Arthritis Care & Research

Vol. 62, No. 5, May 2010, pp 600 – 610
DOI 10.1002/acr.20140
© 2010, American College of Rheumatology
ORIGINAL ARTICLE

The American College of Rheumatology
Preliminary Diagnostic Criteria for Fibromyalgia
and Measurement of Symptom Severity
FREDERICK WOLFE,1 DANIEL J. CLAUW,2 MARY-ANN FITZCHARLES,3 DON L. GOLDENBERG,4
ROBERT S. KATZ,5 PHILIP MEASE,6 ANTHONY S. RUSSELL,7 I. JON RUSSELL,8 JOHN B. WINFIELD,9
10
AND MUHAMMAD B. YUNUS

This criteria set has been approved by the American College of Rheumatology (ACR) Board of Directors as Provisional.
This signifies that the criteria set has been quantitatively validated using patient data, but it has not undergone validation
based on an external data set. All ACR-approved criteria sets are expected to undergo intermittent updates.
As disclosed in the manuscript, these criteria were developed with support from the study sponsor, Lilly Research Labora-
tories. The study sponsor placed no restrictions, offered no input or guidance on the conduct of the study, did not partici-
pate in the design of the study, see the results of the study, or review the manuscript or submitted abstracts prior to the
submission of the paper. The recipient of the grant was Arthritis Research Center Foundation, Inc. The authors received
no compensation. The ACR found the criteria to be methodologically rigorous and clinically meaningful.
ACR is an independent professional, medical and scientific society which does not guarantee, warrant or endorse any
commercial product or service. The ACR received no compensation for its approval of these criteria.

Objective. To develop simple, practical criteria for clinical diagnosis of fibromyalgia that are suitable for use in primary
and specialty care and that do not require a tender point examination, and to provide a severity scale for characteristic
fibromyalgia symptoms.
Methods. We performed a multicenter study of 829 previously diagnosed fibromyalgia patients and controls using
physician physical and interview examinations, including a widespread pain index (WPI), a measure of the number of
painful body regions. Random forest and recursive partitioning analyses were used to guide the development of a case
definition of fibromyalgia, to develop criteria, and to construct a symptom severity (SS) scale.
Results. Approximately 25% of fibromyalgia patients did not satisfy the American College of Rheumatology (ACR) 1990
classification criteria at the time of the study. The most important diagnostic variables were WPI and categorical scales
for cognitive symptoms, unrefreshed sleep, fatigue, and number of somatic symptoms. The categorical scales were
summed to create an SS scale. We combined the SS scale and the WPI to recommend a new case definition of fibromyalgia:
(WPI >7 AND SS >5) OR (WPI 3– 6 AND SS >9).
Conclusion. This simple clinical case definition of fibromyalgia correctly classifies 88.1% of cases classified by the ACR
classification criteria, and does not require a physical or tender point examination. The SS scale enables assessment of
fibromyalgia symptom severity in persons with current or previous fibromyalgia, and in those to whom the criteria have
not been applied. It will be especially useful in the longitudinal evaluation of patients with marked symptom variability.

INTRODUCTION of widespread pain for diagnosis. Widespread pain was
The introduction of the American College of Rheumatol- defined as axial pain, left- and right-sided pain, and upper
ogy (ACR) fibromyalgia classification criteria 20 years ago and lower segment pain.
began an era of increased recognition of the syndrome (1). Over time, a series of objections to the ACR classification
The criteria required tenderness on pressure (tender criteria developed, some practical and some philosophi-
points) in at least 11 of 18 specified sites and the presence
Boston, Massachusetts; 5Robert S. Katz, MD: Rush Univer-
Supported by Lilly Research Laboratories. sity Medical Center, Chicago, Illinois; 6Philip Mease, MD:
1
Frederick Wolfe, MD: National Data Bank for Rheumatic Seattle Rheumatology Associates and Swedish Medical Cen-
Diseases and University of Kansas School of Medicine, ter, Seattle, Washington; 7Anthony S. Russell, MD: Univer-
Wichita; 2Daniel J. Clauw, MD: University of Michigan Med- sity of Alberta, Edmonton, Alberta, Canada; 8I. Jon Russell,
ical School, Ann Arbor; 3Mary-Ann Fitzcharles, MB, ChB: MD, PhD: University of Texas Health Sciences Center, San
Montreal General Hospital and McGill University, Mon- Antonio; 9John B. Winfield, MD: University of North Caro-
treal, Quebec, Canada; 4Don L. Goldenberg, MD: Newton- lina, Chapel Hill; 10Muhammad B. Yunus, MD: The Univer-
Wellesley Hospital, Tufts University School of Medicine, sity of Illinois College of Medicine, Peoria.

