/clinical/,/clinical/cckm-tools/,/clinical/cckm-tools/content/,/clinical/cckm-tools/content/questionnaires/,/clinical/cckm-tools/content/questionnaires/related/,

/clinical/cckm-tools/content/questionnaires/related/name-97148-en.cckm

201606168

page

100

UWHC,UWMF,

Tools,

Clinical Hub,UW Health Clinical Tool Search,UW Health Clinical Tool Search,Questionnaires,Related

The DIRE Score - Predicting Outcomes of Opioid Prescribing for Chronic Pain

The DIRE Score - Predicting Outcomes of Opioid Prescribing for Chronic Pain - Clinical Hub, UW Health Clinical Tool Search, UW Health Clinical Tool Search, Questionnaires, Related


Th me
C
M an
*F inne
Un neso

C inn

D ty of
was
cts w
op
es. T
were reviewed and scored by 6 physicians. Repeat scoring was carried out on a subset of 30 vignettes
after 2 weeks. The main outcome measures were: global impression of compliance and efficacy as
indicated in the medical record and by interviewwith the patient’s treating clinician; and final disposition,
T
pa
the
fro
tro
sup
me
use
Rec
3, 2
Fin
aut
stu
Ad
Pal
245
fair
152
© 2
doi
The Journal of Pain, Vol 7, No 9 (September), 2006: pp 671-681
Available online at www.sciencedirect.com
ie, whether or not opioids were continued or discontinued at the time of last clinical documentation.
Internal consistency of the factors making up the DIRE Score was high (Cronbach’s alpha� .80). Sensitivity
and specificity of the DIRE Score for predicting patient compliance were 94% and 87%, respectively. For
efficacy, sensitivity and specificity were 81% and 76%. For disposition, the sensitivity and specificity were
86% and 73%. Intraclass correlation was 0.94 for interrater reliability and 0.95 for intrarater reliability.
Perspective: Public controversy about the use of long-term opioids for chronic pain fuels physician
ambivalence about the prescribing process. In this initial retrospective study, validity and reliability of
the DIRE Score are demonstrated. The score correlated well with measures of patient compliance and
efficacy of long-term opioid therapy.
© 2006 by the American Pain Society
Key words: Pain, opioids, selection criteria, noncompliance.
wo decades ago, Portenoy and other pain manage-
ment thought leaders laid the groundwork for the
empirical use of opioids in chronic noncancer
in.
12,35,36,43,44,45,47
This concept was a natural outgrowth of
use of opioids for acute and cancer pain. Subsequent data
m multiple case series
17,24,27,42,50,51
and randomized con-
lled trials
3,5,7,8,13,15,16,19,20,25,26,28,29,30,34,37,39,40,41,48
have
ported the use of opioids for this problem; and established
dical bodies have developed guidelines supporting opioid
.
4,11,21
In recent years, however, the acceptance of the use of
opioids for chronic noncancer pain has been tempered
by increasing fears of misuse and safety issues. These
fears have been fueled by media reports of reckless phy-
sician prescribing, diversion, manipulation of controlled-
release opioids, and deaths from overdose.
2,31,38,46
The use of opioids in the setting of acute postoperative
pain, cancer pain, or pain in terminal illness poses few con-
troversies for physicians. The goals of therapy in these con-
ditions are relatively simple and the duration of prescribing
is limited because acute pain and terminal pain have a pre-
dictable end. On the other hand, patients with chronic non-
cancer pain have discomfort that is likely to persist
throughout their lives. Thus, the goals of therapy for the
chronic pain patient are more complex. In addition to pain
control, certain societal values such as being productive,
eived December 11, 2005; Revised February 21, 2006; Accepted March
006.
ancial support for this study was not from any commercial studies. The
hors attest that no conflicts of interest exist with the execution of this
dy.
dress reprint requests to: Miles J. Belgrade, MD, Fairview Pain and
e DIRE Score: Predicting Outco
hronic Pain
iles J. Belgrade,* Cassandra D. Schamber,

airview Pain and Palliative Care Center, University of M
iversity of Minnesota Medical School, Minneapolis, Min
hronic Pain Program, St. Mary’s-Duluth Clinic, Duluth, M
ivision of Biostatistics, School of Public Health, Universi
Abstract: The objective of this retrospective study
tool (the DIRE Score), for use by clinicians, that predi
effective analgesia and be compliant with long-term
assigned to 61 cases from the pain center’s databas
ma
mu
chr
P
liative Care Center, University of Minnesota Medical Center, Fairview;
0 Riverside Ave., Minneapolis, MN 55454. E-mail: mbelgra1@
view.org.
6-5900/$32.00
006 by the American Pain Society
:10.1016/j.jpain.2006.03.001
s of Opioid Prescribing for
d Bruce R. Lindgren

sota Medical Center and Department of Neurology,
ta.
esota.
Minnesota, Minneapolis, Minnesota.
to test the validity and reliability of a scoring
hich chronic noncancer pain patients will have
ioid maintenance treatment. DIRE scores were
hese cases were abstracted into vignettes that
intaining life roles, and achieving a fulfilling lifestyle
st also be considered when evaluating the success of
onic pain treatments.
hysicians who treat pain are eager to have the contro-
671

