Function-centered rehabilitation increases work days in patients with nonacute nonspecific low back pain: 1-year results from a randomized controlled trial.

Abstract

To compare the effect of function-centered treatment (FCT) and pain-centered treatment (PCT) on the number of work days, permanent disability, and the unemployment rate.
Randomized controlled trial.
Inpatient rehabilitation center.
Patients (N=174; 79% male; mean age, 42 y) with previous sick leave of 6 weeks or more.
FCT (4 h/d for 3 wk) emphasized activity despite pain by using work simulation, strength, endurance, and cardiovascular training. PCT (2.5h/d for 3 wk) emphasized pain reduction and included passive and active mobilization, stretching, strength training, and a 4-hour mini back school with education and exercise. Analysis was by intention to treat.
Work days, return to work, rate of patients receiving financial compensation for permanent disability, and unemployment rate. Effect sizes (Cohen d) were defined as small (0.2-0.5), moderate (0.5-0.8), and large (>0.8).
After 1 year, the FCT group had significantly more work days (mean, 118; median, 39.5; interquartile range [IQR], 0-198) than the PCT group (mean, 74; median, 0; IQR, 0-160; Mann-Whitney U test, P=.011). The odds ratio of returning to work in the FCT group relative to the PCT group was 2.1 (95% confidence interval, 1.1-3.9). The differences in unemployment rates and in the numbers of patients receiving compensation for permanent disability were not significant.
FCT is more effective than PCT for increasing work days.

Full text

Page 1

This article was published in an Elsevier journal. The attached copy
is furnished to the author for non-commercial research and
education use, including for instruction at the author’s institution,
sharing with colleagues and providing to institution administration.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright

Page 2

Author's personal copy
ORIGINAL ARTICLE
Function-Centered Rehabilitation Increases Work Days in
Patients With Nonacute Nonspecific Low Back Pain: 1-Year
Results From a Randomized Controlled Trial
Jan Kool, PhD, Stefan Bachmann, MD, Peter Oesch, MSc, Otto Knuesel, MD, Ton Ambergen, PhD,
Rob de Bie, PhD, Piet van den Brandt, PhD
ABSTRACT. Kool J, Bachmann S, Oesch P, Knuesel O,
Ambergen T, de Bie R, van den Brandt P. Function-centered
rehabilitation increases work days in patients with nonacute
nonspecific low back pain: 1-year results from a randomized
controlled trial. Arch Phys Med Rehabil 2007;88:1089-94.
Objective: To compare the effect of function-centered treat-
ment (FCT) and pain-centered treatment (PCT) on the number
of work days, permanent disability, and the unemployment
rate.
Design: Randomized controlled trial.
Setting: Inpatient rehabilitation center.
Participants: Patients (N?174; 79% male; mean age, 42y)
with previous sick leave of 6 weeks or more.
Interventions: FCT (4h/d for 3wk) emphasized activity
despite pain by using work simulation, strength, endurance,
and cardiovascular training. PCT (2.5h/d for 3wk) emphasized
pain reduction and included passive and active mobilization,
stretching, strength training, and a 4-hour mini back school
with education and exercise. Analysis was by intention to treat.
Main Outcome Measures: Work days, return to work, rate
of patients receiving financial compensation for permanent
disability, and unemployment rate. Effect sizes (Cohen d) were
defined as small (0.2–0.5), moderate (0.5–0.8), and large
(?0.8).
Results: After 1 year, the FCT group had significantly
more work days (mean, 118; median, 39.5; interquartile
range [IQR], 0?198) than the PCT group (mean, 74; me-
dian, 0; IQR, 0?160; Mann-Whitney U test, P?.011). The
odds ratio of returning to work in the FCT group relative to
the PCT group was 2.1 (95% confidence interval, 1.1?3.9).
The differences in unemployment rates and in the numbers
of patients receiving compensation for permanent disability
were not significant.
Conclusions: FCT is more effective than PCT for increasing
work days.
Key Words: Exercise therapy; Low back pain; Occupational
diseases; Outcome assessment (health care); Randomized con-
trolled trial; Rehabilitation; Sick leave; Vocational rehabilita-
tion.
© 2007 by the American Congress of Rehabilitation Medi-
cine and the American Academy of Physical Medicine and
Rehabilitation
E
subacute and chronic low back pain (LBP) if compared with
usual care.1,2Choosing the most appropriate type of exercise
and rehabilitation program is difficult. The results of most
pragmatic trials comparing different treatments and methods of
exercises are inconclusive.1,2Rehabilitation programs that fo-
cus on pain reduction and include exercise are still widely used.
Exercise of moderate intensity is supported by passive pain-
modulating treatments such as hot packs, electrotherapy, or
massage. Activity and exercises that increase pain are avoided.
In contrast, rehabilitation based on work hardening and graded
activity programs encourages patients to continue their activi-
