Development and Testing of the Fibromyalgia Diagnostic Screen for Primary Care

Abstract

Background:

The Fibromyalgia Diagnostic Screen was developed for use by primary care clinicians to assist in the diagnostic evaluation of fibromyalgia, a disorder that predominantly affects women.

Methods:

The screen was designed to have a patient-completed questionnaire and a clinician-completed section, which included a brief physical examination pertinent to the differential diagnosis of fibromyalgia. The items in the questionnaire were based on patient focus groups and clinician and patient Delphi exercises, which resulted in a ranking of the most common and troublesome fibromyalgia symptoms. One hundred new chronic pain patients (pain > 30 days) and their primary care physicians completed the screen. The patients were grouped as fibromyalgia or nonfibromyalgia by an independent fibromyalgia specialist, who was blind to screen responses. Logistic regression was used to model the probability of fibromyalgia as a function of physician-reported and patient-reported variables. Best subset regression was used to identify a subset of symptoms that were summed to form a single measure. Receiver operating characteristic (ROC) analysis was then used to select thresholds for continuous variables. The symptom and clinical variables were combined to create candidate prediction rules that were compared in terms of sensitivity and specificity to select the best criterion.

Results:

Two final models were selected based on overall accuracy in predicting fibromyalgia: one used the patient-reported questionnaire only, and the other added a subset of the physical examination items to this patient questionnaire.

Conclusion:

A patient-reported questionnaire with or without a brief physical examination may improve identification of fibromyalgia patients in primary care settings.

Introduction

Chronic pain is increasingly recognized as an important women's health problem across the world.¹ Recent epidemiologic studies have shown that women are at increased risk for many musculoskeletal pain disorders, including fibromyalgia,^2,3 which has a worldwide prevalence of between 0.5% and 5% and affects women approximately 7 times more often than men.^4,5

Fibromyalgia is one of the most common disorders seen in rheumatology practices,⁶ and increasingly more patients with fibromyalgia are coming to their primary care clinicians for an initial evaluation and ongoing care. However, the disorder may be difficult to diagnose in the primary care clinical setting. In a recent survey, patients with fibromyalgia reported that it took an average of 2.3 years and visits to an average of 3.7 physicians before receiving a diagnosis of fibromyalgia.⁷ Patients with fibromyalgia frequently are referred to multiple specialists and have several investigations before a diagnosis of fibromyalgia is established.^7,8

Several factors may contribute to the difficulty in diagnosing fibromyalgia in the primary care setting. First, the diagnosis of fibromyalgia is currently based on the clinical presentation of the patient, with no objective tests to confirm a diagnosis. Second, patients report that it is sometimes difficult to communicate their symptoms to their primary care providers, who may have limited time to assess the multiple symptoms and comorbid disorders associated with fibromyalgia.⁷ Finally, although there has been increased awareness of fibromyalgia with the recent approval by the Food and Drug Administration (FDA) of three medications for the management of fibromyalgia,⁹ many primary care clinicians report insufficient knowledge or training to diagnose fibromyalgia or differentiate fibromyalgia from other pain disorders.⁷

The 1990 American College of Rheumatology (ACR) criteria for fibromyalgia helped to increase recognition of the disorder and stimulate fibromyalgia research. The ACR criteria require at least 3 months of widespread pain defined as axial pain, pain above and below the waist and on the right and left sides of the body. In addition, the criteria include pain in 11 of 18 tender point sites on digital palpation that is performed with an approximate force of 4 kg, which usually causes a whitening of the examiner's nailbed.¹⁰ Although the ACR criteria were intended to be used for classification of patients for clinical and epidemiologic investigations, the authors of the ACR criteria also noted that they could be used for diagnosis of fibromyalgia.¹⁰ Recent surveys of primary care physicians, however, suggest that many are either not aware of the ACR criteria or do not routinely use the criteria to diagnose fibromyalgia in their clinical practice.⁷ The tender point examination has been a barrier for some physicians who have not been trained or have limited time to conduct the examination. Furthermore, although the ACR criteria made no exclusions for the presence of concomitant radiographic or laboratory abnormalities and abandoned the distinction between primary fibromyalgia and secondary concomitant fibromyalgia, it was implicit that clinical examination and clinical judgment exclude other causes of chronic widespread pain.¹¹

