The Clinical Journal of Pain

18:355–365 © 2002 Lippincott Williams & Wilkins, Inc., Philadelphia

Clinical Effectiveness and Cost-Effectiveness of Treatments for

Patients With Chronic Pain
Dennis C. Turk, Ph.D.
University of Washington, Seattle, Washington, U.S.A.

Abstract:
Objective: Chronic pain is a prevalent and costly problem. This review addresses
the question of the clinical effectiveness and cost-effectiveness of the most common
treatments for patients with chronic pain.
Data Sources: Representative published studies that evaluate the clinical effective-
ness of pharmacological treatments, conservative (standard) care, surgery, spinal cord
stimulators, implantable drug delivery systems (IDDSs), and pain rehabilitation pro-
grams (PRPs) are examined and compared. The cost-effectiveness of these treatment
approaches is also considered.
Data Synthesis: Outcome criteria including pain reduction, medication use, health
care consumption, functional activities, and closure of disability compensation cases
are examined. In addition to clinical effectiveness, the cost-effectiveness of PRPs,
conservative care, surgery, spinal cord stimulators, and IDDSs are compared using
costs to return a treated patient to work to illustrate the relative expenses for each of
these treatments.
Conclusions: There are limitations to the success of all the available treatments. The
author urges caution in interpreting the results, particularly in comparisons between
treatments and across studies, because there are broad differences in the pain syn-
dromes and inclusion criteria used, the drug dosages, comparability of treatments, the
definition of “chronic” used, the outcome criteria selected to determine success, and
societal differences. None of the currently available treatments eliminates pain for the
majority of patients. Pain rehabilitation programs provide comparable reduction in pain
to alternative pain treatment modalities, but with significantly better outcomes for
medication use, health care utilization, functional activities, return to work, closure of
disability claims, and with substantially fewer iatrogenic consequences and adverse
events. Surgery, spinal cord stimulators, and IDDSs appear to have substantial benefits
on some outcome criteria for carefully selected patients. These modalities are, how-
ever, expensive. Pain rehabilitation programs are significantly more cost effective than
implantation of spinal cord stimulators, IDDSs, conservative care, and surgery, even
for selected patients. Research is needed to identify which patients are most likely to
benefit from the available treatments and to study combinations of the available treat-
ments since none of them appear capable of eliminating pain or significantly improv-
ing functional outcomes for all treated.
Key Words: Anticonvulsants—Antidepressants—Implantable drug delivery sys-
tems—Opioids—Pain rehabilitation program—Spinal cord stimulation.

Received April 27, 2002; accepted April 27, 2002.

Preparation of this article was supported in part by grants from the
National Institute of Arthritis and Musculoskeletal and Skin Diseases
(AR/AI44724, AR47298), and the National Institute of Child Health
and Human Development/National Center for Medical Rehabilitation
Research (HD33989). Preparation of this article and the special section in
which it appears was supported by an unrestricted educational grant from
Pfizer to the University of Rochester Office of Professional Education.
Address correspondence and reprint requests to Dennis C. Turk,
Ph.D., Department of Anesthesiology, Box 356540, University of
Washington, Seattle, Washington 98195, U.S.A.; e-mail: turkdc@
u.washington.edu
CHRONIC PAIN: NUMBERS OF
PEOPLE AFFECTED
Surveys in several Western countries confirm that
chronic and recurring pain is a significant problem for a
substantial proportion of the population. Verhaak et al.1
reviewed 15 epidemiologic studies and noted that in the
adult population chronic pain ranges from 2% to 40%,
and they concluded that the median point prevalence
(reports of pain at the time of survey, therefore not

355
356
dependent on recollection) of 15% of the adult popula-
tion reported chronic pain. In comparison, epidemiologic
studies in lower income countries (i.e., Nepal, India, Ni-
geria, China, Indonesia, and the Philippines) indicate
comparable point prevalence rates (18.5%), but for back
pain alone.2
Examination of the largest categories of chronic pain
syndromes shows that the vast majority of persons have
symptoms that are not related to demonstrable diseases
such as low back pain and headache.3 In the Netherlands,
over 10,000 new cases of work-related disability are re-
ported annually.4 In the United Kingdom, 12.5% of all
unemployed people cite back pain as the reason.5 An
estimated 2.5 million people reported having back pain
every day of the year.5
Back pain is one of the most common sources of dis-
ability as well as pain. Volinn2 identified seven epide-
miologic studies conducted in Britain, Belgium, Ger-
many, and Sweden that reported on the point prevalence
of low back pain. Weighting the percentages by the
sample size and aggregating across studies reveal that the
rates of back pain in these countries averages approxi-
mately 34%, almost twice that reported in surveys con-
ducted in the less developed countries noted previously.2
Several epidemiologic studies have been published
since the Verhaak et al.1 and Volinn et al.2 reviews. In a
large-scale telephone survey conducted in Australia, ap-
proximately 17% of male respondents and 20% of fe-
male respondents reported the presence of some form of
chronic pain.6 In the Welsh Health Survey more than
30% of respondents reported that they experienced back
pain, and 25% of the adults indicated that they had pain
associated with arthritis.7 In a mail survey conducted in
Sweden, 34.5% of the population indicated the presence
of persistent pain for at least three months.8 A survey
conducted in Scotland found that more than 50% of re-
spondents reported chronic pain.9 The wide range (20%
to 50%) may be due partially to the sampling methods
used, the definition of chronicity, the type of measure,
the focus on specific body locations, the phrasing of the
questions, and the sample size included. Despite the dis-
crepancies, there seems no question that, when asked, a
significant proportion of the population will indicate that
they experience chronic pain.
The figures cited must be viewed with some caution
because response to mail and telephone surveys may lead
to inflated estimates of the prevalence of symptoms. It is
reasonable to consider the behavioral impact of symp-
toms for those who report they have chronic pain—in
particular, are those who report chronic pain impaired or
disabled by these symptoms. The results of the Australia
survey6 noted that approximately 35% of those who
specified that they experienced chronic pain reported that
The Clinical Journal of Pain, Vol. 18, No. 6, 2002
TURK
pain actually interfered with their daily activities. In a
primary care cohort in the United States, 13% of head-
ache and 18% of back pain patients asserted that they
were unable to obtain or maintain full-time work over a
3-year period because of their pain.10
Another way to view the data on chronic pain is to
examine health care use. In the United States, 17% of
patients seen in primary care report persistent pain.11
Based on a national survey of pain specialists conducted
in 1995, the authors12 estimated that 2.9 million Ameri-
cans (1.1% of the population) are treated annually by
health care professionals specializing in chronic pain.
This figure does not, of course, include patients treated
by primary care physicians or specialists who do not
consider themselves pain specialists, nor does it include
visits to practitioners of complementary and alternative
medicine modalities, or self-medication using over-the-
counter preparations.