ver
sol
wh
an
T
tho
ph
the
wh
wi
rev
act
tio
spe
no
M
Th
T
pri
an
Ta nic
SCO
imal o
n-sp
rdan
ry syn
ith se
hy, se
the p
een t
e, tra
of a
ment
erso
res m
o sig
rugs,
to co
r che
dicat
nce,
appoi
For
Tot
Sco
Sco
*Sc
67
sy regarding opioid use in chronic noncancer pain re-
ved. They need a strategy for choosing those patients
o will be most compliant with the prescribing process
d who are most likely to benefit from opioid treatment.
he DIRE Score (Table 1) was developed by the first au-
r in response to a plea for help from a community of
ysicians in northern Minnesota. They were faced with
unexpected departure from practice of one physician
o was treating a large number of chronic pain patients
th opioids. The Minnesota Board of Medical Practice had
iewed his prescribing patterns and had taken corrective
ions. This physician’s colleagues were asking for direc-
n on how to manage complex chronic pain patients and,
cifically, how to determine who should and who should
t be on long-term opioids for chronic pain.
aterials and Methods
e DIRE Score
he DIRE Score is a clinician rating designed for the
ble 1. D.I.R.E. Score: Patient Selection for Chro
RE FACTOR
DIAGNOSIS 1 � Benign chronic condition with min
fibromyalgia, migraine headaches, no
2 � Slowly progressive condition conco
findings. Examples: failed back surge
neuropathic pain.
3 � Advanced condition concordant w
vascular disease, advanced neuropat
INTRACTABILITY 1 � Few therapies have been tried and
2 � Most customary treatments have b
process, or barriers prevent (insuranc
3 � Patient fully engaged in a spectrum
RISK (R � Total of P � C � R � S below)
Psychological: 1 � Serious personality dysfunction or
severe affective disorder, significant p
2 � Personality or mental health interfe
3 � Good communication with clinic. N
Chemical health: 1 � Active or very recent use of illicit d
2 � Chemical coper (uses medications
3 � No CD history. Not drug-focused o
Reliability: 1 � History of numerous problems: me
2 � Occasional difficulties with complia
3 � Highly reliable patient with meds,
Social support: 1 � Life in chaos. Little family support
2 � Reduction in some relationships an
3 � Supportive family/close relationship
EFFICACY SCORE 1 � Poor function or minimal pain relie
2 � Moderate benefit with function im
yet or very low doses or too short of
3 � Good improvement in pain and fu
each factor, rate the patient’s score from 1–3 based on the explanations in the
al score � D � I � R � E.
re 7–13: Not a suitable candidate for long-term opioid analgesia.
re 14–21: Good candidate for long-term opioid analgesia.*
ore cutoffs were not included on the physician rater’s score sheet.
2
mary care physician. It is used to predict efficacy of
algesia and patient compliance with long-term opioid tai
algesic treatment. The DIRE Score is meant to be ap-
ed to patients with chronic noncancer pain who either
already using opioid analgesics or who are being
nsidered for such therapy. It consists of four factors
t are rated separately and then added up to form the
E Score: Diagnosis, Intractability, Risk, and Efficacy.
e Risk factor is further broken down into four subcat-
ories that are individually rated and added together
arrive at the Risk score. The Risk subcategories are:
chological Health, Chemical Health, Reliability, and
cial Support. Each factor is rated on a numerical scale
m 1 to 3, with 1 corresponding to the least compelling
least favorable case for opioid prescribing, and 3 de-
ting the most compelling or favorable case for opioid
scribing. The total score is used to determine whether
atient is a suitable candidate for opioid maintenance
algesia. Scores may range from 7 at the lowest (patient
eives all 1s) to 21 at the highest (patient receives all
. Higher scores are expected to predict a more success-
prescribing process with respect to patient compli-
ce and efficacy of treatment. Appendix 1 provides de-
Opioid Analgesia
EXPLANATION
bjective findings or no definite medical diagnosis. Examples:
ecific back pain.
t with moderate pain, or fixed condition with moderate objective
drome, back pain with moderate degenerative changes,
vere pain with objective findings. Examples: severe ischemic
vere spinal stenosis.
atient takes a passive role in his/her pain management process.
ried but the patient is not fully engaged in the pain management
nsportation, medical illness).
ppropriate treatments but with inadequate response.
al illness interfering with care. Example: personality disorder,
nality issues.
oderately. Example: depression or anxiety disorder.
nificant personality dysfunction or mental illness.
excessive alcohol, or prescription drug abuse.
pe with stress) or history of CD in remission.
mically reliant.
ion misuse, missed appointments, rarely follows through.
but generally reliable.
ntments & treatment.
ew close relationships. Loss of most normal life roles.
roles.
olved in work or school and no social isolation.
pite moderate to high doses.
d in a number of ways (or insufficient info- hasn’t tried opioid
l).
and quality of life with stable doses over time.
-hand column.
The DIRE Score
an
pli
are
co
tha
DIR
Th
eg
to
Psy
So
fro
or
no
pre
a p
an
rec
3s)
ful
an
and f
d life
s. Inv
f des
prove
a tria
nction
right
ls of the scoring strategy for each DIRE Score factor.

Po
A
Un
wa
gra
pri
gu
lec
op
ge
wh
a s
be
ter
tio
Sco
the
tio
oid
rat
Pa
psy
ex
Sta
car
tio
int
ab
na
psy
its
lon
an
pa
T
20
the
we
be
op
(7–
tie
ad
sul
tie
vic
ho
lem
DIR
pa
cas
of
21
ha
Sco
ask
T
the
be
do
to
stu
stu
O
T
im
glo
dis
tim
co
Th
the
ob
pre
pe
bo
me
ad
pre
tie
ha
ou
the
tic
if t
G
po
tio
tiv
scr
wo
tie
rol
an
tie
the
bin
fic
On
nic
ing
the
dis
sen
T
ch
las
sur
pri
du
alt
wa
In
A
OR
pulation Studied
pproval to carry out this study was granted by the
iversity of Minnesota’s Institutional Review Board. A
iver of the HIPAA authorization requirement was
nted by this board due to the minimal risk to the
vacy of protected health information and the safe-
ards to de-identify all of the data. Patients were se-
ted from the outpatient pain management center’s
ioid prescription database using a random number
nerator. This database consists of over 500 patients
o were treated with opioids at the pain center during
creening period. When it was felt that opioids would
continued at relatively stable doses, they were en-
ed in the database to facilitate prescription prepara-
n and tracking. Their charts were reviewed and DIRE
res were assigned by one of the authors (C.D.S.) using
information from their initial pain center consulta-
ns. The decision to initiate opioids and maintain opi-
s was wholly that of the treating clinician in collabo-
ion with the patient. Every patient evaluated at the
in Center underwent both a medical evaluation and a
chological evaluation by a PhD psychologist with pain
pertise. The psychologist provided a Diagnostic and
tistical Manual diagnosis as a starting point for future
e. Each patient also filled out a 15-page health ques-
nnaire prior to their evaluation that was incorporated
o the initial assessment. In addition to information
out pain and medical conditions, the health question-
ire contained information about past psychiatric and
chological history, current mood, anxiety, health hab-
, chemical use, and so on. Every patient started on
g-term opioid therapy signed an opioid agreement
d participated in monitoring as determined by the
in clinic’s policies and procedures.
o obtain an adequate distribution of scores, the first
to 21 patients randomly chosen who fell into each of
higher DIRE Score categories of 12–16 and 17–21
re included in the study. These patients had already
en carefully screened before being placed on chronic
ioids (see above), so patients with a low DIRE Score
11) would not be found among them. To include pa-
nts with low DIRE Scores, a convenience sample of 20
ditional cases was chosen from the inpatient pain con-
tation service’s computerized database. These pa-
nts had been seen by the inpatient consultation ser-
e on multiple occasions and over multiple
spitalizations. They were highly complex and prob-
atic cases with histories of extended opioid use. The
E Scores of these inpatients were calculated from the
tients’ initial pain consultations. A total of 61 patient
es were studied, with 20 or 21 cases falling into each
the following DIRE Score categories: 7–11, 12–16, 17–
. It was felt that about 60 patients would be needed to
ve an adequate number of cases in a full range of DIRE
res. It also seemed an appropriate number of cases to
physicians to review without being too burdensome.
he charts of the 61 patients were reviewed to obtain
required information on their clinical outcomes. To
IGINAL REPORT/Belgrade et al
included in the study, patients had to be on an opioid
ea
in
se that would equal or exceed an equianalgesic dose
oral morphine 30 mg per day for the duration of the
dy. The type of opioid was not differentiated in this
dy.
utcome Measures
hree main clinical outcomes were measured: global
pression of compliance with the prescribing process,
bal impression of efficacy of opioid analgesia, and
position with respect to continuation of opioids at the
e of the last clinical contact. The global impression of
mpliance outcome was rated as good, fair, or poor.
e patients were considered to have poor compliance if
y had recurring signs of misuse including diversion,
taining prescriptions from multiple providers, forging
scriptions, repeated self-escalation of doses, re-
ated episodes of running out of prescriptions early,
rrowing medications from others, administering oral
dication forms in inappropriate ways (eg, intravenous
ministration), concurrent illegal drug use, overdose of
scription medications, or aberrant drug screens. Pa-
nts were considered to have fair compliance if they
d an occasional episode of self-escalation or running
t early, but for the most part were compliant with
ir provider’s recommendations and prescribing prac-
es. Patients were considered to have good compliance
hey had no problems with medication misuse.
lobal impression of efficacy was rated as good, fair, or
or based on two considerations: pain relief and func-
n. The charts were reviewed for the patients’ subjec-
e pain relief descriptions, 0–10 pain scores, and de-
iptions of function. Function was gauged by ability to
rk (work duties and hours), ability to perform activi-
s of daily living, and ability to engage in other life
es such as parenting, care giving, attending school,
d so on. Severe adverse side effects causing the pa-
nts to need to stop the opioids were also considered in
global impression of efficacy. These data were com-
ed with the health care provider’s perceptions of ef-
acy to determine the final score of good, fair, or poor.
e of the authors (C.D.S.) interviewed the treating cli-
ians to obtain their global assessments after review-
the charts. When there was a discrepancy between
chart documentation and the clinician’s impression, a
cussion with the clinician was held to arrive at a con-
sus; but the clinician’s final impression held.
he third outcome measured was disposition-- the
oice to continue or stop the opioids at the time of the
t clinical contact. Duration of opioid use was also mea-
ed. For patients who were transferred back to their
mary care providers on stable doses of opioids, the
ration was calculated only until the time of transfer,
hough it was assumed that the patients’ disposition
s to continue opioid use.
ter- and Intrarater Reliability
short vignette describing the pertinent history for
673
ch of the 61 patients was written from the information
the patients’ records by one of the authors (C.D.S.).