A revision of the ACR criteria in 2010 established a clinical case definition that did not include a physical or tender point examination, but required that other disorders that would otherwise explain the pain be ruled out.¹² Wolfe et al.¹² recommended that all patients being diagnosed have an appropriate clinical assessment and physical examination, which may include an examination of tender point sites, to exclude other diagnoses or to identify coexisting rheumatic disorders that may require treatment. However, neither the 1990 nor the 2010 ACR revised criteria provided guidance about which painful conditions to rule out or what tests or other examinations to perform for the assessment of other conditions.

Studies indicate that primary care physicians sometimes either overlook common rheumatic conditions that are incorrectly labeled as fibromyalgia or miss the diagnosis of fibromyalgia.^13,14 Among the most common diagnoses that were mislabeled as fibromyalgia were treatable inflammatory rheumatologic disorders.¹⁵ Evidence also suggests that primary care physicians may order tests for autoantibodies in patients with symptoms of fatigue and diffuse musculoskeletal pain in the absence of other features, such as joint swelling and typical rash, contributing to low positive predictive values for the tests.¹⁶

Because of the growing numbers of patients requesting an evaluation for fibromyalgia, it has become important to develop tools to help primary care clinicians identify fibromyalgia and commonly associated conditions. Effective screening for fibromyalgia requires a tool that is valid and easy to use in the primary care setting. The goal of this study was to develop and test a diagnostic screen for use in primary care settings to improve the assessment of patients with fibromyalgia.

Materials and Methods

The study was conducted at the University of Cincinnati College of Medicine between July 2008 and October 2009. The University of Cincinnati Institutional Review Board approved the protocol, and all patients provided written informed consent after the study was explained and their questions were answered and before study procedures were initiated. Patients were recruited from the primary care outpatient practice affiliated with the University of Cincinnati.

Fibromyalgia Diagnostic Screen development

The items for the preliminary Fibromyalgia Diagnostic Screen were developed by one of the investigators (L.M.A.), based on clinical and research experience with fibromyalgia patients, with input from co-investigators from primary care (S.B.S.) and rheumatology (L.J.C.). The screen was divided into two sections: a patient-completed symptom and function questionnaire and a physician-completed assessment. The preliminary Fibromyalgia Diagnostic Screen was then field tested for item reduction and validation.

Testing study procedures

Female or male patients were eligible for the study if they were ≥ 18 years of age, met criteria for chronic pain defined as pain > 30 days, and could complete a follow-up appointment with a fibromyalgia specialist. The definition of chronic pain as > 30 days is consistent with another study that developed a screening tool for a different chronic pain condition.¹⁷ Patients were excluded if they had serious, unstable medical or psychiatric conditions that, in the opinion of the investigator, would compromise participation in the study. Patients were also excluded if they were participating in another study within 1 month of study entry. Patients coming to the primary care internal medicine practice at the University of Cincinnati were asked to participate at the time of their clinic visit.

After providing informed consent, patients were asked to complete the Fibromyalgia Diagnostic Screen self-report questionnaire. Their primary care physician completed the Fibromyalgia Diagnostic Screen physician assessment. The patients were then scheduled within 1 month for an evaluation by one of the investigators (L.M.A. or S.B.S.), who are fibromyalgia specialists. The fibromyalgia specialists, who were blinded to the Fibromyalgia Diagnostic Screen responses, performed a history and physical examination, including a review of past medical records and laboratory or other tests, as deemed necessary for diagnostic purposes. The evaluation also included an assessment of the 1990 ACR criteria for fibromyalgia.¹⁰ Based on best evidence and the available information, the fibromyalgia specialists classified the patients into fibromyalgia or nonfibromyalgia categories.

Statistical analysis

The goal of this study was to build parsimonious statistical models that reliably discriminate fibromyalgia from other pain disorders using a simple scoring rule (i.e., would not require a user to apply model coefficients or perform other complex calculations).

The fibromyalgia specialist classification was used as the outcome. Initially, a series of univariate logistic regressions was performed in which each item was used as the single predictor of the diagnostic groups (fibromyalgia vs. nonfibromyalgia). Any item whose univariate test had a p value <0.25 was considered a candidate for the multivariate model-building stage of the analysis.

