ORIGINAL

Introduction: Studies have suggested using thermography as a resource to diagnose fibromyalgia, although there has been no evidence confirming this hypothesis so far. Objective: To evaluate the sensitivity and specificity of computerized infrared thermography as an auxiliary method for diagnosing fibromyalgia. Methods: It is a diagnostic accuracy studywith cross-sectional design. One hundred and three individuals were evaluated for global pain using the Visual Analogue Scale. The measurement of pain at tender points was assessed by algometry, and skin temperature was assessed by thermography. To evaluate sensitivity and specificity, the analysis was performed using the Receiver Operating Characteristic Curve, measured by the area under the curve with their respective confidence intervals. Results: Thermography has not been very sensitive or specific for pain (tender points) and diagnosis of fibromyalgia, according to the evaluation of the Receiver Operating Characteristic Curve, with an area under the curve equal to or lower than 0.75. Conclusion: In this study the thermography was not sensitive and specific as a tool for diagnosing the fibromyalgia syndrome. This study highlights important clinical implications concerning the current methods for diagnosing it, which, despite all efforts, are still subjective and poorly reproducible.


INTRODUCTION
Fibromyalgia syndrome (FMS) is a chronic neurobiological disorder of musculoskeletal and connective tissue disorder, common at any age and sex. It has undefined complex multifactorial etiopathogenesis with gradual acceptance of its validity still questioned. It is difficult to be diagnosed and it is clinically underestimated, constituting a scientific and clinical challenge as an enigmatic puzzle 1,2 . Although people of all ages and both sexes may develop it, women are more likely to have it, and almost 50% of them are aged between 35 and 44. Scientific literature shows FMS prevalence values in the general population between 0.2 and 0.6%, and 2.4 and 6.8% in women 1-8 . FMS is characterized by concomitant symptoms, whose cardinal sign is generalized (pain in four, out of the five regions of the body), persistent musculoskeletal pain (non-inflammatory), associated with fatigue (physical exhaustion), sleep disorders and cognitive problems, with frequent occurrence of additional somatic and psychological symptoms, resulting in widespread damage to quality of life 1,3,4 . High levels of disability are self-reported by patients 5 . Genetic factors, psychosocial variables and environmental stressors also seem to have an influence on FMS [6][7][8] .
Generally, FMS is classified as follows: soft-tissue rheumatism (STR) 9 and it may be primary (when it appears alone) or secondary to other rheumatic diseases 10 ; belonging to chronic overlapping pain conditions (COPCs) 11 ; disorder between the affective spectrum disorders (ASD) 12 ; one of the chronic syndromes with unclear etiology 13 ; among functional somatic syndromes (FSS) and medically unexplained symptoms (MUS) 14 .
Professionals in rheumatology, psychosomatic and analgesic medicine classify FMS as central sensitization, a psychosomatic disorder or a neuropathic pain 8 .
People with FMS also present some abnormalities in the neuroendocrine and autonomic nervous system, neurochemical changes, neuro-immune interactions, with Kumpel  potential involvement in focal neuroinflammatory processes in parts of the diencephalon, contributing to a neurosensitizing action 11,12 .
Although the cause, pathophysiology and mechanisms of FMS are not yet fully understood and are even controversial, the finding of hypersensitivity in multiple stimulation modalities, particularly for unpleasant stimuli in fibromyalgia, suggests that the sensitivity to pain evoked by FMS may be related to an altered hedonic appreciation for sensory stimuli, rather than to peripheral tissue changes 2 . This characterizes FMS as a disorder of pain regulation and central sensitization 3 . This hypothesis is corroborated by brain imaging studies using functional magnetic resonance imaging, in which various disorders of pain processing and regulation (increasing or decreasing pain inhibition) were observed in people with FMS 3 .
Individuals with no identifiable nociceptive input (primary FMS) usually develop regional pain syndrome and, over time, it generalizes. Fibromyalgia syndrome, however, is much more common in individuals with chronic pain attributable to peripheral pain generators, that is, with identifiable nociceptive input (secondary FMS) and, therefore, the peripheral and central aspects of pain should be differentiated 8 .
Diagnostic confounding factors are added, due to the similarity of FMS to other musculoskeletal, neurological, endocrine-metabolic, psychiatric, psychological, and drug-related conditions and symptoms 15 . For example, small fiber neuropathy (the most common neurological disease) underlies 49% of the diseases labeled as fibromyalgia 16 .
Hauser et al. 1 suggest a clinical diagnosis of chronic pain based on a differential assessment in order to provide a more appropriate opinion in relation to FMS.
An accurate diagnosis would be the first decisive step towards more effective care and better treatment results 17 . When patients are recognized based on diagnostic confirmation, both the physician and the patient eliminate a major obstacle to the effective management of the disease. Once the diagnosis has been made, patients report better health satisfaction, less long-term symptoms, and reduced health care use and costs 18 .
In this scenario, using cutaneous Infrared Thermography has contributed to the neuromuscular evaluation of individuals with chronic pain, providing some relevant support in the study of pain. The infrared thermography methodology is non-invasive and the heat offers bidimensional, real-time images without harmful radiation effects. Studies have proposed the use of thermography as a diagnostic resource for fibromyalgia, however, no evidence has been able to confirm this hypothesis so far [19][20][21][22][23] .
As FMS is a polysymptomatic condition, delayed diagnosis is rather common, in addition to excessive testing and inadequate treatment, leading to high costs, such as the ones of some chronic diseases, for instance, diabetes and hypertension. It presents direct costs per patient and indirect ones 24 . Therefore, our aim was to assess the sensitivity and specificity of thermography as an auxiliary method for diagnosing fibromyalgia.