The figures cited here attest to the significant numbers
of people who report that they experience chronic pain
and that pain significantly impacts their lives. Chronic
pain not only affects the individual sufferer but his or her
“significant others.” If we consider spouses, partners,
family members, then the absolute numbers of people in
the population affected by pain expands geometrically,
leaving only a minority of the population untouched.
Few, if any, people will completely avoid an intimate
relation with persistent pain symptoms at some time dur-
ing their lives.
COSTS OF CHRONIC PAIN
Clinicians have an extensive armamentarium available
to treat people with chronic pain—pharmacological
preparations (e.g., opioids, nonsteroidals, anticonvul-
sants, tricyclic antidepressants, NMDA antagonists, topi-
cal preparations), operative procedures, physical modali-
ties (e.g., ultrasound, transcutaneous electrical nerve
stimulation, diathermy), regional anesthesia, neuroaug-
mentation modalities (e.g., spinal column stimulators
[SCSs], implantable drug delivery systems [IDDSs]),
comprehensive pain rehabilitation programs (PRPs)
(e.g., interdisciplinary pain centers, functional restora-
tion programs), and complementary and alternative
medicine modalities (e.g., chiropractic, acupuncture).
The direct costs for the totality of various treatment
approaches are astronomical and underscore the ex-
tremes undertaken to avert the human suffering associ-
ated with chronic pain. Cousins13 suggested that the
costs of health care for patients with chronic pain might
exceed the combined costs of treating patients with coro-
nary artery disease, cancer, and AIDS. In the United
 EFFECTIVENESS OF TREATMENTS
Kingdom, back pain alone is estimated to cost society
$11 billion to $20.6 billion each year14 (note that all cost
estimates used in this article are converted to U.S. dol-
lars). Direct costs associated with migraine in the United
States are estimated to be in excess of $1 billion (1993
rates).15 De Lissovy et al.16 estimated that, in the United
States, the cost of treatment in the first year after failed
back surgery for pain was approximately $18,883 (1997
dollars). Health care expenditures comprise only a rela-
tively small portion of the costs associated with chronic
pain. The majority of the costs are associated with dis-
ability compensation, lost productivity, and lost tax rev-
enue. Frymoyer and Durett17 projected the costs for back
pain, the most prevalent chronic pain syndrome, to ex-
ceed $33.6 billion for health care, $11 to $43 billion for
disability compensation, $4.6 billion for lost productiv-
ity, and $5 billion in legal services. Indirect costs asso-
ciated with migraine are anticipated to exceed $13 billion
each year in the United States (1993 dollars).15 Again
using United States data, patients with rheumatoid arthri-
tis are projected to incur over $14 billion (year 2000
dollars) in medical expenditures and work loss.18 Pain is
costly not only in the United States. In the state of Vic-
toria [Melbourne], Australia, for example, over $151
million was paid out in claims for back pain in 1996–
97.19
The annual health care costs incurred by a chronic
pain patient, excluding costs for surgical procedures,
may range from $500 to as high as $35,400, with the
averages ranging from $12,900 to $18,833 annually
(1988–1997 dollars).16,20,21 The vast majority of chronic
pain patients are managed with medication. Over the past
decade there has been a surge of articles extolling the
virtues of opioids to treat chronic pain that is not asso-
ciated with cancer and a recognition that the negative
consequences previously feared (e.g., addiction, drug di-
version) are not as common as previously thought in this
patient population. In 1999, more than 3 million pre-
scriptions were written for one opioid, Oxycontin (Pur-
due Pharma L.P., Norwalk, CT, U.S.A.).22 The costs of
drugs will vary by location. The cost of one 20-mg tablet
of Oxycontin in two locations in Atlanta, Georgia, in
2000 ranged from $2.22 to $2.56/tablet.23 Straus noted
that a typical patient with failed back surgery syndrome
in two practices in Atlanta might, along with NSAIDS
and other drugs (e.g., muscle relaxants, antidepressants,
anxiolytics, anticonvulsants), be prescribed five 20-mg
tablets of Oxycontin a day. The costs for Oxycontin
alone would exceed $4,600 per year, not including re-
lated physician visits or laboratory work.
There has been growing support for the use of anti-
convulsants, antidepressants, and topical preparations for
neuropathic pain syndromes. The cost for these medica-
357
tions also varies by location. Again, in Atlanta, Straus23
reported that one frequently prescribed anticonvulsant
medication for neuropathic pain, Neurontin (Parke-
Davis, Morris Plains, NJ, U.S.A.), cost $1.19 for each
300-mg tablet. Based on the data provided by Straus,
failed back surgery patients would spend almost $1,300
a year for Neurontin alone. This is a relatively conser-
vative figure because patients may be prescribed up to
four times the dosage stipulated by Straus.
Approximately 317,000 lumbar surgeries are per-
formed each year, primarily for pain, at a conservatively
estimated cost of $15,000 per operation.24 Using these
figures, the cost of lumbar surgery in the United States
would exceed $4.8 billion each year. There are, of
course, many other forms of surgery to alleviate pain
other than lumbar surgery (e.g., carpel tunnel release,
cervical neurectomy).
Based on the Marketdata survey cited earlier,12 only
about 176,850 (6% of those treated by pain specialists) of
chronic pain patients are treated at PRPs. Using the av-
erage figure of $13,28420 for nonsurgical health care
expenditures, seven years as the mean duration of pain,
and the mean number of surgeries 1.725 at an average
cost of $15,000,17 we can estimate the cost of health care
alone before treatment of patients at PRPs to be in excess
of $20 billion. Involvement of multiple disciplines at a
PRP is labor-intensive and costly because each of the
clinicians involved in treatment expects to receive pay-
ment for the services provided. Based on the average cost
of treatment at PRPs ($8,100)12 and the number treated
(176,850), the annual cost of treatment at PRPs would
exceed $1.4 billion (1995 dollars).
Over the past quarter century technological advances
have resulted in a number of sophisticated implantable
devices (i.e., SCS, IDDS) that are used to treat patients
with chronic pain. Segal and Stacey26 suggested that, by
1996, 10,000 SCSs had been implanted worldwide, with
7,000 implanted in the United States. Bell et al.21 pro-
jected that the five-year costs required for treating and
maintaining patients with SCSs in the United States
would equal $76,180. Thus, in the United States by 1996,
more than a half billion dollars had been committed to
these devices, related services, and treatment of adverse
events.
Implantable drug delivery systems are being advo-
cated to treat recalcitrant chronic pain.27 This technology
can be expensive. The initial costs for screening and
hospital and professional charges can range from
$15,000 to $30,000 (1997 dollars). De Lissovoy et al.16
suggested that the five-year costs would range from
$82,893 to $125,102. If the annual costs for medical
management range from $13,000 to $19,000, however,
then the treatment with IDDSs would be expected to
The Clinical Journal of Pain, Vol. 18, No. 6, 2002
358
“break even” in from 4 to 10 years. Similar calculations
for SCSs suggest that the five-year expected costs would
be $76,180 and thus the break-even point would occur in
4 to 6 years.