Best subset selection was used to identify candidate screening models based on the reduced item set from the univariate stage. With this method, models containing a prespecified number of items were examined; for a given number of items, the best models according to the likelihood score (chi-square) statistic were further evaluated.

Because it was considered very desirable to use binary cutoffs for any items that had multiple categories or were essentially continuous (greatly simplifying scoring for an end user), receiver operating characteristic (ROC) curves were generated from the univariate logistic regressions for such variables. These curves graphically display the sensitivity as a function of (1-specificity) for the entire range of possible scores. Points at which the sum of sensitivity and specificity were maximized were identified, and binary variables using these cutoffs were used in place of the original multicategory data.

The investigators reviewed the results from the model-building steps to refine further the set of items based on clinical relevance and ease of obtaining the required information. Based on these results, final scoring rules were established.

Results

Fibromyalgia Diagnostic Screen development

The patient-rated questionnaire items of the preliminary Fibromyalgia Diagnostic Screen are listed in Table 1. The first question contained 12 symptom items drawn from previous work that included patient focus groups and clinician and patient Delphi exercises, which resulted in a ranking of the most common and troublesome clinical domains associated with fibromyalgia.^18,19 In the second question, patients were asked to indicate the site and severity of pain symptoms on a pain scale that included 19 body regions found to best distinguish fibromyalgia from other rheumatologic disorders.²⁰ The third question included items on difficulties with work, family, social, and leisure activities to assess functional impairment. The fourth question collected information about pain duration and impact of physical activity on pain. These questions aimed to identify patients with long duration of pain, which is characteristic of fibromyalgia. In addition, fibromyalgia pain typically gets worse with physical activity, which differentiates it from some other rheumatologic disorders, such as spondyloarthropathy, which typically improves with exercise.¹³ Finally, because some of the symptoms of fibromyalgia may be confused with early manifestations of hepatitis (particularly hepatitis C), patients were queried on risk factors for hepatitis.

Table 1.

Preliminary Fibromyalgia Diagnostic Screen: Patient-rated Questionnaire Items

1. Symptoms rated over the past week on a scale of 0 (none), 1 (mild), 2 (moderate), 3 (severe), 4 (very severe)
Aches or pain	Sleep problems	Attention or concentration problems
Stiffness	Unrefreshing sleep	Sadness or depression
Tenderness to touch	Tiredness or fatigue lasting ≥ 24 hours after minimal activity	Anxiety or worry
Tiredness or fatigue	Memory problems or forgetfulness	Irritable bowel problems (e.g., diarrhea, constipation, or cramping)
2. Pain location and severity over the past week on a scale of 0 (none), 1 (mild), 2 (moderate), 3 (severe), 4 (very severe)
Low back	Right upper arm	Right lower leg
Neck	Left upper arm	Left lower leg
Right shoulder	Right forearm	Abdomen
Left shoulder	Left forearm	Right jaw
Right hip	Right upper leg	Left jaw
Left hip	Left upper leg	Chest
Upper back
3. Experience over the past week on a scale of 0 (without difficulty), 1 (with a little difficulty), 2 (with some difficulty), 3 (with much difficulty), 4 (unable to do so), or not applicable
Do work or school work
Do household-related work
Participate in family or social activities
Take part in leisure activities
4. Pain history (yes or no)
Duration of pain ≥ 3 months?
Duration of pain ≥ 4 years?
Pain gets worse with physical activity or exercise?
5. Hepatitis screen (yes or no)
History of intravenous drug use?
Multiple sex partners?
Tattoos or body piercing with unsterile instruments?
Blood transfusions before 1992/clotting factors before 1987?

