Clinical Rehabilitation 2005; 19: 779
/789

A new method for predicting functional recovery

of stroke patients with hemiplegia: logarithmic
modelling
Tetsuo Koyama, Kenji Matsumoto, Taiji Okuno Department of Rehabilitation Medicine, Nishinomiya Kyoritsu
Rehabilitation Hospital and Department of Physical and Rehabilitation Medicine, Hyogo College of Medicine and
Kazuhisa Domen Department of Physical and Rehabilitation Medicine, Hyogo College of Medicine, Nishinomiya, Hyogo,
Japan

Received 25th August 2004; returned for revisions 14th December 2004; revised manuscript accepted 3rd January 2005.

Objective: To examine the validity and applicability of logarithmic modelling for

predicting functional recovery of stroke patients with hemiplegia.
Design: Longitudinal postal survey.
Subjects: Stroke patients with hemiplegia staying in a long-term rehabilitation facility,
who had been referred from acute medical service 30
/60 days after onset.
Methods: Functional Independence Measure (FIM) scores were periodically
assessed during hospitalization. For each individual, a logarithmic formula that was
scaled by an interval increase in FIM scores during the initial 2
/6 weeks was used for
predicting functional recovery.
Results: For the study, we recruited 18 patients who showed a wide variety of
disability levels on admission (FIM scores 25
/107). For each patient, the predicted
FIM scores derived from the logarithmic formula matched the actual change in FIM
scores. The changes predicted the recovery of motor rather than cognitive functions.
Regression analysis showed a close fit between logarithmic modelling and actual
FIM scores (across-subject R2 /0.945).
Conclusions: Provided with two initial time-point samplings, logarithmic modelling
allows accurate prediction of functional recovery for individuals. Because the
modelling is mathematically simple, it can be widely applied in daily clinical practice.

Introduction and for those paying for it, accurate prediction

In the rehabilitative treatment of stroke patients
with hemiplegia, prediction of functional recovery
is crucial. Accurate prediction facilitates proper
definition of goals of intervention for individual
patients, thus improving the quality and efficiency
of rehabilitation service.1 For providers of services
Address for correspondence: Tetsuo Koyama, Department of
Rehabilitation Medicine, Nishinomiya Kyoritsu Rehabilitation
Hospital, Jurinji-Minamimachi 2-13, Nishinomiya, Hyogo,
Japan 662-0002. e-mail: ytkoyama@bd6.so-net.ne.jp
# 2005 Edward Arnold (Publishers) Ltd
enables effective use of resources by allowing better
estimation of such factors as length of hospitaliza-
tion.2 Thus, for both individual patients and health
care administrators, accurate prediction of func-
tional recovery would provide crucially important
information.
For predicting functional recovery, various
mathematical modelling and other methods have
been employed.3
12 Multivariable linear regression
modelling has proved the most popular.12,13
This type of linear modelling has been useful
for predicting outcome at a specific time-point
10.1191/0269215505cr876oa
780 T Koyama et al.

(e.g., six months after stroke). Stroke patients
typically, however, show nonlinear recovery pat-
terns.14,15 In most stroke cases, patients show rapid
recovery during the initial few months, after which
the pace of recovery to six months from onset
slows towards the final outcome.16 Consequently,
linear modelling is not up to the task of accurately
predicting the prospective outcome. To simulate
the nonlinear aspects of functional recovery, neu-
ral network modelling,17 logistic modelling,18,19
and other types of nonlinear modelling have been
proposed. Although more successfully predictive,
these modelling methods are not widely applied
because of their mathematical complexity. Thus,
for general clinical applicability, there has been a
need for a simpler means of accurately predicting
the progress of recovery.
To explore more simple modelling methods, we
investigated mathematical powers, logarithms,
double-logarithms, and other simple mathematical
functions. Of those, we focused on natural loga-
rithmic functions (ln) because they displayed three
advantages for modelling functional recovery.
First, the progress curves (Figure 1) resembled
actual recovery patterns: if the recovery target is
set at 180 days from onset and assigned a value of
100%, approximately 70% of recovery is registered
at 90 days and subsequent progress occurs at a
reduced rate.1,16,20 Similarly, a logarithmic func-
tion fitted the recovery patterns of upper limb
function of stroke patients.21 Second, based on
scores sampled on two days separated by an
interval, using simple mathematical procedures
(Figure 1), the modelling formula can easily be