ties even if pain increases.2,3It further remains unclear whether
treatments shown to be effective in studies are also effective in
populations with a different cultural background. In a meta-
analysis4evaluating the effect of exercise on sick leave in
patients with nonacute LBP (duration ?6wk), 8 of 14 studies
excluded patients with insufficient knowledge of the national
language. As a consequence, the validity of these results for
populations with a different cultural background is limited.
The major costs of LBP are caused by work absenteeism and
permanent disability. The total expenses of the Swiss Disability
Insurance rose from SFr 4 billion in 1990 to SFr 11 billion in
2005.5Compared with 10 other European countries, Switzer-
land showed the highest rate of increase.6Patients with health
problems who are unlikely to return to their previous work are
referred to the Swiss Disability Insurance, usually by their
physician. The Swiss Disability Insurance uses work adapta-
tion and professional training to keep persons at work. The
possibilities for work adaptation are restricted because many
patients have low education levels and very limited knowledge
of the national language. In the third place, incentives for
employers to prevent LBP-related disability and keep persons
at work are insufficient. Patients with permanent disability
receive a 100% disability allowance from the Swiss Disability
Insurance if their work-related disability exceeds 70%. In most
cases, disability is less than 70%, leading to a 25%, 50%, or
75% disability allowance.
We conducted a pragmatic randomized controlled trial
(RCT) in persons with nonacute nonspecific LBP.7During a
3-week inpatient rehabilitation program, patients received ei-
ther function-centered treatment (FCT) or pain-centered treat-
ment (PCT). We analyzed work absenteeism, and the primary
outcome was the number of work days in the follow-up year.
The results after 3 months have been published previously.7
There was a significant benefit for the FCT group compared
with the PCT group in the number of work days (25.9d vs
15.8d; effect size, .36; P?.029). A follow-up duration of 3
XERCISE, MULTIDISCIPLINARY rehabilitation, and
functional restoration reduce sick days in patients with
From the Department of Rheumatology, Rehabilitation Centre Valens, Valens,
Switzerland (Kool, Bachmann, Oesch, Knuesel); and Departments of Methodology
and Statistics (Ambergen) and Epidemiology (de Bie, van den Brandt), Maastricht
University, Maastricht, the Netherlands.
Supported by the Swiss Federal Office of Health (grant no. 00.00437).
No commercial party having a direct financial interest in the results of the research
supporting this article has or will confer a benefit upon the author(s) or upon any
organization with which the author(s) is/are associated.
Reprint requests to Jan Kool, PhD, School of Health Professions, Institute of
Physiotherapy, Zürich University of Applied Sciences Winterthur, 8401 Winterthur,
Switzerland, e-mail: jan.kool@zhwin.ch.
0003-9993/07/8809-11561$32.00/0
doi:10.1016/j.apmr.2007.05.022
1089
Arch Phys Med Rehabil Vol 88, September 2007

