Infared dermal thermometry for the high-risk diabetic foot.

[Armstrong DG, Lavery LA, Liswood PJ, et al. Infrared dermal thermometry for the high-risk diabetic foot. Phys Ther. 1997;77:169-177.]

Key Words: Charcot's arthropathy, Diabetes mellitus, Foot, Neuropathy, Temperature, Ulceration.

Inflammation is characterized by five cardinal signs: edema (tumor), heat (calor), redness (rubor), pain (dolor), and loss of function. Inflammation may be triggered by a host of factors, including soft tissue or bony injury.[1,2] In the neuropathic extremity, pain and disturbance of function may be absent and thus are poor indicators of inflammation. Inflammation plays a central role in the pathogenesis of the two most devastating sequelae of diabetic peripheral neuropathy: neuropathic ulceration and neuropathic osteoarthropathy (Charcot's joint).[3,4] Although edema and erythema are often present, they are difficult to grade. Infrared dermal thermometry provides an inexpensive means of measuring local skin temperatures to evaluate inflammation in the diabetic foot.[5]

A series of experiments performed at the National Hansen's Disease Center in Carville, La, linked repetitive mild to moderate stress on the insensitive limb with a generalized increase in local skin temperature.[6,7] Beach and Thompson[6] and later Manley and Darby[7] demonstrated that repetitive stress of 3.6 kg/[cm.sup.2] applied to the sole of a rat's foot over the span of approximately 1 week caused an inflammatory response and subsequent ulceration. A controlled repetitive stress experiment performed on the human finger revealed that initially pain became intolerable at approximately 1,000 repetitions of 4.5 kg/[cm.sup.2]. Surface temperatures showed a 5[degrees] F difference between the stressed digits and the adjacent nonstressed digits. The same experiment was repeated 24 hours later. During the second day of testing, pain developed sooner with fewer repetitions. Additionally, skin temperatures were higher, and these temperatures persisted for a longer duration.[8]

The potential clinical utility of foot skin temperature monitoring has been investigated by several authors. Chan and colleagues[9] demonstrated that patients with diabetes and painful neuropathy had higher forefoot skin temperatures than control subjects without diabetes. Stess et all[10] and later Benbow et al[11] suggested that by examining thermographic patterns, patients with diabetes could be screened for risk of ulceration and that high temperatures were predictive of ulceration. These studies concentrated their efforts on the affected extremity. In addition, the devices used for thermometric assessment were bulky, expensive, and, in general, difficult to use or integrate into a typical clinical setting. There are no well-established criteria or discrete temperature levels to determine pathology. Therefore, as a practical application of skin temperature assessment, we have used the contralateral (de, nonaffected) extremity to determine normal or baseline temperature levels. We expect that local skin temperatures will be higher in extremities with pathology (neuropathic ulcers, acute Charcot's arthropathy) and the same in patients without pathology when compared with the contralateral foot. Furthermore, we expect temperatures to return to normal once ulcer or fracture healing is complete. In patients with recurrence of injury, local temperatures will increase prior to reinjury. The purpose of this study was to compare skin temperatures in patients with asymptomatic peripheral sensory neuropathy, patients with new-onset neuropathic ulcers, and patients with Charcot arthropathy at the time of injury, during the course of healing, and prior to reinjury in patients with ulcer recurrence. In each comparison, one foot was compared with the other foot, with the patient serving as his or her own physiologic control.

Materials and Methods

In this study, there were 143 subjects (96 male, 47 female), with an average age of 63.9 years (SD=10.5, range=49-81), who were evaluated at a multidisciplinary tertiary care diabetic foot center between 1993 and 1995. Further descriptive data on the study groups are presented in Table 1. All data were abstracted from medical records during this period. All subjects had type II diabetes mellitus, based on the criteria set forth by the World Health Organization.[12] The subjects were divided into three groups: (1) subjects with peripheral sensory neuropathy and no additional discernible pathology (n=78), (2) subjects with peripheral neuropathy and foot ulcers (n=44), and (3) subjects with peripheral neuropathy and acute neuropathic fractures (Charcot's joint) (n=21). The diagnosis of acute Charcot's arthropathy was made using the criteria of Sanders and Frykberg.[13,14] Semmes-Weinstein monofilaments were used to test sensory perception using the method and criteria described by Birke and Sims.[15]