Figure 1 Model formula and predictive curve. (A) shows a generic structure; (B) shows mathematical procedures to tailor the
generic structure to fit individual degree of recovery. For this, actual FIM scores recorded at two time points (Day A and Day B)
are required. DFIM indicates change in FIM scores between Day A and Day B. Constant in (A) is countervailed in this procedure.
(C) shows the final form of the model formula. Predicted value for Day X can be calculated with this form. FIM, Functional
Independence Measure; ln, natural logarithm.

scaled to fit each individual’s magnitude of recov-
ery. Third, owing to mathematical specificity of
logarithms, the model formula can easily be
calculated (e.g., ln(90)/ln(30)/ln(90/30) /ln(3),
see Figure 1B). To evaluate the practical usefulness
of logarithmic modelling we carried out a long-
itudinal study.
Methods
Patients
Stroke patients with hemiplegia who were ad-
mitted to our long-term rehabilitation hospital
during August 2003 to April 2004 were recruited
into the study. Criteria for inclusion were:
no past history of hemiplegia; capable of indepen-
dent ADL (activities of daily life) before stroke;
wheelchair required for locomotion at admission.
As a result of Japanese health insurance proce-
dures, patients were referred from local community
acute medical services, typically 30
/60 days after
the stroke occurred, and received inpatient care in
our long-term rehabilitation hospital for 30
/180
days. During the prior period of acute medical
hospitalization they received physical therapy.
During long-term rehabilitation hospitalization
they received physical therapy, occupational ther-
apy and speech therapy for a joint total of 120 min
every day. To minimize the influence of variability
of therapeutic regimen, we also limited recruitment
to patients who received treatment from the same
rehabilitation team directed by a single physiatrist
(first author of this article). The protocol was
reviewed and approved by our hospital’s ethical
committee and informed consent was obtained
from all patients.
Assessment of functional recovery
To assess functional recovery, we employed the
Functional Independence Measure (FIM), which
has been widely used in rehabilitation medicine.22
The FIM is derived from scoring 18 items accord-
ing to a seven-point scale (1 /totally dependent,
7/completely independent) to assess functional
independence in ADL. These 18 items are
categorized as self-care (6 items), sphincter
control (2 items), transfers (3 items), locomotion
(2 items), communication (2 items), and social
cognition (3 items). The first four categories
Logarithmic modelling in hemiplegic stroke 781
involve motor functions (FIM-motor) and other
two concern cognitive functions (FIM-cognition).
The total scores score for all 18 items (FIM-
total) is commonly used to assess functional
independence in rehabilitation medicine (totally
dependent in ADL /18, completely independent
in ADL /126).
Using FIM scores, nursing staff assessed the
functional recovery of patients in terms of ADL.
Evaluations were typically recorded a few days
after admission, again at two to six weeks after
admission, and then once a month during hospi-
talization. In our study, to assure reliability of the
evaluations, FIM scores were reviewed at weekly
conferences.
Modelling and evaluation
A generic structure of modelling was given in a
simple natural logarithmic formula (independent
variable /days from onset) (Figure 1A). To tailor
the generic structure to fit each individual’s degree
of functional recovery, we performed calculations
on the total FIM scores at the first two time-points
after admission. For each patient, the increase in
total FIM scores between these two time-points
(DFIM) was used as the basis for scaling a co-
efficient (b) in the generic structure (Figure 1B).
The introduction of this countervailed the constant
in the generic structure. Thus, using the scores at
the initial two sampling points, a generic struc-
ture could be tailored to forecast each patient’s
functional recovery (model formula shown in
Figure 1C).
To assess the fit of the time-course of the
model, FIM scores (FIM-total, FIM-motor and
FIM-cognition) were, on an individual basis,
longitudinally plotted with predicted values for
each patient derived from the model formula. To
assess the general applicability of logarithmic
modelling, using data from all patients, a conven-
tional linear regression analysis was performed to
compare the total FIM scores that were actually
obtained and the predicted values that were
derived from the model formula. For this analysis,
we excluded the scores obtained at the first
two sampling points (indicated by arrowheads
in Figures 2 and 3) to determine the particular b
coefficients.
782 T Koyama et al.