Page 3

Author's personal copy
months is too short for conclusions about the effectiveness of
an intervention aiming at improving work-related outcome.
Consequently, 1-year results are presented that analyze
whether the 3 months of benefit of the FCT group, compared
with that of the PCT group, was maintained 1 year after the
treatment.
This publication presents the results of work days, unem-
ployment, and permanent disability during the follow-up year
in patients with nonacute nonspecific LBP who received 3
weeks of either FCT or PCT.
METHODS
Methods and results regarding compliance and satisfaction
have been more extensively described in the publication about
the results after 3 months.7
Design
We conducted an RCT. Between January 2000 and May
2003, we recruited and treated patients in the center for work-
related rehabilitation in Valens, Switzerland. Randomization
was concealed and assessment of the primary outcome, work
days, was blinded. Independent teams of therapists treated
patients 6 days a week during 3 weeks. Patients could not be
blinded to treatment, but they received no detailed information
about the difference between the 2 treatments. The study was
approved by the ethical committee of Canton St Gallen, Swit-
zerland.
Inclusion and Exclusion Criteria
Eligible patients between 20 and 55 years of age with a
primary diagnosis of nonacute (duration, ?6wk) nonspecific
LBP and at least 6 weeks of sick leave in the previous 6 months
were considered for participation in the study. Patients with a
comorbidity interfering with treatment or working capacity
were excluded. Patients with 2 or more positive predictive tests
for non-return to work were excluded.7,8We included patients
with different nationalities to evaluate the relative effectiveness
of the treatments in subgroups with different cultural and
psychosocial backgrounds.
Treatment
Function-centered treatment.
providing FCT consisted of a rheumatologist, a physical and
occupational therapist trained in ergonomics, a sports therapist,
a social worker, and a nurse. FCT was based on work harden-
ing and functional restoration programs for 4 hours a day. The
primary goal was to increase work-related capacity while em-
phasizing improving self-efficacy. The rheumatologist in-
formed patients about the benign character of nonspecific LBP.
Treatment was based on the patient’s job demands; revealed in
a work-related assessment; and consisted of work simulation,
strength, and endurance training through isokinetic exercise,
cardiovascular training performed by walking and aqua aero-
bics, sports therapy, and self-exercise. Patients were encour-
aged to continue their activities even if their pain increased.
The work certificate after rehabilitation was based on the
patient’s work-related physical capacity and on medical
findings.
Pain-centered treatment.
The PCT team consisted of a
rheumatologist, a physiotherapist, and a nurse, and the primary
goal was pain reduction. The secondary goal was to decrease
disability and improve return to work. The duration of treat-
ment was 3 weeks and 2.5 hours a day. Physical therapy used
individually selected mobilization, stretching, strength training,
and a 4-hour mini back school with education and exercise.
The multidisciplinary team
Low-intensity movement therapy in the pool as well as pro-
gressive muscle relaxation using systematic contraction and
relaxation of specific muscle groups further enhanced relax-
ation. Passive pain-modulating treatments such as hot packs,
electrotherapy, or massage were used daily. In contrast to the
FCT group, patients in the PCT group were told to stop
activities when pain increased. After rehabilitation, the physi-
cian of the rehabilitation center determined the patients’ work-
ing capacity. His work certificate was based on his medical
findings.
In both groups, a rheumatologist prescribed medications
such as analgesics and nonsteroidal anti-inflammatory drugs
and might also apply local infiltrations in the musculature and
other soft tissue of the lumbar region with 5 to 10mL of 0.5%
lidocaine and 40mg of triamcinolone. If required, a psycholo-
gist offered counseling. After rehabilitation, treatment and
sickness certification were at the discretion of the patient’s
primary physician.
Compliance and Satisfaction With Treatment
We reported the evaluation of compliance and satisfaction
with treatment in more detail in the study concerning results
after 3 months.7We monitored patients’ compliance by record-
ing attendance at scheduled appointments and length of stay.
All patients expressed their satisfaction with treatment on a
numeric rating scale from 0 (extremely dissatisfied) to 10
(completely satisfied). We also assessed the therapists’ and
physicians’ compliance with FCT and PCT. A researcher au-
diotaped the verbal information the therapist or physician gave
to patients on 25 consecutive occasions. Seven blinded experts
independently rated the goals formulated, information about
the treatment plan, explanation of the source of the complaints,
and advice about coping with pain. Adequate adherence to the
protocol was arbitrarily defined as an average overall score of
more than 7.5 on a visual analog scale (VAS) from 0 (not at all
according to the treatment protocol) to 10 (perfectly according
to the treatment protocol).
Outcome Measurement
The primary outcome was the number of calendar work days
in the follow-up year. Secondary outcomes were the rate of
patients receiving unemployment benefits or permanent dis-
ability allowances. After 3 and 12 months, we sent question-
naires to the employer, known to be a valid source of infor-
mation about sick leave.9We also sent questionnaires to the
patients’ primary physician who determines fitness for work to
increase data completeness. We compared the information
from the employer and the physician to increase accuracy. The
questionnaires assessed work absenteeism and adaptation of
working hours per day. We accounted for time-reduced work.
For example, a work day with 30% time reduction was counted
as 0.7 work day. Information about disability allowances was
obtained from the Swiss Disability Insurance. Depending on
the level of disability, patients may receive a disability allow-
ance of 25%, 50%, 75%, or 100%. Partial disability allowances
were analyzed in the same manner as time-reduced work.
Primary physicians, employers, and the Disability Insurance
were blinded to the patients’ group assignment because they
were not informed in detail about the applied treatment during
the 3-week rehabilitation.
Health Care Utilization
We assessed the use of health care with questionnaires sent
to the health insurance companies of the patients.
1090
FUNCTION-CENTERED TREATMENT AND WORK DAYS, Kool
Arch Phys Med Rehabil Vol 88, September 2007