The clinician component of the screen (Table 2) included a brief physical examination designed to address some of the key differential diagnoses in fibromyalgia.⁹ The examination consisted of the following assessments: (1) an abbreviated tender point examination, the sites for which were selected based on previous work by White et al.²¹ that found certain tender point sites to be more predictive of fibromyalgia, (2) a brief joint evaluation, assessing key sites for bilateral swollen joints to help with the identification of inflammatory arthritis, (3) a muscle strength test to identify proximal muscle weakness that might be suggestive of inflammatory myopathies, (4) assessment for the presence of connective tissue disorder signs, and (5) assessment for key differentiating signs or symptoms of multiple sclerosis (MS), a disorder that sometimes has fibromyalgia-like symptoms. Laboratory tests that were recommended for all patients to evaluate other possible causes of symptoms or signs included erythrocyte sedimentation rate (ESR), complete blood count (CBC), comprehensive metabolic panel, and thyroid-stimulating hormone (TSH), all of which were recently recommended for the evaluation of fibromyalgia by a panel of fibromyalgia experts.²² Other laboratory tests were collected based on evidence of signs or symptoms of autoimmune or inflammatory disorders, which has been shown to improve the positive predictive values of the tests.¹⁶ Using the findings from the clinician brief assessment, rheumatoid factor assessment was ordered if the patient exhibited bilateral swollen joints, an antinuclear antibody test was done if the patient exhibited any connective tissue sign, creatine phosphokinase testing was ordered for any proximal muscle weakness, and hepatitis C antibody was determined when the patient had risk factors for hepatitis. If the patient experienced symptoms suggestive of MS, additional follow-up by a neurologist was advised.

Table 2.

Preliminary Fibromyalgia Diagnostic Screen: Physician-rated Assessment Items

1. Tender point assessment (bilateral)

Trapezius muscle (midpoint of upper border)

Supraspinatus muscle (at the origins, above the scapular spine near the medial border of the scapula)

Second rib (at the second costochondral junctions, just lateral to the junctions, on the upper surface)

Gluteal muscle (in the upper outer quadrants of the buttocks in the anterior fold of the muscle)

2. Joint evaluation (bilateral swollen or boggy joints on palpation)

Wrists

Metacarpophalangeal joints (MCPs)

Proximal interphalangeal joints (PIPs)

Elbows

3. Muscle strength evaluation for proximal muscle weakness

Deltoid

Biceps

Triceps

Hip flexors

4. Connective tissue signs

Alopecia

Rash (malar, photosensitive, or discoid)

Oral or nasal ulcers

Digital pits or ulcers

5. Multiple sclerosis (MS) signs or symptoms

Visual changes (unilateral partial or complete loss, double vision)

Ascending numbness in a leg or bandlike truncal numbness

Slurred speech (dysarthria)

6. Screen laboratory tests

If any bilateral swollen joints: check rheumatoid factor (RF)

If any connective tissue signs: check antinuclear antibody (ANA)

If any proximal muscle weakness: check creatine phosphokinase (CPK)

If any risk factor for hepatitis: check hepatitis C antibody

If any sign or symptom of MS: consider further MS workup

For all patients: comprehensive metabolic panel, complete blood count (CBC), thyroid-stimulating hormone (TSH), erythrocyte sedimentation rate (ESR)

Fibromyalgia Diagnostic Screen testing

Patient disposition and characteristics

Of 104 patients who were approached to participate and met criteria for the study, 100 completed the patient and physician questionnaires and underwent the follow-up evaluation by the fibromyalgia specialist. Forty-five of the patients were diagnosed with fibromyalgia, 43 of whom also met the ACR criteria for fibromyalgia. Table 3 summarizes the demographic data for the fibromyalgia and nonfibromyalgia patients. There were no significant differences on demographic variables between the fibromyalgia and nonfibromyalgia subgroups. The fibromyalgia specialists identified several causes of pain among the 55 nonfibromyalgia patients, many of whom had multiple pain disorders, including osteoarthritis (n=20, 36%), migraine or chronic headache (n=12, 22%), degenerative disc disease (n=5, 9%), tendonitis or tenosynovitis (n=5, 9%), peripheral neuropathy (n=4, 7%), chronic low back pain (n=4, 7%), joint injuries (n=4, 7%), plantar fasciitis (n=3, 5%), disc herniation (n=2, 4%), myofascial pain (n=2, 4%), hypothyroidism (n=2, 4%), and rheumatoid arthritis (n=2, 4%).

Table 3.