Figure 2 Time-course of actually obtained and predicted FIM scores for patients with left hemisphere lesions (cases 1
/10).
Closed circles show actually obtained FIM-total scores, closed triangles show actually obtained FIM-motor scores, closed
squares show actually obtained FIM-cognition scores and open circles show predictive values derived from the model formula
(Figure 1). Arrowheads indicate initial two sampling time-points for data to tailor the model formula for each individual.
FIM, Functional Independence Measure.
 Logarithmic modelling in hemiplegic stroke 783

Figure 3 Time-course of actually obtained and predicted FIM scores for patients with right hemisphere lesions (cases 11
/18).
Closed circles show actually obtained FIM-total scores, closed triangles show actually obtained FIM-motor scores, closed
squares show actually obtained FIM-cognition scores and open circles show predictive values derived from the model formula
(Figure 1). Arrowheads indicate initial two sampling time-points for data to tailor the model formula for each individual.
FIM, Functional Independence Measure.

Results hemisphere lesion; age 33
/78 (median 67.5) years

old). For both motor and cognitive functions,
Patients these patients showed widely varying levels
We collected and manipulated data for 18 of disability on admission (Table 1, Figures 2
patients (12 male, 6 female; 10 left, 8 right, and 3). Total FIM scores ranged from 25 to 107
784 T Koyama et al.

Table 1 Patients’ profiles

Case Age Gender Hemisphere Lesion Cause of stroke Ope. Intervention Comorbidity

No. 1 82 M Left Corona radiata Infarct (/) OT, PT CAD

No. 2 63 F Left Putamen Hemorrhage (/) OT, PT, ST HT
No. 3 53 M Left Corona radiata Infarct (/) OT, PT, ST HT
No. 4 33 F Left Putamen Hemorrhage (/) OT, PT, ST HT, HL
No. 5 62 M Left Putamen Hemorrhage (/) OT, PT, ST ( /)
No. 6 73 F Left Putamen Infarct (/) OT, PT, ST DM, HT
No. 7 78 F Left Corona radiata Infarct (/) OT, PT ( /)
No. 8 74 F Left MCA Infarct (/) OT, PT, ST Af, CAD, HT
No. 9 74 M Left Putamen Hemorrhage (/) OT, PT, ST HCC
No. 10 65 M Left Prefrontal cortex Hemorrhage (/) OT, PT ( /)
No. 11 74 M Right Prefrontal cortex Hemorrhage (/) OT, PT, ST ( /)
No. 12 50 M Right Corona radiata Infarct (/) OT, PT DM, HT
No. 13 73 M Right Corona radiata Infarct (/) OT, PT DM, HT
No. 14 70 F Right MCA Infarct (/) OT, PT ( /)
No. 15 70 M Right MCA Infarct (/) OT, PT CAD, DM
No. 16 54 M Right Thalamus Hemorrhage (/) OT, PT, ST HT
No. 17 65 M Right MCA Infarct (/) OT, PT HT
No. 18 54 M Right Putamen Hemorrhage (/) OT, PT HT

CAD, coronary artery disease; DM, diabetes mellitus; HCC, hepatic cell carcinoma (post operation); HL, hyperlipidaemia; HT,
hypertension; MCA, middle cerebral artery; Ope., operation (open-skull) during acute medical hospitalization; OT, occupational
therapy; PT, physical therapy; ST, speech therapy.

(median, 63.5), motor FIM scores ranged from 14
to 74 (median, 36), and cognition FIM scores
ranged from 6 to 35 (median, 25). Initial FIM
scores were sampled at from 32 to 77 days (median
50) after occurrence of stroke and the second set
of FIM scores were sampled at from 46 to 104 days
(median 72) after occurrence (indicated by arrow-
heads in Figures 2 and 3). The interval between
these two time-points ranged from 13 to 44 days
(median 32).

Figure 4 Scatterplots showing the relationships between actually obtained FIM-total scores and predicted values derived
from the model formula (see Figures 2 and 3). Data from the two initial sampling time-points for each patient (indicated by
arrowheads in Figures 2 and 3) were excluded from the scatterplots. FIM, Functional Independence Measure.