What is evident from these figures is that chronic pain
is both prevalent and is, regardless of how it is treated,
exceedingly expensive. Despite the high cost of treating
chronic pain, relief for many pain sufferers remains elu-
sive and total elimination of pain is rare. As we shall see
in what follows, although there have been phenomenal
advances in the knowledge of sensory physiology,
anatomy, and biochemistry, along with the development
of potent analgesic medications and other innovative
medical and surgical interventions, we have not elimi-
nated pain and disability as problems for a significant
portion of the population.
A number of criteria have been used to evaluate the
effectiveness of different pain treatment approaches and
modalities. Self-reports of pain and adverse events are
the most commonly used endpoints. Other criteria exam-
ined include functional activities, “return to work”
(RTW), health care utilization, and, to a lesser extent,
reduction in disability compensation. In the remainder of
this article I will examine the clinical effectiveness and
cost-effectiveness of the most common treatments used
for patients with chronic pain.
PAIN REDUCTION
The first-line treatment of pain consists of a host of
pharmacological agents. Pain medications are the second
most prescribed drugs (after cardiac-renal drugs) during
visits to physicians’ offices and emergency rooms,28 ac-
counting for 12% of all medication prescribed during
ambulatory office visits in the United States.29 Pharma-
ceutical industry data indicate that over 312 million pre-
scriptions for analgesic medications were written in 2000
in the United States, more than one prescription for an-
algesics for every man, woman, and child (Merck, per-
sonal communication, November, 2001)!
Despite their frequent use, currently available medi-
cations do not eliminate pain. For example, the average
pain reduction for patients placed on “long-term opioids”
is approximately 32%, when effects reported in studies
are weighted by sample size.31 I placed long-term opioid
therapy in quotation marks because the duration of the
majority of the published trials ranges from a week to
several months.31
A word of caution here and throughout this article
where I reported on averaged data across studies. Often
diverse pain syndromes, medications, surgical proce-
dures, and studies conducted in different countries are
included in the aggregation. This combination of factors
The Clinical Journal of Pain, Vol. 18, No. 6, 2002
TURK
may mask the results of the efficacy of different treat-
ments with different samples.
At a recent consensus meeting of specialists who treat
neuropathic pain, tricyclic antidepressants, anticonvul-
sant drugs, and topical preparations were viewed as the
treatments of choice.32 Seldom, however, do these phar-
macological agents reduce pain below a rating of 4 on an
11-point visual analog scale (e.g., 0–10 with 10 being the
highest level of pain) with 30% to 40% of patients re-
porting at least 50% reduction in pain.33,34 A common
outcome criterion used in drug trials is the “number
needed to treat” (NNT) to achieve a 50% reduction in
pain beyond what would have been achieved with a pla-
cebo. Several meta-analyses have reported using this cri-
terion for a range of antidepressant and anticonvulsant
medications for different pain syndromes, primarily neu-
ropathic pain. Pooling the data from these trials reveals
that the NNTs to achieve 50% pain reduction was 2.9.33
That is, for every three people receiving an antidepres-
sant or anticonvulsant for their pain, one will experience
at least a 50% pain relief that would not have occurred
with a placebo. Conversely, two of the three treated pa-
tients will have less than a 50% reduction in pain. I will
return to the criteria on which to base conclusions about
clinical effectiveness later. As is the case for opioids, the
duration of trials for antidepressants and anticonvulsants
is relatively brief, usually less than 3 months.
Obviously, medications do not eradicate all pain for
the majority of patients treated. This does not mean that
these drugs are not beneficial, only that we need to be
cautious in what outcomes can reasonably be expected.
The data also suggest that there is a need to consider
treatment combinations that may potentially improve
outcomes since these drugs do not completely eliminate
the problem of pain.
Persistence of back pain, and, to a lesser extent, other
chronic pain syndromes, frequently leads to surgery. A
number of studies,35–38 however, reveal that significant
pain may persist after surgery. For example, Lehmann et
al.37 showed that 75% of patients who had spinal fusion
for back pain continued to report pain after surgery;
Dvorak et al.35 studied 575 back pain patients who had
surgery herniated disks and noted that 70% continued to
report back pain up to 17 years after surgery; and North
et al.38 noted that 66% of patients who underwent repeat
surgery for back pain continued to experience pain five
years after surgery.
Several studies of SCSs have reported impressive re-
sults in pain reduction for carefully selected patients with
pain of long duration. For example, in a well-designed
study, North et al.39 reported on a long-term follow-up
(mean 7 years) for a consecutive series of patients. They
 EFFECTIVENESS OF TREATMENTS
noted that 52% of 171 patients receiving permanent im-
plants reported at least 50% relief of pain and 60% in-
dicated that they would undergo the procedure again
even though they still continued to have at least some
pain. In a systematic review, Turner et al.40 reviewed 39
studies of SCSs for low back pain and concluded that, on
average, 59% of patients had at least a 50% reduction in
pain.
More recently, however, Van Buyten et al.41 reported
that, four years after implantation with an SCS, 61% of
patients continued to report that their pain ranged from
“uncomfortable” to “horrible”; 37% fewer of the patients
rated their pain in this category at follow-up than before
implantation. One caution in interpreting these results:
the follow-up excluded 22 patients who had the stimu-
lators removed, 20 who could not be contacted, and an-
other 22 who had died (none attributed to the SCS).
Burchiel et al.42 presented one-year follow-up data and
reported that 40% of implanted patients indicated at least
a 50% reduction in pain. The absolute percentage of
average change in pain severity, as reported on a visual
analog scale (range 0–10), was only 18.6%, decreasing
from 7.31 to 5.95. These data raise a question regarding
the meaning of categorical ratings of percent pain reduc-
tion. The reduction in pain of 18.6% would not reach the
criterion of clinical significance—a 33% reduction.43
Moreover, the data indicate that patients still continued
to experience substantial pain severity after receiving the
implant.
Significant reductions of pain have also been reported
for IDDSs for both chronic pain associated with cancer
and other pain syndromes (e.g., back pain). Good to ex-
cellent results have been reported in most published stud-
ies. For example, Hassenbusch et al.44 and Paice et al.45
both reported a mean pain reduction of approximately
60% after epidural infusions in mixed samples of pa-
tients with different pain syndromes. In a small study (n
⳱ 16), Kumar et al.46 reported a mean 57.5% reduction
in pain at a follow-up of over two years, with 44% (n ⳱
7) patients reporting greater than 50% reduction in pain.