600

speaking fees. pain index [WPI]) and characteristic fibromyalgia symp- tion. and/or honoraria (less than $10. including an index of pain extent (widespread toms. experienced with fibromyalgia patients and the fibromyal- Zeneca.000 each) cian. the importance of symptoms that had not been To accomplish these objectives. we collected an extensive set of became increasingly known and appreciated as key fibro. fibromyalgia diagnosis and helpful in following patients longitudinally. a severity scale. DLG. diagnostic criteria. First. primary care use. gia tender point examination. 1035 North Em. previous diagnosis of fibromyalgia. In addi. 2009. there was little variation in symptoms among fibromyalgia based on patient self-report. Thus. Fitzcharles has received consultant fees. In the second stage. models were developed for points obscured important considerations and erroneously the surrogate classification criteria. cognitive symp. 9 used only the ACR classification criteria diagnosis. UCB. We recruited study phy- myalgia and a control group of rheumatic disease patients sicians by selecting randomly from a list of 113 rheuma- with noninflammatory disorders to address the issues of tologists who were members of the ACR and who indi- fibromyalgia diagnosis and symptom severity. These two considerations suggested the need for rate report. not an assistant. PM. We also included 5 physicians with diagnostic criteria for fibromyalgia. They had to be Forest.000 severity. We required the fibromyalgia study patients to have a Dr. Forest. speak. Patients who improved or whose symp. complete physician assessment from Forest. and Lilly. MAF. Clauw has received consultant fees. and Boehringer Ingelheim. physicians contributing valid patients to phase 1 of the Submitted for publication August 24. Dr. Second. the ACR marily report phase 1 data except for comparisons of clas- classification criteria set such a high bar for diagnosis that sification rates and severity variables.Diagnostic Criteria for Fibromyalgia 601 cal. It was not a requirement of diagnosis poria. these criteria are not known fibromyalgia expertise selected from the authors (FW. reduced set of variables in a physician questionnaire for- mous criteria (9).000 each) from Pfizer. Address correspondence to Frederick Wolfe. we employed a 2-stage considered by the ACR Multicenter Criteria Committee design. Dr. see if a shortened. DJC. and has given must have been diagnosed with fibromyalgia by the same expert testimony for medicolegal and various law firms and examining rheumatologist prior to the date of study assess- received a fee. and (more than $10. They were enrolled as and/or honoraria (less than $10. tory controls within a 4-month period. and has given ex. KS 67214. but to count was rarely performed in primary care where most represent an alternative method of diagnosis. controls. and 15 diagnosed some . We required that the physi- speaking fees. grate severity scale– based symptoms in these new clinical was performed incorrectly (2). Pfizer. 2) to inte- fibromyalgia diagnoses occurred. or other characteristics. we conducted a multicenter study of patients with a diagnosis of fibro. Participating physicians had to be certain that they Dr. Jon Russell has received consultant fees. study. MD. been diagnosed on clinical grounds or by the ACR classi- tional Data Bank for Rheumatic Diseases. The objec. fatigue. Many physicians did not criteria. requirements and methods.000 each) from Pfizer. speaking fees (less than $10. Dr. The purpose of the phase 2 study was to myalgia diagnosis. know how to examine for tender points and some simply these criteria should be suitable for use in primary care refused to do so (3).000 each) from Johnson & they appeared in the clinic for usual care (not by being Johnson and UCB. built on the characteristic features of fibromyalgia. and the extent of somatic symptoms (4 –7). From the resultant data. and Abbott Canada. a number of fibromyalgia experts believed that tender toms.000 each) from Roche and BMS. In addition. I. UCB. and when performed. Goldenberg has received consultant fees. mat that could be reduced to a single page suitable for There was still another important problem with fibro. would see 10 fibromyalgia patients and 10 noninflamma- and/or honoraria (less than $10. will be the subject of a sepa- patients. pleted a short instructional questionnaire on the Internet pert testimony for various medicolegal and law firms and and satisfactorily completed a brief examination on study received a fee. Dr. ASK. Yunus has received speaking fees (more ment. Lilly. Suite 288. practical physician questionnaire toms and tender points decreased could fail to satisfy the would work as well in categorizing fibromyalgia as the ACR 1990 classification definition. Russell has forms. and Jazz. Anthony S. accepted in revised form February 10. and/or honoraria (less than $10. patient and physician variables from 514 patients and myalgia features: for example. JBW). Schering-Plough. Patients diagnosed with fibromyalgia could have than $10. ASR. it became increasingly clear that the tender point meant to replace the ACR classification criteria. 6 used only clinical diagnosis. Study subjects and physicians. Consequently. Pfizer. fication criteria (10). cated an interest in participating in the study after an tives of this study were: 1) to identify non–tender point e-mail solicitation.000 each) from Am- gen/Wyeth. Survey criteria. 2010.org. Winfield has received consultant recalled) and without consideration of current diagnosis. Dr. SUBJECTS AND METHODS With all of these considerations in mind. In the first stage. and 3) to in practice has often been a symptom-based diagnosis. Pierre Fabre. Of the 30 arthritis-research. IJR. and linked the disorder to peripheral muscle abnormality (8). develop a fibromyalgia symptom severity (SS) scale. It was not clear how to longer. and Jazz.000 each) from Pfizer. a broad-based severity scale that could differentiate among patients according to the level of fibromyalgia symptoms. Fibromyalgia subjects each) from Lilly. more detailed phase 1 assessments. E-mail: fwolfe@ to have satisfied the ACR classification criteria. speaking fees. we pri- categorize or assess these patients. Lilly. and Astra. All of the physicians com- ing fees.000) from Forest Laboratories. Cypress Biosciences. some physicians considered that fibromyalgia was tients and controls were assessed by physicians with a a spectrum disorder and was not well served by dichoto. and the patient. and that patient forms could only be completed by received consultant fees (less than $10. and/or honoraria (more than $10. speaking fees. Na. fees. Lilly. Wichita. 315 additional pa- Finally.

Among the expert physicians. Two physicians enrolled fewer than 10 pa. 4 used the ACR classification criteria past week?.7%) for one or more protocol comfort in the upper abdomen (stomach).602 Wolfe et al patients using clinical methods and some patients using lem to fatigue is a major problem. fever. 2008. muscle weakness. and 2 used both methods. wheezing (asthma). trouble problems. paresthesias. or fibromyalgia problem. As with the fibromyalgia pa. problems abdomen. pain or cramps in the lower abdomen (colon). 2. Symptoms were as. indicated if the patients had the following symptoms: continuous. or patient’s form for completeness and request completion of odors. sensitivity to bright lights. indigestion or belching. fatigue. dry mouth. performed the ACR tender point count (0 –18) (1) and mined by study analyses (see below). tendonitis. From the ing.e. headache. generally mild or intermit. pain. respectively. We instructed study sites to of the hands. heartburn. legs. group of physician participants there were 10 “experts. black or tarry stools (not from iron). nervousness (anxiety). white. Patients indicated the number and were no more than 10 years younger or 10 years older of medications they used in the last month to help control than the fibromyalgia case. yellow skin or eyes (jaundice). with painful disorders such as degenerative neck and back pain anchors from waking up unrefreshed is no problem to syndromes or regional disorders. de- trols of the same sex within 10 years of age by sex (n ⫽ 96) pression. vomit- excluded subjects for invalid control diagnoses (n ⫽ 12) ing.. or discomfort in the blad- tient forms. The categories were deter. hearing difficulties. reference list consisted of: muscle pain. considerable problems. rheumatoid arthritis). trouble with anxiety or depression. and reported the extent of morning stiffness. tients. waking up unrefreshed is a major problem. trouble thinking or remembering. Physicians also and/or at a moderate level. and 3 ⫽ severe. reaction. 10 ⫽ very certain). 2009. the control subjects must have had a prior control We also asked patients to indicate which of the follow- diagnosis. diarrhea. mouth sores. cognitive sessed using the following words: pain. Physi- (11). muscle pain. In phase 1. joint pain. depres- awaking unrefreshed. easy bruis- 514 subjects (mean 17 and median 20 per center). fatigue. weakness of muscles. loss of appetite. and overall (global) severity using the same tent. tender missing items before the patient left the clinic. 4 – 6. explosive watery bowel movements. cognitive symptoms. often present categorical scoring as the patients did. or other nonrheumatic ears. and April 30. dizziness. loss of hair. and for failure to match fibromyalgia patients with con. swelling Study variables: phase 1. sun sensitivity (unusual skin cal literature. we enrolled 610 patients from 32 referring shortness of breath. faintness. or change in taste. constipation. anxiety. 2) how tients as having few or no somatic symptoms. pain/cramps in the with sleep. 2 ⫽ moderate. irritable bowel . headache. feet. and 0 –10 scale (0 ⫽ very uncertain. insomnia. Physicians were instructed not to look at the pa. 4 used problem has sleep (i. fatigue. active cancer. or with emotional upset. 1–3. muscle pain. low back pain. Patients also or similar disorders who had not been diagnosed previ. completed the Health Assessment Questionnaire II func- ously as having fibromyalgia and who were of the same sex tional disability scale (12). dry eyes. tiredness (fatigue). with anchors from sleep is no problem to sleep diagnosis. and 3 for values of 0.. neck pain. After exclusions. numbness/tingling/burning. ringing in defined neuropathic causes of pain. life-disturbing problems. The over the past week?. nausea. is a major problem. ing symptoms they experienced in the last 3 months: order (e. seizures or by the time of study closure. frequent urination. or ankles (not due to arthritis). fatigue severe enough to limit daily activity. These areas were those previously described as Physician variables: phase 1. They completed categorical scales for pain. Physicians’ staff was asked to check each der region. Physicians were asked part of the Regional Pain Scale (renamed here as the WPI) not to look at the forms completed by the patient. completed the same painful body region scale as their Patients completed 7 categorical scales for symptoms patients did. Patients with any inflammatory rheumatic dis. liver problems. physicians between December 2. We also analyzed the WPI as a categorical variable.” and blue skin color changes in fingers on exposure to cold defined as having published on fibromyalgia in the medi. joint swelling. sleep problem. blurred vision or problems focusing. cluded because of insufficient site enrollments. patients completed 4 and muscle tenderness. 1. (frequent. hives or welts. resting at night) been for you in the clinical diagnosis. and their patients were ex. loud noises. The scale We then provided the physicians a list of symptoms for questions and anchors were 1) severity of pain over the last reference purposes and asked them to categorize the pa- week. or a great deal of symptoms. waking 1 ⫽ slight or mild problems. unrefreshed. and 4) how much of a problem has Control subjects were patients with noninflammatory waking up unrefreshed been for you in the last week?. pervasive. dizziness. We summed the positive replies to create a 0 –56 count of somatic symp- Patient variables: phase 1. irritable bowel syndrome. or frequent sore throats. violations. have patients complete their forms before seeing the phy. diarrhea tients. lymph nodes. rash. In addition. a moderate much of a problem has fatigue or tiredness been for you number of symptoms. toms scale. with anchors from fatigue is no prob. loss of causes for pain were excluded from the study. They ⱖ7 of the WPI.g. pain. chest pain. cians indicated their certainty of the prior diagnosis on a with categories 0. itching. pain or dis- We excluded 96 patients (15. with anchors from no pain to severe pain. Patients were asked to indi. We also constipation. severe). sleep disturbance. cate in which of 19 body areas they had pain during the last week. there were convulsions. red. 1 and 5 patients each. fractures. over the past week that were scored as: 0 ⫽ no problem. fullness. painful urination. irritable bowel syndrome. osteoarthritis. interstitial cystitis. not sunburn). sician. fluid-filled blisters. 3) how much of a the ACR methods. and overall severity of your arthritis sion. visual analog scales that were scored as 0 –10.