Th
tio
an
Th
tio
giv
im
gn
sic
pa
we
T
pri
cia
tie
gn
int
Re
De
F
me
at
fro
12
ran
T
2.
tal
do
pa
T
in
tin
Th
cau
Du
T
in
tie
(n
op
ha
Th
the
mo
(29
wh
dis
DI
T
ou
sho
Re
C
hig
DIR
we
Sco
did
the
Ta
Ca
Ne
Ab
Ne
He
Mu
Ch
Tem
Ota
Pai
Pos
Ch
Ch
Un
Ta
Th
Poo
Fac
Poo
Op
a
Op
O
U
P
O
A
Ta
S
1 (p
2 (f
3 (g
Sco
1 (p
2 (f
3 (g
Ou
0 (a
67
e vignettes were used to present the clinical informa-
n about the patients to five primary care physicians
d one neurologist and pain specialist (author M.J.B.).
ree of the primary care physicians were family practi-
ners and two were internists. The physicians were
en the DIRE Score sheet (Table 1) and were given min-
al preliminary instruction on how to score it. The vi-
ettes were presented in a randomized order. The phy-
ians scored the seven different DIRE factors for each
tient and the DIRE Scores were then calculated and
re compared to evaluate interrater reliability.
wo weeks after the initial scoring session, 3 of the
mary care physicians and the pain management spe-
list repeated the process of scoring a subset of 30 pa-
nts who had been randomly chosen from the 61 vi-
ettes. These scores were compared to calculate
rarater reliability over time.
sults
mographics
orty patients were female and 21 were male. The
an age by gender was 44 (SD� 12.5) years for females
the time of the initial evaluation, with a range in age
m 26 to 81. The mean age for males was 44.3 (SD �
.6) years at the time of the initial evaluation, with a
ge of 18 to 77 years of age.
he breakdown of patient diagnoses is shown in Table
There was a preponderance of chronic musculoskele-
back and neck pain (25/61 patients), followed by ab-
minal pain (9/61 patients) and neuropathic pain (8/61
tients).
he reasons for discontinuation of opioids are shown
Table 3. Of the 28 patients whose opioids were discon-
ued, only 3 were stopped because of poor efficacy.
e remaining 25 patients were taken off opioids be-
se of poor compliance.
ration of Opioid Management
ble 2. Distribution of Diagnoses Among
ses
DIAGNOSIS NO.PATIENTS
ck/back/radicular lower extremity pain 25
dominal pain 9
uropathic pain 8
adache 6
ltiregional pain 5
est wall pain 1
poral-mandibular joint pain 1
lgia 1
nful inflammatory skin condition 1
tamputation stump pain 1
ronic peri-anal fistula 1
ronic knee infection pain 1
clear diagnosis 1
4
he median duration of time from initiation of opioids
the pain clinic to the point of last contact for all pa-
1 (s
nts who continued to receive opioids was 37.5 months
� 34). The median duration of time from initiation of
ioids to the point of discontinuation in patients who
d been taken off of opioids was 12 months (n � 27).
e median duration of time from initiation of opioids to
point of last contact for all patients combined was 26
nths (n� 61). Twelve of the 41 screened clinic patients
.2%) were discontinued due to noncompliance,
ereas 15 of the 20 unscreened patients (66.7%) were
continued due to noncompliance.
RE Score Distribution
he distribution of DIRE Scores among the three main
tcomes—compliance, efficacy, and disposition—is
wn in Table 4.
liability
ronbach’s alpha for the DIRE Score was .80, denoting
h internal consistency. The Diagnosis subscore of the
E was inversely related to the total score; and if it
re removed completely, the alpha value of the DIRE
re would increase to 0.85. The Intractability subscore
not correlate very highly with the other items or with
total score. As a result, it had no impact on the alpha
ble 3. Reasons for Discontinuation of Opioid
erapy
REASON FOR DISCONTINUATION NO.PATIENTS
r efficacy 3
titious disorder 2
r reliability with clinic follow-up 1
ioid misuse (multiple providers, lost
nd forged prescriptions)
7
ioid abuse/dependence
verdose 4
nauthorized route 2
olysubstance abuse 4
pioid abuse 1
berrant urine drug screen 4
ble 4. Mean DIRE Scores
BY COMPLIANCE
CORE NMEAN SD MINIMUM MAXIMUM
oor) 31 10.7 2.29 8 17
air) 5 13.6 2.07 11 16
ood) 25 16.1 1.66 11 19
BYEFFICACY
re N Mean Std Dev Min Max
oor) 32 11.2 2.8 8 18
air) 21 15.5 2.29 10 19
ood) 8 14.8 2.05 11 17
BYDISPOSITION
tcome N Mean Std Dev Min Max
ctive) 33 14.7 2.81 8 19
The DIRE Score
top) 28 11.3 2.78 8 18

