Infrared skin temperature measurement cannot be used to detect myofascial tender spots

902

Infrared Skin Temperature Measurement Cannot Be Used to Detect Myofascial Tender Spots Mohan Radhakrishna, MD, Robert Burnham, MD, MSc ABSTRACT. Radhakrishna M, Burnham R. Infrared skin temperature measurement cannot be used to detect myofoscial tender spots. Arch Phys Med Rehabil 2001;82:902-5. Objective: To determine the relationship between skin temperature and pressure tolerance in patients with myofascial pain. Design: Blinded, criterion standard. Setting: Community physiatry clinic. Patients: Sixteen consecutive female patients with myofascial pain or fibromyalgia with shoulder girdle symptoms above the T4 level for at least 3 months. No patient met the exclusion criteria of recent trauma to the area or therapy within 48 hours. Interventions: Skin temperature was measured by using a hand-held infrared thermometer over 36 points arranged in a grid on the upper and midtrapezius. Pressure threshold was then assessed at each point by using a pressure threshold meter. A second, blinded examiner then examined each patient to find any myofascial tender spots and noted within which square on the grid they occurred. Main Outcome Measures: The correlation between temperature and pressure threshold and the temperature differences between tender and nontender areas. Results: A nonsignificant correlation of .023 (p ⫽ .57) was found between temperature and pressure threshold. The mean temperature of the tender spots was 32.1°C. No significant difference existed between tender spot temperature and temperature of nontender points (32.1°C, p ⫽ .653) or contralateral points (32°C, p ⫽ .893). Conclusions: Skin temperature, measured with a hand-held infrared thermometer, cannot be used to diagnose and follow treatment progress of myofascial tender spots, because skin temperature over tender spots does not correlate with pressure sensitivity. Key Words: Myofascial pain syndromes; Pressure; Rehabilitation; Skin temperature. © 2001 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation YOFASCIAL PAIN SYNDROME (MPS) is a poorly recognized clinical entity that is often made more comM plex by litigation or insurance claims. This situation makes it

From the Division of Physical Medicine and Rehabilitation, Department of Medicine, University of Alberta, Edmonton, Alb, Canada. Accepted in revised form August 7, 2000. Supported by the University of Alberta Orthopedic Research Fund. Presented as a poster at the Cervical Spine Research Society Meeting, Seattle, WA, December 16 –18, 1999. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated. Reprint requests to Mohan Radhakrishna, MD, Dept of Physical Medicine and Rehabilitation, Glenrose Rehabilitation Hospital, 10230 111 Ave, Edmonton, Alb, T5G 0B7 Canada. 0003-9993/01/8207-6195$35.00/0 doi:10.1053/apmr.2001.23875

Arch Phys Med Rehabil Vol 82, July 2001

important to develop objective, ideally noninvasive, methods of MPS diagnosis. The diagnosis involves trigger points— tender, hyperirritable, taut bands of muscle fibers capable of referring pain elsewhere. Diagnosis of soft-tissue pain has been greatly complicated by the plethora of related terms: fibromyositis, myositis chronica, chronic rheumatism, and myofascitis, among others, as outlined by Travell and Simons.1 Difficulty with terminology and poorly defined diagnostic criteria have made experimental and clinical comparisons problematic. Furthermore, use of such terms as tender spots, tender points, and active and latent trigger points has confounded the process of defining MPS. Studies2-4 indicate that the prevalence of myofascial trigger points in patients with a regional pain complaint range from 30% to 93%. Although the methods and definitions in these studies were inconsistent, the prevalence of myofascial trigger points in all 3 investigations is impressive. Unfortunately, detection of these points relies only on clinical examination— the effectiveness of which is quite variable and depends on the training and experience of the examiner.5-7 The invention and validation of a pressure threshold meter has provided a means of quantifying (in kg/cm2) the amount of pressure a person can tolerate over a given part of the body.8,9 Although this technique has become the criterion standard for myofascial trigger point detection, it is not completely objective and is limited by patient honesty and consistency. Another objective method that has achieved much attention in detecting trigger points has been thermography. This approach stems from the belief that autonomically mediated vasodilatation forms a component of trigger point pathology. The science of thermography uses infrared cameras to capture body surface heat emission and thus measures skin temperature. Uematsu et al10,11 and Weinstein and Weinstein12 have been the investigators in a number of studies showing temperature symmetry. In Uematsu’s11 1988 survey of 90 healthy volunteers and 40 body regions, he found thermal asymmetry of less than 0.4°C in all areas except the fingers and toes. Between-side deviance beyond this limit is thought to be abnormal and to indicate a pathologic process. Proponents of thermography cite its safety and noninvasiveness as key attributes. According to Fischer and Chang,13 skin overlying trigger points is 1°C warmer than its asymptomatic contralateral side. Several studies and reviews13-16 have pointed out the use of thermography in finding trigger points and have supported the theory that trigger points are warmer than nontender areas. Nonetheless, many investigators17,18 have questioned its accuracy— partly because it has been difficult for thermographers to delineate standards of practice.19 Much of the research has been criticized20 for improper and vaguely described design, small numbers of cases and control subjects (if any), and a lack of blinding of the interpreters. In addition, for the office-based physician, thermography can be quite inaccessible. Various types of hand-held skin thermometers have found uses in clinical electrophysiology, and the thermocouple transducer has been shown by Nemcovsky et al21 as being useful for following up progress in patients with temporal mandibular joint pain. The tympanic infrared thermometer used in the present experiment