These results are particularly impressive since the pa-
tients who were included had had inadequate pain con-
trol after conservative approaches, long-term use of opi-
oids, and implantation of an SCS.
It is important to be alert to the mixed diagnosis of
patients treated across and within studies using the same
treatment, since some syndromes may be more or less
responsive to the treatment. For example, in contrast to
the Hassenbusch et al.44 and Paice et al.45 studies, Has-
senbusch et al.47 found that pain reduction was only 39%
for a sample of patients with neuropathic pain who were
implanted, and, in a mixed sample of predominantly neu-
ropathic pain, Anderson and Burchiel48 found that only
359
11 of 22 patients achieved at least a 25% reduction in
pain.
The reduction of pain after treatment at PRPs has been
reported to be statistically significant in several meta-
analyses25,49,50, with one meta-analysis25 reporting the
mean pain reduction for patients treated at PRPs as 37%.
The majority of patients continue to experience consid-
erable pain. The rates of reduction in pain across differ-
ent treatment modalities are displayed in Fig. 1. It is
important to acknowledge once again that none of the
available treatments for chronic pain have been demon-
strated to eliminate all pain for all patients.
Interestingly, the pain reduction achieved at PRPs is
accompanied by a significant decrease (63%) in prescrip-
tion pain medication.25 A number of SCS studies report
significant reductions in consumption of analgesics. For
example, Ohnmeiss et al.51 reported that, two years after
implantation, 84% of patients decreased or eliminated
opioids. North et al.39 reported that 58% of patients
treated with SCSs demonstrated a reduction in, or elimi-
nation of, analgesic medication consumption, and in a
retrospective study, Kumar et al.52 reported that 40% of
patients no longer used prescription analgesics. In con-
trast, however, a prospective study by Burchiel et al.42
found that only 7% of patient reported elimination of
opioid use one year after the implant. In one recent study,
SCSs have been reported to reduce consumption of opi-
oids by 35%.41 One study53 noted a 76.7% decrease in
oral pain medication after implantation of an intraspinal
infusion system. The medication-sparing effects for im-
plantable therapies and PRPs can provide significant sav-
ings, in costs for medication and health care.
IATROGENIC COMPLICATIONS AND
ADVERSE EVENTS
At least as important as reduction in pain severity
is the potential for iatrogenic complications and ad-
verse events to result from treatment. Selection of any
FIG. 1. Mean percentage of pain reduction after treatments for
chronic pain.
The Clinical Journal of Pain, Vol. 18, No. 6, 2002
360
intervention must balance the positive outcomes with
potential negative consequences. Each of the treatments
for chronic pain patients, perhaps with the exception of
PRPs, has the potential for undesirable consequences.
Long-term use of medication raises concerns about
tolerance, tolerability, drug diversion (i.e., opioids), and
side effects including neurotoxicity (e.g., 33,54). There are
always contraindications based on the patient’s medical
condition, age, childbearing status, co-occurring medical
conditions, and other medications prescribed.
Studies have reported that a significant percentage of
chronic pain patients treated with surgery report that
their pain is worse after surgery.36–38 Malter et al.55 re-
ported complication rates of 18% and 7% for back pain
patients having lumbar surgery involving fusion and with
laminectomy or discectomy alone, respectively. Reop-
eration rates were 18% and 15% for fusion and nonfu-
sion patients. Up to 33% of back surgeries are repeated
because of problems associated with previous surgery.56
Subsequent operations do not guarantee resolution of
pain, with some studies acknowledging the poor results
achieved for reoperations.38,57 Bell et al.21 suggested that
there is a 10% probability of a repeat surgery in every
succeeding year after lumbar surgery.
In a systematic review of 13 outcome studies for
SCSs, Turner et al.40 reported that on average, 42% of
patients experienced complications requiring interven-
tions related to the procedure itself or malfunctions of the
device. Although many of the complications are minor,
some require substantial medical or surgical intervention
and, consequently, cost.
PHYSICAL FUNCTIONING
Outcomes regarding changes in pain severity are de-
pendent on patients’ self-reports. Many factors can bias
patients’ responses. Moreover, the relation between pain
reports and functional behavior is modest. As a conse-
quence, evaluation of the effectiveness of pain treatment
approaches should consider functional outcomes such as
improvement in physical activity and, when appropriate,
RTW, along with reductions in pain severity.
Inspection of the pharmacological studies for pain re-
veals a striking lack of attention to functional outcomes.
For example, in one systematic review of antidepressants
for neuropathic pain,34 only 1 of 20 studies included any
indication of improvements in physical functioning after
initiation of treatment. The emphasis in the pharmaco-
logical literature is almost exclusively on pain relief and
adverse events. There are no data demonstrating that sig-
nificant numbers of patients return to work and minimal
data on improvement in functioning after long-term opi-
oids,58 tricyclic antidepressants,33,35,59–61 or anticonvul-
The Clinical Journal of Pain, Vol. 18, No. 6, 2002
TURK
sant medication.59–62 Similarly, outcomes studies for re-
gional anesthesia ignore RTW as an outcome. Only 1 of
14 studies examining the efficacy of regional blockage
for complex regional pain syndrome included functional
measures as outcomes.63
Similar to the pharmacological studies, outcomes re-
search for IDDSs rarely attends to the impact of treat-
ment on functioning. In the 18 studies described in a
recent review of intrathecal morphine,64 only 4 included
outcomes related to functional activities. When func-
tional outcomes are reported, the results are not particu-
larly impressive. For example, Paice et al.45 noted that
only 22.8% of a mixed sample of cancer and noncancer
patients with pain treated with IDDS reported “great”
increases in activities of daily living; however, 24.6%
reported no change, and 3.8% indicated a decrease in
functional activities.
In contrast to pharmacological, regional anesthesia,
and IDDS studies, investigators evaluating the effective-
ness of SCSs, surgery, and PRPs have, more frequently,
considered changes in functional outcomes along with
pain severity. For example, in a retrospective study of
SCSs, Van Buyten et al.41 noted a 26.6% improvement in
daily activities at follow-up. Kumar et al.46 found a
37.5% reduction in the number of patients reporting re-
strictions in their activities after implantation of an
IDDS. Interestingly, Gallon65 noted that 58% of back
pain patients who had surgery described themselves as
worse on a measure of physical functioning, compared
with 30% of patients who indicated they were more dis-
abled after standard care. In their meta-analysis, Flor et
al.25 reported a 43% increase in physical activities after
treatment in PRPs.
Many studies reporting on changes in physical activi-
ties have relied on patients’ self-reports. A more objec-
tive measure of function is RTW. There are caveats that
need to be taken into consideration, however, when con-
sidering RTW rates. The actual return to work may have
little to do with the readiness of a person to resume
job-related activities.66 For example, the mean duration
of pain for patients treated at PRPs is 7 years,25 and
someone who has been away from work for such a long
period may not have a job to return to and may find skills
outdated, making return to the previous job difficult.
