The first subject discussed was
the course of training, which was as-
sessed positively by the physicians
and therapists. They were able to op-
erate the Training Assistant and to
confirm that the patients were able to
handle the program. As far as validity
is concerned, this confirmation must
be examined in detail to be certain
that the patients were indeed able to
Figure 5: Recruitment of patients with total knee replacements.
676 Eisermann et al. Am. J. Phys. Med. Rehabil. ●Vol. 83, No. 9
operate the Training Assistant and
that the criterion “course of training”
in the on-line questionnaire has been
correctly evaluated in terms of their
abilities.
The arrangement of training was
also assessed positively. Physicians
and therapists alike regarded the ar-
rangements in general as attractive
and the movement descriptions as
vivid. They emphasized the impor-
tance of the video animations and the
audio sequences. However, it should
be mentioned that for some of the
patients, the movement descriptions
were not adequate.
Assessment of the training’s ef-
fectiveness was similar throughout,
to the extent that all regarded the
computer-aided training as being just
as effective as conventional training.
In fact, three participants (one physi-
cian, two therapists) thought that the
computer-aided training was more
effective because the patients were
more meticulous and conscientious
in carrying out the multimedia exer-
cises led by a computer-animated
trainer moving in real time. The phy-
sicians and therapists confirmed the
patients’self-assessments. The great
majority of the patients described the
computer-aided training as helpful
for their performance and recovery.
Training supervision was also re-
garded positively in the respect that
preparation of patients for the com-
puter-aided training was unproblem-
atic and with duration of 30 mins
claimed no more time than the prep-
aration required for conventional
training. Admittedly, it might be
helpful for the training if the thera-
pist carrying out the patient’s prepa-
ration were the same one who was
available during the training or su-
pervising it.
Corresponding with the positive
trend of enthusiasm for the training
shown by the patients, the physicians
and therapists would also continue to
make use of the Training Assistant,
for their in-, out-, and ambulatory
patients.
Randomized, Nonblinded,
Controlled Trial
Patients with Total Hip Replace-
ments. Patients with total hip re-
placement who fulfilled the inclusion
criteria were randomized to receive
either computer-aided training or
conventional training during their
inpatient rehabilitation. Table 1
shows that the baseline characteris-
tics of these patients were similar in
the two groups.
Both forms of training showed
significant improvements in scores
used for the result evaluation until
discharge. Effect sizes
15,16
ranged
from 0.67 to 1.34 (cases), respec-
tively, from 0.76 to 1.34 (controls)
(Table 2). There was no statistically
significant difference between these
two groups on Harris Hip Score,
Staffelstein Score, FIM score, and
FFbH.
At the 6-mo follow-up, all pa-
tients with total hip replacement who
had completed the inpatient rehabil-
itation program were mailed a ques-
tionnaire and asked to report their
functional status using the FFbH for
patients with osteoarthritis. Nine pa-
tients did not return the question-
naire (Fig. 4).
Six months after the program,
the average functional capacity of the
case group was 72.7 (SD ⫽22.8)
compared with a rating of 37.4 (SD ⫽
16.8) in the same patients before
treatment. This was a statistically sig-
nificant improvement (Wilcoxon’sZ
⫽⫺6.6, P⬍0.001). The control
group score increased in a very sim-
ilar way from 38.3 (SD ⫽19.2) to
74.8 (SD ⫽23.0). There was no effect
and no statistically significant differ-
ence in improvement between groups
(Kolmogorov-Smirnov Z ⫽0.639,
P⫽0.809).
Looking back on their training
program during the inpatient reha-
bilitation, patients marked very good
on a 5-point rating scale (2 ⫽very
good; ⫺2⫽very bad) for both com-
puter-aided training using the Train-
ing Assistant and conventional train-
ing. The average case group rating
was 1.26 (SD ⫽0.59) compared with
a rating of 1.21 (SD ⫽0.73) in the
control group. Also, 22 of 65 cases
and 22 of 57 controls indicated very
good. There was no statistically sig-
nificant difference between the two
groups.
Patients with Total Knee Replace-
ments. Patients with total knee re-
placement who fulfilled the inclusion
criteria were randomized to receive
either computer-aided training or
conventional training during their
TABLE 1
Baseline characteristics of patients with total hip
replacements
Cases
(n⫽74)
Controls
(n⫽64)
Comparison
(PValue)
Sex 60.8% women 70.3% women NS
Age, yrs 67.8 69.3 NS
Employed 9.5% 6.3% NS
Living alone 31.1% 31.3% NS
No. of days from operation to
rehabilitation admission
22.5 23.5 NS
Rehabilitation length of stay, days 19.8 19.8 NS
Median admission HHS 62 65 NS
Median admission FIM™score 117 116.5 NS
Median admission FFbH score 35.7 33.3 NS
NS, not significant (P⬎0.05); HHS, Harris Hip Score; FFbH, Hanover
Functional Ability Questionnaire.