SKIN TEMPERATURE CANNOT PREDICT MYOFASCIAL TENDER SPOTS, Radhakrishna

903

provides more reliable and comparably valid measures of skin temperature than a plate thermister or skin infrared thermometer; its validity was high as assessed by the correlation with intramuscular temperature.22 Our hypothesis was that this hand-held thermometer could be used to measure skin temperature over trigger points and to determine if these areas are indeed warmer than nonpathologic zones. The temperatures were compared with the pressure thresholds over a grid-delineated area of the trapezius. Trigger points were determined by a second, blinded examiner by using the following criteria: pain on gentle pressure and presence of the “jump sign.” These responses were chosen because they are the most replicable23 and would therefore include the most points. In accordance with Fischer,24 we called the trigger points tender spots in the present article. A positive result would give physicians a new, inexpensive, objective, office-based assessment tool. Because this study is, to our knowledge, the first blinded, prospective trial involving temperature and myofascial pain, we believe that negative results would show that skin temperature measurement likely has no role in myofascial pain diagnosis. METHODS Subjects The present experiment was conducted at a community physiatry clinic with subjects recruited from the clinic population. The University of Alberta Health Research Ethics Board approved the protocol and consent form before the study started. During a 1-month period (March 1999), each person having a prior diagnosis of myofascial pain or fibromyalgia was approached about enrolling in the study. One investigator (RB) interviewed the subjects and confirmed their eligibility. Inclusion criteria were the presence of posterior shoulder girdle pain of longer than 3 months’ duration that was not explained by any other pathology from clinical, radiologic, or electrophysiologic examinations. Exclusion criteria were the presence of trauma to the area within 2 weeks or from therapy within 48 hours. Seventeen consecutive patients (16 women, 1 man) meeting these criteria consented to take part in the study. Procedures Each subject donned a hospital gown and sat with the upper back exposed and uncovered with hair or clothing in a draft-

Fig 1. Sample grid over upper back.

Fig 2. Measuring temperature by using skin thermometer.

free, 20°C room for 15 minutes. A grid was drawn on the subject’s upper back with a water-soluble marker. The grid encompassed the upper and middle trapezius, an area known for its preponderance of myofascial trigger points. The borders of the grid were the 7th cervical vertebra superiorly, the acromia laterally, and the superior angles of the scapula inferiorly. Three horizontal rows were created with the lower 2 (rows B, C) being 1 inch high. The superior row, A, comprised the remaining space, usually 0.5 inches. The trapezius was divided into 7 equal squares on each side of midline for rows B and C, and 4 squares for row A (fig 1). A blinded investigator (MR) then took the skin temperature at the center of each square by using an infrared thermometera (fig 2). By using a pressure threshold meter,b the same examiner then measured the pressure threshold of the center of each square. Pressure was increased slowly by using Fischer’s25 technique at a rate of 1kg/s until the subject reported the onset of pain. Before measurements were performed on the back, 1 point on a thenar eminence was used as a practice trial so subjects could feel the pressure threshold meter. Subjects were instructed to let the examiner know when they first began to feel discomfort. In each row, for both temperature and pressure threshold, measurement began at the rightmost square and proceeded from right to left across each row. A second examiner (RB), who was unaware of the pressure threshold or temperature measurements, then examined the subject for the location of any tender spots to ascertain whether clinically detected tender spots were warmer than nontender spots and contralateral points. Tender spots were defined as areas deemed painful by the subject and on which manual pressure elicited the “jump sign.” The squares on the grid where tender spots were found were recorded, but the temperature and painful-pressure measurements for those spots were not repeated. Statistical Analyses Because precedent research in this area was lacking, an exact sample size calculation could not be made. Therefore, anticipated effect size and variability values were not available. An estimated sample size of 85 was determined assuming one would require at least a moderate correlation between skin temperature and soft-tissue tenderness, a statistical power of .80, and an alpha level of .05.26 Given that 36 squares were to be sampled per subject, a minimum of 3 subjects was required. Arch Phys Med Rehabil Vol 82, July 2001