Table 1.
Study Group Characteristics

                                Gender          Age (y)

Group                   n       M       F       X       SD

Asymptomatic sensory
neuropathy              78      50      28      65.9    11.1
Neuropathic ulcer       44      37      7       61.9    10.1
Charcot's arthropathy   21      9       12      60.5    6.8

                        Time With Type I
                        Diabetes Mellitus (y)

Group                   X       SD

Asymptomatic sensory
neuropathy              1 5.7   4.7
Neuropathic ulcer       15.8    4.9
Charcot's arthropathy   17.0    5.8

Subjects were excluded if they had Charcot's joint and concomitant ulceration or a soft tissue or bone infection. Subjects with peripheral arterial occlusive disease (PAOD) were excluded. The working diagnosis of lower-extremity PAOD was made by a combination of one or more clinical signs and symptoms and one or more criteria based on measurements obtained by noninvasive vascular studies. Clinical signs and symptoms included claudication, pain at rest, or absent pulses. Noninvasive criteria included a transcutaneous oxygen measurement of [is less than] 40 mm Hg,[16] an ankle-brachial index of [is less than] 0.80,[17,13] or an absolute toe systolic pressure of [is less than] 45 mm Hg.[19,20] Twenty-two patients were excluded from study using these criteria.

We evaluated skin temperatures with the Exergen DT 1001 infrared skin temperature probe.(*) Temperatures were recorded (in degrees Fahrenheit) from the digital analog display on the top of the device. The temperature probe displays values in increments of 0.1[degrees] F. The probe's infrared lens measures an approximately 1.0-[cm.sup.2] area of skin. The device, which continuously autocalibrates, is held approximately 0.5 cm from the skin surface during measurement and is accurate to within [+ or -] 0.2 [degrees] F.[21] Temperature measurements were made after subjects were allowed to rest for 15 minutes in the examination room. Ambient air temperature was thermostatically controlled at 70 [degrees] + 2 [degrees] F during the test period. Readings were recorded from six sites on the soles of both feet. These sites were the hallux; the first metatarsocuneiform joint; the cuboid; and the first, third, and fifth metatarsal heads. Reliability of the measurements was not assessed in this study.

All patients with acute Charcot's arthropathy and ulcers were initially treated with total contact casting using the technique described by Kominsky.[22] Casts were checked at regular intervals and evaluated for proper fit. Casts of the patients with ulcers were changed weekly and were discontinued when the ulcer was completely epithelialized. The duration of casting for patients with Charcot's arthropathy was dependent on cast comfort and integrity (3 weeks maximum). Casting was discontinued based on clinical, radiographic, and dermal thermometric signs of quiescence (de, resolution of the acute symptoms of Charcot's arthropathy). Following casting, patients with Charcot's arthropathy progressed to removable cast walkers and then to accommodative shoes with anklefoot orthoses, as required. Following transition into therapeutic shoes, patients with Charcot's arthropathy were routinely followed at 2-month intervals. Patients with ulcers were followed at 2-, 4-, and then 8-week intervals following ulcer healing and introduction to therapeutic shoes. Patients with Charcot's arthropathy were followed for an average of 19.0 months (SD=4.0, range=12-27). Patients with ulcers were followed for 24.3 months (SD=6.1, range=13-36), and patients with asymptomatic sensory neuropathy were followed for 21.9 months (SD=6.4, range=12-37).

Data Analysis

Each patient's contralateral extremity served as his or her own control for the symptomatic extremity. The anatomic site of ulceration or Charcot's fracture was matched on the contralateral extremity. A two-tailed Student's t test for matched samples was used to compare temperature differences between limbs within each study group. An unpaired t test was used to compare temperatures among the three groups. An alpha level of .05 was used in this study.[23]

Results

There were differences in skin temperature in both the patients with Charcot's arthropathy (P [is less than] .0001) and the patients with neuropathic ulcers (P [is less than] .0001) when temperatures were compared with those of the corresponding site on the contralateral limb. There was no difference in skin temperatures between feet in the patients with asymptomatic sensory neuropathy (P [is greater than] .05) (Tab. 2).