METHODS
It as diagnostic accuracy study with cross-sectional design. Measurement instruments are constant objects of investigations for information validity (measurement) 25 . In this sense, investigations of diagnostic accuracy (precision) may contribute to the assumption of explicit evidence of validation that relate to the potential discriminative and predictive ability of a test. Sensitivity and specificity are two key characteristics of measures of diagnostic accuracy 26 .
Accuracy measures are commonly used in diagnostic instruments, and it is suggested that evaluations of an instrument's discriminating capacity can be enlightening,

Study location and sample
Patients were recruited from the patient registry of a rehabilitation reference center (target population of other studies) 26,27 located in the south of the city of São Paulo, patients with a medical diagnosis of fibromyalgia in the age group of 25 to 60 years and patients without a diagnosis of fibromyalgia matched for age and sex. The sample was initially composed of 103 female individuals, randomized into two groups: Fibromyalgia Syndrome Group, with skeletal muscle pain (n=80), whose individuals had FMS, and Control Group, no history of skeletal muscle pain (n=23), whose individuals were healthy.
Patients answered a questionnaire to check if they met all the inclusion criteria of this study and signed the Informed Consent Form. Age-matched female subjects with a history of generalized musculoskeletal pain for at least 3 consecutive months and pain in at least 11 or more of the 18 tender points and with a clinical diagnosis of fibromyalgia were included in the Fibromyalgia Syndrome Group. In this study, 18 tender points were evaluated, according to the criteria of the American College of Rheumatology.
The study did not include individuals diagnosed with infectious or contagious neurological disease, whose clinical condition evolved into loss of urinary sphincter control; with previous cancer diagnosis; in severe neurological conditions; with acute orthopedic injuries or illnesses; unable to walk; who had undergone previous surgery 2 months before the start of data collection and who had suffered previous acute myocardial infarction 6 months before the start of data collection.
All assessments were carried out at the Exercise Physiology Laboratory at the Polyclinic of the Brazil Adventist University (UNASP). The global pain assessment was performed using the Visual Analogue Scale (VAS) 28 , pain quantification at tender points was assessed using algometry (Pain Test™ FPX Compact Digital Algometer Pain Diagnostic Gage 20Lb x. 25Lb; 10kg X 100Gr, Italy) according to the parameters recommended in literature 29 , and the skin temperature was assessed through thermography.