Page 4

Author's personal copy
Statistics
We performed a power calculation before the study started
(power, .80; type I error, .05) indicating that 90 patients per
group were needed to detect a difference of 40 workdays
(standard deviation [SD], 95). We used SPSSaand Statabfor
statistical analysis. Analysis was based on the intention-to-treat
principle.10
We compared the median number of work days during the
follow-up year in the FCT and PCT groups with a Mann-
Whitney U test. Dispersion was determined by means of the
interquartile range (IQR) representing the 25th and 75th per-
centile values. We determined effect sizes (Cohen d) defined as
small (0.2–0.5), moderate (0.5–0.8), and large (?0.8). We
analyzed the influence of baseline differences and covariates on
the number of work days in the 2 treatment groups. Covariates
of interest were the duration of sick leave before treatment, age,
cultural background, education, workload, and job qualifica-
tion. Because a large proportion of patients in this study had
zero work days during the follow-up year, we used negative-
binomial logistic hurdle regression.11This approach is partic-
ularly useful for the analysis of count data with an excess of
zero counts. In a first step, logistic regression was used for
analyzing the proportion of patients returning to work, defined
as 1 or more days of work. Results are given as odds ratios
(ORs) of returning to work in which the odds of returning to
work for patients in the FCT group are compared with those of
the PCT group.
In the second step of the analysis, the negative-binomial part
of the model is used for analyzing the number of work days
among those patients who have returned to work, defined as
having worked during at least 1 day after treatment. The result
parameter of the second step of the analysis is the incidence
rate ratio (IRR). Here an incidence rate is proportional to the
expected number of work days in a certain period conditional
on having had at least 1 work day. The IRR is the quotient of
2 incidence rates. For example, the incidence rate for the FCT
group is compared with the incidence rate for the PCT group.
We analyzed the effect of treatment and the influence of
covariates on receiving a disability allowance with ordinal
regression by using a proportional odds model with treatment
as independent factor. We determined the ORs for unemploy-
ment after 1 year in the FCT compared with the PCT group.
Logistic regression was used to evaluate the influence of treat-
ment and covariates on the unemployment rate.
RESULTS
Participants
Figure 1 shows the participants of the study, and table 1
displays the baseline comparability of the 2 groups. Because
recruitment rate was lower than expected, the duration of the
study was prolonged for 4 months, and 174 instead of the
initially planned 180 patients were included. There were no
significant differences between the groups for most variables
with the exception that more persons in the FCT group were
involved in litigation.
Protocol Compliance
All patients attended at least 90% of the scheduled treat-
ments, and treatment duration was comparable (table 2). Pro-
tocol compliance of therapists was above the predefined min-
imum level of 7.5 on a 10-point VAS in both treatment groups
(see table 2). Side effects did not reduce compliance with the
research protocol except in 1 patient in the FCT group who
insisted on getting hot packs and massage for pain relief. All
other patients were treated as intended. During rehabilitation, 1
patient was diagnosed with a necrosis of the femoral head and
referred for surgery. Both patients were included in the anal-
ysis. Patients’ satisfaction with treatment was the same in the
FCT and PCT group, indicating that the effort to keep patients
unaware of any expected treatment advantage was successful.
Outcome Measurements
Work days.
87% of the employers and 81% of the primary physicians.
Differences in reporting between physicians and employers
occurred in 12% of the cases. The research assistant contacted
the involved persons and resolved these discrepancies. We
retrieved the number of work days and the time restriction in
the 1-year follow-up period for 82 of 87 (94%) and 84 of 87
(97%) of the patients in the FCT and PCT groups, respectively.
The number of work days accounting for time-reduced work
was significantly larger in the FCT group (mean ? SD,
118?134; median, 39.5; IQR, 0?198) compared with the PCT
group (mean, 74?114; median, 0; IQR, 0?160; Mann-Whit-
ney U test, P?.011). The effect size was .35, representing a
small effect.
Negative-binomial logistic hurdle regression confirmed the
effect of FCT compared with PCT. The first part of the analysis
showed that the treatment effect was caused by a larger pro-
portion of patients who returned to work in the FCT group
(59.8%) than in the PCT group (41.4%). The OR of returning
to work in the FCT group compared with the PCT group was
2.11 (95% confidence interval [CI], 1.150?3.853; P?.016).
The second part of the negative-binomial logistic hurdle re-
gression showed no difference in number of working days
We obtained completed questionnaires from
Fig 1. Flowchart of subjects through each stage of the study from
initial screening to 1-year follow-up assessment. Abbreviation: FU,
follow-up.
1091
FUNCTION-CENTERED TREATMENT AND WORK DAYS, Kool
Arch Phys Med Rehabil Vol 88, September 2007

