A Clinical Investigation into the Relationship between

Increased Periwound Skin Temperature and Local Wound

Infection in Patients with Chronic Leg Ulcers
Advances in Skin & Wound Care, August 2010
Clinical Topic: Skin/wound Expires: 8/31/2012

care

CE Collection

Add to Planner

Take Test

A Clinical Investigation into the Relationship between Increased Periwound Skin

Temperature and Local Wound Infection in Patients with Chronic Leg Ulcers
Marjorie Fierheller MSc, BScN, RN
R. Gary ibbald BSc, MD, FRCPC (Med Derm), MACP, FAAD, MAPW

Advances in Skin &

Wound Care: The
Journal for Prevention
and Healing - Featured
Journal
August 2010
Volume 23 Number 8
Pages 369 - 379

Request
Permissions

Abstract
PURPOSE: To enhance the clinician's knowledge about the relationship between
increased periwound skin temperature and local wound infection in patients with
chronic leg ulcers.
TARGET AUDIENCE: This continuing education activity is intended for physicians and
nurses with an interest in skin and wound care.
OBJECTIVE: After participating in this educational activity, the participant should be
better able to:
1. Interpret research findings on chronic wound assessment including skin
temperature assessments.
2. Examine the study reported in this article for appropriate use of periwound skin
temperature assessment.
3. Analyze this study's findings regarding the relationship between skin temperature
and chronic wound infection.
ABSTRACT:
OBJECTIVE: Increased local temperature is a classic sign of wound infection, and its
quantitative measurement has the potential to assist with assessment and diagnosis of
chronic deep wound and surrounding skin infection at the bedside. Evidence
supporting such use in chronic wound care is very limited. This clinical pilot study was
conducted in an attempt to quantify the relationship between increased periwound

skin temperature and wound infection, as well as validate use of a handheld infrared
thermometer for the wound care practitioner.
DESIGN, SETTING, AND PARTICIPANTS: Using a cross-sectional design, 2 groups of
participants were recruited from a chronic wound clinic: without wounds (n = 20) and
with chronic leg ulcers (n = 40). Participant and wound characteristics were
documented. All skin temperatures were documented using a handheld infrared
thermometer under consistent environmental conditions within the clinic. Data
analysis was based on the difference ([DELTA]) in skin temperature (in degrees
Fahrenheit) between a target or wound site and an equivalent contralateral control
site. Wound infection was identified using the combination of a validated assessment
tool and clinical judgment. Supplemental semiquantitative bacterial swabs were
collected from all wounds.
OUTCOME MEASURES: Descriptive statistics were analyzed using the chi-squared
calculation. A Pearson r calculation of test-retest skin temperature data collected from
nonwounded participants initially determined reliability of the infrared thermometer.
Correlation of increased periwound skin temperature to wound infection was
determined by calculation of a 1-way analysis of variance.
MAIN RESULTS: The infrared thermometer was found to be reliable (r = 0.939, P = .
000 at a 95% confidence interval). A statistically significant relationship between
increased periwound skin temperature and wound infection was identified (F =
44.238, P = .000 at a 95% confidence interval). Neither patient nor wound
characteristics were significantly different between the participants with noninfected or
infected wounds.
CONCLUSION: The results of this study demonstrate that incorporating quantitative
skin temperature measurement into routine wound assessment provides a timely and
reliable method for a wound care practitioner to quantify the heat associated with
deep and surrounding skin infection and to monitor ongoing wound status. Study
limitations may reduce transferability of these findings to wound types other than
chronic leg ulcers. Further research is needed to support and strengthen these results.

INTRODUCTION/LITERATURE REVIEW
Delayed healing of chronic wounds can prolong patient distress and discomfort,
increase risk of complication, and significantly add to healthcare cost.1,2 Studies
relate the presence of chronic wound infection to reduced tensile strength in healing
tissue and delayed healing.3-5 Although failure to heal may be the result of many
coexisting factors, infection is often a significant contributor.6 Timely diagnosis and
treatment of infection are necessary to prevent further tissue damage and optimize
healing potential.6-10
Ideally, optimal management of wound infection begins with a thorough clinical
assessment,10,11 and diagnosis is then based on clinical judgment supplemented if
needed with microbiological data.6,12 In reality, the best methods to accurately
assess and interpret signs and symptoms of infection in chronic wounds continue to be
debated13 and assessment is considered an "imprecise science,"14 dependent on the
experience and diagnostic skill of an individual clinician.15 Lack of consensus among
researchers and wound care practitioners contributes to a gap in related practice.
After reading this article, clinicians should be better able to interpret research findings
on chronic wound assessment including skin temperature assessments, examine the
data reported in this article for appropriate use of periwound skin temperature
assessment, and analyze the relationship between skin temperature and chronic
wound infection.