September 2004 Computer-Aided Training 677
inpatient rehabilitation too. Table 3
shows that the baseline characteris-
tics of these patients in the study
were similar in the two groups.
Both forms of training showed
significant improvements in scores
and items used for the result evalua-
tion until discharge. Effect sizes,
11,12
which were computed using the stan-
dard deviation of admission (pre)
scores range from 0.73 to 1.40
(cases), respectively, from 0.96 to
1.16 (controls) (Table 4). There was
no statistically significant difference
between these two groups on Hospi-
tal for Special Surgery Score, Staffel-
stein Score, FIM score, and FFbH.
At the 6-mo follow-up, as was
done for patients with hip arthro-
plasty, all patients with knee arthro-
plasty who had completed the inpa-
tient rehabilitation program were
mailed a questionnaire and asked to
report their functional status using
the FFbH for patients with osteoar-
thritis. One patient was deceased, and
five patients did not return the ques-
tionnaire (Fig. 5).
Six months after the program,
the average functional capacity of the
case group was 76.9 (SD ⫽16.8) as
compared with a rating of 46.4 (SD ⫽
14.4) in the same patients before
treatment. This was a statistically sig-
nificant improvement (Wilcoxon’sZ
⫽⫺6.0, P⬍0.001). The control
group score also increased signifi-
cantly from 48.3 (SD ⫽16.7) to 70.6
(SD ⫽20.6). Differences between fol-
low-up and admission scores showed
a small effect to the credit of the case
group (effect size ⫽0.38). However,
statistically speaking, there was no
significantly better improvement for
the case group (Kolmogorov-Smir-
nov Z ⫽1.134, P⫽0.153).
At the 6-mo follow-up, patients
with total knee replacement marked
very good on a 5-point rating scale (2
⫽very good) for both computer-
aided training and usual training.
The average case group rating was
1.26 (SD ⫽0.81) compared with a
rating of 1.28 (SD ⫽0.74) in the
control group. Also, 23 of 54 cases
and 27 of 60 controls indicated very
good. There was no statistically sig-
nificant difference between the two
groups.
Relationship of Demographic Vari-
ables to Functional Gain. One pur-
pose of our analysis was to under-
stand the influence of basic
demographic variables on change in
functional ability within the case
group. For this analysis, FFbH
change was made the dependent vari-
able, whereas sex, age, educational
level, and indication where desig-
nated as the independent variables.
TABLE 2
Effect sizes for patients with total hip replacements
Scores/Measured Values
Cases (n⫽74) Controls (n⫽64)
Diff T2-T1
a
SD T1
b
ESpre
c
Diff T2-T1
a
SD T1
b
ESpre
c
StS hip 14.8 11.3 1.31 17.1 12.8 1.34
HHS 14.8 13.1 1.13 16.2 15.2 1.07
FIM™2.6 3.9 0.67 3.4 4.5 0.76
FFbH 22.5 16.8 1.34 23.4 19.2 1.22
Distancewalked 640.8 269.9 2.37 671.3 224.1 3.00
Flexion 6.6 12.7 0.52 5.2 14.8 0.35
StS, Staffelstein Score; HHS, Harris Hip Score; FFbH, Hanover Functional Ability Questionnaire.
a
Difference between discharge and admission scores.
b
Standard deviation of admission scores.
c
Effect size standardized with standard deviation of admission (pre) scores.
TABLE 3
Baseline characteristics of patients with total knee
replacements
Cases
(n⫽68)
Controls
(n⫽68)
Comparison
(PValue)
Sex 66.2% women 79.4% women NS (P⫽0.061)
Age, yrs 70.2 69.7 NS
Employed 4.4% 2.9% NS
Living alone 36.8% 35.3% NS
No. of days from operation to
rehabilitation admission
21.6 24.1 NS
Rehabilitation length of stay,
days
20.1 19.9 NS
Median admission HSS 59.5 58.5 NS
Median admission FIM™118 118 NS
Median admission FFbH 47.2 48.6 NS
NS, not significant (P⬎0.05); HSS, Hospital for Special Surgery Score;
FFbH, Hanover Functional Ability Questionnaire.