Page 5

Author's personal copy
between both groups among those patients who had at least 1
working day; the IRR was not significant (IRR?1.10; 95% CI,
0.776?1.568; P?.586).
Because of baseline differences between the 2 treatment
groups, litigation was entered into the model. In addition, we
investigated the effect of the potential covariates sick leave
before treatment, age, sex, education, job qualification, and
cultural background. Including covariates into the model did
not change the results. The logistic regression part of the model
showed a significant positive effect of FCT compared with
PCT. The OR of returning to work was 2.57 for patients in the
FCT compared with the PCT group (P?.011) (table 3). There
was a negative effect for litigation, longer sick leave before
treatment, and southeast European cultural background in both
groups. Education, workload, and job qualification were not
associated with the number of working days. The negative-
binomial regression part of the model showed that significantly
Table 1: Patient Characteristics Before Treatment
Characteristics FCT (n?87)PCT (n?87)
Mean age ? SD (y)
Sex (men/women)
Diagnosis, n (%)
Low back and leg pain
(ICD-10 M 54.4)
Low back pain (ICD-10 M
54.5)
Pain medication, n (%)
Mean BMI ? SD
Marital status, n (%)
Married
Single
Divorced
Living arrangement, n (%)
Alone
With partner or family
Living alone, family lives in
original country
Mean children ? SD (n)
Citizenship status, n (%)
Swiss citizen
Permanent immigrant
citizen
Permit depending on
employment
Limited permit (?1y)
Cultural background, n (%)
Switzerland
Southeast Europe
Southwest Europe
Education, n (%)
6 years (primary school)
7?9 years
?9 years
No professional education
Unemployed, n (%)
Qualification at last job, n (%)
Unskilled worker
Skilled worker
Foreman
Independent worker
Mean salary ? SD (SFr)
Heavy work: workload ?10kg,
n (%)
Mean work satisfaction ? SD
Mean sick leave 2y before
treatment ? SD (calendar
days)
Work-related litigation,* n (%)
41.6?8.4
69/18
42.5?8.4
68/19
73 (84) 71 (81)
14 (16) 16 (19)
67 (77)
26.7?4.2
62 (71)
27.2?4.0
63 (72)
13 (15)
11 (13)
67 (77)
10 (11.5)
10 (11.5)
15 (17)
67 (77)
5 (6)
16 (18)
68 (78)
3 (3.5)
2.0?1.32.0?1.4
38 (44)
40 (46)
35 (40)
42 (48)
8 (9)9 (10)
1 (1) 1 (1)
38 (44)
31 (35)
18 (21)
35 (40)
39 (45)
13 (15)
11 (13)
66 (76)
10 (11)
38 (44)
18 (21)
14 (16)
66 (76)
7 (8)
42 (48)
20 (23)
38 (44)
38 (44)
8 (9)
3 (3)
4501?1271
68 (78)
42 (48)
33 (38)
8 (9)
4 (4)
4522?1248
68 (78)
1.8?2.1
184?156
2.4?2.9
199?135
16 (18)9 (10)
Abbreviations: BMI, body mass index; ICD, International Classifica-
tion of Diseases.
*Mann-Whitney U test, P?.039.
Table 2: Treatment Duration, Compliance, and Satisfaction (no
significant differences)
VariableFCTPCT
Mean treatment duration ? SD (d)
Protocol compliance therapists
(range, 0–10), median (IQR)
Treatment compliance patients
(range, 0–8), median (IQR)
Patient satisfaction (range, 1–7),
median (IQR)
Treatment
Explanations about complaints
Advice about coping with
complaints
Increased ability to control
complaints
22.2?3.7
8.7 (7.3–9.6)
22.3?3.8
9.2 (7.7–10)
8 (5–8)7 (5–8)
6 (4–7)
5 (3–7)
6 (4–7)
6 (4–7)
5 (3–6)
6 (5–7)
5 (3–7) 5 (3–7)
Table 3: The Effect of FCT and Covariates on Work Days During
the Follow-Up Year Proportion of Patients Returning to Work
(logistic regression)
VariableOR95% CIP
Treatment FCT vs PCT
Litigation
Previous sick leave (mo)
Male
Cultural background SE
Europe/Switzerland
Cultural background SW
Europe/Switzerland
Age (y)
Education 6y/?6y
Job qualification
unskilled/other
2.566
0.224
0.910
0.858
0.303
(1.242–5.301)
(0.068–0.737)
(0.841–0.986)
(0.312–2.359)
(0.125–0.734)
.011
.014
.021
.767
.008
0.513(0.178–1.485).219
0.993
0.308
0.298
(0.951–1.037)
(0.061–1.555)
(0.080–1.110)
.746
.154
.071
No. of Work Days Among Those Patients Who Returned to Work
(negative binomial regression)
VariableIRR95% CI P
Treatment FCT vs PCT
Litigation
Previous sick leave (mo)
Male
Cultural background SE
Europe/Switzerland
Cultural background SW
Europe/Switzerland
Age (y)
Education 6y/?6y
Job qualification
unskilled/other
1.165
0.707
0.962
1.361
0.668
(0.819–1.657)
(0.318–1.568)
(0.918–1.009)
(0.880–2.104)
(0.419–1.062)
.396
.393
.111
.166
.088
0.909(0.523–1.578).734
1.005
0.620
1.168
(0.984–1.027)
(0.268–1.433)
(0.683–2.000)
.654
.263
.570
Abbreviations: OR, odds ratio of returning to work; SE, southeast;
SW, southwest.
1092
FUNCTION-CENTERED TREATMENT AND WORK DAYS, Kool
Arch Phys Med Rehabil Vol 88, September 2007