Table 2.
Skin Temperatures by Pathology and Contralateral Control

                         Skin Temperature ([degrees] F)

                         Right Foot             Left Foot

Group                    X    SD   Range        X   SD   Range

Asymptomatic sensory
neuropathy

Metatarsal 1            81.0  4.8  68.0-90.9   80.9  4.5  72.4-91.2

Metatarsal 3            81.5  4.5  72.9-91.7   81.9  4.2  72.3-89.9

Metatarsal 5            80.9  4.4  72.6-90.3   81.3  4.3  72.2-92.3


                        Affected Foot         Contralateral Foot

Group                   X     SD   Range      X     SD   Range

Neuropathic ulcer       89.4  2.9  78.5-94.0  83.8  2.8  77.2-90.2

Charcot's arthropathy   90.4  3.6  84.0-95.4  82.1  4.1  73.1-89.1


                         Mean
                         Difference
Group                   (95% Cl)(a)       P

Asymptomatic sensory
neuropathy

Metatarsal 1             0.1               NS(b)
                         (-0.4-0.7)
Metatarsal 3              0.4              NS
                         (-0.1 -0.8)
Metatarsal 5              0.4              NS
                         (0.0-0.7)

Group

Neuropathic ulcer         5.6           [is less than] .0001
                         (4.8-6.4)
Charcot's arthropathy     8.3           [is less than] .0001

(a) Cl = confidence interval

(b) NS= not significant

All patients in the neuropathic ulcer group had grade I ulcers of the forefoot, based on the Meggitt-Wagner classification.[24,25] Skin temperatures were highest at the ulcer site in 95% (n=42) of the patients with neuropathic ulcers (Tab. 3). Bilateral skin temperatures taken at the previous ulcer site 12 months after returning to shoe use showed no difference between the previous ulcer site and the corresponding site on the opposite extremity (84.9 [degrees] + 3.4 [degrees] F versus 85.1 [degrees] + 3.1 [degrees] F). Following resolution of ulceration and return to shoe use, five patients (11.4% of the total sample) experienced reulceration at the site of the previous ulceration. These wounds appeared at a mean 12.2 months (SD=6.4) following return to the use of therapeutic shoes (Tab. 4). All wounds were Meggitt-Wagner grade I in depth. Skin temperatures taken during the visit prior to reulceration were higher on the preulcerative limb than on the contralateral limb (89.6 [degrees] [+ or -] 1.2 [degrees] F versus 82.5 [degrees] [+ or -] 2.9 [degrees] F, P=.003). There was no difference in temperatures at the time the patients returned to shoe use between the two groups (83.8[degrees] [+ or -] 2.9 [degrees] F versus 84.1 [degrees] [+ or -] 3.7 [degrees] F).

Table 3.
Ulcer Location and Skin Temperature

                         Affected Foot      Contralateral Foot
                         [degrees] F        [degrees] F

Group             n       X    SD   Range        X   SD   Range

Hallux            8     87.2  3.7  78.5-90.0   84.0  2.8  80.3-89.3

Metatarsal 1     15     89.3  2.7  83.9-92.5   84.0  3.5  77.2-90.2

Metatarsal 2-4   11     90.6  2.6  85.5-93.0   83.6  2.4  78.3-87.2

Metatarsal 5     10     90.2  2.3  85.7-94.0   83.8  2.3  79.7-87.2

                         Mean
                         Difference
Group                   (95% Cl)(a)       P

Hallux                   32              [is less than] .03
                         (0.5-6.0)
Metatarsal 1             5.3             [is less than] .0001
                         (4.1-6.5)
Metatarsal 2.4           7.1             [is less than] .0001
                         (6.0-8.1)
Metatarsal 5             6.3             [is less than] .0001
                         (4.7-8.0)

(a) Cl = confidence interval
Table 4.
Skin Temperatures Before and During Reulceration

                               Affected Foot
                               [degrees] F

                                X       SD    Range

Visit before reulceration (n=5) 89.6    1.2  87.8-91.1

Reulceration visit (n=5)        90.1    1.9  87.1-92.1

All patients with neuropathic
ulcers 1 year after return to
permanent shoe use (n=44)       84.9    3.4  76.6-91.0