678 Eisermann et al. Am. J. Phys. Med. Rehabil. ●Vol. 83, No. 9
Table 5 shows that there were no
significant or clinically meaningful
differences in the change of the aver-
age functional ability measure.
DISCUSSION AND
CONCLUSIONS
Computer-aided training is a
new tool in orthopedic rehabilitation.
The patient trains and transmits
training variables to the therapist.
The therapist analyzes, adapts and
transfers new training configurations
to the patient. They communicate
with each other by mutual exchange
of certain kinds of data. Little is
known about practicability and effec-
tiveness of tele-rehabilitation systems
or to what extent the patients would
accept them as an alternative to con-
ventional therapy. This study was
performed to evaluate just such a
program—the Universal Training As-
sistant—implemented in the every-
day operation of a rehabilitation
clinic.
The computer-aided training
should at least lead to equal treat-
ment results when compared with a
self-training program in groups su-
pervised by a therapist as part of the
rehabilitation setting. The findings of
our present study demonstrate that
treatment with the Training Assistant
improves functional status, activities
of daily living, functional indepen-
dence, and range of motion and re-
duces pain just as well as conven-
tional training. Training was effective
and safe. Adverse events were not re-
ported across the groups. For pa-
tients with total hip replacement or
total knee replacement, the comput-
er-aided training can be regarded as
equivalent to the conventional train-
ing relating to the results of the re-
habilitation program.
Regarding acceptance, the com-
puter-aided training was assessed in a
positive manner. For all criteria, only
average values between good and very
good were seen. All patients, regardless
of sex, age, and educational level were
able to use the system without major
problems and were compliant. The re-
sult of the group discussion was that
both physicians and therapists recog-
nize the Training Assistant as a new
instrument to assist rehabilitation.
Overall, the study showed that
this new approach is feasible and
practicable in principle. It has led to
results that correspond to those that
TABLE 4
Effect sizes for patients with total knee replacements
Scores/Measured Values
Cases (n⫽68) Controls (n⫽68)
Diff. T2-T1
a
SD T1
b
ESpre
c
Diff. T2-T1
a
SD T1
b
ESpre
c
StS knee 16.4 12.2 1.34 15.7 14.2 1.11
HSS 14.8 10.7 1.38 14.2 12.2 1.16
FIM™2.9 4.0 0.73 2.6 2.7 0.96
FFbH 20.1 14.4 1.40 16.1 16.7 0.96
Distance walked 569.7 263.1 2.17 571.9 166.7 3.43
Flexion 10.4 18.3 0.57 9.4 18.3 0.51
StS, Staffelstein Score; HSS, Hospital for Special Surgery Score; FFbH, Hanover Functional Ability Questionnaire.
a
Difference between discharge and admission scores.
b
Standard deviation of admission scores.
c
Effect size standardized with standard deviation of admission (pre) scores.
TABLE 5
Average functional ability measure change by sex, age,
educational level, and indication in the case group
Variables
Pretreatment to Posttreatment
Change in Hanover Functional
Ability Questionnaire
Mean
Standard
Deviation
Comparison
(PValue)
Sex 0.109
Female (n⫽90) 19.8 16.7
Male (n⫽52) 23.9 14.6
Age, yrs 0.418
⬍65 (n⫽38) 23.6 15.6
65–74 (n⫽72) 19.8 15.5
⬎74 (n⫽32) 22.1 17.9
Education level 0.922
9-yrs of elementary school (n⫽121) 21.6 16.2
Higher levels (n⫽16) 22.2 16.0
Indication 0.327
Total hip replacement (n⫽74) 22.5 17.7
Total knee replacement (n⫽68) 20.1 14.1
September 2004 Computer-Aided Training 679
can be achieved by self-training.
Therewith, reported findings from re-
cently published exploratory trials
2–4
were confirmed.
It calls for complementary re-
search in inpatient and outpatient
settings to determine whether these
results are sufficient to justify a wid-
ening of the system. In addition, fur-
ther studies would have to examine
which particular patient groups
would profit most by employing such
a tele-rehabilitation system.