904

SKIN TEMPERATURE CANNOT PREDICT MYOFASCIAL TENDER SPOTS, Radhakrishna

SPSS, version 8.0,c was used for statistical calculations: specifically, the correlation between pressure tolerance and skin temperature for individual and pooled data, the independent 2-tailed t test when comparing tender spot values with the remaining points, and paired t test values for comparing tender spots with their contralateral points. RESULTS Seventeen consecutive subjects agreed to participate in the study. Each had 36 data points corresponding to the number of squares on the trapezius. Data from 1 male subject (fibromyalgia) were excluded because he reported pain at the minimal pressure level at all points on the grid. The remaining 16 women (11 myofascial pain, 5 fibromyalgia) had their data analyzed in 2 ways. First, a correlation was made between temperature and pressure for each subject; this correlation was repeated with all data pooled. Second, the temperatures and pressure thresholds of clinically detected tender spots were compared with their contralateral locations and with the nontender sites as a whole. Individual correlations varied from ⫺.49 to .30 (table 1). For 3 subjects, these correlations were statistically significant, using a probability level of p less than .05. When pooling data, the values could not simply be averaged because skin temperature varied between individuals. For each subject, the minimal temperature and pressure were subtracted from each of the other values to derive relative values. Data from all subjects were then pooled for a total of 576 points. The overall group correlation was .023 (p ⫽ .57). The second aim of the present study was to discover if the 146 tender spots found on clinical examination were warmer than the nontender points. This information was analyzed in 2 ways. First, the temperature of the tender spots was compared with the same location on the contralateral side. When both positions were tender, data were included in both categories. The mean temperature of both the tender and contralateral spots was 32.1°C (p ⫽ .46) (table 2). Importantly, the pressure threshold for tender spots was 1.8kg/cm2, compared with 2.1kg/cm2 for the contralateral side (p ⫽ .001). Although the tender spots were no warmer than

Table 1: Subject Demographics and Correlation Between Temperature and Pressure Threshold Subject

Age (yr)

Side of Symptoms

Diagnosis

Correlation

p

Pooled Data 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

32 40 38 53 30 44 47 35 36 46 36 49 43 38 47 38

Right Left Bilateral Left Left Left Left Left Right Bilateral Bilateral Bilateral Right Right Right Bilateral

MPS MPS FMS MPS MPS MPS MPS MPS MPS FMS FMS FMS MPS MPS MPS FMS

.023 .307 ⫺.357 .058 ⫺.494 .069 ⫺.178 ⫺.118 ⫺.086 .242 ⫺.195 ⫺.070 ⫺.020 .082 ⫺.244 .177 ⫺.360

.57 .07 .03* .74 ⬍.01* .69 .30 .49 .62 .16 .26 .68 .90 .63 .15 .30 .03*

Abbreviation: FMS, fibromyalgia syndrome; MPS, myofascial pain syndrome. * p ⬍ .05. Arch Phys Med Rehabil Vol 82, July 2001

Table 2: Comparison of Mean Temperatures and Pressure Thresholds of Tender and Nontender Spots

Temperature (°C) Pressure threshold (kg/cm2)

Tender Spots (n ⫽ 146)

Nontender Spots (n ⫽ 430)

t

32.1 ⫾ 0.1

32.0 ⫾ 0.1

.449

1.8 ⫾ 0.1

2.2 ⫾ 0.0

⫺.51

p

.653 ⬍.001

NOTE. Data are mean ⫾ standard error of the mean (SEM).