Demographic Characteristics of Patients

Characteristic n	Fibromyalgia 45	Nonfibromyalgia 55
Sex
Female, n (%)	43 (95.6)	47 (85.5)
Age, years
Mean (SD)	51.2 (10.1)	52.4 (12.6)
Race
White (non-Hispanic/Latino), n (%)	36 (80)	44 (80)
Black (non-Hispanic/Latino), n (%)	7 (15.6)	10 (18.2)
Black (Hispanic/Latino), n (%)	1 (2.2)	0
Hispanic/Latino, n (%)	1 (2.2)	0
> 1 race, n (%)	0	1 (2.2)
Education
Less than high school graduate, n (%)	1 (2.3)	0
High school graduate or GED, n (%)	10 (22.7)	4 (7.3)
Partial college, n (%)	14 (31.8)	18 (32.7)
College graduate, n (%)	8 (18.2)	16 (29.1)
Graduate degree, n (%)	11 (25.0)	17 (30.9)
Marital status
Married, n (%)	21 (46.7)	36 (65.5)
Never married, n (%)	8 (17.8)	7 (12.7)
Divorced, n (%)	13 (28.9)	9 (16.4)
Separated, n (%)	2 (4.4)	0
Widowed, n (%)	1 (2.2)	3 (5.5)
Vocational status
Homemaker, n (%)	0	1 (1.8)
Unemployed (not due to illness), n (%)	2 (4.4)	2 (3.6)
Unemployed (due to illness/disabled), n (%)	11 (24.4)	7 (12.7)
Employed part-time, n (%)	4 (8.9)	9 (16.4)
Employed full-time, n (%)	21 (46.7)	26 (47.3)
Retired, n (%)	6 (13.3)	10 (18.2)
Other, n (%)	1 (2.2)	0

GED, general educational development; SD, standard deviation.

Other disorders associated with pain that affected single nonfibromyalgia patients were local muscle strain, radiculopathy, spinal stenosis, Arnold-Chiari malformation, vertebral fracture, scoliosis, chondromalacia, mixed connective tissue disease, polymyositis, chronic abdominal pain status post bypass surgery, reflex sympathetic dystrophy, MS, myotonia, pseudotumor cerebri, temporomandibular disorder, foot overpronation, and foot rash. Twenty-one of the 45 fibromyalgia patients had comorbid pain disorders, including osteoarthritis (n=9, 20%), migraine (n=6, 13%), degenerative disc disease (n=3, 7%), spinal stenosis (n=2, 4%), tendonitis (n=2, 4%), and the following disorders that occurred in single individuals with fibromyalgia: peripheral neuropathy, sciatica, rotator cuff tear, reflex sympathetic dystrophy, Morton's neuroma, Arnold-Chiari malformation, temporomandibular disorder, disc herniation, and chronic abdominal pain status post bowel obstruction. Two patients with rheumatoid arthritis who were also diagnosed with fibromyalgia were not included in the analysis because the clinician screen examination specifically evaluates for joint swelling characteristic of rheumatoid arthritis.

Development of scoring rules for final Fibromyalgia Diagnostic Screen patient questionnaire

Univariate analysis showed that two of the symptom history items, the duration of pain ≥ 3 months and pain that gets worse with physical activity or exercise, were significantly predictive of fibromyalgia. Also, all the symptom domain items except irritable bowel problems, stiffness, and sleep problems were significantly (p<0.05) associated with fibromyalgia. To reduce the symptom item information to a manageable single index, we used best subset selection. The following items formed the best item set and captured key symptom domains without redundancy: tenderness, tiredness, unrefreshing sleep, memory problems, sadness/depression, and anxiety/worry. The sum of the intensity of these six items was then computed, and this total score was again used as a univariate predictor of diagnosis. From ROC analysis of the resulting regression, a threshold of 8 (of a possible 24) was found to maximize the sum of sensitivity and specificity.

We analyzed various thresholds based on the number and severity of pain locations, but this approach did not necessarily capture the concept of widespread pain, as it was possible that a number of painful regions could be clustered into one region of the body and not represent widespread pain that is characteristic of fibromyalgia. The approach that best captured the concept of widespread pain was to divide the body into five body regions (axial, right upper body, right lower body, left upper body, and left lower body) and eliminate three sites: abdominal pain and right and left jaw pain. We then needed to determine a threshold level of pain within one or more locations within each region to count that region as positive. We found that sensitivity and specificity were best balanced by setting the threshold for each region liberally, with only mild pain in at least one location within the region being sufficient to count the region as positive. Then from a maximum of five regions, we tested various cutoffs for the number of regions that should constitute widespread pain. The presence of at least mild pain in at least three of the five regions was found to be most predictive of fibromyalgia.