                                 Contralateral Foot
                                  [degrees] F

                                  X     SD    Range

Visit before reulceration (n=5)   82.5  2.9  78.9-85.3

Reulceration visit (n=5)          83.7  1.5  82.0-85.3

All patients with neuropathic
ulcers 1 year after return to
permanent shoe use (n=44)         85.1  3.1  77.8-90.3

                                 Mean
                                 Difference
                                 (95% Cl)(a)       P


Visit before reulceration (n=5)  7.1             .003
                                  (4.1-10.1)
Reulceration visit (n=5)         6.5              [is less than] .0001
                                  (5.2-7.7)
All patients with neuropathic
ulcers 1 year after return to
permanent shoe use (n=44)         0.2             NS(b)
                                 (-0.5-0.9)

(a) Cl =confidence interval.

(b) NS=not significant.

In the group of patients with asymptomatic sensory neuropathy, there was no difference in the sites evaluated at 12 months' follow-up (Tab. 5). One patient in this group developed an ulcer at 9 months' follow-up. The skin temperature gradient at the affected site at the visit prior to reulceration was 2.9 [degrees] F (85.9 [degrees] versus 83.0 [degrees] F). Skin temperature at the site of ulceration was somewhat greater than at the site on the contralateral limb (89.9 [degrees] versus 81.2 [degrees] F). The predominant sites of acute Charcot's arthropathy were in the tarsometatarsal calcaneocuboid/talonavicular articulations (Tab. 6). Skin temperatures at these sites were higher than at the corresponding sites contralaterally. The site of maximum skin temperature correlated with the anatomic site of maximum bony involvement (radiographically) in all patients. There were no ulcerations or recurrence of Charcot's arthropathy during the group's mean 19.0-month (SD=4.0-month) follow-up period. This group showed no difference in the three sites evaluated 12 months following quiescence of acute Charcot's arthropathy and subsequent return to prescription shoe use (Tab. 6). Data regarding follow-up for all three groups are illustrated in the Figure.

[Figure ILLUSTRATION OMITTED]

Table 5.
Temperature at 1-Year Follow-up for
Subjects With Asymptomatic Sensory Neuropathy

                         Skin Temperature ([degrees] F)

                         Right Foot             Left Foot

                         X    SD   Range        X   SD   Range

Location
(n = 78)

Metatarsal 1            80.3  5.1  72.0-90.1   80.8  4.7  68.0-91.1

Metatarsal 3            81.6  4.6  72.9-90.4   81.3  4.2  72.6-89.7

Metatarsal 5            82.6  4.1  70.1-89.6   82.1  3.9  69.7-89.9

                         Mean
                         Difference
                        (95% Cl)(a)       P

Location
(n = 78)

Metatarsal 1             0.5               NS(b)
                         (-0.1-1.1)
Metatarsal 3              0.4              NS
                         (-0.2-0.9)
Metatarsal 5              0.4              NS
                         (0.1-1.2)

(a) Cl=confidence interval.

(b) NS = not significant.
Table 6.
Location of Charcot's Arthropathy and Skin Temperature

                         Affected Foot      Contralateral Foot
                         [degrees] F        [degrees] F

Location          n       X    SD   Range        X   SD   Range

Lisfranc's joint  10    91.1  3.2  86.7-95.4   82.3  3.8  73.1-95.9

Chopart's joint    9    89.6  4.2  84.0-95.0   81.6  4.7  76.1-89.1

Ankle joint        2    90.2  2.9  88.1-92.2   82.9 90.1  79.2-86.5

                         Mean
                         Difference
Location                 (95% Cl)(a)       P

Lisfranc's joint          8.8             [is less than] .0001
                         (6.9-10.7)
Chopart's joint           8.0             [is less than] .0001
                         (5.7-10.4)
Ankle joint               7.3             [is less than] .0001
                         (- 13.0-27.6)

(a) Cl=confidence interval.

(b) NS=not significant.

Discussion

The results of this study indicate that, at the time of diagnosis, there were differences in skin temperatures in the affected lower extremity of patients with Charcot's joint and patients with neuropathic ulcers compared with the temperatures of their contralateral lower extremity. In the patients with diabetes, peripheral sensory neuropathies, and no additional pathology, the temperatures of the right and left lower extremities were not different. All patients who experienced reulceration during the follow-up period (11.4% of the total sample) showed elevated skin temperature gradients (de, increased temperature differences between lower extremities) on the visit prior to reulceration. This finding suggests that elevated skin temperature gradients may be predictive of future ulceration. To our knowledge, no previous work has quantified specific temperature gradients using infrared dermal thermography.