their contralateral counterparts, their identification as tender spots was confirmed with algometry. Tender spots were also compared with the remaining points as a group. The average temperature and pressure threshold of nontender points were 32.1°C and 2.2kg/cm2, respectively. As with the contralateral points, the difference between pressure thresholds was significant (p ⬍ .001) whereas the temperature difference was not (p ⫽ .33) (table 3). DISCUSSION The goal of the present study was to show the relationship, if any, between skin temperature and pressure tolerance and to determine if the temperature of skin overlying myofascial tender spots was warmer than comparable nontender areas. Our pooled pressure threshold and temperature correlation was a nonsignificant .023. When subjects were analyzed individually, the correlation coefficients varied from ⫺.49 to .30. Interestingly, 3 subjects had significant negative correlations, which may imply that tender spots could be cooler than nontender spots in some individuals, a previously unreported finding. These 3 subjects did not appear to be different from the others, either demographically or clinically. There did not appear to be any trend between diagnosis type (fibromyalgia vs myofascial pain) and correlation value. The large number of nontender points should not have negated any potential correlation for the pooled data because, if there truly was a link, one would expect nontender points to be cooler as a group than the tender points. The lack of correlation between temperature and pressure sensitivity indicates that skin temperature cannot be used to follow treatment progress. To ascertain whether tender spots, as found on clinical examination, were more sensitive and/or warmer than their nontender counterparts, 2 analyses were conducted. Compared with their contralateral points and with the nontender points as a whole, tender spots did not have a different temperature but did have a lower pressure threshold. Although small, the 0.3kg/cm2 difference was sufficient to be statistically significant. Fischer16 stated that a difference of 2kg is sufficient for myofascial trigger point diagnosis. The low value found in our subjects can be partially explained by the fact that the lowest measure on the pressure threshold meter was 1.0kg/cm2, and this level was ascribed for many tender areas when, in fact, the

Table 3: Comparison of Temperature and Pressure Tolerance of Tender Spots and Their Contralateral Points

Temperature (°C) Pressure threshold (kg/cm2)

Tender Spots (n ⫽ 146)

Contralateral Spots (n ⫽ 146)

32.1 ⫾ 0.1

32.1 ⫾ 0.1

1.8 ⫾ 0.1

2.1 ⫾ 0.1

NOTE. Data are mean ⫾ SEM.

t

⫺.134

p

.894

⫺4.401 ⬍.001

SKIN TEMPERATURE CANNOT PREDICT MYOFASCIAL TENDER SPOTS, Radhakrishna

actual threshold may have been lower. Furthermore, only pain and the presence of the “jump sign” were necessary for a point to be termed a tender spot. Although it is a modification of the trigger point criteria of Travell and Simons,1 this approach remains common in clinical practice. These criteria were used to ensure that all trigger points were detected (the remaining criteria occur only inconsistently5-7). Lowering the criteria for tender point selection may, however, have increased the pressure threshold of the group as a whole, making it similar to nontender points. The lack of correlation between temperature and pressure tolerance does not agree with the work of Fischer and Chang13 who found that hot spots in 11 patients with unilateral low back pain had a lower pain threshold than the contralateral side. Nor do they jibe with Kruse and Christiansen’s assertion15 that hot spots on myofascial pain sufferers are more sensitive than “cold spots” on asymptomatic volunteers. A 1990 unblinded survey16 involving 50 patients with various diagnoses such as radiculopathy, trigger points, and muscle spasm indicated that hot spots on a thermogram correlated with lower pressure threshold. These discrepancies may exist because different patient populations were studied. The present sample included only women with shoulder girdle pain. In addition, those studies applied thermography first and then tested the pressure threshold over the hot spots and only 1 control point. These hot spots were assumed to match the location of the patient’s pain. Our results concur with those of Sherman et al27 who indicated that many musculoskeletal painful areas were undetectable by thermography. They also conform to Swerdlow and Dieter19 who found that the high false-positive and negative rates of thermography made it impossible to accurately predict trigger points. This is despite the fact that our definition of trigger points was much less stringent than theirs. The pressure threshold of our subjects was dramatically lower (2.1kg/cm2 in nontender locations, n ⫽ 430) than that of Fischer’s asymptomatic group (3.7 ⫾ 1.9kg/cm2),24 implying that patients with a chronic pain condition have a lower pain threshold than able-bodied subjects. This raised perception of pain may also explain the 0.3kg/cm2 pressure threshold difference between tender and nontender points. To our knowledge, the present study is the only study in which a hand-held thermometer was used to measure the temperature of myofascial trigger points. Earlier myofascial research has used thermography to measure skin temperature over an entire region and then to direct attention to any hot spots. Our technique, though labor intensive, correlates well with computerized video thermography28 and with the plate thermistor, the standard office method of assessing skin temperature.22 It may be useful for future research because it combines the accuracy of the plate thermistor with the speed of thermography. CONCLUSIONS The present study indicates that the hand-held infrared skin thermometer cannot be used to identify myofascial tender spots because skin temperature over tender spots does not correlate with pressure sensitivity. References 1. Travell J, Simons D. Myofascial pain and dysfunction: the trigger point manual. Baltimore: Williams & Wilkins; 1983. 2. Fricton JR, Kroening R, Haley D, Siegert R. Myofascial pain syndrome of the head and neck: a review of clinical characteristics of 164 patients. Oral Surg 1985;60:615-23. 3. Skootsky SA, Jaeger B, Oye RK. Prevalence of myofascial pain in general internal medicine practice. West J Med 1989;151:157-60. 4. Gerwin RD. A study of 96 patients examined both for fibromyalgia and myofascial pain. J Musculoskel Pain 1995;3 Suppl 1:121.