Gardner et al7,16 developed a Clinical Signs and Symptoms Checklist (CSSC), with
the purpose of standardizing assessment of chronic wounds for signs and symptoms of
infection. This tool includes a combination of both classic9 and secondary17 signs and
symptoms of chronic wound infection. Erythema (rubor), increased temperature
(calor), pain (dolor), edema (tumor), and purulent discharge are considered to be
classic signs and symptoms of infection.7,9 Cutting and Harding17 identified
increased serous discharge, delayed healing, discolored granulation tissue, pocketing
of wound base, foul odor, and wound breakdown as more subtle secondary signs and
symptoms of infection often found in chronic wounds granulating by secondary
intention.
Interrater reliability of the CSSC was analyzed using percent agreement and [kappa]
coefficient. According to the [kappa] statistics, heat (as detected by the assessors
hand), foul odor, and discoloration of granulation tissue had moderate agreement
(0.40-0.60), whereas increasing pain, edema, wound breakdown, delayed healing, and
friable granulation tissue had "almost perfect" agreement (0.80-1.00). Although the
Gardner et al16 study is limited by a relatively small sample size (n = 36), diagnosis
of infection based on bacteriology alone, and inclusion of a variety of different wound
types, their tool has been validated for assessment of chronic wound infection and is
frequently cited in related literature.6,9,14,18-20
Sibbald et al10 developed a guide to standardize assessment of infection in chronic
wounds. Signs and symptoms of superficial wound bacterial damage (increased
bacterial burden, critical colonization, local infection, covert infection) include 3 or
more ofnon-healing wound status, exudate, red and friable granulation tissue, necrotic
debris, and/or smell (NERDS). Signs and symptoms of deep or surrounding skin
wound infection include 3 ormore of increased wound size or increased temperature,
os (probing to bone), new areas of breakdown, erythema and/or edema, exudate, and
smell (STONEES). Although these superficial indicators of chronic wound bacterial
damage parallel the secondary signs and symptoms identified by Cutting and
Harding,17 those related to deep infection more closely reflect classic signs and
symptoms.
The complex, polymicrobial environment found in chronic wounds has potential to
delay healing and promote bacterial damage and infection.6 Although not generally
considered to be diagnostic of infection, assessment of the type, amount, and nature
of bacterial burden is important to supplement clinical decision making and direct
appropriate treatment.13 Debate is ongoing regarding the best methods to accurately
determine bacteriology within chronic wounds and accurately interpret the meaning of
these findings.2,18,21,22
Quantitative biopsy is considered the criterion standard for collection of bacterial
culture specimens, but it primarily identifies bacteria in the deep wound tissue as the
superficial compartment is only a few millimeters in thickness. It is, however, time
consuming, invasive, costly, and not routinely used in clinical practice.13 The more
easily obtained, inexpensive, and relatively accurate, semiquantitative wound swabs
are commonly used in clinical practice settings1,23-25 (Figure 1).
Figure 1. SENSITIVITY OF SEMIQUANTITATIVE WOUND SWABS
A criterion standard for measurement of skin temperature has not yet been
established.26 The subjective nature of skin temperature changes detected by an
examiner's hand during physical assessment limits its diagnostic accuracy.27
Conventional mercury or electronic thermometers are difficult to attach to the body
surface, require significant amounts of time to equilibrate, and are prone to low
readings because of poor surface contact.28 A handheld infrared thermometer
(Figure 2), has the potential to provide an objective, quantitative measurement of
skin surface temperature.28