Economic factors (e.g., the job market) will also influ-
ence whether someone returns to gainful employment
after treatment. Finally, administrative decisions regard-
ing appropriateness of RTW may be primary, with little
consideration given to the presence of the patients’
symptoms or physical capacities. These limitations, how-
ever, apply equally to all treatments, so even though
there are significant limitations, RTW rates can be used
to compare different pain treatments.
 EFFECTIVENESS OF TREATMENTS
Rates of RTW after lumbar surgery have been re-
ported to be as low as 20%.38 One recent study reported
a 50% RTW rate after radiofrequency facet joint dener-
vation in the treatment of low back pain.67 At first this
outcome seems quite impressive; however, the group of
patients treated with sham therapy showed the same rate
of RTW. These sobering results underscore the necessity
of developing and using appropriate placebo treatments.
Studies of patients who have been implanted with
SCSs suggest that from 5% to 31% eventually return to
work.38,39,46,68 One of the few studies on the results of
IDDSs on the RTW criterion indicates that none of the
patients (n ⳱ 16) returned to gainful employment.46
These outcomes can be compared with the RTW rates
reported for patients treated at PRPs that range from 48%
to 65%.25 The outcomes after treatment at PRPs are quite
impressive given the long pain duration (mean ⳱ 7
years) for treated patients.
CLOSURE OF DISABILITY CLAIMS
Disability payments may exceed medical costs for per-
sons with chronic pain by a factor of five. An important
outcome, at least from a societal perspective, is closure
of disability claims. Once again, I need to invoke a cau-
tion. As in the case for RTW, closure of disability claims
may depend on administrative decisions and not on pain
per se. Only outcomes studies for PRPs and one for
SCS42 have reported on changes in disability claims after
treatment. Flor et al.25 note an approximately 50% reduc-
tion in rates of disability after treatment at PRPs. This rate
can be compared with the study of SCSs conducted in the
United States42 that reported a 20% reduction (not statisti-
cally significant) in disability status one year after implant.
Recently, Thomsen et al.69 used social records instead
of self-reports to evaluate the efficacy of a PRP, obtain-
ing data on disability and welfare costs for a period of 6
months before entry to a 4-month waiting list and at a
9-month follow-up after termination. The authors identified
significant reductions in social transfers (welfare benefits,
sickness benefit, and pensions). These investigators noted a
63% decline in benefits during the follow-up period.
COST BENEFITS AND COST-EFFECTIVENESS
At a time when resources for health care are limited
there is a growing concern about not only the clinical
effectiveness of treatment but also the cost-effectiveness.
As the data presented earlier attest, there is no question
that chronic pain syndromes are exceedingly costly. The
expenditures for chronic pain, in addition to actual health
care, include welfare and disability payments, lost tax
revenue, lost productivity, and expenses required to train
replacement workers.
361
Health care utilization
One way to consider cost-effectiveness is to evaluate
reduction in health care utilization. The results of the
meta-analysis by Flor et al.25 indicate that, after treat-
ment at PRPs, patients required one third the number of
surgical interventions and hospitalizations compared
with patients treated by alternative medical and surgical
care. Beyond the professional costs and facility charges
for medical and surgical treatments, an additional factor
contributing to the comparative cost benefits for PRPs, in
contrast with neuroaugmentative modalities and long-
term opioid therapy, is that minimal medical monitoring
of patients treated at PRPs is required.
Simmons et al.20 reported a 62% reduction of medical
costs after treatment at a PRP. If we extrapolate the sav-
ings reported by Simmons et al. to the 2,318 patients
successfully treated in PRPs reported in the meta-
analysis by Flor et al.,25 factoring the cost of treatment at
the PRP, a savings of over $18 million in medical ex-
penses would accrue during the first year after treatment.
These figures are for the first year’s medical expenses
only and do not reflect savings in reduced disability pay-
ments, gains in productivity, and gains in revenue from
taxes paid by the successfully treated patients.
Of course, reduction in medication is not a goal of
pharmacological treatments. Rather the intent is to have
patients take appropriate and adequate doses of medica-
tion to reduce their pain and thereby improve the quality
of their lives.
Using the figure of 176,000 patients treated at PRPs12
and the figures for health care expenditures described
above, I estimate that the medical cost savings during the
first year after treatment at PRPs at more than $1.87
billion. Bell et al.21 suggest that the annual medical costs
for failed back syndrome patients is $12,900 (including
hospitalization, physician visits, medications, alternative
therapies, diagnostic procedures, rehabilitation, and other
therapies). Using the figure of 68% reduction in health
care expenses in the year after treatment at a PRP re-
ported by Simmons et al.,20 I estimate that the cost sav-
ing for health care expenditures in years subsequent to
the first year after treatment at $8,772 per treated patient
per year. The average age of patients treated at PRPs is
44,25 and, assuming a mean life expectancy of 75 years,
we can estimate $45 billion in savings in health care
expenditures (without correcting for inflation rates dur-
ing those 30 years). Stieg and Turk 70 calculated
$350,000 (corrected for inflation to 1999 dollars) as the
cost for each compensation-covered male subject with a
back injury on the job lost to permanent disability. Sys-
tematic comparisons of the cost-effectiveness across dif-
ferent modalities need a common metric. Cost effective-
ness can be viewed as:
The Clinical Journal of Pain, Vol. 18, No. 6, 2002
362 TURK
Cost-Effectiveness =
Cost of treatment
Percentage Who Achieve Outcome
This formula can be used to examine any outcome of
interests (e.g., pain reduction, improvement in quality of
life, patient satisfaction). When costs are not involved it
is referred to as the cost benefit rather than cost-
effectiveness. To illustrate the use of this formula I will
compare the cost-effectiveness of several treatments in
returning people to work, making use of the data pro-
vided in several publications (i.e., 23,25,38).
I graphically presented the cost of SCSs, standard
care, surgery, and PRPs in returning one treated patient
to work. Examination of the data presented in Fig. 2
reveals that treatment of patients at PRPs would be 6.29,
15, and 25 times more cost effective than surgery, con-
servative care, or implantation of SCSs, respectively.
These estimates, moreover, do not take into consider-
ation the costs incurred for iatrogenic problems that may
follow surgery or SCSs. Iatrogenic problems can be
prevalent and costly. For example, as noted previously in
their review of studies of SCSs for back pain, Turner et
al.40 concluded that 42% of patients who had an SCS
implanted experienced at least one complication, with
the costs in some cases exceeding $19,000. Moreover,
these relative cost ratios do not take into consideration
routine follow-up care and monitoring after treatment.