Procedure and data collection
The collection was carried out in two moments. In the first one, fibromyalgia patients had a clinical consultation to assess the stability and monitoring of the disease, In the second moment, all participants were submitted to pain assessment using the Visual Analogue Scale (VAS). All participants were submitted to assessment of pain perception thresholds through algometry. To perform the algometry, the researcher used a 90° approximation angle (formed between the stimulation surface and the stimulated point).
Algometry was always performed by a single evaluator, who received training before data collection. For evaluation, the participant was instructed to say "stop" as soon https://doi.org/10.7322/abcshs.2021113.2146 In order to avoid thermal variation, the examination room was kept free of air current and exposure to ultraviolet rays, without temperature variations. The room was air-conditioned at a temperature of 23°C±0.5°C and illuminated with cold-light by means of fluorescent lamps, with humidity of 45% and air speed <0.2m/s and, preferably, laminar flow 31 .
The infrared thermography protocol adopted in this research was guided by the guidelines of the American Academy of Thermology (AAT) Neuro Musculoskeletal Thermography, considering the preparation of the participants, the conditions of the examination room and the process of capturing and recording images 32 . The images and results were transformed into a PDF file, filed in a folder for each participant to perform the results spreadsheet. The interpretation of the thermograms was processed and analyzed by a specific software, using the Rainbow colorimetric scale, with a temperature range of 23° to 35°C and an emissivity standard of 0.98.

Data analysis
To assess sensitivity and specificity, the analysis was performed using the ROC curve, indicating the area under the curve (AUC) with its respective confidence intervals (95%). Among the precision indices to summarize the ROC curves, AUC is the most commonly used. The area under the curve summarizes the "global" location of the entire ROC curve and can be interpreted as the probability that an affected individual chosen at random will be classified as more likely to be affected than an unaffected individual chosen at random. It refers to the average sensitivity value for all possible specificity values, being especially useful in a comparative study of two diagnostic instruments 26 .
Data analysis was performed using descriptive statistical techniques with absolute and relative distributions or in mean and standard deviation, as appropriate. Data

RESULTS
The sample initially consisted of 103 individuals, 23 healthy ones without fibromyalgia (Control Group) and 80 with fibromyalgia (Table 1).
Pressure sensitivity and specificity obtained from 18 tender points were analyzed using algometry. For the sensitivity and specificity of the exam to be considered good, the AUC should be greater than 0.75. Table 2 shows that the pressure obtained at the trapezius, epicondyle, trochanter, gluteus medius and medial surface of the knee had AUC greater than 0.75. The most sensitive and specific tender point was the gluteus medius, with AUC of 0.87 and 95% CI (0.79-0.94). Table 3 shows that the pressure obtained at the occipital, epicondyle and medial surface of the knee tender points showed AUC greater than 0.75. The medial surface of the knee and the epicondyle presented similar AUC=0.81.
The sensitivity and specificity of the temperature obtained from the 18 tender points were analyzed using thermography. For the sensitivity and specificity of the exam https://doi.org/10.7322/abcshs.2021113.2146 to be considered good, the AUC should be greater than 0.75. Table 4 shows the temperature obtained at the anatomical points analyzed, it was not possible to obtain an AUC equal to 0.75, with the trapezius being the tender point with the highest AUC (0.66), with a confidence interval of 95 % (95% CI) 0.51-0.81. The trochanter had the lowest AUC 0.48 (95% CI 0.31-0.64). Table 5 shows the temperature obtained at the anatomical points analyzed, it was not possible to obtain an AUC equal to 0.75. The trapezius was the tender points with the highest AUC 0.68, with 95% CI (0.51-0.85). The epicondyle had a lower AUC (0.51), with 95% CI 0.32-0.69.