Previous work[10,11] has demonstrated that high foot temperatures are useful in identifying patients with diabetes who are at risk of ulceration. The ideal treatment of the high-risk diabetic foot is prevention. The clinical changes that precede ulceration of neuropathic fracture in persons with diabetes, however, are often subtle and difficult to appreciate. The potential benefit of dermal thermometry is that nonpalpable differences in skin temperature can be detected and immediate action can be taken, such as total contact casting, use of therapeutic footwear, bed rest, or a decrease in daily activity to prevent limb-threatening complications.

After a neuropathic ulcer heals, 20% to 58% of patients develop another ulcer within a year.[5,26,27] This reulceration is thought to be due to the persistence of subclinical inflammation caused by repetitive trauma associated with areas of moderate to high pressures. Newly healed tissues are vulnerable to breakdown by even minor stress.[2,7,8] As skin temperatures cool and inflammation subsides, the tissues are able to accept additional stress without breakdown.[28] The site in question should be off-weighted (ie, off-loaded to reduce planter pressures by accommodating the at-risk foot in various types of prosthetic devices or casts) until the skin temperature of the affected foot equilibrates with the skin temperature of the contralateral foot and then should be monitored during subsequent follow-up visits. An increase in local skin temperatures after wound healing should raise suspicions in members of a foot care team. At this time, the team may consider modifying footwear or insoles and also reinforcing patient education regarding activity level. The low rate of reulceration in our study compared with previous studies probably reflects use of temperature gradients to guide us in more aggressive intervention.

Skin temperature measurements may also be useful for monitoring the progression of Charcot's fractures through the major clinical phases. In the acute stage of the deformity, the affected area is often grossly erythematous, edematous, and warm to the touch.[29] In the postacute phase, when gross signs of inflammation have resolved, premature return to activity can trigger another acute episode. Skin temperature measurements can be used to detect subtle temperature changes that may persist for months after a palpable difference can no longer be perceived. Equilibration of temperatures, combined with clinical and radiographic signs of resolution, can be used as a benchmark to determine when guarded weight bearing should begin.

Limitations

There are several limitations to this study. The retrospective design limits truly uniform follow-up, particularly following resolution of ulceration or quiescence of Charcot's arthropathy. All patients studied had adequate peripheral arterial perfusion. The role that ischemia plays in thermometric response to inflammation, therefore, cannot be discussed within the context of these data. Additionally, patients with bilateral neuropathic injuries were not evaluated. It may be inferred that patients with bilateral injuries will have commensurate inflammation and increased temperatures, thus diminishing the quality of the contralateral limb as a control and therefore the utility of dermal thermometry. Clearly, this subject warrants further investigation, prospectively monitoring large cohorts of patients with neuropathic or ischemic injuries before, during, and after ulceration and acute Charcot's arthropathy.

Conclusion

These data suggest that infrared dermal thermography may provide valuable information to the clinician to assist in the detection, treatment, and prevention of neuropathic lower-extremity sequelae. Because dermal thermometry is simple, noninvasive, and relatively inexpensive, in the future patients at risk for ulceration may be able to use a temperature probe on a daily basis at home to detect sites of inflammation in the same manner that home glucometers are used to monitor glucose levels. Patients with diabetes may be able to learn to adjust their activity just as they do their insulin. In addition, skin temperature measurements may be used to monitor the effectiveness of footwear or other prosthetic devices designed to decrease high-pressure areas as well as to monitor the progress of therapy for ulcers and Charcot's arthropathy.

(*) Exergen Corporation, 51 Water St, Watertown, MA 02172.

This article was presented in abstract format at the 32nd annual scientific symposium of the European Association for the Study of Diabetes Vienna, Austria; September 1, 1996.

This article was submitted June 20, 1996, and was accepted October 7, 1996.

References

[1] Swartz MH. The physical examination. In: Swartz MH, ed. Physical Diagnosis. Philadelphia, Pa: WB Saunders Co; 1989:73-82.