Assessment of model fit
For each individual, the pattern of increase in
the predicted values that were derived from the
model formula was very similar to the total FIM
scores that were actually obtained: so close in fact,
that the correspondence in some cases (3, 6, 8, 10,
14, 15 and 16) was almost identical (Figures 3 and
4). Actual total FIM scores comprised two main
components: measures of motor and cognitive
ability. Close observation of the time-courses of
these subcomponents showed that the main con-
tribution to the growth patterns of total FIM
scores was mainly from the motor subcomponents.
In sharp contrast, changes in cognitive subcompo-
nents were, in most cases, minimal (Figures 3 and
4). This finding indicates that the model formula
simulates the recovery pattern of motor rather
than cognitive components.
For cases 1, 4 and 12, the predicted values
exceeded the actually obtained total FIM scores.
Dissociation between actual and predicted values
tended to be greater towards the high end (/120)
of the total FIM range. The predicted values for
case 11 also exceeded the actual FIM scores. This
case was exceptional, being the only patient who
scored large changes in FIM-cognition soon after
admission. Nonetheless, even in this case, the
growth pattern of FIM-motor scores was compar-
able to other cases and similar to logarithmic
curve.
Model fit was then assessed using group data.
Regression analysis comparing actual data and
predicted values revealed that the model formula
accurately predicted actually obtained FIM-total
scores (Figure 4; R2 /0.945).
Discussion
A logarithmic function was applied to simulate the
time-course of functional recovery of stroke pa-
tients with hemiplegia. Based on this, we developed
a new model formula that, using FIM results
sampled from two points in time during recovery,
could be applied accurately to predict the pattern
of functional recovery in individual recovery.
Among patients with a wide variety of motor and
cognitive disability, the model formula accurately
predicted actual functional recovery during hospi-
Logarithmic modelling in hemiplegic stroke 785
Clinical messages
“/ Logarithmic modelling accurately predicts
functional recovery of stroke patients with
hemiplegia.
“/ Provided with two initial time-point sam-
plings, logarithmic modelling can be tai-
lored to forecast each patient’s functional
recovery.
“/ The modelling is mathematically simple
enough to be adopted in daily clinical
practice.
talization. Thus, the new model formula based on
logarithmic function could be a powerful tool for
predicting functional recovery of stroke patients
with hemiplegia.
Modelling using raw FIM-total scores
FIM assessment was originally based on an
ordinal rather than an interval scale. Subsequently,
Rasch analysis has provided a model for convert-
ing the ordinal scale of raw FIM-total scores into
an interval scale.23 Although after Rasch conver-
sion the data showed shows a logistic curve, raw
FIM-total scores tend to show an almost linear
relationship with converted values within the
range from 25 to 120. Within this range, raw
FIM-total scores have been widely employed as
interval values in many previous studies.24,25
Accordingly, to keep our model simple, we emp-
loyed raw FIM scores as the basis for mathema-
tical modelling.
Validity of logarithmic modelling using two
time-point samplings
During actual treatment, naturally each patient
and those giving them care are intensely interested
in the particular prospects of functional recovery.
Few studies, however, have focused on individual
time-course and degree of functional recovery.1 In
this study, for each patient, we longitudinally
sampled FIM scores at 4
/7 time-points during
hospitalization. Observation of this data indicated
that recovery patterns assessed by FIM scores
could be modeled as logarithmic function. More-
over, using data for individual patients that were
sampled at two time-points during recovery, the
786 T Koyama et al.
model accurately predicted the actual results later
obtained for the individuals. We know of no other
prediction modelling studies that provide useful,
simple, individual-based mathematical modelling.
When forecasting the functional recovery of an
individual stroke patient, a single physiatrist often