In the Thomsen et al.69 study cited previously, the
authors evaluated the clinical and cost-effectiveness of a
PRP in Denmark. This study is particularly noteworthy
because the authors had access to social records and
could evaluate health care costs and disability without
depending on patient self-reports. The authors found that
in the 6 months before treatment at a PRP the average
patient spent $263 per month for health care. In the 9
months after treatment the patients averaged $182 per
month (a difference of $81 per month from the before-
treatment cost). This reflects a 31.1% decrease in health
care consumption after treatment.
Thomsen et al.69 did not report on the age of the pa-
tients included in their study, but if we accepted the data
FIG. 2. Cost to return one treatment patient to work.
The Clinical Journal of Pain, Vol. 18, No. 6, 2002
aggregated across 65 studies reported by Flor et al.25 as
representative, the mean age of treated patients would be
44 years. To consider accrued cost saving, I will use 19
years, since this allows us to follow patients until the
treated patients reaches age 65. Based on the figures
reported by Thomsen et al., the anticipated average an-
nual savings in health care would be $972 per patient.
Since the cost of the treatment reported was very low,
$99 (substantially lower than reported in the United
States), the “break-even point” would be achieved in
about six weeks. Multiplying the annual rate by 19 years,
the expected savings in health care expenditures for each
patient would be $18,468. The number of patients avail-
able at the follow-up was 122. Again, using the 19-year
figure, the estimated health care savings for patients
treated at the PRP would be in excess of $2.25 million.
Of course, we must subtract the cost of the treatment of
the 131 patients—$99/patient × 131 treated patients ⳱
$12,969—but even after subtracting the costs of the
treatment at the PRP, the health care savings would ex-
ceed $2.16 million. We can consider the reductions in
social transfer for 19 years as well. Working through the
arithmetic, the total reduction would amount to $58,892
per patient treated. Using the number of patients avail-
able at follow-up (122), the figure totals more than $7
million.
Combining the savings in health care and social trans-
fers reported by Thomsen et al.69 indicates that the av-
erage savings per patient treated at this PRP over a period
of 19 years would be equivalent to $77,360, and for the
122 (and factoring the cost of treatment of 131 patients)
the calculation would lead to substantial savings, in ex-
cess of $9 million.
CONCLUSIONS
Pain is not a monolithic entity such as a fracture or
deficiency of some essential nutrient. Pain is, rather, a
concept used to focus and label a group of sensations,
thoughts, emotions, and behaviors. Since there are many
facets to pain, it should be obvious that no single out-
come measure captures all of the relevant issues. For this
reason, outcomes assessment must look at a variety of
criteria to adequately describe the effects of any
treatment.
Concerns about the cost of health care and, conse-
quently, the cost-effectiveness ratios of treatments, have
become almost as important as clinical effectiveness.
Simply asking whether any treatment is effective is not
the appropriate question. Rather, it is important to focus
on the results for a set of outcomes and to compare these
results with alternative treatments. The need for multiple
criteria to evaluate treatment outcome is underscored by
 EFFECTIVENESS OF TREATMENTS
the reports that success in one area may not necessarily
be related to success in others. For example, some pa-
tients may report significant pain reduction but not their
return to work. Patients and third-party payers may in-
terpret these results quite differently.
I need to reiterate an issue that I raised earlier. Inter-
preting results by aggregating outcomes across and
within studies, including different pain syndromes, treat-
ments, dosages of medications, criteria of success, coun-
tries, follow-up periods, and criteria used to determine
success, should be undertaken with due caution.
The available data do suggest that PRPs improve over-
all functioning of chronic pain sufferers. These improve-
ments are not just evident on self-report measures; they
are shown on objective criteria, such as employment sta-
tus, medication use, utilization of the health care system,
and closure of disability cases. Careful selection of pa-
tients treated with SCSs also appears to lead to improve-
ments of pain and functioning, although there is a cost of
iatrogenesis. The results for RTW, keeping in mind the
caution noted in regard to this outcome, are less impres-
sive. The PRPs appear to provide the best outcome on
rates of RTW and disability, with almost 50% of treated
patients returning to gainful employment.25 Moreover, a
substantial proportion of patients have closure of disabil-
ity claims after treatment.25 The PRPs have the potential
to save substantial sums in both health care expenditures
and disability payments. Unfortunately, studies of the
most common treatments, medication, rarely consider
outcomes other than improvements in pain severity and
prevalence of adverse events and complications associ-
ated with the procedures.
The characteristics of patients treated by pain special-
ists differ in important ways from patients treated in
primary care facilities. Typically, populations treated by
specialists demonstrate a high prevalence of depression,
with averages approximately 50%.71 In contrast, in a
survey of pain patients treated in a health maintenance
organization, Von Korff et al.72 reported that the preva-
lence of major depression ranges from 6% to 10% de-
pending on the type of pain. In epidemiologic surveys
conducted by Crook et al.,73,74 chronic pain patients re-
ferred to pain specialists were compared with persistent
pain patients in the community who were not referred to
specialty pain treatment facilities. Examination of differ-
ences identified (Table 1) led Weir et al.75 to conclude
that “the picture that emerges is one of a pain clinic
patient who is isolated from meaningful relationships,
whose pain interferes with all aspects of his life, who has
lost his energy and determination, is unemployed, is in
constant pain and experience the threat, loss and harm of
his pain problem.” Given this rather dismal description,
363
TABLE 1. Differences between patients with chronic pain
treated in the community and by pain specialists*
Patients treated by pain specialists demonstrate:
Higher prevalence of work-related injuries
Higher levels of emotional distress
Greater health care utilization
More constant pain
More negative attitudes about the future
Higher prevalence of opioid intake
Higher prevalence of past surgery
Greater functional impairment
*Adapted from Crook et al.73,74
the results for the pain treatments reviewed, especially
those provided by pain specialists, are quite remarkable.
Despite the skepticism and criticisms raised by third-
party payers, the body of literature available provides
substantial evidence that currently available treatments
can significantly reduce the severity of pain experienced
by a large number of people. Functional outcomes,
health care consumption, and reduction in disability also
can be favorably affected. This is especially evident in
the results reported for PRPs. None of the available treat-
ments on its own, however, seems capable of eliminating
all pain for all pain sufferers. As a consequence we must
be realistic in the message that we convey to patients
about the most likely outcomes regarding pain.
There is a trade off between pain reduction and iatro-
genic complications and adverse events. With the ex-
ception of PRPs, all of the other commonly used treat-
ments for chronic pain have some risk associated with
them. Health care providers must carefully balance the
risk against the potential benefits of each treatment
when selecting from among available treatment options.
The prudent health care provider will be cautious in se-
lecting patients for treatments based on potential for
positive outcomes and adverse events and iatrogenic
complications.