Figure 2. INFRARED THERMOMETERThe clinician uses a handheld infrared

thermometer to measure skin surface temperature. Photo courtesy of
Canadian Association of Wound Care CD Series, 2001.
Human skin has an important role in thermoregulation through its extensive and
extremely labile vascular bed.29 Surface 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.30,31
A cuplike mirror in the nose of the infrared thermometer "traps" radiation emitted
from a target site on the skin, transfers it through an optical system, and converts it
to an electrical signal that can be read on an LCD display. This type of instrument is
automatically programmed and calibrated to provide several readings per second.30
The temperature of any site on the human body is said to be comparable to a
symmetrical site under normal circumstances.32 No reference range exists for body
surface temperatures,33 which can vary widely between person to person and by
body location. The differences in skin temperature between a specific target site and a
symmetrical contralateral reference point, such as the opposite foot or leg, are utilized
for the data to be meaningful.31,33,34
Skin temperature is subject to the influence of both internal (physiological) and
external (environmental) factors and, as such, can be significantly affected by
variations in ambient temperature, surface moisture, location on the body, proximity
to blood flow, or compromised blood flow (Figure 3).31
Figure 3. PROPERTIES OF HUMAN SKIN TEMPERATURE
An online literature search of MEDLINE, PubMed, CINAHL, ScienceDirect, Blackwell
Synergy, Medscape, and World Wide Wounds was conducted between June 2004 and
March 2007 and repeated in October 2009, with a focus on chronic wounds and/or leg
ulcers. Keywords included assessment and/or diagnosis of chronic wound infection;
signs and symptoms of chronic wound infection, chronic wound
microbiology/bacteriology; skin temperature; infrared thermometry, dermal
thermometry; plus skin temperature and wound infection.
Horzic et al35 compared skin temperatures adjacent to healing postsurgical wounds (n
= 30) with those recorded preoperatively. Increased skin temperature was detected
for the first 3 days and gradually decreased between days 4 and 8, corresponding to
the inflammatory stage of healing. Persistence of increased skin temperature after the
third postoperative day was considered to be a sign predictive of infection and
disturbed healing.35 Increased skin temperature measured with an infrared
thermometer has been found to be an early predictor of subsequent postoperative
sternal wound infection(n = 150).36 Skin temperature elevation sustained past 18
weeks after uncomplicated total knee replacement may also indicate development of
infection.37
Clinical application of a handheld infrared thermometer has been validated for
assessment of the heat associated with inflammation in diabetic neuropathic (Charcot)
foot problems34 and diabetic foot risk assessment for patients and practitioners.38
Armstrong et al39 investigated correlation of increased skin temperature with severity
of diabetic foot infection and clinical outcome (n = 332). They were unable to
demonstrate a relationship between skin temperature and poor clinical outcome but
found that a 10[degrees] F or greater differential between limbs had a significantly
lower clinical response (P = .0007). An elevated white blood cell count was associated
with nearly twice the increased risk of poor outcome, but no correlation was found to
increased skin temperature. The suggestion was made that although white blood cell
count reflects a systemic response to infection, skin temperature reflects a local
response.39
Foot and leg ulcers (n = 112) that demonstrated an elevated skin temperature

recorded with a handheld infrared thermometer were 8 times more likely to be

diagnosed with deep wound infection in a recent prospective cross sectional study.40
STUDY PURPOSE
As increased local temperature is one of the classic signs of wound infection, its
quantitative measurement may have the potential to aid in assessment and diagnosis
of chronic wound infection at the bedside. Infrared thermometry is currently used for
this purpose in some chronic wound care settings, but strong research evidence to
support such use is difficult to find. This clinical study was conducted in an attempt to
quantify the relationship between increased periwound skin temperature and wound
infection in chronic leg ulcers. It was also a validation study for the potential use of a
handheld infrared thermometer to assess chronic leg ulcers for the presence of
infection.
METHODS AND PROCEDURES
A cross-sectional study design provided both an economical and time-efficient method
of investigation. Central Ethics Board approval was obtained, and consenting
participants were purposefully recruited from within the patient population of a chronic
wound clinic using specific inclusion/exclusion criteria (Figure 4). Participants with

febrile illness, active inflammatory disorders, pedal pulses or absent,

or with a history of peripheral vascular disease (ankle brachial
pressure index <0.8 mm Hg) were excluded. The level of significance
was set at .05 and confidence interval at 95%.
Figure 4. INCLUSION/EXCLUSION CRITERIA
Group 1 (nonwounded): Confirming reliability of the infrared thermometer used for
this study and identification of a "normal" skin temperature range were necessary. For
this purpose, participants without wounds (n = 20) were recruited, and their
demographic characteristics recorded (Figure 5).
Figure 5. PARTICIPANT CHARACTERISTICS
Under consistent environmental conditions, skin temperature data were obtained from
both target and contralateral control sites (dorsum of the foot, anterior ankle, 10 cm
above the ankle, and 10 cm below the knee) on the nonwounded study participants
(group 1a). These sites were marked with a pen, and taking of skin temperatures was
then repeated by a second blinded researcher within 5 minutes' time (group 1b).
These 2 sets of skin temperature readings provided test-retest data with which to
calculate correlation.
Group 2 (with wounds): Participants with chronic leg ulcers (n = 40) were also
recruited, and their demographic and wound characteristics recorded (Figures 5 and
6).
Figure 6. WOUND CHARACTERISTICS
Under consistent environmental conditions, skin temperature data were obtained from
both periwound and contralateral control sites. Subjects with wounds (n = 40) were
initially stratified into 20 subjects with greater than 2[degrees] F difference and
20subjects with less than 2[degrees] F difference between their highest periwound
skin temperature and an equivalent contralateral control site. A 3[degrees] to
4[degrees] F difference in skin temperature had been observed to be associated with
signs and symptoms of infection in a variety of chronic wound types within our large

chronic wound practice.