[2] Hall OC, Brand PW. The etiology of the neuropathic planter ulcer. J Am Podiatr Assoc. 1979;69:173-177.

[3] Tredwell J. Pathophysiology of tissue breakdown in the diabetic foot. In: Kominsky SJ, ed. Medical and Surgical Management of the Diabetic Foot. St Louis, Mo: Mosby Year Book; 1994:97-112.

[4] Todd WF, Laughner T, Samojla BG. The diabetic foot. In: Robbins JM, ed. Primary Pediatric Medicine. Philadelphia, Pa: WB Saunders Co; 1994:213-243.

[5] Bergtholdt HT. Temperature assessment of the insensitive foot. Phys Ther. 1979;59:18-22.

[6] Beach RB, Thompson DE. Selected soft tissue research: an overview from Carville. Phys Ther. 1979;59:30-35.

[7] Manley MT, Darby T. Repetitive mechanical stress and denervation in planter ulcer pathogenesis in rats. Arch Phys Med Rehabil. 1980;51: 171-175.

[8] Brand PW. The insensitive foot (including leprosy). In: Jahss M, ed. Disorders of the Foot and Ankle. 2nd ed. Philadelphia, Pa: WB Saunders Co; 1991:2173-2175.

[9] Chan AW, MacFarlane IA, Bowsher DR. Contact thermography of painful diabetic neuropathic foot. Diabetes Care. 1991;14:918-922.

[10] Stess RM, Sisney PC, Koss KM, et al. Use of liquid crystal thermography in the evaluation of the diabetic foot. Diabetes Care. 1986;9:267-272.

[11] Benbow SJ, Chan AW, Bowsher DR, et al. The prediction of diabetic neuropathic planter foot ulceration by liquid crystal contact thermography. Diabetes Care. 1994;17:835-839.

[12] Second Report on Diabetes Mellitus. Geneva, Switzerland: World Health Organization; 1980.

[13] Sanders LJ, Frykberg RG. Charcot's joint. In: Levin ME, O'Neal LW, Bowker JH, eds. The Diabetic Foot. 2nd ed. St Louis, Mo: Mosby-Year Book; 1993:48-63.

[14] Sanders LJ, Frykberg RG. The Charcot foot. In: Frykberg RG, ed. The High-Risk Foot in Diabetes Mellitus. New York, NY: Churchill Livingstone Inc; 1991:325-335.

[15] Birke JA, Sims DS. Plantar sensory threshold in the ulcerated foot. Lepr Rev. 1986;57:261-267.

[16] Bacharach J, Rooke T, Osmundson P, Glovizzki P. Predictive value of transcutaneous oxygen pressure and amputation success by use of supine and elevation measurements. J Vasc Surg. 1992;15:558-563.

[17] LoGerfo FW, Coffman JD. Vascular and microvascular disease of the foot in diabetes. N Engl J Med. 1984;311:1615-1619.

[18] Carter S. Elective foot surgery in limbs with arterial disease. Clin Orthop. 1993;289:228-236.

[19] Orchard TJ, Strandness DE. Assessment of peripheral vascular disease in diabetes: report and recommendation of an international workshop. Diabetes Care. 1993;83:685-695.

[20] Apelqvist J, Castenfors J, Larsson J. Prognostic value of ankle and toe blood pressure levels in outcome of diabetic foot ulcers. Diabetes Care. 1989;12:373-378.

[21] Dermatemp Infrared Scanner Reference. Watertown, Mass: Exergen Corporation; 1996.

[22] Kominsky SJ. The ambulatory total contact cast. In: Frykberg RG, ed. The High-Risk Foot in Diabetes Mellitus. New York, NY: Churchill Livingstone Inc; 1991:449-455.

[23] Gehan EA, Lemak NA. Statistics in Medical Research. New York, NY: Plenum Publishing Corp; 1994.

[24] Meggitt B. Surgical management of the diabetic foot. Br J Hosp Med. 1976;6:227-332.

[25] Wagner FW. The dysvascular foot: a system for diagnosis and treatment. Foot Ankle. 1981;2:64-122.

[26] Helm PA, Walker SC, Pulliam GF. Recurrence of neuropathic ulcerations following healing in a total contact cast. Arch Phys Med Rehabil. 1991;72:967-970.