Currently, few data are available that are consistent in
identifying the characteristics of patients who would
most likely benefit from any of the pain treatment meth-
ods available. Studies are needed that answer the ques-
tion: what treatments delivered in what ways are most
effective for patients with what set of characteristics with
the least iatrogenic complications and adverse events?
Successful answers to this question will permit more
clinically effective and cost-effective ways to treat the
difficult population of patients with chronic pain.
REFERENCES
1. Verhaak PFM, Kerssens JJ, Dekker J, et al. Prevalence of chronic
benign pain disorder among adults: a review of the literature. Pain
1998;77:231–9.
2. Volinn E. The epidemiology of low back pain in the rest of the
The Clinical Journal of Pain, Vol. 18, No. 6, 2002
364
world. A review of surveys in low- and middle-income countries.
Spine 1997;22:1747–54.
3. Frolund F, Frolund C. Pain in general practice. Scand J Primary
Healthcare 1986;4:97–100.
4. Lousberg R. Chronic pain. Multiaxial diagnostics and behavioral
mechanisms. Maastricht: University of Maastricht, 1994 (thesis).
5. Elliott AM, Smith BH, Penny KI, et al. The epidemiology of
chronic pain in the community. Lancet 1999;354:1248–52.
6. Blyth FM, March LM, Brnabic AJM, et al. Chronic pain in Aus-
tralia: a prevalence study. Pain 2001;89:127–34.
7. National Assembly for Wales. Welsh Health Survey. Cardiff: Gov-
ernment Statistical Service, 1999.
8. Beregman S, Herrstrom P, Hogstrm K, et al. Chronic musculo-
skeletal pain, prevalence rates and sociodemographic associations
in a Swedish population study. J Rheumatol 2001;28:1369–77.
9. Back Care, 2001. Available at: http://www.back pain.org/pr-
facts.htm.
10. Stang P, Von Korff M, Galer BS. Reduced labor force participation
among primary care patients with headache. J Gen Intern Med
1998;13:296–302.
11. Gureje O. Persistent pain and well-being: a World Health Organi-
zation study in primary care. JAMA 1998;280:147–51.
12. Marketdata Enterprises. Chronic pain management programs: a
market analysis. Valley Stream, NY: Marketdata Enterprises,
1995.
13. Cousins MJ. Foreword. In: Fordyce WE, ed. Back pain in the
workplace: management of disability in nonspecific conditions.
Task Force Report. Seattle: IASP Press, 1995, p. ix.
14. Maniadakis N, Gray A. The economic burden of back pain in the
UK. Pain 2000;84:95–103.
15. Hu XH, Markson LE, Lipton RB, et al. Burden of migraine in the
United States: disability and economic costs. Arch Intern Med
1999;159:813–8.
16. De Lissovoy G, Brown RE, Halpern M, et al. Cost-effectiveness of
long-term intrathecal morphine therapy for pain associated with
failed back surgery syndrome. Clin Ther 1997;19:96–112.
17. Frymoyer J, Durett C. The economics of spinal disorders. In: Fry-
moyer J, ed. The adult spine. Philadelphia: Lippincott-Raven,
1997:143–150.
18. Lubeck PA. Review of the direct costs of rheumatoid arthritis.
Pharmacoeconomics 2001;19:811–8.
19. Buchbinder R, Jolley D, Wyatt M. Population-based intervention
to change back pain beliefs and disability: three-part evaluation.
BMJ 2001;322:1516–20.
20. Simmons J, Avant W, Demski J, et al. Determining successful pain
clinic treatment through validation of cost effectiveness. Spine
1988;13:24–34.
21. Bell G, Kidd D, North R. Cost-effectiveness analysis of spinal cord
stimulation in treatment of failed back surgery syndrome. J Pain
Symp Manage 1997;13:286–95.
22. Red Book. Drug Topics. Top 200 brand-name drugs by prescrip-
tion. Montvale, NJ: Medical Economics, March 6, 2002.
23. Straus BN. Chronic benign pain syndromes—the cost of interven-
tion. Spine; in press.
24. National Center for Health Statistics. 1997 national hospital dis-
charge survey. Series 13, no. 144. Washington, DC: U.S. Dept. of
Health and Human Services, Center for Disease Control, 1997.
25. Flor H, Fydrich T, Turk DC. Efficacy of multidisciplinary pain
treatment centers; a meta-analytic review. Pain 1992;49:221–30.
26. Segal R, Stacey BR. Spinal cord stimulation revisited. Neurol Res
1998;20:391–6.
27. Krames ES, Olson K. Clinical realities and economic consider-
ations: patient selection in intrathecal therapy. J Pain Symptom
Manage 1997;14(Suppl):S3–13.
28. Schappert SM. Ambulatory care visits to physicians offices, hos-
pital outpatient departments, and emergency departments: United
States, 1996. Vital and Health Statistics, series 13 (134),
1998;1–80.
29. National Center for Health Statistics. National ambulatory medical
The Clinical Journal of Pain, Vol. 18, No. 6, 2002
TURK
care survey, 1998. Washington, DC: U.S. Dept. of Health and
Human Services, 1998.
30. Ref. deleted during editing.
31. Turk DC, Loeser JD, Monarch ES. Chronic pain: purposes and
costs of interdisciplinary pain rehabilitation programs. TEN: Cost
Neurosci; in press.
32. Dworkin RH, Allen RR, Argoff CR, et al. Guidelines for the phar-
macologic management of chronic neuropathic pain. Submitted
for publication.
33. Collins SL, Moore A, McQuay HJ, et al. Antidepressants and
anticonvulsants for diabetic neuropathy and postherpetic neuralgia:
a quantitative systematic review. J Pain Symptom Manage. 2000;
20:449–58.
34. McQuay HJ, Tramer M, Nye BA, et al. A systematic review of
antidepressants in neuropathic pain. Pain 1996;68:217–27.
35. Dvorak J, Gauchat M, Valach L. The outcome of surgery for
lumbar disc herniation: I. A 4–17 years’ follow-up with emphasis
on somatic aspects. Spine 1988;13:1418–22.
36. Franklin GM, Haug J, Heyer NJ, et al. Outcome of lumbar fusion
in Washington State Workers’ Compensation. Spine 1994;19:
1897–904.
37. Lehmann TR, Spratt KF, Tozzi JE, et al. Long-term follow up of
lower lumbar fusion patients. Spine 1987;12:97–104.
38. North RG, Ewend MG, Lawton MT, et al. Failed back surgery
syndrome: 5-year follow-up after spinal cord stimulator implanta-
tion. Neurosurgery 1991;28:692–9.
39. North RB, Kidd DH, Zahurak M, et al. Spinal cord stimulation for
chronic, intractable pain: experience over two decades. Neurosur-
gery 1993;32:384–94.
40. Turner JA, Loeser JD, Bell KG. Spinal cord stimulation for chronic
low back pain: a systematic literature synthesis. Neurosurgery
1995;37:1088–96.