while the faxed to the National Data Bank for Rheumatic Diseases. fatigue. dry mouth. insomnia. For reference. mood. indicated the presence or absence of (StataCorp). de- pression. but could not score the individ. pain/cramps in the abdomen. heartburn. 4 – 6. if accuracy is the more direct. wheezing. Based on the study results described below. The phase 2 form was simplified and completed only by the physician. and chi-square tests. for processing. and bladder spasms. numbness/tingling. Statistical methods. somatic symptoms. Spearman’s coefficients were used for correlation ual regions.6% had 3– 8 missing data points. Raynaud’s phenomenon. Figure 1. Phase 2 study and variables. Breiman argues that the decrease in predictive physician was contacted immediately for corrections. cognitive problems. chest pain. examining. seizures. a complete set. closure. unrefreshed sleep. question about widespread pain. distribution of study variables using a probability density In phase 1 analyses. the examining ent things. crease in accuracy is thought to be a better measure: “. waking unrefreshed. the criteria and severity scales would probably come from the RESULTS short set. The study was approved by the Via Christi The complete set included all of the study variables. easy bruising. 81. a moderate number of statistical package. we considered 3 groups of classifier function. version 2. terms of their importance as predictors. The reader can think of the probability density variables: a short set of variables. 12. Conflict of interest and ethics. The mean de- with scoring as described above. Classification tree analysis used the rpart recursive Missing data were rare: in phase 1. and irritable bowel syn. an intermediate set. They were reviewed for ally predicts. and the R statistical package. cognition. Figure 1 describes the 1. but they measure differ- missing data. Random forest analysis was used to determine variable drome. muscle tenderness. thinking or remembering problem. The short set included the WPI and cate. dry eyes. itching. stable and meaningful indica- possible. Kansa. The out-of-bag error rate is a robust symptoms.Diagnostic Criteria for Fibromyalgia 603 syndrome. and provided a rating of the extent of somatic importance and “out-of-bag” error rates (10. constipation. anticipated that. pain in the upper abdomen.8% had 1 missing data point. rash. we created an SS scale by summing the 0 –3 scores of somatic symp- toms. sets as to their classifying value and to understand what It contained the following items: a categorical WPI and a might be lost in shortening the set of variables. The complete set allowed us to compare the 3 in phase 1 whose clinical diagnosis was fibromyalgia and . the Gini Index. compensation. oral ul- cers. diarrhea. vomiting. nausea. Data were analyzed using Stata. We analyzed the intermediate set under the consideration that the individual symptoms might also be Demographics. At the time of study visual analog scale. 7–10. and if such data were found. nervousness. blurred vision. as indi- 0 –3. painful urination.000 for their work in interview- cognitive problems. hair loss. muscle pain. and it represents a ranking of variables in bance. and function as a “smoothed-out” version of a histogram. or ⱖ11). Examining physicians gorical scales for pain. hives/welts. The study authors received no the short set plus all of the individual somatic symptoms. regression ana- vided with a list of the widespread pain regions (scored as lyses with dummy variables. loss of appetite. fatigue/tiredness.7% of patients had partitioning R analysis programs to determine preliminary no missing data.11) using the R symptoms (few or no symptoms. . and fatigue into a 0 –12 scale. because of the need for simplicity. cated. On completion. 315 valid patients had been enrolled. Physicians performed and recorded a tender analyses. sleep disturbance. hearing difficulties. Physicians enrolled 258 valid patients important. dizziness. Comparisons between groups used t-tests. Physicians were pro. We Institutional Review Board. Missing patient data and physician data that tor of variable importance (personal communication)” were not correctable within one week were left as missing. The methods and rules for physician and patient recruitment were the same for the second phase. Variable importance is list of 41 symptoms was supplied on the questionnaire. and completing study forms for 10 pa- sleep. fever. and cut points for criteria variables (14). Distribution of key fibromyalgia (FM) variables in con- myalgia expertise were not recruited. were each compensated $2. patients trols and patients with current or prior FM (phase 1). . except that the authors with known fibro. we included categorical scales for sleep distur. The intermediate set included all of the variables in tients and 10 controls. study forms were the Gini Index reflects the overall goodness of fit. In addition. frequent urination. predictive accuracy depends on how well the model actu- Wichita. loss of/change in taste. ringing in ears. version 10. muscle weakness.81. described by the mean decrease in accuracy criterion or Finally. and fatigue. and unrefreshed ing. sun sensitivity. VAS ⫽ did not complete questionnaires. The two are related. or a great deal of symptoms).1 point examination. (13). a measure of misclassification error. shortness of breath. headache.