According to the authors, the
Universal Training Assistant has a
great deal of potential in outpatient
applications. In an ideal set-up, the
patient learns to operate the system
as an inpatient and continues to use
it after dismissal. The patient com-
pletes the specified training program
and is analyzed by the therapist re-
garding performance of training. The
patient then offers the therapist feed-
back based on objective achievements
and subjective assessments. Addi-
tional tools like Web cams or accel-
erometers for controlling the exer-
cises during outpatient rehabilitation
have to be developed and integrated
in the system. The existence of such
tools would increase the compliance
levels of the patients and, at the same
time, reduce the risk that the system
will be put in the closet like many
ortheses. Experiences made with car-
diologic patients have shown that
computer-supported, remote-con-
trolled training has a very positive
effect on the patients’motiva-
tion.
17,18
Therefore, future studies
should test the effects that orthope-
dic tele-rehabilitation systems have
on motivation levels.
REFERENCES
1. Lathan CE, Kinsella A, Rosen MJ, et al:
Aspects of human factors engineering in
home telemedicine and telerehabilitation
systems. Telemed J 1999;5:169 –75
2. Burdea G, Popescu V, Hentz V, et al:
Virtual reality–based orthopedic telereha-
bilitation. IEEE Trans Rehabil Eng 2000;
8:430 –2
3. Palsbo SE, Bauer D: Telerehabilitation:
Managed care’s new opportunity. Manag
Care Q 2000;8:56 –64
4. Liu L, Miyazaki M: Telerehabilitation
at the University of Alberta. J Telemed
Telecare 2000;6 (suppl 2):47–9
5. Bortz J, Döring N: Forschungsmetho-
den und Evaluation. Berlin/Heidelberg,
Springer, 1995
6. Torbati T, Schladitz G: Evaluation of
course and results of indoor rehabilita-
tion measures with the Staffelstein Score
after total hip arthroplasty. Orthopä-
dische Praxis 2001;37:236 –42
7. Harris WH: Traumatic arthritis of the
hip after dislocation and acetabular
fractures: Treatment by mold arthro-
plasty. An end result study using a new
method of result evaluation. J Bone Joint
Surg (Am) 1969;51:737–55
8. Ranawat CS, Shine JJ: Duocondylar
total knee arthroplasty. Clin Orthop 1973;
94:185–95
9. Stineman MG, Hamilton BB, Granger
CV, et al: Four methods for characteriz-
ing disability in the formation of function
related groups. Arch Phys Med Rehabil
1994;75:1277–83
10. Langen EG de, Fommelt P, Wied-
mann KD, et al: Evaluation of functional
independence in rehabilitation by the
Functional Independence Measure (FIM).
Rehabilitation 1995;34:4 –11
11. FIM Funktionale Selbständigkeits-
messung, german version (FIM-Arbe-
itskreis Deutschland, Österreich,
Schweiz). München, Internationale Vere-
inigung zum Assessment in der Reha-
biliation, 1997
12. Kohlmann T, Raspe H: The Hannover
Functional Ability Questionnaire for mea-
suring back pain-related functional limi-
tations (FFbH-R). Rehabilitation 1996;
35:1–8
13. Kohlmann T, Richter T, Heinrichs K,
Peschel U, Knahr K, Kryspin-Exner I. En-
twicklung und Validierung des Funk-
tionsfragebogens für Patienten mit Ar-
throsen der Hüft- und Kniegelenke
(FFbH-OA). In: Schliehe F, Schunter-
mann MF, editors. 8th Rehabilitation-
swissenschaftliches Kolloquium. Reha-
Bedarf–Effektivität–Ökonomie; 03/8–10/
99; Norderney, Germany. Frankfurt/M.:
WDV Wirtschaftsdienst: 1999:40 –2
14. Khan ME, Anker M, Paatel BC, et al:
The use of focus groups in social and
behavioural research: Some methodolog-
ical issues. World Health Stat Q 1991;44:
145–9
15. Cohen J: Statistical Power Analysis
for the Behavioral Sciences. Hillsdale, NJ,
Erlbaum, 1988
16. Maier-Riehle B, Zwingmann C: Effect
size variations in the single group pre-
post study design: A critical view. Reha-
bilitation 2000;39:189 –99
17. Tegtbur U, Jung K, Markolsky U, et
al: Entwicklung eines chipkartenges-
teuerten Heimergometertrainings für die
Reha-Phase. Herz/Kreisl 2000;32:334
18. Gerling J, Denkler P, Haase I: Com-
puter-based cardiac tele-rehabilitation,
in: Second World Congress of the Inter-
national Society of Physical and Rehabil-
itation Medicine (ISPRM), Prague, May
18–22, 2003, Abstracts. Prague, ISPRM,
2003, pp 221
680 Eisermann et al. Am. J. Phys. Med. Rehabil. ●Vol. 83, No. 9