Correlation Between Wound Temperature Obtained With an Infrared Camera and Clinical Wound Bed Score in Venous Leg Ulcers

Introduction. The measurement of skin and wound bed temperature in chronic wounds may be a useful way to optimize the assessment and diagnosis of chronic wound infection. The aim of this clinical research trial was to correlate the wound bed score, validated by Falanga in 2006, to wound bed and perilesional skin temperature with an easy-to-use, handheld, noninvasive thermometer. Materials and Methods. In this study, the authors recruited 18 patients affected by venous insufficiency and lower leg ulcers. A total of 24 chronic wound bed and perilesional skin ulcers were assessed using an infrared camera (FLIR T620 Thermal Imager, FLIR Systems Boston, MA). At the same visit, an operator blinded to the thermal image results made a wound bed score to make a clinical evaluation of the lesion. Results. The wound bed temperature range after dressing removal was between 31°C and 35°C, and the perilesional skin temperature range was between 31°C and 34°C. The wound bed score range was between 5-14 (14 patients > 10; 11 patients ≤ 10). The study data showed an increasing relationship between the wound bed score and the wound bed temperature according to several studies that have demonstrated 33°C is the critical temperature level required for normal cellular activity. The correlation between the wound bed score and the perilesional skin temperature is weaker compared to other measurements. Conclusion. The results obtained in this preliminary research suggest that this correlation is worth being further investigated with a larger dataset.

Human skin has a relevant role in thermoregulation through its extensive vascular bed. The skin temperature of the human body is the result of a thermal balance between energy supplied from the core, and perfusion and energy lost to the environment via radiation, conduction, convection, and evaporation.¹ Skin temperature is influenced by physiological and environmental factors and can be affected by variations in ambient temperature, surface moisture, body site, and blood flow.²

A standard criterion for measuring skin temperature has not yet been established.³ Conventional mercury or electronic thermometers are difficult to apply to the body surface, require significant amounts of time to equilibrate, and are prone to low readings because of poor surface contact. A handheld infrared thermometer has the potential to provide an objective, quantitative measurement of skin surface temperature.⁴

The temperature of any site of the body can be comparable to a symmetrical site under normal circumstances, but no reference range exists for body surface temperatures, which can differ from person to person and according to location on the body.⁵ 

This clinical research trial was performed to correlate the wound bed score, validated by Falanga in 2006⁶ (Table 1), to wound bed and perilesional skin temperature by means of an easy-to-use, handheld, noninvasive infrared camera in patients with venous leg ulcers. 

In this study, the authors recruited 18 patients affected by venous insufficiency and lower leg ulcers by means of clinical and instrumental signs, using specific inclusion/exclusion criteria (Table 2). A total of 24 chronic wound beds and perilesional skin areas were assessed using an infrared camera (Figures 1A and 1B) (FLIR T620 Thermal Imager, FLIR Systems, Boston, MA). The camera has a thermal sensitivity (noise equivalent temperature difference) < 0.04°C at 30°C and an accuracy within + 2% of reading (Figure 2). At the same visit, an operator blinded to the thermal image readings registered a wound bed score to allow a clinical evaluation of the lesion. The wound bed score validated by Falanga⁶ can be considered a useful tool in clinical and research settings, because it seems to have validity in predicting complete wound closure in wounds treated with either standard therapies or advanced modalities. Wound size assessment was achieved by using a validated 3-dimensional imaging system⁷  (Silhouette,  Aranz Medical, New Zealand). 

The wound surface area ranged from 5 cm² to 51.5 cm², with an average of 34.6 cm². Wound management was the same for all patients evaluated and was performed according to standard treatment with a compression multilayer system and moist wound healing.

Temperature measurements were taken in an air-conditioned facility (mean room temperature 22°C) after a 15-minute period of patient rest immediately after dressing removal.

The aim of the study was to correlate the mean wound bed temperature, assessed by using an infrared camera, to the items of the wound bed score that referred to the wound bed (ie, healing edges, black eschar, greatest wound depth/granulation tissue, pink wound bed) and the mean perilesional skin temperature to the items of the wound bed score for the surrounding skin (ie, exudate, edema, periwound dermatitis, callus).

Statistical analysis
The statistical analysis was performed in MATLAB (Mathworks, Natick, MA). The least-square algorithm was used to obtain a linear regression model of the temperatures of the wound bed (T_wb) or perilesional skin (T_ps) versus the wound bed score. The validity of assuming a linear correlation, expressed by the Pearson’s coefficient analysis, was checked by the Kolmogorov-Smirnov test at a 95% confidence level to determine if the variables were bivariate normal distributed. This test allowed the correlation analysis to be checked for robustness to outliers and skewness of data. Besides linearity, the monotonicity of data was benchmarked by the Spearman’s rank-order coefficient.