9 56. physician fibromyal.4 Patient symptom count (0–48) 22. patients having the greatest symptom severity with prior fibromyalgia generally occupying the severity scale mid- Diagnosis and diagnostic methods.4 3.9%.7 24.9 ⫾ 0.3 ⫾ 2. 74.5 ⫾ 2. versus 52. ACR classification point between current fibromyalgia and controls (Table 1).517).1.9 3.0 0.1%) with current fibromyalgia (ACR clas.1) 67 (13.3 ⫾ 2. except for somatic symptoms. for the count of patient-endorsed somatic symp- clinical diagnosis was used in 36. P ⫽ 0.0 0.0 ⫾ 0.5 Widespread pain. of pain medications 3. Based groups for key variables.3 ⫾ 3.770).3 7.0 ACR 1990 classification criteria positive. and control diagnosis 9. (ACR classification criteria negative.9 0. % patients 92.2 No.7 1.4 ⫾ 4.1 7.0 ⫾ 2.1 1. prior fibromyalgia 8.001). % physicians 93.3 Tender point count (0–18) 15.2 years. † Sum of physician somatic symptoms. and the SS noses.0%) with prior fibromyalgia gia had the WPI shifted somewhat to the left (Figure 1A).3 ⫾ 2. % patients 92.1 ⫾ 3.2 ⫾ 8. the physician somatic symptom scale.9 ⫾ 8.6 ⫾ 12.2 ⫾ 5.7 4. categorical (0–3) 2.2 ⫾ 0.9 0.3 ⫾ 2.6 Patient VAS unrefreshed sleep (0–10) 7.6 6.035). VAS ⫽ visual analog scale.6 1.8 1. P ⫽ 0.8 ⫾ 3.0 ⫾ 3. prior fibromyalgia patients had scores that were patients who had been previously diagnosed with fibro.7 ⫾ 8.4 8.9 physicians and the 20 clinical rheumatologists (P ⫽ 0.0 ACR 1990 classification criteria positive. To compared with patients previously diagnosed by the ACR determine variables that best identify fibromyalgia and to classification criteria (18. while prior fibromyal- sis positive).1) Widespread pain index (0–19) 11. ACR ⫽ American College of Rheumatology.4.9 ⫾ 4.604 Wolfe et al Table 1.2 ⫾ 3.2% versus 9.5 ⫾ 3.0% control subjects. toms. Patients previously diagnosed by clinical criteria were more likely to be classified as prior fibromyalgia (38.9 Symptom severity scale (0–12)† 8. on prior diagnosis and ACR classification criteria status: followed by unrefreshed sleep (Figure 1C). percentage of non-Hispanic whites (86. The tender point count (Figure on these data.2 ⫾ 2.0 3.8 4. trol subjects.0) 251 (48.6 3.1 ⫾ 2.8% versus 85.640).0 ⫾ 2.5 1.1 Widespread pain.5 ⫾ 2.1 3.5 HAQ-II score (0–3) 1.7.7 0.1 Patient global severity.1%) who were 25% of patients considered to have fibromyalgia by their neither current nor prior fibromyalgia patients (control physicians do not satisfy ACR classification criteria for subjects) (Table 1).2 3.0 ACR 1990 classification criteria positive.4 1. Characteristics of patients by fibromyalgia status.4% of fibromyalgia diag.0 physicians Physician global severity. physician cognition. 67 patients (13.3 ⫾ 0. and 251 patients (48. the current fibromyalgia versus 14. we divided the subjects into ACR . The proportion examine the predictive power of study variables without of patients who were controls or had prior or current the use of tender points.3 ⫾ 2. Using a 0 –10 physician certainty of fibromyalgia.2 years.0 0. rent fibromyalgia patients had similar distributions of so- sification criteria positive. of patients (%) 196 (38.732). Fibromyalgia subjects were fibromyalgia did not differ between the group of 10 expert slightly older than controls (mean ⫾ SD age 54.3 ⫾ 12. the mean certainties were: fibromy.6% of fibromyalgia diagnoses and However. % physicians 93. At the time of the study examination.9 59. There was a clear difference in clinical findings and symp- P ⫽ 0.5 ⫾ 1. P ⫽ 0.9 * Values are the mean ⫾ SD unless otherwise indicated. and that they appear to have an intermediate prior diagnosis scale.4 9.7 31. categorical (0–3) 2.1 2. Selected clinical characteristics of patients with current or prior fibromyalgia and controls in phase 1* Current Prior Variable fibromyalgia fibromyalgia Controls No.7 1.3%) Misclassification rates and fibromyalgia classifiers.0 Patient VAS sleep (0–10) 6.4 1.0%. Prior and cur- 196 patients (38. we categorized patients into 3 groups based 1D) demonstrates the clearest distinction between groups.3 4.3 2.9 ⫾ 1. % patients or 100. 256 who were control subjects. Figure 1 shows differences between of controls satisfied the ACR classification criteria.8 18. these data show that approximately gia diagnosis positive). physician waking unrefreshed. Taken as a whole.4 Physician somatic symptoms (0–3) 2. physician fibromyalgia diagno.4 5.1 tom severity among the groups. but did not differ by the percentage of males (8.4 ⫾ 3. P ⬍ 0.8 Patient VAS fatigue (0–10) 7. somewhat closer to current fibromyalgia patients than to myalgia satisfied the ACR classification criteria and 2.7 5.3 ⫾ 2. severity position between fibromyalgia patients and con- algia 9. or education level (mean ⫾ SD 14.9 ⫾ 2.2 ⫾ 3.0 ⫾ 2.4 ⫾ 4.0 0. criteria were used in 63. and physician fatigue.7 ⫾ 0.9%.9 ⫾ 2.3 ⫾ 0.5% of scale.0 Patient VAS pain (0–10) 6. matic symptom counts (Figure 1B).2 Physician widespread pain index (0–19) 11. HAQ-II ⫽ Health Assessment Questionnaire II.