va
sho
(Ta
tor
we
ph
I
mi
sub
fix
ou
cor
Dif
car
me
an
pa
Sco
wi
Va
F
wa
ph
on
tio
the
ba
inc
T
co
Ta
no
co
ing
fac
(P
T
co
co
lat
tot
oid
era
T
to
an
Sco
ma
the
op
W
for
mu
no
DIR
acr
of
pa
en
an
O
L
for
pli
.00
inc
to
lik
les
clin
Se
T
usi
DIR
or
rec
Sco
dia
an
hig
the
87
ou
ve
an
an
we
tiv
Ta
P
MS
DW
GO
DT
GP
MJ
Ta
DI
DI
D
Dia
Intr
Ris
Ris
hea
Ris
Ris
Sup
Effi
OR
lue. The other 5 items, Risk (P�C�R�S) and Efficacy,
wed a strong positive correlation with the total score
ble 5). Although the Diagnosis and Intractability fac-
s did not correlate well with the total DIRE Score, they
re retained because of their clinical importance in the
ysician decision-making process (see comment below).
nterrater and intrarater reliability was analyzed using a
xed linear model (SAS procedure MIXED), for which the
jects were the random effect and the physicians were
ed effects. For interrater reliability, 6 physicians carried
t DIRE Scores on the 61 case vignettes. The intraclass
relation of the DIRE Score among the raters was 0.94.
ferences in the mean DIRE Score between the primary
e physician raters and the neurologist/ pain manage-
nt specialist were not significant except for one rater
d that rater was still within one DIRE Score point of the
in specialist (Table 6). Intraclass correlations of the DIRE
re tested on 30 case vignettes with 4 physician raters
th repeat scoring after a 2-week interval was 0.95.
lidity
ace validity of the elements making up the DIRE Score
s felt to be strong, based on feedback from primary
ysicians and resident physicians who were instructed
the use of this tool in lectures, seminars, and rota-
ns. The DIRE Score has been made available for use on
Veterans Administration Medical Center intranet
sed on positive feedback from physicians who have
orporated the tool into their practices.
he association between disposition, efficacy, and
mpliance and the individual DIRE factors is shown in
ble 7. Using the �
2
test for trend, all factors but Diag-
sis showed a significant relationship (P � .001) with
mpliance. Intractability was the only factor not show-
a significant association (P � .05) with efficacy. All
tors showed a significant relationship with disposition
� .05) except for Diagnosis.
he nonparametric Spearman correlation coefficient
mparing the total DIRE Score to global assessment of
mpliance was 0.76 (P � .001), indicating a high corre-
ion between the two. The correlation between the
al DIRE score and global assessment of efficacy of opi-
ble 5. Cronbach’s Alpha After Deleting Each
RE Factor (Cronbach’s Alpha for the Total
RE Score � .80)
IRE FACTOR CORRELATION WITH TOTAL ALPHA
gnosis �0.65 .85
actability 0.35 .81
k-psychological 0.69 .75
k-chemical
lth
0.66 .75
k-reliability 0.82 .71
k-social
port
0.68 .75
cacy 0.60 .77
IGINAL REPORT/Belgrade et al
analgesia was 0.58 (P � .001), which indicated mod-
te correlation between the two (Table 8A). *Ph
he nonparametric Wilcoxon Rank Sum Test was used
compare the two disposition categories (continued
d discontinued opioid prescribing) with the DIRE
re. The mean DIRE Score for the patients that re-
ined on opioids (14.7) was significantly higher than
mean DIRE Score for the group that was taken off of
ioids (11.3) (P � .001) (Table 8B).
e also carried out validity testing of the DIRE Score
2 broad diagnostic subgroups: those with mainly
sculoskeletal pain (N � 25) and those with mainly
nmusculoskeletal pain (N� 36). The correlation of the
E Score with compliance and efficacy was significant
oss both diagnostic categories (Table 8A). Comparison
the DIRE Score with disposition for musculoskeletal
in and nonmusculoskeletal pain showed similar differ-
ces between patients who were continued on opioids
d those who were discontinued (Table 8B).
dds Ratios
ogistic regression results for the DIRE Score adjusted
age and gender showed odds ratios of 2.36 for com-
ance, 1.45 for efficacy, and 0.65 for disposition (P �
1 for all values). This means that for each 1 point
rease in DIRE Score, a patient is 2.36 times more likely
be in a higher compliance category, 1.45 times more
ely to be in a higher efficacy category, and 0.65 times
s likely to be taken off of the opioid by the treating
ician.
nsitivity and Specificity
he results from the logistic regression were evaluated
ng receiver operating characteristic (ROC) curves. A
E Score of 13 was found to be the best cut-off point at
below which chronic opioid therapy would not be
ommended. Prior to the statistical analysis, the DIRE
re empirical cut-offs were set at low (7–11), interme-
te (12–16), and high (17–21). However, the outcome
alysis suggested a clear separation between low and
h DIRE Scores that warranted a single cut-off point. At
cut-off point of 13, there was 94% sensitivity and
% specificity with poor versus good/fair compliance
tcomes; 81% sensitivity and 76% specificity with poor
rsus good/fair efficacy outcomes; and 86% sensitivity
d 73% specificity with final disposition. The sensitivity
d specificity of each outcome were similar—ie, there
re similar numbers of false positives and false nega-
es for each outcome.
ble 6. Mean DIRE Score by Physician Rater
HYSICIAN RATER MEAN (SD)
12.7 (3.5)*
13.2 (3.8)
13.3 (3.6)
13.2 (3.7)
13.3 (3.7)
B (pain specialist) 13.5 (3.8)
675
ysician raters who are significantly different (P � .05) from M.J.B.

Table 7. Chi-square Comparisons of Each DIRE Factor by Compliance, Efficacy, and Disposition
TABLE 7A COMPLIANCE (1 � POOR,2� FAIR,3� GOOD)
Dia
Intr
Ris
Ris
Ris
Ris
Effi
Dia
Intr
Ris
Ris
Ris
Ris
Effi
Dia
676 The DIRE Score
ITEM SCORE 1(N� 31) 2 (N � 5) 3 (N � 25) P*
gnosis 1 5 (16.1) 2 (40) 4 (16) .242
2 26 (83.9) 2 (40) 17 (68)
3 0 (0) 1 (20) 4 (16)
actability 1 14 (45.2) 1 (20) 2 (8) �.001
2 16 (51.6) 4 (80) 17 (68)
3 1 (3.2) 0 (0) 6 (24)
k-psychological 1 23 (74.2) 2 (40) 3 (12) �.001
2 6 (19.4) 3 (60) 14 (56)
3 2 (6.5) 0 (0) 8 (32)
k-chemical health 1 12 (38.7) 0 (0) 0 (0) �.001
2 16 (51.6) 4 (80) 11 (44)
3 3 (6.7) 1 (20) 14 (56)
k-reliability 1 22 (71) 1 (20) 1 (4) �.001
2 5 (16.1) 2 (40) 2 (8)
3 4 (12.9) 2 (40) 22 (88)
k-social support 1 16 (51.6) 2 (40) 0 (0) �.001
2 14 (45.2) 1 (20) 16 (64)
3 1 (3.2) 2 (40) 9 (36)
cacy 1 21 (67.7) 0 (0) 2 (8) �.001
2 9 (29.0) 5 (100) 22 (88)
3 1 (3.2) 0 (0) 1 (4)
TABLE 7B GLOBAL IMPRESSION OF EFFICACY
ITEM SCORE 1(N� 32) 2 (N � 21) 3 (N � 8) P*
gnosis 1 8 (25) 2 (9.5) 1 (12.5) .044
2 24 (75) 15 (71.4) 6 (75)
3 0 (0) 4 (19.1) 1 (12.5)
actability 1 12 (37.5) 4 (19.1) 1 (12.5) .072
2 17 (53.1) 15 (71.4) 5 (62.5)
3 3 (9.4) 2 (9.5) 2 (25)
k-psychological 1 22 (68.8) 5 (23.8) 1 (12.5) .004
2 7 (21.9) 10 (47.6) 6 (75)
3 3 (9.4) 6 (28.6) 1 (12.5)
k-chemical health 1 10 (31.3) 2 (9.5) 0 (0) .031
2 17 (53.1) 7 (33.3) 7 (87.5)
3 5 (15.6) 12 (57.1) 1 (12.5)
k-reliability 1 21 (65.6) 2 (9.5) 1 (12.5) �.001
2 4 (12.5) 3 (14.3) 2 (25)
3 7 (21.9) 16 (76.2) 5 (62.5)
k-social support 1 17 (53.1) 1 (4.8) 0 (0) .003
2 11 (34.4) 13 (61.9) 7 (87.5)
3 4 (12.5) 7 (33.3) 1 (12.5)
cacy 1 22 (68.8) 0 (0) 1 (12.5) �.001
2 9 (28.1) 20 (95.2) 7 (87.5)
3 1 (3.1) 1 (4.8) 0 (0)
TABLE 7C DISPOSITION (0 � ACTIVE,1� STOP)
ITEM SCORE 0(N� 33) 1 (N � 28) P*
gnosis 1 6 (18.2) 5 (17.9) .255
2 22 (66.7) 23 (82.1)
3 5 (15.2) 0 (0)