[27] Uccioli L, Faglia E, Monticone G, et al. Manufactured shoes in the prevention of diabetic foot ulcers. Diabetes Care. 1995;18:1376-1378.

[28] Brand PW. The diabetic foot. In: Ellenberg M, Rifkin H, eds. Diabetes Mellitus: Theory and Practice. 3rd ed. New York, NY: Medical Examination Publishing; 1983:803-828.

[29] Sandrow RE, Torg JS, Lapayowker MS, Resnik EJ. Use of thermography in the early diagnosis of neuropathic arthropathy of the feet in diabetics. Clin Orthop. 1972;88:31-33.

Invited Commentary

I appreciate the opportunity to comment on this clinically relevant study. The authors should be commended for sharing their work with the physical therapy community. The many physical therapists who are involved in the care of patients with high-risk feet should find application for this work in their practice.

This study met the authors' purposes of quantifying temperature differences between feet with neuropathic ulcers and acute neuropathic fractures compared with contralateral feet before and 1 year after healing and documenting temperature differences in feet immediately prior to reulceration. Analysis of a more complete set of serial data would have provided additional information on the pattern and timing of temperature differences. These findings support earlier work based on observations and case study.[1-3] This study validates the use of a relatively inexpensive hand-held radiometer in clinical decision making in the management of neuropathic foot problems related to diabetes.

In physical therapy practice, pain is a primary guide in determining the intensity, duration, and effectiveness of treatment. Patients with high-risk feet, however, have an inappropriate or lost pain response. Clinicians should be proactive in screening the feet of patients with diabetes. Evidence of sensory loss, poor circulation, or high foot pressure or a previous history of ulcer or neuropathic fracture identifies the foot at high risk.[4] The feet of these patients should be checked when changes in physical activity or footwear are made. Thermometry offers an objective measure for evaluating patient progress. In particular, this study illustrates the value of thermometry in evaluating weight-bearing status, footwear, and activity level during healing.

This study demonstrates the importance of early recognition and aggressive conservative treatment of neuropathic ulcers and fractures. Neuropathic ulcers and fractures are too often misdiagnosed or undertreated. The authors demonstrate the value of thermometry only within the context of a comprehensive treatment program, which has off-weighting devices and custom-made footwear readily available and a staff that shares an understanding of the mechanics of neuropathic foot problems.

In this study, patients were followed at routine intervals after healing and introduction to shoes. Scheduled follow-up has been recommended for the prevention of ulceration and reulceration in the high-risk patient.[4] Some patients, however, still experienced reulceration. Reulceration may occur because wounds have not fully healed. After closure (de, the time period in which patients begin walking in shoes) the wound is undergoing the remodeling phase of healing. This phase of healing lasts 6 months to a year, during which time type III collagen is replaced by stronger type I collagen and connective tissue fibers are realigned.[5] Moderate, intermittent stress on the wound is a stimulus for tissue remodeling, but high, continuous stress may injure tissues.! During this period, additional efforts should be made to empower patients and family caregivers with the knowledge to examine the feet, feel for areas of high temperature, and immediately report problems to a clinician. The authors also believe that patients should be given skin thermometers for daily monitoring.

The infrared thermometer used in this study can be read using either the Fahrenheit or Celsius scale. The authors used the Fahrenheit mode, which provides higher resolution but is awkward when comparing data from studies commonly reporting in metric units. The cost of the infrared scanners is about $800. Less expensive thermometers are available that have similar accuracy but require direct contact with the skin and have a slower recording time. These less expensive devices may be suitable for patient use. In the absence of skin thermometers, the hand can identify differences in skin temperature of 1 [degrees] to 2 [degrees] C within and between feet.[4-5]

Absolute temperature measurement has not been recommended for monitoring skin temperature in the feet because of its large diurnal variation. Serial comparison of temperature differences between inflamed and normal tissues may also vary due to environmental and physiologic factors. Equilibrating the feet for 15 minutes at a room temperature of 21 [degrees] to 22 [degrees] C has been recommended to maximize and standardize the contrast between inflamed and normal feet.[6]

The issue of how many degrees of temperature difference is significant for clinical decision making was only indirectly addressed in the article. In this study, the largest mean difference (95% confidence interval) between the feet of the patients with asymptomatic neuropathy was 1.2 [degrees] F (about 0.7 [degrees] C). Previous studies[1,2] have shown differences of 1 [degrees] C to be indicative of tissue inflammation related to repetitive stress on the foot. The authors should comment on what temperature difference they believe is clinically significant.