cognition.2 6.4% (14. (variable group) and whether the WPI was used as a con. NA ⫽ not applicable. whereas removing the WPI and to create an SS scale. Figure 2A shows that the WPI and (84. Random forest out-of-bag misclassification rates for physician and patient variable groups in the diagnosis of ACR 1990 classification criteria–positive fibromyalgia (phase 1)* Error rate Error rate Error rate Error rate (no WPI. muscle tenderness were the most important variables in again with improvement in the phase 2 data.Diagnostic Criteria for Fibromyalgia 605 Table 2. Random forest analysis provides in- negative cases (all controls satisfying the ACR criteria). importance levels (somatic symptoms. tive cases (excluding prior fibromyalgia).3%. with the tender point count and the WPI. and with very slightly better results if tion rates when muscle variables (muscle pain. with correlation coefficients of 0. In these In addition. cognition. complete group ⫽ all study variables. 4 – 6. Using variables (Table 2 and Figure 2B). short group ⫽ WPI and categorical scales for pain. fatigue. we turned back to the interme- but leaving all muscle pain and muscle tenderness vari.7 NA Intermediate 14.6%.4 NA NA NA Intermediate 6. no (no WPI. cut points for study variables. we examined the complete set of variables.1% (Table 2). For comparison. tween 16. classification criteria–positive (all patients satisfying the the WPI and muscle data (Table 2) and had the lowest ACR classification criteria) and ACR classification criteria– misclassification rate.0%. Removing all of the muscle variables and the WPI resulted in a misclassification rate of 16. the correct clas- Depending on the number of variables in the model sification rates for phase 1 ran between 89. waking unre- These data show that misclassification rates obtainable freshed. Overall. % Complete 12. in the comparison of ACR classification analyses. cognitive problems. was less strongly correlated. These data pro- some gain obtained by including information about mus- vide an estimate of the lowest misclassification rates of cle pain or muscle tenderness. 0. splitting it into categories at 0.1 14.4%. We cre- Based on the information in Table 2. fatigue. and ⱖ7. and fatigue. the misclassification rate was 18. using the ACR classifi- . mood. There was the classification of cases and noncases. 1–3.9 10. formation on variable importance and misclassification after excluding patients who physicians designated as hav. with higher rates being obtained in phase 2 analyses. these variables are strongly correlated tain ACR classification criteria–related variables are be. rates. be identified with an acceptable rate of error by the use of We next performed analyses to determine misclassifica- the WPI alone. Using the The SS scale and fibromyalgia diagnostic criteria. We then used random forest analyses to partitioning to the data and determined that values of the rank physician predictors of fibromyalgia by their impor. We examined best identified control cases according to their ACR clas- the WPI as a continuous variable (0 –19) and also after sification criteria status. physician study variables When all of the patients were considered. these data show surrogate criteria obtainable under optimum data mining that patients who satisfy the ACR classification criteria can conditions.0% if muscle pain is al.2–10. muscle muscle tenderness or muscle pain is assessed. pain intensity. we the short variable list (that never contains muscle pain or identified 6 categorical rating scale variables that had high muscle weakness).6%) as physician diagnosis 6. As under optimum data mining conditions that do not con. and unrefreshed sleep. shown in Table 3. an criteria–positive versus ACR classification criteria–nega- intermediate set.437. model (7. sleep problems. ACR ⫽ American College of Rheumatology. waking unrefreshed.0 10. tinuous or a categorical variable. and mood).0% and 92. We then applied recursive cation criteria.1% and 18. somatic symptoms. To same intermediate model but keeping muscle pain re. From Figure 2B. no Error rate Variable group (continuous WPI) (categorical WPI) muscle variables) muscle tenderness) (no WPI) Physician evaluator.346 and lowed). we selected the ated an SS scale by summing the 0 –3 scores of somatic variables contained in the intermediate WPI categorical symptoms. tenderness) and the WPI were removed from the models.3% (Table 2). WPI ⱖ7 best identified fibromyalgia cases and values ⱕ6 tance and to obtain misclassification rates.3 Short 8. WPI ⱖ7 per- misclassified fibromyalgia cases and controls at rates of formed about as well (83. % Complete 6.4 NA NA Patient evaluator.9 NA * Prior fibromyalgia cases are excluded from the analysis. and a short set (Table 2). but does not provide information about the optimum ing fibromyalgia but who did not satisfy the ACR classifi. build and identify an alternative definition of fibromyalgia duced the error rate to 14.3% error rate) for further study because that into a 0 –12 scale.3 18. diate categorical model that excluded the WPI and muscle ables resulted in a misclassification rate of 10.1%) in the classification of fibromyalgia in phase 1. WPI ⫽ widespread pain index.1 NA Short 13.4 7. model represented the best practical model that contained When applied to study patients.3 16. intermediate group ⫽ short group variables plus all individual somatic symptoms. sleep disturbance. respectively.

6. for prior fibromyalgia was 6. 53. and for controls was 3. We used the SS scale and the WPI to recommend a new case definition of fibromyalgia: (WPI ⱖ7 AND SS ⱖ5) OR (WPI 3– 6 AND SS ⱖ9) (Table 4). Spearman’s correlation coefficients for all study subjects (phase 1) Tender point Widespread Variable count pain index Tender point count 1.773 (Table 3). This definition recognizes that fibromyalgia is more than just a high WPI scale by requir- ing an SS scale score ⱖ5.504 0. Other variables are present/ absent conditions. but they measure different things (13).606 Wolfe et al Table 3. the SS scale appropriately follows the Figure 3. including of cases correctly compared with the clinician’s correct the widespread pain index (WPI) and muscle symptoms.524 0. Variable importance (physician variables) in distin.680).1% of ACR classification criteria–positive cases would not be diag- nosed as having fibromyalgia.6% cian evaluation by random forest analysis in phase 1. cation criteria definition of fibromyalgia.548 0.000 0. and it represents a ranking of variables in terms of their importance as predictors. the mean ⫾ SD SS scale score for fibromyalgia was 8. Using this definition. As Figure 2.1% of prior fibromyalgia patients would be diagnosed as having fibromyalgia. Overall.773 1. and non-fibromyalgia. 9. while the predictive accuracy depends on how well the model actually predicts. shown in Table 5. using the ACR classification criteria as guishing fibromyalgia from controls in intermediate detail physi. The two are related.733) and the tender point count (0. The SS scale was strongly correlated with the WPI (0. and 14.346 0. provided that there is sufficient body pain. suggesting 2.1% of controls would be diagnosed as having fibromyalgia. Fibromyalgia case definition: diagnostic criteria. B. and recognizes that a high level (SS ⱖ9) of symptoms should be sufficient for diagnosis.5% using the recommended definition.592 Fatigue† 0.680 0. the fibromyalgia rate among all of the study subjects would increase from 38.534 Pain 0. A symptom severity scale score ⬎6 identifies patients satisfying that it can categorize fibromyalgia symptom severity. which was correlated with the WPI at 0. .604 Sleep problems 0.3 ⫾ 2. The Gini Index reflects the overall goodness of fit. and was the strongest correlate of WPI after the tender point count.1% to 45.622 0. able importance (physician variables) in distinguishing fibromy- algia from controls in intermediate detail physician evaluation in We also used the SS scale to aid in the diagnosis of phase 1 by random forest analysis.0 ⫾ 2. prior fibromyalgia.3% of cases.683 Waking unrefreshed† 0.546 Mood 0.1%.2 (Table 1).773 Widespread pain index 0. The mean decrease in accuracy is thought to be a better measure.575 0. the new diagnostic criteria in 92. Vari. the gold standard. Variable importance is described by the mean decrease in accuracy criterion or the Gini Index.437 * Five-component symptom severity scale (includes mood).733 Symptom severity scale* 0. IBS ⫽ irritable bowel syndrome.631 0.6. Distribution of severity scores using the American Col- lege of Rheumatology (ACR) 1990 classification criteria definition definitions of ACR classification criteria–positive fibromy- of fibromyalgia (FM) by category of FM diagnosis in phases 1 and algia. As shown in Figure 3.713 Somatic symptoms† 0. C ⫽ categorical scale.0 ⫾ 2. A. † Component of 4-component symptom severity scale.000 Symptom severity scale 0. diagnosis rate of 84.672 0. excluding the Regional Pain Scale (WPI) and muscle symptoms.656 Cognition† 0. the diagnostic criteria diagnose 82.