Ea
T
gn
car
est
DIR
du
Di
T
pir
pa
cia
a p
Lik
Ta
Intr
Ris 9 (27
7 (51
7 (21
Ris 2 (6.
7 (51
4 (42
Ris 6 (18
4 (12
3 (69
Ris 5 (15
9 (57
9 (27
Effi 8 (24
4 (72
1 (3.
Not
*Th
Ta
Co
D
All
No
(
Mu
(
*Th
Ta
All
No
Mu
*Th
OR
se of Use
he primary care physicians who scored the case vi-
ettes were asked to estimate the time it would take to
ry out a DIRE Score on a patient they knew. The mean
imate was 102 seconds. This indicates that using the
E Score is quick and would not burden the clinician
ring a routine office visit.
ble 7. Continued
TABLE 7C
ITEM SCORE 0
actability 1
2 2
3
k-psychological 1
2 1
3
k-chemical health 1
2 1
3 1
k-reliability 1
2
3 2
k-social support 1
2 1
3
cacy 1
2 2
3
e: Counts are given with percentages in parentheses.
e P value is from the �
2
test for trend.
ble 8A. Correlation of DIRE Score With
mpliance and Efficacy
IRE SCORE FOR:
COMPLIANCE EFFICACY
CORRELATION*PCORRELATION*P
subjects (n � 61) 0.76 �.001 0.58 �.001
nmusculoskeletal
n � 36)
0.75 �.001 0.59 �.001
sculoskeletal
n � 25)
0.75 �.001 0.52 .007
e nonparametric Spearman correlation was used.
ble 8B. Comparison of DIRE Score by Dispositio
MEAN DIRE SCORE FOR: ACTIVE
subjects (n � 61) 14.7 (n � 33)
nmusculoskeletal (n � 36) 14.1 (n� 17)
sculoskeletal (n � 25) 15.4 (n� 16)
IGINAL REPORT/Belgrade et al
e P value was obtained from the nonparametric Wilcoxon rank sum test.
scussion
his investigation is a preliminary validation of an em-
ically derived assessment tool that selects or excludes
tients for chronic opioid analgesia. It fits with physi-
ns’ intuitive understanding of the elements that make
atient likely to be reliable with higher risk therapies.
e other assessment tools, it helps objectify the deci-
n-making process so as to make it more dispassionate,
en-handed, and consistent.
e have shown that primary care physicians were con-
rdant in their scoring of the chronic pain cases. We
ve also shown that the scores of primary care physi-
ns were also concordant with those of a pain manage-
nt specialist. The DIRE Score was strongly correlated
th compliance with opioid treatment. It correlated
derately with efficacy of opioid treatment. That 1 in-
ument correlates with both compliance and efficacy
y reflect the importance of good treatment compli-
ce in order to achieve efficacy.
he retrospective nature of this study is 1 important
itation. Scoring patients from abstracted case vi-
DISPOSITION (0 � ACTIVE,1� STOP)
33) 1 (N � 28) P*
.2) 12 (42.9) .007
.7) 15 (53.6)
.2) 1 (3.6)
.3) 19 (67.9) .007
.5) 6 (21.4)
.2) 3 (10.7)
1) 10 (35.7) .001
.5) 14 (50)
.4) 4 (14.3)
.2) 18 (64.3) �.001
.1) 5 (17.9)
.7) 5 (17.9)
.2) 13 (46.4) .008
.6) 12 (42.9)
.3) 3 (10.7)
.2) 15 (53.6) .04
.7) 12 (42.9)
0) 1 (3.6)
DISPOSITION
P*DISCONTINUE DIFFERENCE
11.3 (n� 28) 3.4 �.001
11.2 (n� 19) 2.9 .006
11.6 (n� 9) 3.8 .019
677
sio
ev
W
co
ha
cia
me
wi
mo
str
ma
an
T
lim
n
(N�
5 (15
2 (66
6 (18

gn
ph
fro
fer
ula
the
yse
groups. The population studied consisted of a prepon-
derance of patients with musculoskeletal neck back and
limb pain (41%). We were able to show that DIRE Scores
in
an
ou
we
I
the
de
Th
lat
rel
mo
cri
wi
oid
tha
Dia
oth
pe
R
ha
im
fec
the
me
tio
sel
to
dis
no
da
ge
op
the
mo
ing
pre
rep
use
too
pa
de
pa
pa
ho
an
ment and Documentation Tool (PADT) which is a struc-
tured progress note that highlights key outcomes when
following patients who are on chronic opioid analgesic
the
he
pre
tio
rep
too
me
dru
ille
mu
Sim
the
ad
de
Co
I
sho
too
for
sho
wi
ov
nic
pa
DIR
his
too
Ac
T
the
Gr
Tra
rev
Sco
ass
Re
1.
De
ins
in
44
2.
pa
3.
M,
de
tre
67
the broad diagnostic categories of musculoskeletal
d nonmusculoskeletal pain both correlated well with
tcomes of efficacy and compliance, and compared
ll with final disposition of opioid prescribing.
ntractability was not correlated with efficacy of opioid
rapy. This is understandable, because efficacy is by
finition difficult to achieve in an intractable condition.
e Diagnosis subscore of the DIRE was inversely corre-
ed with the total DIRE Score and did not strongly cor-
ate with outcome although it correlated with efficacy
derately well. The reason for retaining Diagnosis as a
terion for selecting patients for opioid therapy rests
th the importance of not prescribing long-term opi-
s in patients without a diagnosis or with a condition
t is not associated with moderate to severe pain. The
gnosis factor of the DIRE Score might correlate with
er outcomes that weren’t measured, such as results of
er review or regulatory scrutiny.
eviews of opioid therapy for chronic noncancer pain
ve mostly favored such therapy, showing consistent
provement in pain and acceptable adverse ef-
ts.
6,9,22
The randomized controlled trials upon which
se reviews draw suffer from a short duration of treat-
nt and a tendency to not show improvement in func-
nal status. In our sample, only 8 of the 41 randomly
ected patients demonstrated good efficacy according
their treating clinicians; and 28 of the patients were
continued from opioid therapy due to poor efficacy,
ncompliance, or frank abuse of the drugs. While these
ta are not our main outcome measures, they do sug-
st that the number of patients for whom long-term
ioids will be very successful is small. This underscores
need for careful selection of patients and ongoing
nitoring of treatment goals. Another scored screen-
instrument for opioid therapy was presented in a
liminary report by Adams et al.
1
This is a 26-item self-
ort questionnaire to detect likelihood of opioid mis-
. This report did not measure predictive validity of the
l. A recent study by Webster and Webster presents a
ferences
Adams LL, Gatchel R, Robinson RC, Polatin P, Gajraj N,
schner M, Noe C: Development of a self-report screening
trument for assessing potential opioid medication misuse
chronic pain patients. J Pain Symptom Manage 2004;27:
0-459 20
rapy.
32
The PAD-T is a documentation tool that can
lp inform physicians by tracking outcomes. It does not
dict efficacy or compliance and is not a patient selec-
n tool. The DIRE Score is intended to complement, not
lace, other established screening and monitoring
ls
1,14,32
and to complement careful patient manage-
nt using opioid prescribing agreements and urine
g screens.
10,18,23,33
If a patient is found to be using
gal substances on a urine drug screen, then action
st be taken no matter what the DIRE Score may be.
ilarly, if there is a violation of an opioid agreement in
course of treatment, the physician is obligated to
dress and manage that aberrant behavior indepen-
nt of a DIRE Score.
nclusions
n this retrospective investigation, the DIRE Score was
wn to be valid and reliable. It is a rapid assessment
l for selecting patients with chronic noncancer pain
long-term opioid analgesic therapy. The DIRE Score
wed very good correlation with patient compliance
th opioid analgesia; and moderate correlation with
erall efficacy of opioid therapy (from the treating cli-
ian’s perspective) in patients with chronic noncancer
in. A prospective randomized controlled study of the
E Score applied directly to patients rather than case
tories is needed to further establish the role of this
l in the clinical setting.
knowledgments
he authors wish to thank the following physicians for
ir participation in this study: Garrick B. Olsen, MD,
ace C. Peterson, MD, Mary Beth Schleevogt, MD, Doris
n-Stoebe, MD, and Donald Weinshenker, MD. They
iewed the case vignettes and assigned the DIRE
res. We also wish to thank Colleen Curran, BA, for her
istance with the statistical analysis of the data.
Albert T, Adams D: Oxycontin crackdown raises physician,
tient concerns. American Medical News. June 25, 2001, P 1
Allan L, Hays H, Jensen N-H, Le Polain de Waroux B, Bolt
Donald R, Kalso E: Randomized crossover trial of trans-
rmal fentanyl and sustained-release oral morphine for
ating chronic non-cancer pain. Pain Practice 1(4):374 Dec
s of the DIRE Score for individual pain diagnostic scoring patients. Passik et al proposed the Pain Assess-
ettes is not the same as scoring a patient known to the
ysician at the time of evaluation. Cases were drawn
m a population of patients enrolled in a tertiary re-
ral center and may not be representative of the pop-
tion seen in primary care clinics. Another limitation is
size of the study, which did not permit separate anal-
8
tient self-report questionnaire, the Opioid Risk Tool,
signed to predict aberrant behaviors in opioid-treated
tients.
49
They prospectively followed a cohort of 185
in patients over 12 months and found a high likeli-
od of aberrant behaviors among high-scoring patients
d a low likelihood of aberrant behaviors among low-
The DIRE Score
01