The wound bed temperature range after dressing removal was between 31°C and 35°C, and the perilesional skin temperature range was 31°C-34°C. The wound bed score range was 5-14 (14 patients > 10; 11 patients ≤ 10).

The correlation of data was determined as follows. Figure 3 shows the scatter plot for the wound bed score and the temperature of the wound bed. The coefficient of determination (R² = 0.65) indicated that about 65% of the variations could be explained by the linear regression model. However, a linear correlation could not be assessed by Pearson’s coefficient since the Kolmogorov-Smirnov test rejected the hypothesis that the dataset represented a normal distribution. A possible explanation could be ascribed to the small number of items in the dataset. Therefore, instead of linearity, the authors benchmarked the monotonicity of data by the Spearman’s rank-order coefficient. This coefficient, ρ, is 0.805 with a P value of 2 x10^-6 (confidence level 95%). This result proves a monotonic increasing relationship between the wound bed score and the temperature of the wound bed. 

The dataset composed of the wound bed score and the temperature of the perilesional skin returned ρ = 0.55 and P value = 0.005 (confidence level 95%). Although in this case the monotonicity was weaker, the result suggests that this correlation is worthy of further investigation with a larger dataset. Figure 4 shows the scatter plot for the wound bed score and the temperature of the perilesional skin. The regression model is far from being linear (R² = 0.35). 

Data from this study showed an increasing relationship between the wound bed score and the wound bed temperature in accordance with several studies that have demonstrated 33°C is the critical temperature level required for normal cellular activity.⁸

The correlation between the wound bed score and the perilesional skin temperature was weaker. This result suggests that this correlation is worthy of further investigation with a larger dataset, so the authors will continue the study to implement the sample size and hopefully reach a better correlation.

In the chronic wound, increased local temperature is one of the classic signs of wound infection and inflammation, and its quantitative measurement may be useful to optimize the assessment and treatment of lesions. Horzic and coauthors⁹ demonstrated the persistence of increased skin temperature in an area adjacent to healing postsurgical wounds after the third postoperative day was considered to be a predictive sign of infection and delayed healing. Increased skin temperature measured with an infrared thermometer has been found to be an early predictor of postoperative sternal wound infection.¹⁰

Armstrong and colleagues⁸ investigated the correlation between increased skin temperature and the severity of diabetic foot infection and clinical outcomes in 232 patients by using a handheld infrared thermometer. They found that a 10°F or greater differential between limbs had a significantly lower clinical response (P = 0.0007). A recent prospective cross-sectional study demonstrated foot and leg ulcers with an elevated skin temperature recorded with a handheld infrared thermometer were 8 times more likely to be diagnosed with a deep wound infection.¹¹

Wound bed temperature plays a key role in wound healing. It has been demonstrated that when the temperature of the wound bed falls below the core body temperature, healing can be delayed due to lack of collagen deposition and a reduction in late-phase inflammatory cells and fibroblasts.¹²

In vitro studies have shown that 33°C is the critical level at which neutrophil, fibroblast, and epithelial cell activity decreases.¹³ Some authors have suggested pressure ulcer surface areas undergo a faster reduction when the wound bed is 36°C-38°C.¹⁴ McGuiness and associates¹⁵ found that wound bed temperatures immediately after dressing removal were marginally below the 33°C threshold deemed necessary for cellular activity (mean 32.6°C; range: 25.3°C-37.3°C). Fierheller and colleagues¹⁶ demonstrated a strong relationship between infection and the quantitative measurement of increased periwound skin temperature.

In the authors’ opinion, it could be useful in further studies to clarify the importance of temperature changes in making an early diagnosis of deeper or spreading infection. A handheld infrared camera can assist the wound care practitioner with early identification of hard-to-heal wounds, allowing for timely intervention and facilitating the monitoring of ongoing treatment responses.

Affiliations: Department of Dermatology, University of Pisa, Italy; Department of Chemistry and Industrial Chemistry, University of Pisa, Italy; and National Research Council of Italy, C.N.R.

Correspondence:
Marco Romanelli, MD, PhD
Wound Healing Research Unit
Department of Dermatology
University of Pisa
Via Roma 61
56124 Pisa, Italy
m.romanelli@med.unipi.it

Disclosure: This study was supported by the EU-funded FP7 ICT-317894 SWAN-iCare project.