takes many clinical parameters into consideration.
These include: initial motor and cognitive impair-
ment levels,26,27 initial day of rehabilitation,28
recovery rate,29 site and size of lesion,30,31 age,32
psychological status,33 unilateral spatial neglect,34
co-morbidities35 and other factors.36 Most of these
previous prediction studies have attempted to
integrate multiple factors into the model. Our
study, however, uses only FIM scores sampled on
different days with an interval of 2
/6 weeks
between them. The results show that, processed
through our logarithmic equation, these data
enable powerful and accurate forecasting of func-
tional recovery. Since initial patterns of recovery
could be affected by any of the multiple factors
mentioned above, the FIM scores of individual
patients are likely to be influenced by some or all
of these factors.
Simplicity of logarithmic modelling
With the goal of developing a new forecasting
technique to predict functional recovery, we tested
several mathematical functions in our preliminary
analyses. Taking a lead from a previous study that
employed logarithmic transformations of FIM-
total scores to model functional outcome,12 we
tested, among other manipulations, various dou-
ble-logarithmic functions. In fact, in some cases,
preliminary models using double-logarithmic func-
tions did fit actual data slightly better than the
model formula that we are presenting here. Even
so, we preferred not to employ double-logarithmic
modelling because of its complexity.
Focusing on logarithmic modelling, we at-
tempted to adjust the clause for ‘days from onset’
to improve the model fit. We attempted adjustment
based on the clinical observation that the start of
functional recovery varies from case to case
depending on site, size, and age of lesion. Our
preliminary analyses, however, revealed that the
contribution of such adjustments was minimal.
Thus, to keep things simple, we applied logarithmic
modelling without any adjustments (Figure 1).
The model formula is simple and structurally
flexible (Figure 1C). For consistency, in this study,
we used data from the first and second FIM
samplings after admission. Any pair of periodic
samplings, however, are suitable for defining the
coefficient (b) of the model formula. The flexibility
of the model formula enables easy re-estimation if
predictive and actual values deviate. This simpli-
city and flexibility means that the model formula is
suitable for wide clinical application.
Possible limitations of logarithmic modelling
In this study, we customized the individual’s
model formula by using scores from two FIM
samples: based on results of assessment done with
an intervening period of from 13 to 44 days, data
from this sampling pair were collected at between
32 and 104 days after the occurrence of stroke. The
model was effective within these sampling para-
meters. Further studies are needed to find out the
limits of applicability to FIM data collected at
earlier or later phases of affliction.37 It is promising
that case 15 (Figure 3), using data collected
relatively soon (33 days) after stroke occurrence
and with a short sampling interval (13 days),
provided accurate prediction. This modelling
might be useful even at earlier stages of illness
and during shorter periods of hospitalization.
Close observation of the time-course data
plotted for each individual revealed that factors
for change in the predictive model were the motor
components rather than cognitive components.
Thus the model may not be applicable for patients
whose clinical manifestations are mainly cognitive
rather than motor (e.g., patients with subara-
chnoid haemorrhage).38,39 Time-course plotting
also revealed a tendency for predicted values to
exceed the actual data towards the high end of the
FIM-total range. In view of the linearity of raw
FIM-scores (as discussed above) this might imply
that the model formula is best utilized when
predictive values range from 25 to 120.
Applicability of logarithmic modelling
Our study samples yielded data on patients who
varied widely in age, lesion characteristics, and
levels of motor and cognitive disabilities. The
results that we obtained indicate that the logarith-
mic model formula (Figure 1C) could be effectively
applied for various types of hemiplegic stroke