4.
ica
chr
ww
Ma
5.
Z,
chr
clin
6.
for
7.
Lac
tro
plu
tor
ceb
8.
HW
an
mo
do
op
27
9.
saf
pa
10
10
uri
19
11
Inc
tre
Ma
12
an
13
ox
60
14
tio
of
15
Sw
con
ran
chr
16
Th
gio
rel
stim
An
17
AL
no
18
cin
19
op
Pa
20
Ka
ran
21
niz
Gu
cag
22
chr
saf
23
Bra
ico
ap
24
tra
AJ
25
An
bje
lon
tis
stu
26
effi
de
eta
hip
27
tie
52
28
de
in
of
20
29
sev
30
ske
can
31
ab
ag
32
Kle
do
op
33
rev
og
ma
34
ny
con
of
76
35
pa
OR
The American Academy of Pain Medicine and The Amer-
n Pain Society: The use of opioids for the treatment of
onic pain: A Consensus Statement. Available at: http://
w.painmed.org/ or http://www.ampainsoc.org/ Accessed
rch 9, 2005
Arkinstall W, Sandler A, Goughnour B, Babul N, Harsanyi
Darke AC: Efficacy of controlled-release codeine in
onic non-cancer pain: a randomized, placebo-controlled
ical trial. Pain 1995;62:169-78
Ballantyne JC, Mao J: Medical progress: opioid therapy
chronic pain. N Engl J Med 2003;349:1943-1953
Caldwell JR, Hale ME, Boyd RE, Hague JM, Iwan T, Shi M,
outure PG: Treatment of osteoarthritis pain with con-
lled release oxycodone or fixed combination oxycodone
s acetaminophen added to nonsteroidal antiinflamma-
y drugs: a double blind, randomized, multicenter, pla-
o controlled trial. J Rheumatol 1999;26:862-869
Caldwell JR, Rapoport RJ, Davis JC, Offenberg HL, Marker
, Roth SH, Yuan W, Elliot L, Babul N, Lynch PM: Efficacy
d safety of a once-daily morphine formulation in chronic,
derate-to-severe osteoarthritis pain: results from a ran-
mized, placebo-controlled, double-blind trial and an
en-label extension trial. J Pain Symptom Manage 2002;23:
8-291
Chou R, Clark E, Helfand M: Comparative efficacy and
ety of long-acting oral opioids for chronic non-cancer
in: a systematic review. J Pain Symptom Manage 2003;26:
26-1048
. Eskridge KD, Guthrie SK: Clinical issues associated with
ne testing of substances of abuse. Pharmacotherapy
97;17:497-510
. Federation of State Medical Boards of the United States,
: Model policy for the use of controlled substances for the
atment of pain. Available at: http://fsmb.org/ Accessed
rch 9, 2005
. France RD, Urban BJ, Keefe FJ: Long-term use of narcotic
algesics in chronic pain. Soc Sci Med 1984:19:1379-1382
. Gimbel JS, Richards P, Portenoy RK: Controlled-release
ycodone for pain in diabetic neuropathy. Neurology 2003;
:927-934
. Gourlay DL, Heit HA, Almahrezi, A: Universal precau-
ns in pain medicine: a rational approach to the treatment
chronic pain. Pain Med 2005;6:107-112
. Hale ME, Fleischmann R, Salzman R, Wild J, Iwan T,
anton RE, Kaiko RF, Lacouture PG: Efficacy and safety of
trolled-release versus immediate-release oxycodone:
domized, double blind evaluation in patients with
onic back pain. Clin J Pain 1999;15:179-183
. Harke H, Gretenkort P, Ladleif HU, Rahman S, Harke O:
e response of neuropathic pain and pain in complex re-
nal pain syndrome I to carbamazepine and sustained-
ease morphine in paitents pretreated with spinal cord
ulation: a double-blinded randomized study. Anesth
alg 2001;92:488-495
. Haythornthwaite JA, Menefee LA, Quatrano-Piacentini
, Pappagallo M: Outcome of chronic opioid therapy for
n-cancer pain. J Pain Symptom Manage 1998;15:185-194
. Heit HA, Gourlay DL: Urine drug testing in pain medi-
e. J Pain Symptom Manage 2004;27:260-267
. Huse E, Larbig W, Flor H, Birbaumer N: The effect of
IGINAL REPORT/Belgrade et al
ioids on phantom limb pain and cortical reorganization.
in 2001;90:47-55
36
in
. Jamison RN, Raymond SA, Slawsby EA, Nedeljkovic SS,
tz NP: Opioid therapy for chronic noncancer back pain: a
domized prospective study. Spine 1998;23:2591-2600
. Joint Commission on Accreditation of Healthcare Orga-
ations: Approaches to Pain Management: An Essential
ide for Clinical Leaders, Joint Commission Resources, Chi-
o, 2003
. Kalso E, Edwards JE, Moore RA, McQuay HJ: Opioids in
onic non-cancer pain: systematic review of efficacy and
ety. Pain 2004;112:372-380
. Katz NP, Sherburne S, Beach M, Rose RJ, Vielguth J,
dley J, Fanciullo GJ: Behavioral monitoring and urine tox-
logy testing in patients receiving long-term opioid ther-
y. Anesth Analg 2003;97:1097-1102
. Kell MJ, Musselman DL: Methadone prophylaxis of in-
ctable headaches: pain control and serum opioid levels.
PM 1993;3:7-14
. Kjaersgaard-Andersen P, Nafei A, Skov O, Madsen F,
dersen HM, Kroner K, Gjoderum O, Pedersen L, Brane-
rg PE: Codeine plus paracetamol versus paracetamol in
ger-term treatment of chronic pain due to osteoarthri-
of the hip: a randomized, double-blind, multi-centre
dy. Pain 1990;43:309-318
. Lloyd RS, Costello F, Eves MJ, James IG, Miller AJ: The
cacy and tolerability of controlled-release dihydroco-
ine tablets and combination dextropropoxyphene/parac-
mol tablets in patients with severe osteoarthritis of the
s. Curr Med Res Opin 1992;13:37-48
. Mahowald ML, Singh JA, Majeski P: Opioid use by pa-
nts in an orthopedics spine clinic. Arthritis Rheum 2005;
:312-321
. Maier C, Hildebrandt J, Klinger R, Henrich-Eberl C, Lin-
na G: Morphine responsiveness, efficacy and tolerability
patients with chronic non-tumor associated pain- results
a double-blind placebo-controlled trial (MONTAS). Pain
02;97:223-233
. Moran C: MST continuous tablets and pain control in
ere rheumatoid arthritis. Br J Clin Res 1991;2:1-12
. Moulin DE, Iezzi A, Amireh R, Sharpe WK, Boyd D, Mer-
y H: Randomised trial of oral morphine for chronic non-
cer pain. Lancet 1996;347:143-147
. Passik SD: Responding rationally to recent reports of
use/diversion of oxycontin (letter). J Pain Symptom Man-
e 2001;21:359-360
. Passik SD, Kirsh KL, Whitcomb L, Portenoy RK, Katz NP,
inman L, Dod, SL, Schein JR: A new tool to assess and
cument pain outcomes in chronic pain patients receiving
ioid therapy. Clin Ther 2004;26:552-561
. Passik SD, Schreiber J, Kirsh KL, Portenoy RK: A chart
iew of the ordering and documentation of urine toxicol-
y screens in a cancer center: do they influence patient
nagement? J Pain Symptom Manage 2000;19:40-44
. Peloso PM, Bellamy N, Bensen W, Thompson GT, Harsa-
i Z, Babul N, Darke AC: Double blind randomized placebo
trol trial of controlled release codeine in the treatment
osteoarthritis of the hip or knee. J Rheumatol 2000;19:
4-771
. Portenoy RK: Chronic opioid therapy for nonmalignant
in: from models to practice. APS J 1992; 1:285-288
679
. Portenoy RK, Foley KM: Chronic use of opioid analgesics
non-cancer pain: report of 38 cases. Pain 1986;25:171-186