Page 6

Author's personal copy
more patients returned to work from the FCT group (59.8% in
the FCT, 41.4% in the PCT group). Among those patients who
returned to work for 1 or more days, neither treatment nor any
of the evaluated covariates were significantly associated with
the number of working days (see table 3).
Disability and Unemployment
After 1 year, 32 of 87 patients in the FCT group and 38 of
87 patients in the PCT group received a permanent disability
allowance, either full (100%) or partial (25%?75%). Ordinal
regression showed that receiving a higher disability allowance
was independent of treatment (P?.199). Receiving a higher
disability allowance was significantly associated with increas-
ing age but not with litigation, sex, workload, education, job
qualification, sick leave before treatment, and cultural back-
ground. Including these covariates did not change the results.
There was no difference in the unemployment rate after 1
year in the FCT and PCT group (FCT 37/87 [43%] vs PCT
45/87 [52%]; OR?.69; 95% CI, 0.38?1.26; P?.225). A
higher unemployment rate was generally observed in patients
who were older (P?.010), came from southeast Europe
(P?.002), or had a lower education level (P?.026). Unem-
ployment was independent of litigation, previous sick leave,
sex, and job qualification. Logistic regression with unemploy-
ment as the dichotomous dependent variable, treatment as main
effect, and correction for covariates did not change the results
(P?.245).
Health Care Use in the Follow-Up Period
We assessed health care use in the follow-up period by
means of a questionnaire sent to the health insurance provider.
The return rate was 78%. Interventions after rehabilitation were
comparable in the FCT and PCT group. Diagnostic procedures,
mainly radiography, magnetic resonance imaging, and com-
puted tomography, were used in 44% and 39% of the patients,
respectively. Medications, mainly analgesics and antidepres-
sants, were used by 90% and 94% of the patients. Physical and
occupational therapy was prescribed to 83% and 78% of the
patients.
DISCUSSION
This is the first study in Switzerland showing a significant
increase in the mean number of work days during the follow-up
year after FCT compared with PCT (118d vs 74d, effect size,
.35).
The results show that the effect during the first 3 months was
maintained until 1 year after treatment. Although the formal
effect size statistic only indicates a small effect, gaining over
40 days of additional work can be considered a substantial
improvement given the large cost per lost work day. Work days
during the follow-up year in both treatment groups were neg-
atively influenced by litigation, longer sick leave before treat-
ment, and southeast European cultural background without an
interaction with treatment. There was no significant effect of
treatment on the unemployment rate or on the number of
patients receiving a permanent disability allowance.
Strengths of this study were the relatively large number of
patients, applicability of the treatment to patients with other
cultural backgrounds, and limited knowledge of the Swiss
national language. The population under study was represen-
tative for the population in Switzerland at risk to develop
permanent LBP disability. The studied group was characterized
by a long duration of LBP and sick leave and a large proportion
of predominantly male workers with a low education level. We
did not exclude patients with minimal knowledge of the Ger-
man or Italian language. The results of this study are important
because a recent meta-analysis concluded that it remained
unclear whether exercise reduces work absenteeism.12Com-
pared with a study in Sweden by Lindström et al,3our study
included fewer patients with other nationalities (58% vs 75%).
In our study, the off-work duration was longer and pain inten-
sity was higher (5.6 points vs 3.3 points, on a 10-point scale).
Our results are in accordance with other research indicating
that exercise is safe and reduces disability.13
Study Limitations
Limitations of our study are that we did not use a workplace
intervention, which might have improved the results. We did
not assess psychologic comorbidity and several other factors
that potentially influenced the outcome. Randomization as used
in this study remains essential to generate comparable groups
regarding known and unknown predictive factors. The rehabil-
itation team of the FCT group also included a social worker.
Part of the difference in outcome could be caused by this
factor. There was no difference between groups in the use of
psychologic counseling. Less than 10% of the patients in both
groups were referred and received on average two 30-minute
sessions.
Blinding was not possible in patients and in the members of
the multidisciplinary teams. The best available alternative was
to keep patients unaware of the treatment of the other group
and of any expected treatment advantage. The excellent treat-
ment compliance in both groups and the comparable patient
satisfaction indicate that this attempt was probably successful.
The effect of the FCT on the number of disability allowances
and the unemployment rate after 1 year was not significant.
This is in accordance with the 3 months of results and with the
results of a previous meta-analysis.4Insufficient power may be
an important reason for the nonsignificant difference in the
unemployment rate and in the number of patients receiving a
disability allowance. The power of this study to detect a 20%
reduction in disability allowances is only .19 corresponding
with a type II error probability of 81%. Accounting for each
day of work absence with €155 (U.S. $21314), the savings
during the follow-up year are €6200 (U.S. $8520). The cost of
rehabilitation paid by health insurance is €220 (U.S. $303) per
day or €4900 (U.S. $6749) per patient in the FCT and PCT
groups.
The unemployment rate after 1 year was high in both groups.
The duration of unemployment benefits is limited to 2 years.
After this period, patients may either receive a disability al-
lowance or social benefits. The difference in the proportion of
patients at work between groups was maintained until 1 year
after treatment. Therefore, we plan a 3-year follow-up and
cost-effectiveness analysis.
Further research is needed to identify the essential elements
of treatment based on risk-factor assessment during the acute
phase of LBP.15Outpatient programs that may be more cost-
effective must be developed. In view of the fact that the
majority of patients in this study did not return to work, a
further reduction of work-related disability is urgently needed.
Improvements in outcome may be reached by earlier interven-
tions and by combining rehabilitation and workplace interven-
tions. In addition, politicians and insurance companies in Swit-
zerland should establish legal and financial incentives for
workers, employers, and involved insurances to reduce long-
term work-related disability. These incentives are nearly non-
existent in Switzerland.
1093
FUNCTION-CENTERED TREATMENT AND WORK DAYS, Kool
Arch Phys Med Rehabil Vol 88, September 2007