sun sensitivity.4 95.0 83.3 79. fatigue/tiredness. waking unrefreshed. left Jaw.6 Classifying patients according to ACR 1990 classification criteria status ACR 1990 criteria vs.6 95. painful urination. WPI ⱖ7 89. muscle weakness. The final score is between 0 and 12. diarrhea. depression.1 ACR 1990 criteria vs. right Hip (buttock. A severity score ⱖ7 (WPI 3– 6 AND SS ⱖ9). ringing in ears.5 scale (SS ⱖ7) Diagnostic criteria ((WPI ⱖ7 AND SS ⱖ5) OR SS ⱖ9) 92. thinking or remembering problem. right 2) SS scale score: Fatigue Waking unrefreshed Cognitive symptoms For the each of the 3 symptoms above.6 87. When applied to the ACR classi- correctly classifies 92. This Table 5. * Somatic symptoms that might be considered: muscle pain. nervousness. dry eyes. physician diagnosis 84. vomiting. easy bruising.Diagnostic Criteria for Fibromyalgia 607 Table 4. FM patients 74. dry mouth. frequent urination. 3) The patient does not have a disorder that would otherwise explain the pain. right Abdomen Lower arm. constipation. right Lower leg.9 ACR 1990 criteria vs. dizziness. left Chest Neck Upper arm.2 88.2%. categorical SS 84. considerable problems. Fibromyalgia diagnostic criteria Criteria A patient satisfies diagnostic criteria for fibromyalgia if the following 3 conditions are met: 1) Widespread pain index (WPI) ⱖ7 and symptom severity (SS) scale score ⱖ5 or WPI 3–6 and SS scale score ⱖ9. loss of appetite. pain/cramps in the abdomen. left Upper back Shoulder girdle. left Lower arm. trochanter).0 75. nausea.5 FM diagnostic criteria ACR 1990 classification criteria vs. fever. hair loss. clinical fibromyalgia. and bladder spasms. heartburn.2 90. right Jaw. FM ⫽ fibromyalgia.5 ACR 1990 criteria vs. generally mild or intermittent 2 ⫽ moderate. indicate whether the patient has:* 0 ⫽ no symptoms 1 ⫽ few symptoms 2 ⫽ a moderate number of symptoms 3 ⫽ a great deal of symptoms The SS scale score is the sum of the severity of the 3 symptoms (fatigue. left Lower leg. categorical SS scale (SS ⱖ7) * ACR ⫽ American College of Rheumatology. diagnostic criteria 88.3 87. WPI ⫽ widespread pain index. itching. pain in the upper abdomen. . Percentage of patients correctly classified according to ACR 1990 classification criteria status and diagnostic criteria* Phase 1 Phase 2 Prior FM All Prior FM All Comparison excluded patients excluded patients ACR classification criteria positive. numbness/tingling. hives/welts.3 89. because Figure 3 suggests that sever.3% of cases identified by the new fication criteria. cognitive symptoms) plus the extent (severity) of somatic symptoms in general. wheezing. In how many areas has the patient had pain? Score will be between 0 and 19.2 90. WPI ⱖ7 plus muscle pain 89. rash.2 vs. WPI ⱖ7 plus muscle tenderness 92.1 82. SS ⫽ symptom severity.8 ((WPI ⱖ7 AND SS ⱖ5) OR SS ⱖ9) ACR 1990 classification criteria vs. right Upper leg. life-disturbing problems Considering somatic symptoms in general. shortness of breath. Ascertainment 1) WPI: note the number areas in which the patient has had pain over the last week. indicate the level of severity over the past week using the following scale: 0 ⫽ no problem 1 ⫽ slight or mild problems. often present and/or at a moderate level 3 ⫽ severe: pervasive. loss of/change in taste. oral ulcers. blurred vision. correct classification falls to 79. Raynaud’s phenomenon. seizures. headache. Shoulder girdle. diagnostic criteria definition: (WPI ⱖ7 AND SS ⱖ5) OR ity scale criteria might be applicable.9 84. continuous. irritable bowel syndrome. hearing difficulties. trochanter). left Upper leg.8 84.6 92.2 90. 2) Symptoms have been present at a similar level for at least 3 months. right Lower back Upper arm. insomnia. chest pain.6 95. left Hip (buttock.

5% in the first phase of assessments. The validation sample have fibromyalgia because they do not meet criteria. measurement of fibromyalgia symptom severity to patients The utility of such diagnostic criteria is very dependent on who do and do not currently satisfy ACR fibromyalgia the clinical setting. arises because fibromyalgia diagnosis is based on severity fication criteria compared with 74. fatigue. The loss of a tender point or a painful region the study. as the definition of fibromyalgia is those proposed by Yunus et al in 1981 (15).625.5 ⫾ 2. These ACR classifi- In creating the SS scale we chose not to use the mood cation criteria performed well in specialty clinics and variable. and 44. somatic symptoms 0. waking volve a tender point count or an adequately executed unrefreshed 0. although they were considered to have fibromyalgia by their physicians.619. where pa- tients continue to have their condition even though they may subsequently not meet diagnostic criteria. 13. Practically. fibromyalgia differs from rheumatoid arthritis or teria to identify ACR classification criteria cases worked systemic lupus erythematosus. As shown in Figure 3. It diagnosed as having fibromyalgia. including improvement. In this was slightly different in the 2 phases (Figure 3). and fatigue 0. criteria status: 42. difficult to assess and because it might be a resultant The diagnostic criteria suggested here (Table 4) are not feature rather than a primary feature of the illness. However. However. We also selected the SS scale. they have not been widely embraced in primary care. 76. we were cognizant of the need to make the diagnostic tom count to measure fibromyalgia symptom severity. DISCUSSION We suggest the use of the SS scale to quantify fibromy- algia symptom severity as a workable solution to this prob- We are aware of the inherent problems in the diagnosis of lem. However. though we omitted mood. but who do not meet criteria versus patients who do not rent nor prior fibromyalgia patients. it was correlated with the SS they were designed to address certain realities. thereby linking current other disease was thought by the examiner to be account- status to previous diagnosis.6 ⫾ 1. abdominal pain. The initial ACR fibromyalgia classifi- criteria. the distribution of SS scale somatic syndromes that lack objective physical or labora- scores defines the 3 diagnostic categories and provides tory features or well-characterized pathologic findings.9 ⫾ 1. we identified 2 ferentiating power. Using the newly allowed fewer tender points in the presence of many proposed categorical definition for fibromyalgia. the case definition shown in Table 3.8 and 2. First.6. and they stressed the importance of symptoms. although unrefreshed sleep is better. Practically. These authors shifted in the new criteria definition. best identified patients diagnosed with the no power to correct classification and we did not include ACR classification criteria. Among patients here provide a solution for this classification dilemma. patients who have satisfied the ACR cation criteria stipulated that chronic widespread pain classification or diagnostic criteria at one time can be was present for 3 months and it was understood that no followed with the symptom scale. can result in fail- based on prior diagnosis and current ACR classification ing to meet classification or diagnostic criteria. we suggest that either scale can be variables that best defined fibromyalgia and its symptom used. None of this provide a measure of fibromyalgia symptom severity. the symptoms. as tender count when performed. and which strongly correlated with the tender point count and headache had variable importance. illnesses that do not in- slightly better in phase 2. and somatic symp. a 5-component SS scale that included of fibromyalgia has changed somewhat with increasing mood did not perform as well as a 4-component scale. variables. We categorized the 315 patients into 3 groups for any reason.3% the conundrum involves patients who have fibromyalgia with prior fibromyalgia. the mean ⫾ SD respective scores were tients in both phases of the study did not satisfy the ACR 8. and the knowl- . it provided marginally better dif- In developing new diagnostic criteria. but it many respects. composite variable composed of physician-rated cognitive In addition. in the 1990 ACR classification criteria. were very useful in providing some patient homogeneity We chose not to include mood because we judged mood for clinical trials. respect.0% met ACR classi. recognition of the importance of cognitive problems and Unrefreshed sleep was a better measure than sleep distur- somatic symptoms. mean ⫾ SD SS scale score for fibromyalgia was 8.621. these criteria are similar in concept to is easy enough to use them appropriately.725 and with the components of the SS scale as clinical diagnosis in primary care does not ordinarily in. Instead. volve diagnosis based on symptom severity. irritable bowel syndrome. Using the SS scale that approximately 25% of diagnosed fibromyalgia pa- score cutoff of ⱖ7. Neither the ACR Phase 2 validation. Al- meant to replace the ACR classification criteria. The WPI. the scale at 0. Second. and cri. In addition. they added the SS scale. in sensitivity analyses. although it had importance in Figures 2A and B. in developing criteria for clinical diagnosis problems. ing for the chronic widespread pain.4% who were neither cur.6% of fibromyalgia diagnoses. a them in the diagnostic criteria.0. we used algia diagnostic criteria: (WPI ⱖ7 AND SS ⱖ5) OR (WPI simple categorical scales for the symptom scale and com- 3– 6 AND SS ⱖ9). Of the binary spectrum: the WPI and the composite SS scale. To that end. The ACR classification criteria were classification criteria nor the diagnostic criteria suggested used in 74. actually be used in clinical practice.8 ⫾ 1.612. follows: cognition 0. classification criteria. unrefreshed sleep.4. One of the important findings of the current study was and for non-fibromyalgia was 2.2% with current fibromyalgia. factors that were not even considered bance and we recommend its use in the SS scale. In guarantees that the scales will be used appropriately. We also used the SS scale alone to pressed the WPI to a similar categorical scale. We criteria simple enough and easy enough so that they would used the WPI and the SS scale together to define fibromy.608 Wolfe et al falloff is to be expected.