37
Tra
an
pla
38
ua
39
Ra
con
lat
tio
40
K,
cen
41
Go
ox
for
Sym
42
Tra
Sym
43
int
Co
mo
44
op
ag
45
chr
19
46
Ma
47
AA
the
19
48
pa
Ne
49
in o
oid
50
ox
thr
51
in p
Ma
Ap
Factors
T
mine the extent to which the patient’s diagnosis is suffi-
ciently compelling or advanced to warrant an aggressive
pharmacological approach. Benign conditions that
pre
ch
pa
dit
co
sco
dro
or
rop
ad
ad
T
to
ha
ho
of
or
Int
the
Int
ap
co
me
pa
wh
T
tie
Chemical Health, Reliability, and Social Support.
The Psychological Health category evaluates a pa-
tient’s psychological and psychiatric health. A score of 1
is g
or
car
sch
ch
giv
ali
to
tio
T
lat
Ris
cal
to
no
ten
co
tie
T
ex
me
wh
pa
low
68
sent with few objective findings like myofascial pain,
ronic tension-type headache, and nonspecific low back
in are assigned a score of 1. Moderately advanced con-
ions with physical or radiographic findings that are
ncordant with moderate to severe pain are given a
re of 2. Examples include failed lumbar surgery syn-
me, moderately advanced degenerative disc disease
spinal stenosis, and mild to moderate peripheral neu-
athies. A Diagnosis score of 3 would be assigned to
vanced conditions such as ischemic vascular disease,
vanced neuropathies, or severe spinal conditions.
he Intractability factor is an assessment of the extent
which all appropriate treatments have been ex-
usted or are inaccessible and also how engaged or
w passive a patient is in the treatment process. A score
1 is given to patients having the least intractable pain
the lowest motivation in the treatment process. The
ractability score is 2 if a patient has tried multiple
rapies but is not actively utilizing those therapies. The
ractability score is also 2 if there are many barriers to
plying appropriate therapies such as insurance non-
verage or geographic isolation from pain manage-
nt services. An Intractability score of 3 is assigned to
tients who have tried most available treatments and
o continue to actively utilize the helpful measures. me
iven to patients with serious personality dysfunction
mental illness that greatly interferes with medical
e. A patient with severe personality dysfunction or
izophrenia with active paranoia would get a Risk-Psy-
ological score of 1. A Risk-Psychological score of 2 is
en to patients with depression, anxiety, or a person-
ty that mildly interferes with care. A score of 3 is given
patients with no mental illness or personality dysfunc-
n who communicate well with clinic staff.
he Chemical Health category evaluates a patient’s re-
ionship with substances that have abuse potential. A
k-Chemical Health score of 1 represents active chemi-
dependency or substance abuse. A score of 2 is given
patients who have a chemical dependency history
w in remission, or who are chemical copers—ie, they
d to seek medicine or other substances as a way of
ping with all their problems without overt abuse. Pa-
nts with no chemical dependency score a 3.
he Reliability category rates patients based on past
periences with the prescribing clinician or well-docu-
nted information from other health care providers
o have treated the patient. It is a global assessment of
st compliance with medications, appointments, fol-
ing directions, and following through with treat-
he Diagnosis factor requires the clinician to deter-
is subdivided into 4 categories: Psychological Health,
. Raja SN, Haythornthwaite JA, Pappagallo M, Clark MR,
vison TG, Sabeen S, Royall RM, Max MB: Opioids versus
tidepressants in postherpetic neuralgia: a randomized,
cebo-controlled trial. Neurology 2002;59:1015-1021
. Roche T: The potent perils of a miracle drug. Time Jan-
ry 8, 2001, p. 47
. Roth SH, Fleischmann RM, Burch FX, Dietz F, Bockow B,
poport RJ, Rutstein J, Lacouture PG: Around-the-clock,
trolled-release oxycodone therapy for osteoarthritis-re-
ed pain: placebo-controlled trial and long-term evalua-
n. Arch Intern Med 2000;160:853-860
. Rowbotham MC, Twilling L, Davies PS, Reisner L, Taylor
Mohr D: Oral opioid therapy for chronic peripheral and
tral neuropathic pain. N Engl J Med 2003;348:1223-1232
. Salzman RT, Roberts MS, Wild J, Fabian C, Reder RF,
ldenheim PD: Can a controlled release oral dose form of
ycodone be used as readily as an immediate-release form
the purpose of titrating to stable pain control? J Pain
ptom Manage 1999;18:271-279
. Simpson RK Jr, Edmondson EA, Constant CF, Collier C:
nsdermal fentanyl for chronic low back pain. J Pain
ptom Manage 1997;14:218-224
. Taub A: Opioid analgesics in the treatment of chronic
ractable pain of non-neoplastic origin. In: Kitahata LM,
llins D, eds. Narcotic analgesics in anesthesiology. Balti-
re: Williams and Wilkins, 1982:199-208
pendix 1. Explanation of DIRE Score
0
. Tennant FS, Robinson D, Sagherian A, Seecof R: Chronic
ioid treatment of intractable non-cancer pain. Pain Man-
e 1988;Jan/Feb:18-36
. Tennant FS, Uelman GF: Narcotic maintenance for
onic pain: medical and legal guidelines. Postgrad Med
83;73:81-94
. Tough P: The alchemy of oxycontin. The New York Times
gazine, July 29, 2001, p32
. Urban BJ, France RD, Steinberger DL, Scott DL, Maltbie
: Long-term use of narcotic-antidepressant medication in
management of phantom limb pain. Pain 1986;24:191-
7
. Watson CPN, Babul N: Efficacy of oxycodone in neuro-
thic pain: a randomized trial in postherpetic neuralgia.
urology 1998;50:1837-1841
. Webster LR, Webster RM: Predicting aberrant behaviors
pioid-treated patients: preliminary validation of the opi-
risk tool. Pain Medicine 2005;6(6):432-442
. Ytterberg SR, Mahowald ML, Woods SR: Codeine and
ycodone use in patients with rheumatic disease pain. Ar-
itis Rheum 1998; 41(9):1603-1612
. Zenz M, Strumpf M, Tryba M: Long-term opioid therapy
atients with chronic nonmalignant pain. J Pain Symptom
nage 1992;7:69-77
he Risk factor is designed to predict how likely a pa-
nt is to comply with the opioid prescription process. It
The DIRE Score
nt recommendations. A Risk-Reliability score of 1 is