patients. Our new model is valuable for its
simplicity and applicability on an individual basis.
Using FIM scores that were sampled at two
different time-points, using a regular pocket calcu-
lator (without a log function) and a logarithm
look-up table (see Appendix), within minutes it is
possible to come up with a prediction for each
individual’s functional status for a particular day.
Thus, the model formula, based on simple loga-
rithmic function, could be adopted in everyday
clinical practice for predicting the functional
recovery of stroke patients with hemiplegia.
References
1 Tilling K, Sterne JA, Rudd AG, Glass TA, Wityk
RJ, Wolfe CD. A new method for predicting
recovery after stroke. Stroke 2001; 32: 2867
/73.
2 Dam M, Tonin P, Casson S et al . The effects of
long-term rehabilitation therapy on poststroke
hemiplegic patients. Stroke 1993; 24: 1186
/91.
3 Allen CM. Predicting the outcome of acute stroke: a
prognostic score. J Neurol Neurosurg Psychiatry
1984; 47: 475
/80.
4 Reding MJ, Potes E. Rehabilitation outcome
following initial unilateral hemispheric stroke. Life
table analysis approach. Stroke 1988; 19: 1354
/58.
5 Gladman JR, Harwood DM, Barer DH. Predicting
the outcome of acute stroke: prospective evaluation
of five multivariate models and comparison with
simple methods. J Neurol Neurosurg Psychiatry
1992; 55: 347
/51.
6 Falconer JA, Naughton BJ, Dunlop DD, Roth EJ,
Strasser DC, Sinacore JM. Predicting stroke
inpatient rehabilitation outcome using a
classification tree approach. Arch Phys Med Rehabil
1994; 75: 619
/25.
7 Gompertz P, Pound P, Ebrahim S. Predicting stroke
outcome: Guy’s prognostic score in practice.
J Neurol Neurosurg Psychiatry 1994; 57: 932
/35.
8 Mauthe RW, Haaf DC, Hayn P, Krall JM.
Predicting discharge destination of stroke patients
using a mathematical model based on six items from
the Functional Independence Measure. Arch Phys
Med Rehabil 1996; 77: 10
/13.
9 Lai SM, Duncan PW, Keighley J. Prediction of
functional outcome after stroke: comparison of the
Orpington Prognostic Scale and the NIH Stroke
Scale. Stroke 1998; 29: 1838
/42.
10 Sanchez-Blanco I, Ochoa-Sangrador C,
Lopez-Munain L, Izquierdo-Sanchez M,
Fermoso-Garcia J. Predictive model of functional
Logarithmic modelling in hemiplegic stroke 787
independence in stroke patients admitted to a
rehabilitation programme. Clin Rehabil 1999; 13:
464
/75.
11 Thommessen B, Bautz-Holter E, Laake K.
Predictors of outcome of rehabilitation of elderly
stroke patients in a geriatric ward. Clin Rehabil
1999; 13: 123
/28.
12 Inouye M. Predicting outcomes of patients in Japan
after first acute stroke using a simple model. Am J
Phys Med Rehabil 2001; 80: 645
/49.
13 Giaquinto S, Buzzelli S, Di Francesco L et al . On
the prognosis of outcome after stroke. Acta Neurol
Scand 1999; 100: 202
/208.
14 Partridge CJ, Johnston M, Edwards S. Recovery
from physical disability after stroke: normal
patterns as a basis for evaluation. Lancet 1987; 1:
373
/75.
15 Jorgensen HS, Nakayama H, Raaschou HO,
Vive-Larsen J, Stoier M, Olsen TS. Outcome and
time course of recovery in stroke. Part II: Time
course of recovery. The Copenhagen Stroke Study.
Arch Phys Med Rehabil 1995; 76: 406
/12.
16 Sonoda S, Chino N, Domen K, Saitoh E. Changes
in impairment and disability from the third to the
sixth month after stroke and its relationship
evaluated by an artificial neural network. Am J Phys
Med Rehabil 1997; 76: 395
/400.
17 Oczkowski WJ, Barreca S. Neural network
modelling accurately predicts the functional
outcome of stroke survivors with moderate
disabilities. Arch Phys Med Rehabil 1997; 78:
340
/45.
18 Stineman MG, Maislin G, Fiedler RC, Granger CV.
A prediction model for functional recovery in
stroke. Stroke 1997; 28: 550
/56.
19 Lofgren B, Gustafson Y, Nyberg L.
Cross-validation of a model predicting discharge
home after stroke rehabilitation. Validating stroke
discharge predictors. Cerebrovasc Dis 2000; 10:
118
/25.
20 Calautti C, Baron JC. Functional neuroimaging
studies of motor recovery after stroke in adults: a
review. Stroke 2003; 34: 1553
/66 Epub 2003 May 8.
21 Goodwin N, Sunderland A. Intensive, time-series
measurement of upper limb recovery in the
subacute phase following stroke. Clin Rehabil 2003;
17: 69
/82.
22 Linacre JM, Heinemann AW, Wright BD, Granger
CV, Hamilton BB. The structure and stability of the
Functional Independence Measure. Arch Phys Med
Rehabil 1994; 75: 127
/32.
23 Wright BD, Linacre JM, Smith RM, Heinemann
AW, Granger CV. FIM measurement properties and
Rasch model details. Scand J Rehabil Med 1997; 29:
267
/72.
788 T Koyama et al.
24 Ring H, Feder M, Schwartz J, Samuels G.
Functional measures of first-stroke rehabilitation
inpatients: usefulness of the Functional
Independence Measure total score with a clinical
rationale. Arch Phys Med Rehabil 1997; 78: 630
/35.
25 Kwon S, Hartzema AG, Duncan PW, Min-Lai S.
Disability measures in stroke: relationship among
the Barthel Index, the Functional Independence
Measure, and the Modified Rankin Scale. Stroke
2004; 35: 918
/23.
26 Ween JE, Alexander MP, D’Esposito M, Roberts
M. Factors predictive of stroke outcome in a
rehabilitation setting. Neurology 1996; 47: 388
/92.
27 Zinn S, Dudley TK, Bosworth HB, Hoenig HM,
Duncan PW, Horner RD. The effect of poststroke
cognitive impairment on rehabilitation process and
functional outcome. Arch Phys Med Rehabil 2004;
85: 1084
/90.
28 Novack TA, Satterfield WT, Lyons K, Kolski G,
Hackmeyer L, Connor M. Stroke onset and
rehabilitation: time lag as a factor in treatment
outcome. Arch Phys Med Rehabil 1984; 65: 316
/19.
29 Mayo NE, Korner-Bitensky NA, Becker R.
Recovery time of independent function post-stroke.
Am J Phys Med Rehabil 1991; 70: 5
/12.
30 Chaudhuri G, Harvey RF, Sulton LD, Lambert
RW. Computerized tomography head scans as
predictors of functional outcome of stroke patients.
Arch Phys Med Rehabil 1988; 69: 496
/98.
31 Saeki S, Ogata H, Hachisuka K, Okubo T,
Takahashi K, Hoshuyama T. Association between
location of the lesion and discharge status of ADL
in first stroke patients. Arch Phys Med Rehabil 1994;
75: 858
/60.
32 Bagg S, Pombo AP, Hopman W. Effect of age on
functional outcomes after stroke rehabilitation.
Stroke 2002; 33: 179
/85.
33 Paolucci S, Antonucci G, Pratesi L, Traballesi M,
Grasso MG, Lubich S. Poststroke depression and
its role in rehabilitation of inpatients. Arch Phys
Med Rehabil 1999; 80: 985
/90.
34 Katz N, Hartman-Maeir A, Ring H, Soroker N.
Functional disability and rehabilitation outcome in
right hemisphere damaged patients with and
without unilateral spatial neglect. Arch Phys Med
Rehabil 1999; 80: 379
/84.
35 Pettersen R, Dahl T, Wyller TB. Prediction
of long-term functional outcome after stroke
rehabilitation. Clin Rehabil 2002; 16: 149
/59.
36 Kwakkel G, Wagenaar RC, Kollen BJ, Lankhorst
GJ. Predicting disability in stroke
/a critical review
of the literature. Age Ageing 1996; 25: 479
/89.
37 Asberg KH, Nydevik I. Early prognosis of stroke
outcome by means of Katz Index of activities of
daily living. Scand J Rehabil Med 1991; 23: 187
/91.
38 Hellawell DJ, Taylor R, Pentland B. Persisting
symptoms and carers’ views of outcome after
subarachnoid haemorrhage. Clin Rehabil 1999; 13:
333
/40.
39 Svensson E, Starmark JE. Evaluation of individual
and group changes in social outcome after
aneurysmal subarachnoid haemorrhage: a long-
term follow-up study. J Rehabil Med 2002; 34:
251
/59.
 Logarithmic modelling in hemiplegic stroke 789

Appendix
A quick reference for logarithmic function

/

x Ln (x )

1.0 0.000
1.1 0.095
1.2 0.182
1.3 0.262
1.4 0.336
1.5 0.405
1.6 0.470
1.7 0.531
1.8 0.588
1.9 0.642
2.0 0.693
2.2 0.788
2.4 0.875
2.6 0.956
2.8 1.030
3.0 1.099
4.0 1.386
5.0 1.609
6.0 1.792
7.0 1.946
8.0 2.079
9.0 2.197
10.0 2.303

ln, natural logarithm.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.