From the statistical results in combination with clinical considerations, we evaluated a set of models that involved only patient-reported pain location data, symptom history, and symptom domains. The final model from these sources of information uses the following decision rule (Fibromyalgia Diagnostic Screen-Patient): Fibromyalgia is predicted to be present if the patient reports pain for ≥ 3 months duration, pain that gets worse with physical activity or exercise, at least mild pain in locations found in at least three of the five body regions (regardless of which particular regions are affected), and a sum of intensities of eight or more over the set of six key symptom domains. Table 4 presents the sensitivity, sensitivity, and accuracy of the Fibromyalgia Diagnostic Screen-Patient.

Table 4.

Accuracy of Proposed Models of Fibromyalgia Diagnostic Screen in Predicting Fibromyalgia Diagnosis

Model	Sensitivity	Specificity	Sensitivity + Specificity	Accuracy
Fibromyalgia Diagnostic Screen -Patient	78%	78%	1.56	78%
Fibromyalgia Diagnostic Screen -Patient and negative clinician examination (all components)	46%	87%	1.33	69%
Fibromyalgia Diagnostic Screen -Patient and 2 of 8 positive tender points	68%	87%	1.55	79%
Fibromyalgia Diagnostic Screen -Patient and negative joint examination	76%	81%	1.57	79%
Fibromyalgia Diagnostic Screen -Patient and 2 of 8 positive tender points and negative joint examination	71%	85%	1.56	79%
Fibromyalgia Diagnostic Screen -Patient and all laboratory results normal	63%	83%	1.46	74%
Fibromyalgia Diagnostic Screen -Patient and ESR <40	78%	81%	1.59	80%
Fibromyalgia Diagnostic Screen -Patient and TSH<1.5 ULN	73%	81%	1.55	78%
Fibromyalgia Diagnostic Screen -Patient and ESR <40 and TSH<1.5 ULN	73%	82%	1.56	78%
Fibromyalgia Diagnostic Screen -Patient and 2 of 8 positive tender points and negative joint examination and ESR <40 and TSH<1.5 ULN	61%	92%	1.53	77%

ESR, erythrocyte sedimentation rate; TSH, thyroid-stimulating hormone; ULN, upper limits of normal.

Supplemental models with the addition of clinician-rated items and laboratory tests were subsequently tested to determine how these items affected the accuracy of the Fibromyalgia Diagnostic Screen (Table 4). The addition of clinician examination items and selected laboratory tests improved the specificity of the screen but decreased sensitivity in most models. The addition of the tender point examination and the joint examination to the Fibromyalgia Diagnostic Screen-Patient resulted in a slight improvement in the overall accuracy. Of the laboratory tests, the ESR added to the Fibromyalgia Diagnostic Screen-Patient also slightly improved the overall accuracy of the screen.

The Appendix presents the final items for the final Fibromyalgia Diagnostic Screen-Patient as well as the supplemental clinician examination items and laboratory tests (Fibromyalgia Diagnostic Screen-Clinician) that can be added to the Fibromyalgia Diagnostic Screen-Patient.

Discussion

The Fibromyalgia Diagnostic Screen is a screening tool that was developed and tested for use in primary care settings to facilitate the assessment of patients with fibromyalgia. The screen has two components: a brief, patient-rated questionnaire (Fibromyalgia Diagnostic Screen-Patient) and an abbreviated clinician examination and laboratory tests (Fibromyalgia Diagnostic Screen-Clinician) (Appendix). The Fibromyalgia Diagnostic Screen-Patient had good sensitivity and specificity for fibromyalgia when used to assess patients coming to primary care clinicians with pain lasting > 30 days. The addition of the clinician items improved the specificity of the screen somewhat but only slightly improved the overall accuracy of the screen. Therefore, primary care clinicians may elect to use the patient-rated question alone in their initial assessment of patients with pain who may have fibromyalgia in order to take advantage of the higher sensitivity of the patient screen. The clinician may add the examination and laboratory items as desired to guide the differential diagnosis of fibromyalgia. The operating characteristics of the Fibromyalgia Diagnostic Screen with and without components of the clinician examination will be evaluated further in a follow-up study that will also compare the screen to the recently published 2010 Wolfe criteria for fibromyalgia, which eliminated the tender point examination and combined measures of widespread pain and symptom severity.¹²