The relevance of this study extends beyond the high-risk feet of patients with diabetes. Similar foot problems develop in other neuropathic conditions such as peripheral nerve injuries, spine bifida, hereditary sensory neuropathy, and leprosy. Thermometry may also have application in sensate patients where objective monitoring of inflammation is useful during physical rehabilitation. Thermometry may enhance the evaluation of postoperative wounds, arthritic joints, musculoskeletal injuries, or compensation cases.

References

[1] Brand PW. Repetitive stress on insensitive feet: the pathology and management of planter ulceration in neuropathic feet. Social and Rehabilitation Service Grant No. RC-75-MPO; 1975.

[2] Bergtholdt HT, Brand PW. Temperature assessment and planter inflammation. Lepr Rev. 1976;47:211-219.

[3] Harris JR, Brand PW. Patterns of disintegration of the tarsus in the anaesthetic foot. J Bone Joint Surg [Br]. 1966;48:4-16.

[4] Birke JA, Sims DS. The insensitive foot. In: Hunt GC, McPoil TC, eds. Physical Therapy of the Foot and Ankle. 2nd ed. New York, NY: Churchill Livingstone Inc; 1995:159-207.

[5] Weiss EL. Connective tissue in wound healing. In: McCulloch JM, Kloth LC, Feedar JA, eds. Wound Healing Alternative in Management. 2nd ed. Philadelphia, Pa; FA Davis Co; 1995:16-31.

[6] Bergtholdt HT. Temperature assessment of the insensitive foot. Phys Ther. 1979;59:18-22.

Jim Birke, PhD, PT, CPed Chief Physical Therapy and Foot Department Gillis W Long Hansen's Disease Center Carville., LA 70721 (jbirke@)concentric. net)

Author Response

We appreciate Dr Birke's thoughtful commentary on our article. His comments bring up several important points concerning the use of dermal thermometry. With regard to his query regarding a specific recommendation of the number of degrees we deem clinically significant, we typically will take action when skin temperature differences exceed 3 [degrees] to 4 [degrees] F (2 [degrees] C).

Indeed, 11% of the patients in the neuropathic ulcer group experienced reulceration a mean of 12 months following healing. This time period is less than that reported in other published works.[1,2] We believe that this finding was due in no small part to a treatment bias present in our study. In our diabetic foot specialty clinic, patients who have elevated skin temperatures without ulceration are typically off-loaded with a total contact cast, a healing sandal, or a removable cast walker as a prophylactic measure until temperatures equilibrate. In the patients who experienced reulceration, we believe that a glaring indicator (in the form of a focally elevated skin temperature) was present at the visit prior to reulceration but was overlooked.

We, too, believe that this modality is clinically relevant beyond this particular clinical realm. We frequently use the device to monitor progression of inflammation in virtually all areas of our respective practices.

We greatly appreciate Dr Birke's commentary and look forward to future dialogues in Physical Therapy.

References

[1] Helm PA, Walker SC, Pulliam GF. Recurrence of neuropathic ulcerations following healing in a total contact cast. Arch Phys Med Rehabil. 1991;72:967-970.

[2] Uccioli I., Faglia E, Monticone G, et al. Manufactured shoes in the prevention of diabetic foot ulcers. Diabetes Care. 1995;18:1376-1378.

DG Armstrong, DPM, is Assistant Professor, Department of Orthopaedics, University of Texas Health Science Center, 7703 Floyd Curl Dr, San Antonio, TX 78284 7776 (USA) ([e-mail]). Address all correspondence to Dr Armstrong.

LA Lavery, DPM, is Assistant Professor, Department of Orthopaedics, University of Texas Health Science Center, and Diabetic Foot Research Group.

PJ Liswood, DPM, is in clinical practice, Diabetic Foot Centers of America, 11717 Le Havre Dr, Potomac, MD 20854.

WF Todd, DPM, is in clinical practice, Diabetic Foot Centers of America.

JA Tredwell, DPM, is Director, Diabetic Foot Centers of America.