given to patients with a track record of poor reliability
and a score of 2 is given to patients with occasional prob-
lems, but who are generally reliable. A score of 3 is given
to patients who have proven themselves highly reliable.
The Social Support score assesses both the patient’s
social support system and the patient’s ability to function
in life roles such as work, school and parenting. A score
of 1 is given to patients with a life in chaos and with loss
of
ad
sco
fun
tie
wo
T
of
ba
po
in
effi
op
im
pro
an
tim
ua
on
len
Ap
P
ch
leg
dia
am
ten
nu
de
low
eq
tha
me
rop
Sh
lea
cau
an
S
ren
say
ex
ity issues are limiting her ability to deal with her pain.
Although she has anger toward the medical system, she
is reliable with appointments and medication use. She
has no chemical dependency history.
She is married and has two children. Her husband was
with her at her first appointment and is supportive. Her
mother lives next door to her and is supportive as well.
She has had to stop working because of pain.
D
I
R
3,
E
D
E
giv
co
lik
are
co
pa
pro
sco
tio
reg
she
pre
to
Ris
sub
em
int
dis
1);
en
wa
be
ch
she
Ris
tem
rat
mo
on
ag
dis
W
ab
DIR
alg
red
681ORIGINAL REPORT/Belgrade et al
most normal life roles. They may have no permanent
dress, be socially isolated, and have lost their job. A
re of 2 is given if some but not all social support and
ction is maintained, and a score of 3 is given to pa-
nts who maintain close relationships and continue to
rk and contribute to society.
he Efficacy factor is the clinician’s global assessment
pain improvement attributed to opioid analgesics. It is
sed on a combination of the patient’s subjective re-
rts of pain improvement in addition to improvement
function with opioids. A score of 1 represents poor
cacy and/or low function despite moderate doses of
ioid analgesics. A score of 2 is given for moderate
provement in pain and function. A score of 3 is for
ven efficacy with good improvement in both pain
d function while maintaining stable opioid doses over
e. A score of 2 is given if the efficacy cannot be eval-
ted because the patient has not used opioids or has
ly been on low doses of opioids (less than the equiva-
t of morphine 30mg oral per day).
pendix 2. Case Vignette Example
atient #27 is a 37-year-old female with a history of
ronic low back pain radiating down the backs of both
s. She has had multiple lumbar surgeries and has a
gnosis of failed back surgery syndrome. Physical ex-
ination is normal except for loss of a left Achilles
don reflex and right L5 sensory loss. She has tried a
mber of medications, including NSAIDs, tricyclic anti-
pressants, carbamazepine, tramadol, alprazolam, and
-dose opioids (ie, less than 30 mg oral morphine
uivalent per day). She has not used higher doses other
n for acute postoperative pain. Nonmedication treat-
nts have included epidural steroids, acupuncture, chi-
ractic, physical therapy, homeopathy, and hypnosis.
e is not actively utilizing the techniques that she
rned. She cannot participate in a pain program be-
se her level of disability is too great and her endur-
ce is too low.
he feels victimized by the medical system and is cur-
tly in litigation against her back surgeon because she
s that the hardware used in her lumbar fusion was
perimental. Her primary physician feels that personal-
iagnosis � 2
ntractability � 2
isk � 2�3�3�2 � 10 (Psychological � 2, Chemical �
Reliability � 3, Social � 2)
fficacy � 2
IRE Score � 16
xplanation of DIRE Score choices: This patient was
en a Diagnosis score of 2 because she has a stable
ndition with multiple surgical interventions that have
ely resulted in altered mechanics of the spine. There
also some chronic neurological changes that may
ntribute to or suggest the presence of neuropathic
in. So her condition is more than minimal, but not
gressive or severely impaired. Intractability is given a
re of 2 because she has tried quite a few interven-
ns: surgeries, medicines, alternative treatments, and
ional anesthetic procedures with little success. Yet
is not actively engaged in a self-care program and
sents barriers to initiating a rehabilitation effort due
severe deconditioning and pain-related disability. The
k score adds up to 10 based on the following risk
categories: Risk-Psychological is 2 because of the
otional barriers of anger against the surgeon that
erferes with successful adaptation. She is not severely
abled psychologically (ie, does not warrant a score of
nor is she free of any psychological factors that influ-
ce her care and response to treatment (ie, does not
rrant a score of 3). Risk-Chemical Health score is 3
cause there are no medication issues or history of
emical dependency. Risk-Reliability score is also 3 as
has been reliable with appointments and treatment.
k-Social score is 2 given her intact family support sys-
, but loss of life roles such as work. Efficacy score is
ed a 2 because this patient has not yet had a trial of
derate doses of opioid therapy. If she had been tried
long-acting opioids or higher doses of short-acting
ents, she would receive a score of 1 due to extreme
ability.
ith the DIRE Score totaling 16, this patient is well
ove the cut-off in favor of prescribing opioids. The
E Score predicts that she will comply with opioid an-
esic therapy and that it could be effective for her with
uced pain and improved function.