Several limitations related to the development of the Fibromyalgia Diagnostic Screen should be considered. First, there is no objective gold standard for the diagnosis of fibromyalgia. The designation of fibromyalgia or nonfibromyalgia in this study was determined by the clinical judgment of the fibromyalgia specialists and based on best evidence available, including information from the patient history and examination. The reliance on the specialist review as the diagnostic standard inevitably introduces imprecision in the screening process. In addition, although the specialists were blinded to the responses on the questionnaires, they were not blinded to the content of the questionnaires, which may have inadvertently influenced the specialist examination.

Second, many patients with fibromyalgia have comorbid pain disorders, which complicated the development of the screen and the evaluation of patients. In the present study, we excluded 2 patients from analysis who were diagnosed with comorbid rheumatoid arthritis because the clinician screening examination included an assessment of joint swelling, characteristic of rheumatoid arthritis. In practice, the presence of a condition such as rheumatoid arthritis would not necessarily exclude fibromyalgia.²² Third, for the sake of brevity, we did not present all the possible models we tested for the screen, and it is possible that we did not consider all possible combination of items in the models we tested. Fourth, our selection of the items for the preliminary screen may have missed other, more important differentiating characteristics of fibromyalgia. However, we based our selection of items on available data on important clinical domains of fibromyalgia. Finally, our patient sample was drawn from a single university-based primary care practice. The performance of the screen in a larger, more diverse primary care population is unknown and will be tested in a follow-up multisite study.

Conclusions

This study provides evidence for the validity of the Fibromyalgia Diagnostic Screen, a tool capable of accurately screening for fibromyalgia in primary care patients with pain duration > 30 days. Based on the findings of the study, the Fibromyalgia Diagnostic Screen-Patient has good sensitivity and specificity and may be used alone to screen patients for fibromyalgia. Clinician-rated items may be added to the patient-rated screen to aid in the evaluation of patients if desired by the clinician. This screen tool may help to increase awareness of fibromyalgia and facilitate the identification of patients with fibromyalgia in primary care settings. Additional studies are needed to validate the instrument further in more diverse settings and to assess the impact of using the Fibromyalgia Diagnostic Screen on primary care clinician practices and patient outcomes.

Footnotes

Acknowledgments

We thank Mark Bibler, M.D., and Greg Kennebeck, M.D., who completed the physician portion of the screen on primary care patients from their outpatient practice affiliated with the University of Cincinnati. We acknowledge our research staff, Megan Clayton Stovall and Elizabeth N. Mariutto at the University of Cincinnati, Cincinnati, Ohio. Finally, we thank the patients for their participation in this study. This work was supported by an Investigator-Initiated Research grant from Pfizer, Inc.

Disclosure Statement

L.M.A. has received consulting fees from Astra Zeneca, Cypress Bioscience, Inc., Eli Lilly and Company, Forest Laboratories, Inc., Grunenthal, Johnson & Johnson, Sanofi Aventis, Takeda, and Pfizer, Inc., and has received research support from Boehringer Ingelheim, Cypress Bioscience, Inc, Eli Lilly and Company, Forest Laboratories, Inc., Novartis, and Pfizer, Inc. L.J.C. has received research support from Pfizer, Inc.

Appendix: The Final Fibromyalgia Diagnostic Screen

Scoring of the Fibromyalgia Diagnostic Screen-Patient:

Positive screen for fibromyalgia if Yes to all of the following:
1. At least mild pain in at least 1 site within at least 3 out of 5 areas of the body	Yes	No
2. Duration of pain 3 months or longer	Yes	No
Pain gets worse with physical activity or exercise	Yes	No
3. Sum of 8 or more in symptom severity	Yes	No

References

Collett

, Berkley

. Task Force on Fact Sheets for the Global Year Against Pain. Pain in women, 2007/8. Pain, 2007; 132,Suppl 1:S1–S2.