41. Van Buyten J-P, Van Zundert J, Vueghs P, et al. Efficacy of spinal
cord stimulation: 10 years of experience in a pain center in Bel-
gium. Eur J Pain 2001;5:259–307.
42. Burchiel KJ, Anderson VC, Brown FD, et al. Prospective, multi-
center study of spinal cord stimulator for relief of chronic back and
extremity pain. Spine 1996;21:2786–94.
43. Farrar JT, Portenoy RK, Berlin JA, et al. Defining the clinically
important difference in pain outcome measures. Pain 2000;88:
287–94.
44. Hassenbusch SJ, Stanton-Hicks MD, Soukup J, et al. Sufentanil
citrate and morphine/bupivacaine as alternative agents in chronic
epidural infusions for intractable noncancer pain. Neurosurgery
1991;29:76–82.
45. Paice JA, Penn RD, Shott S. Intraspinal morphine for chronic pain:
a retrospective, multicenter study. J Pain Symptom Manage 1996;
11:71–80.
46. Kumar K, Kelly M, Pirlot T. Continuous intrathecal morphine
treatment of chronic pain of nonmalignant etiology: long-term ben-
efits and efficacy. Surg Neurol 2001;55:79–88.
47. Hassenbusch SJ, Stanton-Hicks M, Covington EC, et al. Long-
term intraspinal infusions of opioids in the treatment of neuro-
pathic pain. J Pain Symptom Manage 1995;10:527–43.
48. Anderson VC, Burchiel KJ. A prospective study of long-term in-
trathecal morphine in the management of chronic nonmalignant
pain. Neurosurgery 1999;44:289–301.
49. Guzman J, Esmail R, Karjalainenen K, et al. Multidisciplinary
rehabilitation for chronic low back pain: systematic review. BMJ
2001;7301:1511–5.
50. Morley S, Eccleston, C, Williams A. Systematic review and meta-
analysis of randomized controlled trials of cognitive behavior
therapy and behavior therapy for chronic pain in adults, excluding
headache. Pain 1999;80:1–13.
51. Ohnmeiss DD, Rashbaum RF, Boddanffy GM. Prospective evalu-
ation of spinal cord stimulation in patients with intractable leg
pain. Spine 1996;21:1344–51.
52. Kumar K, Nath R, Wyatt GM. Treatment of chronic pain by epi-
dural spinal cord stimulation: a 10-year experience. J Neurosurg
1991;75:402–7.
 EFFECTIVENESS OF TREATMENTS
53. Doleys D, Coleton M, Tutak U. Use of intraspinal infusion therapy
with non-cancer pain patients: follow up and comparison of work-
er’s compensation vs. non-worker’s compensation patients. Neu-
romodulation 1998;1:149–59.
54. Breivik H. Opioids in cancer and chronic non-cancer pain
therapy—indications and controversies. Acta Anesthesiol Scand
2001;45:1059–66.
55. Malter AD, McNeney B, Loeser JD, et al. 5-year reoperation rates
after different types of lumbar spine surgery. Spine 1998;23:
814–20.
56. Long DM, Filtzer DL, BenDebba M, et al. Clinical features of the
failed-back syndrome. J Neurosurg 1988;69:61–71.
57. Friedlieb O. The impact of managed care of the diagnosis and
treatment of low back pain: a preliminary report, Am J Med Qual
1994;9:24–9.
58. Turk DC. Clinician attitudes about prolonged use of opioids and
the issue of patient heterogeneity. J Pain Symptom Manage 1996;
11:218–30.
59. Sindrup S, Jensen TS. Efficacy of pharmacological treatments of
neuropathic pain: an update and effect related to mechanism of
drug action. Pain 1999;83:389–400.
60. Sindrup SH, Jensen TS. Pharmacologic treatment of pain in poly-
neuropathy. Neurology 2000;55:915–20.
61. Sindrup SH, Jensen TS. Pharmacotherapy of trigeminal neuralgia.
Clin J Pain 2001;18:22–7.
62. McQuay H, Carrol D, Jadad AR, et al. Anticonvulsant drugs for
management of pain: a systematic review. BMJ 1995;311:
1047–52.
63. Kingery WS. A critical review of controlled clinical trials for pe-
ripheral neuropathic pain and complex regional pain syndromes.
Pain 1997;73:123–39.
64. Prager JP. Neuraxial medication delivery: the development and
maturity of a concept for treating pain associated with failed back
surgery syndrome. Spine; in press.
65. Gallon R, Perception of disability in chronic back pain patients: a
long-term follow up. Pain 1989;37:67–75.
365
66. Turk DC. Transition from acute to chronic pain: role of demo-
graphic and psychosocial factors. In: Jensen TS, Turner JA, Wie-
senfeld-Hallin Z, eds. Proceedings of the eighth world congress on
pain: progress in pain research and management. Seattle: IASP
Press, 1997:185–213.
67. Leclaire R, Fortin L, Lambert R, et al. Radiofrequency facet joint
denervation in the treatment of low back pain: a placebo-controlled
clinical trial to assess efficacy. Spine 2001;26:1411–6.
68. Kupers R, Van den Oever R, Van Houdenhove B, et al. Spinal cord
stimulation in Belgium: a nation-wide survey on the incidence,
indications and therapeutic efficacy by the health insurer. Pain
1996;56:211–7.
69. Thomsen A, Sorensen J, Sjogren P, et al. Chronic non-malignant
pain patients and health economic consequences. Eur J Pain
2002;6:341.
70. Stieg RL, Turk DC. Chronic pain syndrome: the necessity of dem-
onstrating the cost-benefit of treatment. Pain Manage 1988;1:
58–63.
71. Banks SM, Kerns RD. Explaining high rates of depression in
chronic pain: a diathesis-stress framework. Psychol Bull 1996;119:
95–110.
72. Von Korff M, Dworkin S, LeResche L, et al. Epidemiology of
temporomandibular disorders: II. TMD pain compared with other
common pain sites. In: Dubner R, Gebhart G, Bond M, eds. Pro-
ceedings of the fifth world congress on pain. Amsterdam: Elsevier,
1988:506–11.
73. Crook J, Tunks E, Rideout E, et al. Epidemiologic comparison of
persistent pain sufferers in a specialty pain clinic and in the com-
munity. Arch Phys Med Rehabil 1986;67:451–5.
74. Crook J, Weir R, Tunks E. An epidemiologic follow-up survey of
persistent pain sufferers in a group family practice and specialty
pain clinic. Pain 1989;36:49–61.
75. Weir R, Browne GB, Tunks E, et al. A profile of users of specialty
pain clinic services: predictors of use and cost estimates. J Clin
Epidemiol 1992;45:1399–415.
The Clinical Journal of Pain, Vol. 18, No. 6, 2002