DISCUSSION
The aim of this study was to analyze the sensitivity and specificity of thermography for auxiliary use as a diagnostic method for fibromyalgia. The main result of this study showed that the thermography was not very sensitive or specific for the diagnosis of fibromyalgia. For the sensitivity and specificity of the test to be considered Brioschi et al. 33 claim that 87% of patients with chronic pain do not have an anatomical substrate demonstrable by routine imaging exams to explain their pain, which corresponds, in most cases, to dysfunctions of the neuro-musculoskeletal system, justifying the use of other diagnostic and imaging resources in the study of the patient's pain and in their clinical and therapeutic direction, such as infrared imaging.
Making the diagnosis of fibromyalgia has been shown to be still much discussed.
In 1990, the ACR approved the criteria for fibromyalgia, as well as for its classification 34 .
In 1990 and 2010, the ACR approved and updated criteria for fibromyalgia as well as for its classification 34 . The set of criteria has been quantitatively validated using patient data but has not yet undergone validation based on a set of external data, so these criteria are subject to updates. It is noteworthy that in 2015, the ACR started to provide approval only for classification criteria, no longer considering the financing or approval of diagnostic criteria" 30 . In this study, a difference in skin temperature was found between control subjects and those with fibromyalgia in most tender points. The biggest AUC was for the trapezius.
However, for the diagnosis of manifested pain, AUC was more expressive.
Thermography has already been shown to be sensitive to expose skin temperature variation and can be an important method in detecting temperature variations in tender points of patients with fibromyalgia. This is because in these anatomical points, temperature variation can occur both for increase and decrease, since, if an inflammation due to muscle tension occurs, it can result in an increase in temperature, or, if there is a reduction in blood flow due to muscle spasm, this may cause a reduction in temperature 36 .
Lima et al. 37 carried out a study with 50 female patients, between 20 and 83 years old, with reports of chronic pain, having the whole body exam done, which brought Another study evaluating the back temperature of patients with unilateral muscle spasm using thermography showed that there is a significant temperature difference between individuals who experience pain in one hemisphere of the back in relation to control subjects and the contralateral hemisphere. The authors attribute this temperature difference to the vasomotor condition 38 .
In this sense, evaluating the temperature sensitivity of patients with fibromyalgia compared to controls, patients immersed the forearm in cold water at 1ºC, with a significant difference in relation to core body temperature in relation to peripheral temperature of the forearm. The temperature among patients with fibromyalgia was significantly lower and had a shorter tolerability time. There was a correlation between the previous pain and the temperature after the procedure 23 . Gomes et al. 38 claim that thermography has high specificity, and the important thing is a high sensitivity to identify chronic pain. The IR image does not demonstrate the presence of pain, but the vasomotor changes in the same projection areas. For this reason, it must be preserved as part of information to always be integrated with other clinical data. https://doi.org/10.7322/abcshs.2021113.2146 Apparently, although some studies have shown this difference in skin temperature between fibromyalgia patients and control subjects, in this study, thermography has not proved to be a fully viable method to be applied in the auxiliary diagnosis of fibromyalgia.
The clinical limitations of this study refer to the fact that only an assessment of skin temperature was performed. It is possible that, if more evaluations were carried out, the results could be different. On the other hand, the present study has important clinical applications, mainly for the current diagnostic methods of fibromyalgia, which are still subjective and not very reproducible, when checking other technologies for diagnostic aid.

Conclusion
In this study, no statistically significant differences were found between the control groups and those with fibromyalgia syndrome, with no consistent information supporting the use of thermography as a diagnostic tool for fibromyalgia syndrome. It helps to explain a still important technological gap for a more objective diagnosis of fibromyalgia. https://doi.org/10.7322/abcshs.2021113.2146   Table 3: Area under the pressure curve obtained at the tender points, with the pain manifested by the patient with fibromyalgia *significant at p<0.05