Greenspan

, Craft

, LeResche

et al. Studying sex and gender differences in pain and analgesia: A consensus report. Pain, 2007; 132,Suppl 1:S26–S45.

Fillingim

, King

, Ribeiro-Dasilva

, Rahim-Williams

, Riley

3rd . Sex, gender, and pain: A review of recent clinical and experimental findings. J Pain, 2009; 10:447–485.

White

, Harth

. Classification, epidemiology, and natural history of fibromyalgia. Curr Pain Headache Rep, 2001; 5:320–329.

Wolfe

, Ross

, Anderson

, Russell

, Hebert

. The prevalence and characteristics of fibromyalgia in the general population. Arthritis Rheum, 1995; 38:19–28.

Goldenberg

, Burckhardt

, Crofford

. Management of fibromyalgia syndrome. JAMA, 2004; 292:2388–2395.

Choy

, Perrot

, Leon

et al. A patient survey of the impact of fibromyalgia and the journey to diagnosis. BMC Health Serv Res, 2010; 10:102.

Hughes

, Martinez

, Myon

, Taïeb

, Wessely

. The impact of a diagnosis of fibromyalgia on health care resource use by primary care patients in the UK: An observational study based on clinical practice. Arthritis Rheum, 2006; 54:177–183.

Arnold

. The pathophysiology, diagnosis, and treatment of fibromyalgia. Psychiatr Clin North Am, 2010; 33:375–408.

10.

Wolfe

, Smythe

, Yunus

et al. The American College of Rheumatology 1990 criteria for the classification of fibromyalgia: Report of a multicenter criteria committee. Arthritis Rheum, 1990; 33:160–172.

11.

Vanderschueren

, Van Wambeke

, Morlion

. Fibromyalgia: Do not give up the tender point count too easily: Comment on the article by Wolfe et al. Arthritis Care Res, 2010; 62:1675author reply 1676–1678.

12.

Wolfe

, Clauw

, Fitzcharles

et al. The American College of Rheumatology preliminary diagnostic criteria for fibromyalgia and measurement of symptom severity. Arthritis Care Res, 2010; 62:600–610.

13.

Fitzcharles

, Esdaile

. The overdiagnosis of fibromyalgia syndrome. Am J Med, 1997; 103:44–50.

14.

Fitzcharles

, Boulos

. Inaccuracy in the diagnosis of fibromyalgia syndrome: Analysis of referrals. Rheumatology, 2003; 42:263–267.

15.

Zih

, Da Costa

, Fitzcharles

. Is there a benefit in referring patients with fibromyalgia to a specialist clinic? J Rheumatol, 2004; 31:2468–2471.

16.

Suarez-Almazor

, Gonzalez-Lopez

, Gamez-Nava

, Belseck

, Kendall

, Davis

. Utilization and predictive value of laboratory tests in patients referred to rheumatologists by primary care physicians. J Rheumatol, 1998; 25:1980–1985.

17.

Portenoy

. for the ID Pain Steering Committee. Development and testing of a neuropathic pain screening questionnaire: ID Pain. Curr Med Res Opin, 2006; 22:1555–1565

18.

Arnold

, Crofford

, Mease

et al. Patient perspectives on the impact of fibromyalgia. Patient Educ Couns, 2008; 73:114–120.

19.

Mease

, Arnold

, Choy

et al. Fibromyalgia syndrome module at OMERACT 9: Domain construct. J Rheumatol, 2009; 36:2318–2329.

20.

Wolfe

. Pain extent and diagnosis: Development and validation of the regional pain scale in 12,799 patients with rheumatic disease. J Rheumatol, 2003; 30:369–378.

21.

White

, Harth

, Speechley

, østbye

. A general population study of fibromyalgia tender points in noninstitutional adults with chronic widespread pain. J Rheumatol, 2000; 27:2677–2682.

22.

Arnold

, Clauw

, McCarberg

. for the FibroCollaborative. Improving the recognition and diagnosis of fibromyalgia. Mayo Clin Proc, 2011; 86:457–464.