Page 7

Author's personal copy
CONCLUSIONS
Compared with PCT, FCT significantly increased the aver-
age number of work days during the follow-up year. The
benefit was 40 days, and the effect size was .35. In both
treatment groups, work days were negatively influenced by
litigation, longer sick leave before treatment, and southeast
European cultural background. Treatment had no effect on the
unemployment rate or the number of patients receiving a per-
manent disability allowance, but this result must be interpreted
with caution because the study was underpowered for this
outcome.
Acknowledgment:
hence was not registered in a trial registry.
The study enrolled subjects before 2005 and
References
1. Guzman J, Esmail R, Karjalainen K, Malmivaara A, Irvin E,
Bombardier C. Multidisciplinary rehabilitation for chronic low
back pain: systematic review. BMJ 2001;322:1511-6.
2. Schonstein E, Kenny DT, Keating J, Koes BW. Work condition-
ing, work hardening and functional restoration for workers with
back and neck pain. Cochrane Database Syst Rev 2003;(1):
CD001822.
3. Lindström I, Öhlund C, Eek C, et al. The effect of graded activity
on patients with subacute low back pain: a randomized prospec-
tive clinical study with an operant-conditioning behavioral ap-
proach. Phys Ther 1992;72:279-90; discussion 291-3.
4. Kool J, de Bie R, Oesch P, Knusel O, van den Brandt P, Bach-
mann S. Exercise reduces sick leave in patients with non-acute
non-specific low back pain: a meta-analysis. J Rehabil Med 2004;
36:49-62.
5. Buri M. Disability Insurance Statistics. Berne: Swiss Federal
Social Insurance Office; 2005.
6. Marin B, Prinz C. Facts and figures on disability welfare. A
pictographic portrait of an OECD report. Vienna: European Cen-
tre of the OECD; 2003.
7. Kool J, Oesch P, Bachmann S, et al. Increasing days at work using
function-centered rehabilitation in nonacute nonspecific low back
pain: a randomized controlled trial. Arch Phys Med Rehabil
2005;86:857-64.
8. Kool J, Oesch P, Bachmann S. [The predictive validity of nation-
ality for functional capacity and treatment outcome] [German].
Schweiz Arzteztg 2000;81:2656-69.
9. van Poppel MN, de Vet HC, Koes BW, Smid T, Bouter LM.
Measuring sick leave: a comparison of self-reported data on sick
leave and data from company records. Occup Med (Lond) 2002;
52:485-90.
10. Hollis S, Campbell F. What is meant by intention to treat analysis?
Survey of published randomised controlled trials. BMJ 1999;319:
670-4.
11. Gurmu S. Semi-parametric estimation of hurdle regression models
with an application to Medicaid utilization. J Appl Econ 1997;12:
225-42.
12. Hayden JA, van Tulder MW, Malmivaara AV, Koes BW. Meta-
analysis: exercise therapy for nonspecific low back pain. Ann
Intern Med 2005;14:765-75.
13. Rainville J, Hartigan C, Martinez E, Limke J, Jouve C, Finno M.
Exercise as a treatment for chronic low back pain. Spine J 2004;
4:106-15.
14. Knuti R, Schwaniger U. Arbeit und Gesundheit. In: Kocher G,
Oggier W, editors. Gesundheitswesen Schweiz 2004-2006. So-
lothurn: KSK; 2001. p 27-33.
15. Grotle M, Vollestad NK, Brox JI. Screening for yellow flags in
first-time acute low back pain: reliability and validity of a Nor-
wegian version of the Acute Low Back Pain Screening Question-
naire. Clin J Pain 2006;22:458-67.
Suppliers
a. Version 13; SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL
60606.
b. Version 9; StataCorp, 4905 Lakeway Dr, College Station, TX
77845.
1094
FUNCTION-CENTERED TREATMENT AND WORK DAYS, Kool
Arch Phys Med Rehabil Vol 88, September 2007