However. causes of chronic widespread pain. . the SS scale. Although our criteria are unable to provide a useful classification for the goal was to develop simple. There. it seems possible that the ACR the patient subsequently not satisfy ACR classification or classification criteria might be withdrawn. but does not include the WPI. The criteria and severity scale also pro. we collapsed the diagnostic criteria had a least 1 tender point and 96. Based on the results in Table 2. Should tic criteria perform well. well as data from physicians. fibromyalgia symp. it should be of these criteria in primary care.3% questionnaire to categorical scales and changed the format had 3 or more tender points. we were concerned that a simple Physicians who are used to the ACR classification crite- questionnaire would lose too much information. In designing this study. the patients with fibromyalgia diagnosed by physicians bromyalgia. was 12. followup study in the primary care setting be accom- sified 84. However. second. with marked symptom variability. the SS scale can also be used at any time regardless classified by the ACR classification criteria. especially useful in the longitudinal evaluation of patients vide room for those who are uncomfortable with the fibro. had symptoms for at least 3 months. The SS diagnostic information. It is larity (16). the results of the classification that may occur.7% compared with 6. we did not study the performance through the use of the SS scale. We recommend that a remembered that physician diagnosis only correctly clas.6% of other rheumatic conditions.7% of the patients in the study satisfying the little gain from the complete data sets. This simple clinical case def- but not satisfying fibromyalgia criteria at the same time. rightly criticized for circu. is also implicit in the proposed diagnostic criteria. Of inition of fibromyalgia correctly classifies 88.Diagnostic Criteria for Fibromyalgia 609 edge that these are key factors will increase recognition of myalgia concept. patient data did not work as toms are accorded appropriate importance by the provi. The shift in the conceptualization of fibromy. diagnosis of fibromyalgia that are suitable for use in pri- algia. well in correct classification as physician data (Table 2). as we expect. created a de facto case definition by imposing important for physicians to perform an appropriate clini- the 11 tender point rules on top of a crude definition of cal assessment to exclude other diagnoses. posed. the time- algia that occurred in the clinic and in research studies. 98.e.2% with physician data. we used a complex series of questions and evaluation examination criterion in the new diagnostic criteria. frame for the physician assessment of WPI and the SS scale however. it is clear that the ACR classification This study has a number of limitations. not the patient who makes the diagnosis. the cri. all of the phase 2 in Table 5 suggest that the criteria would work patients being diagnosed should have a physical examina- satisfactorily in the final format. This dance between the older ACR classification criteria case difference is related to the fact that it is the physician and definition and the current definition that we have pro. we have developed a case definition and current symptom severity status and change in status with.1% of cases. gia will have had a similar level of symptoms for at least 3 tion somewhat toward important symptoms: first. We would like to point out that implicit in the 1990 ACR Readers might wonder: why is it so difficult to make new classification criteria was the requirement that clinical fibromyalgia criteria? The central problem in fibromyalgia examination and clinical judgment had excluded other criteria is the absence of a gold standard or case definition. It will be have suggested. and such an exclusion The ACR classification criteria. Even though the new criteria to a form that could be used in the clinic. too. and does not of diagnostic status. and/or to iden- widespread pain (17). while the proposed diagnostic cri. The new diagnostic cri- variable importance analyses. diagnostic criteria for fibromyalgia: (WPI ⱖ7 AND SS ⱖ5) out the contradiction of having a diagnosis of fibromyalgia OR (WPI 3– 6 AND SS ⱖ9). The SS scale alone provides some require a physical or tender point examination. diagnostic criteria. We also did not test the criteria among those with teria. identified 82.. It scale enables assessment of fibromyalgia symptom severity provides information as to symptom severity and it allows in persons with current or previous fibromyalgia. and in fibromyalgia to be seen as part of a continuum. plished. The diagnostic criteria and teria require that patients diagnosed as having fibromyal- SS scale we have proposed shift the fibromyalgia defini. Subsequent studies as well as the tify potential coexisting rheumatic diseases that may re- data of this study showed the importance of (improved) quire treatment themselves. Table 4. provided no clear case definition. In the current was 1 week. which may include examination of tender point sites. How- in phase 1. the diagnos- SS scale can document baseline symptom severity. which showed ever. as they can simply report the WPI and such important symptoms. be done in the future. even with shift of definition. we derived an empirical case definition from the determine the level of symptoms. which is included in tion. sion of the SS scale. something that can be somewhat ameliorated mary and specialty care. practical criteria for clinical 25% of fibromyalgia cases we identified as prior fibromy.1% of cases course. as some those to whom the criteria have not been applied. If. The best misclassification rate obtainable with patient data The suggested diagnostic criteria create a small discor. In addition. quantitative measures of body pain (18 –22). Results from do not include a physical examination criterion. the SS scale can be used to measure In summary. all of variables that comprise the SS scale in characterizing fi. ria may be uncomfortable with the absence of a physical fore. following ways: following a diagnosis of fibromyalgia by other rheumatic conditions) to determine the rate of mis- ACR classification or diagnostic criteria. teria can be satisfied by a high level of symptoms if the Although we collected data directly from patients as WPI score is not high enough. and of the key Although not a study requirement for enrollment. and we recommend that this patients correctly (Table 5). We used this timeframe to more accurately study. The patient population should We envision the use of the diagnostic criteria in the include those with relevant differential diagnoses (i. months. in agreement with the ACR classification criteria.