905

5. Njoo KH, Van der Does E. The occurrence and inter-rater reliability of myofascial trigger points in the quadratus lumborum and gluteus medius: a prospective study in non-specific low back pain patients and controls in general practice. Pain 1994;58:317-23. 6. Gerwin RD, Shannon S, Hong C-Z, Hubbard D, Gevirtz R. Identification of myofascial trigger points: inter-rater agreement and effect of training [abstract]. J Musculoskel Pain 1995;3 Suppl 1:55. 7. Gerwin RD, Shannon S, Hong C-Z, Hubbard D, Gevirtz R. Inter-rater reliability: myofascial trigger point examination. Pain 1997;69:65-73. 8. Reeves JL, Jaeger B, Graff-Radford SB. Reliability of the pressure algometer as a measure of myofascial trigger point sensitivity. Pain 1986;23:313-21. 9. Delaney GA, McKee AC. Inter and intra-rater reliability of the pressure threshold meter in measurement of myofascial trigger point sensitivity. Am J Rehabil 1993;72:136-9. 10. Uematsu S. Symmetry of skin temperature comparing one side of the body to the other. Thermology 1985;1:4-7. 11. Uematsu S, Edwin DH, Jankel WR, Kozikowski J, Trattner RT. Quantification of thermal asymmetry. J Neurosurg 1988;69:552-5. 12. Weinstein SA, Weinstein G. Computerized electronic thermography in the evaluation of muscle pain. Clin Thermogr 1989;1:38-48. 13. Fischer AA, Chang CH. Temperature and pressure threshold measurements in trigger points. Thermology 1986;1:212-15. 14. Fischer AA. Documentation of myofascial trigger points: a follow-up study. Arch Phys Med Rehabil 1988;69:286-91. 15. Kruse RA Jr, Christiansen JA. Thermographic imaging of myofascial trigger points: a follow-up study. Arch Phys Med Rehabil 1992;73:819-23. 16. Fischer AA. Application of pressure algometry in manual medicine. J Manual Med 1990;5:145-50. 17. American Academy of Neurology, Therapeutics and Technology Assessment Subcommittee. Assessment: thermography in neurologic practice. Neurology 1990;40:523-5. 18. Racz GB, Krane D. Thermography. In: Raj PP, editor. Pain Medicine. St. Louis (MO). Mosby; 1996. p 92-7. 19. Swerdlow B, Dieter JN. An evaluation of the sensitivity and specificity of medical thermography for the documentation of myofascial trigger points. Pain 1992;48:205-13. 20. Awerbuch MS. Thermography—its current status in musculoskeletal medicine [review]. Med J Aust 1991;154:441-4. 21. Nemcovsky CE, Benvenisti A, Gazit E. Variation of skin surface temperature over the masseter muscles in patients with myofascial pain following occlusal splint treatment. J Oral Rehabil 1995;22: 769-73. 22. Burnham R, McKinley R, Vincent D. A comparison of the reliability, validity and responsiveness of three types of skin surface thermometers. Muscle Nerve 1999;22:1306-7. 23. Simons DG. Clinical and etiological update of myofascial pain from trigger points. J Musculoskel Pain 1996;4:93-121. 24. Fischer AA. Algometry in the daily practice of pain management. J Back Musculoskel Rehabil 1997;8:151-63. 25. Fischer AA. Pressure algometry over normal muscles. Standard values, validity and reproducibility of pressure threshold. Pain 1987;30:115-26. 26. Cohen J. Statistical power analysis for the behavioral sciences. rev ed. New York (NJ): Academic Pr; 1977. 27. Sherman RA, Barja RH, Bruno GM. Thermographic correlates of chronic pain: analysis of 125 patients incorporating evaluations by a blinded panel. Arch Phys Med Rehabil 1987;68:273-9. 28. Sherman RA, Woerman AL, Karstetter KW. Comparative effectiveness of videothermography, contact thermography, and infrared beam thermography for scanning relative skin temperature. J Rehabil Res Dev 1996;37:377-86. Suppliers a. First-Temp Genius Instant Clinical Thermometer; Intelligent Medical Systems, Carlsbad Research Centre, 6339 Paseo Del Lago, Carlsbad, CA 92009. b. Pain Diagnostics and Thermography Co, 17 Wooley Ln E, Great Neck, NY 11021. c. SPSS Inc, 233 S Wacker Dr, 11th FI, Chicago, IL 60606. Arch Phys Med Rehabil Vol 82, July 2001