MD. MD. Validation of the International Classification of Functioning. Fibromyalgia syndrome. 1996. J Rheumatol 2009. FACR. 21. Masi AT. CPI. J Rheu- matol 2009. Sivalingam tation. Alloway.. Arnold LM. Bombardier C. I.. MD.342:906 –9. Chen. 11. . MD. Fitzcharles MA. Anthony S. Amjad Roumany. MD. should reports of increased mortality influence management? Curr Rheumatol Rep 2005. et al. Wolfe. study physicians and their questionnaire submissions that Fibromyalgia syndrome module at OMERACT 9: domain con- insured study quality. When to diagnose fibromyalgia. J Rheumatol 2009. Atkinson EJ. MD. Simon LS. MD. and Karen E. MD. J Rheumatol 2003. 10. Robert W. Wilke. Fibromyalgia syndrome: review of clinical presen- Kathleen P. 2005. MD. Terence W. of the Health Assessment Questionnaire II: a revised version Torralba. Choy EH. Box. Buskila D. The authors thank Kimberly Harp of the National Data 4. Condemi. Silman AJ. 17. Hieblinger R. Clauw. Philip the continuum versus discrete disorder hypothesis. Coenen M. MD. Trock. William S. editors.20:485–501. Koneru. DO. Jon Russell. Cieza A. Wolfe F. Winfield. Glass Bank for Rheumatic Diseases for her superb work with JM. Jon Russell. MD. Yunus. Kimberly Carter Cer. criteria were developed? [editorial]. Rheumatology (Oxford) 2007. Sudhakar Tumuluri. Cohen ML. Steven J. Alghafeer. Goldenberg. Wolfe F. Kivitz. ville S. Russell. for the OMERACT Fibromyalgia Working Group. Clauw. North Am 1994. pain: is an improved classification possible? J Rheumatol Mease.75:397– 402. et al.34: che. 13. 7.36:2318 –29. MD.36:2330 – 4. integrity of the data and the accuracy of the data analysis. Lilly Research Laboratories did not participate in the design of Joint Bone Spine 2008. MD. FRCPC. Zagar. Sri L. Mease P. MD. Pain extent and diagnosis: development and valida- Theresa A. The following physicians partici. Wolfe F. Lindamood.799 patients with rheu- matic disease. Lawrence-Ford.75:6 –21. Smythe HA. critically for important intellectual content. Jasek. 16. to all of the data in the study and takes responsibility for the 11:151–71. Wolfe had full access with matched normal controls. A. p. Quintner JL. Fitzcharles.46: MD. FRCPC. Michaud K. MD. New York: Springer. Anthony S. Buskila D. MD. Jane H. Robert J. 22. Finn JD. tion of the regional pain scale in 12. Turkiewicz. Katz. Lynne S. Lee.50:3296 –305. Bennett R. Bartfai G. McBeth J. 2. Croft PR. Goldenberg. Nasef. Widespread Acquisition of data. et al. Katz. Aromaa A. Feigenbaum SL. MD. J Rheumatol 2007. Silman AJ. Winfield. Analysis and interpretation of data. 5. Content and criterion validity of the preliminary evaluations: Micha Abeles. Rothenberg. struct. Fibromyalgia syndrome. Miller KA. Is the report of widespread body pain Goldenberg DL. submitted abstracts. et al. Alan J. Boonen A. MD. Wolfe F. Berk RA. Meador. 160 –72. MD. Fitzcharles. Dao. Semin Arthritis Rheum 1981.42:263–7. Russell. 12. MD. MD. Bennett RM.30:369 –78. the study. berg. Winkelmann A. Inaccuracy in the diagnosis of et al.36:831– 6. Winfield. Elizabeth Reinitz. MD. Ndudi Oparae. Starz. Investigating the determinants of international differ- fibromyalgia syndrome: analysis of referrals. Rheumatology ences in the prevalence of chronic widespread pain: evidence (Oxford) 2003. MD.610 Wolfe et al ACKNOWLEDGMENTS of diagnostic and clinical features of fibromyalgia among fam- ily physicians. John B. Anthony S.23:1628 –32. Frances Y. MD. Russell R. Ara H. MD. PhD. Mar- pated in the study by enrolling patients and performing tin SA.14:238 – 41. MD. MD. Stucki G. Yunus. 1136 – 8. Karina D. of the Health Assessment Questionnaire. 8. see the results of the study. Macfarlane GJ. Ibrahim S. MD. MD. Wolfe. Rheum Dis Clin Dikranian. Peterson. MD. Car- MD. The American College of Rheumatology associated with long-term increased mortality? Data from the 1990 criteria for the classification of fibromyalgia: report of Mini-Finland Health Survey. Corey W. 1. Fibromyalgia: where are we a decade after the American College of Rheumatology classification FACP. Fu Flora Bai. 2000. Clauw DJ. veny. Russell. Jones GT. Golden- tautology. MD. Arthritis Rheum Walitt. core dataset for clinical trials in fibromyalgia syndrome. Wendy W. Walker. Macfarlane GJ. 1415–25. Arnold LM. MD. Pye SR. 2004. Straniero. J Rheumatol Suppl 2005. Mease P. Outcome and predictor relationships in Utsinger. Macfarlane GJ. Clauw DJ. 14. and treatment. Michaud K. Golombek. Crofford LJ. 9. Wolfe. Sherif A.11:R67. Ann Rheum Dis 2009. Therneau TM. Data mining and knowledge discovery handbook. Clauw. MD. Data mining within a regression framework. Arthritis Rheum 2002. Mease P. Wolfe F. or review the manuscript or 20. Arnold LM. Silverman. Gonter. David H. MD. a problem of Study conception and design. Leung. Fam Pract 1997. Choy EH. The significance of the sleeping-waking brain ROLE OF THE STUDY SPONSOR for the understanding of widespread musculoskeletal pain and fatigue in fibromyalgia syndrome and allied syndromes. All authors were involved in drafting the article or revising it 15. Awareness 68:690 –5. MPH. 18. 3. Anthony M. Baer. Soforo. from the European Male Ageing Study. In: Maimon O. MD. Levin. Crofford L. Neil J. Yunus MB. Rochester: Mayo Foundation. Arthritis Rheum 1990. Yunus M. Kanagasegar. 19. Jeff A. Disability and Health core set for chronic widespread pain from the perspective of fibromyalgia patients. Neumann L. John J. MD. Clauw DJ. pathogenesis. Arthritis Res REFERENCES Ther 2009. Ekaterina V. Peter D. Lan X. MD. Roland Staud. Shvartzman P. Carpenter. Knekt P. and all authors ap- Primary fibromyalgia (fibrositis): clinical study of 50 patients proved the final version to be published. Pincus T. Stuart L. Lancet 1993. Nicholas R. Calabro JJ. I. the Multicenter Criteria Committee. fibromyalgia and rheumatoid arthritis: evidence concerning MD. Moldofsky H. Wu FC. Crofford LJ. Patel. et al. Flint. Rokach L. Jr. MPH. outcome measures. Dr. An introduction to recursive par- AUTHOR CONTRIBUTIONS titioning using the RPART routines. MD. Michael W. Boulos P. Robert G.33: 46:805–7. MD. Chronic widespread pain and fibromyalgia: Mease. 6. MD. Jr.7:339 – 41. Macfarlane GJ. Brian T. Mikkelsson M. Kathryn H. MD. 231–55. Andrew M. Schollum J. Development and validation Rakesh C. MD. MD. Sibirski D.