Diagnosis of wounds as noninfected or infected was determined by assessment of
clinical signs and symptoms using the CSSC plus clinical judgment and supplemented
with semiquantitative bacterial wound swab results. Subsequently, 18 wounds were
identified as noninfected (group 2a), and 22 as infected (group 2b).
STATISTICAL ANALYSIS
Statistical analysis was conducted using Microsoft EXCEL for Windows, SPSS version
14 for Windows, or the Online Statistical Calculator
(http://www.physics.csbsju.edu/stats/Index.html ). Assistance and direction
from a statistician were obtained.
Calculations for sensitivity, specificity, and positive and negative predictive values of
the clinical signs and symptoms of infection were made. Chi-squared contingency
tables were used to compare participant and wound characteristics.
Group 1 (nonwounded): Continuous/parametric statistics were obtained from skin
temperature readings in the nonwounded population (groups 1a and 1b) using the
test-retest design and analyzed to determine reliability of the infrared thermometer.
Values obtained from the left leg were subtracted from those obtained from the right
leg to give a difference in temperatures in degrees Fahrenheit. All calculations were
based on this value. Mean, SD, and reliability/correlation coefficient or Pearson r were
calculated.
Group 2 (wounded): Interval/nonparametric statistics were obtained from skin
temperature readings in the study population with wounds. Wounds were identified as
either noninfected (group 2a, n = 18) or infected (group 2b, n = 22) as per the study
protocol. One-way analysis of variance permitted calculation of correlation between
categorical/parametric (infection) and interval/nonparametric (temperature) data.
RESULTS
The nonwounded participants tended to be younger and have less concomitant illness
or medication use than those with wounds. Forty-five percent (n = 9) identified a
medical history and concomitant medication use for conditions including previous
carcinoma (n = 2), thyroid dysfunction (n = 2), diabetes (n = 2), anemia (n = 1), and
hypercholesterolemia (n = 1), either alone or in combination. In addition, obesity (n =
2), venous stasis (n = 2), and lymphedema (n = 1) were identified comorbidities.
Twenty percent (n = 4) exhibited a mild to moderate level of hypertension at the
study visit of between 130/80 and 160/82 mm Hg.
Mean skin temperature readings ranged from 89.9[degrees] F (SD, 3.0[degrees] F) at
the foot to 90.7[degrees] F (SD, 2.2[degrees] F) at 10 cm below the knee and did not
vary significantly between the 4 sites tested. The mean difference in skin temperature
between sites for the first researcher was 0.388[degrees] F (SD, 1.10[degrees] F),
and for the second researcher, 0.333[degrees] F (SD, 1.02[degrees] F) (Figure 7).
Figure 7. TEST-RETEST SKIN TEMPERATURE DATA (NONWOUNDED)
The infrared thermometer was found to be reliable in this relatively "healthy"
nonwounded population under consistent environmental conditions (r = 0.939, P = .
000 at a 95% confidence interval) (Figure 7). This handheld instrument was easy to
use at the bedside.
No significant difference in participant or wound factors was found between the
wounded noninfected and infected groups (Figures 5 and 6).

All (100%) of the wounded noninfected group 2a had a diagnosis of venous disease.
Sixty-seven percent (n = 12) noted a medical history and concomitant medication use
for conditions including hypertension (n = 5), arthritis (n = 4), diabetes (n = 3), heart
disease (n = 3), and thyroid dysfunction (n = 1), either alone or in combination. In
addition, obesity (n = 1), previous deep vein thrombosis (n = 1), bilateral hip
replacements (n = 1), and bilateral knee replacements (n = 1) were identified
comorbidities. Fifty-six percent (n = 10) exhibited a mild to moderate level of
hypertension at the study visit of between 130/80 and 168/98 mm Hg.
Ninety-one percent (n = 20) of the wounded infected group 2b had a diagnosis of
venous stasis disease, whereas 1 participant had a chronic postsurgical and 1 had a
chronic postradiation leg ulcer. Seventy-three percent (n = 16) noted a medical history
and concomitant medication use for conditions including hypertension (n = 6),
coronary artery disease (n = 4), arthritis (n = 3), anemia (n = 3), thyroid dysfunction
(n= 1), asthma (n = 1), renal artery stenosis (n = 1), and previous radiation
treatment (n = 1), either alone or in combination. In addition, obesity (n = 2),
previous deep vein thrombosis and/or clotting disorder (n = 2), bilateral knee
replacements (n= 1), bilateral hip replacements (n = 1), and smoking (n = 1) were
identified comorbidities. Thirty-six percent (n = 8) exhibited mild hypertension at the
study visit of between 130/80 and 144/80 mm Hg.
The wounded noninfected group 2a presented with an average of 3 (range, 0-7) * signs
and symptoms of infection, whereas the infected group (group 2b) averaged 8 (range,
4-12)*. Signs and symptoms of infection most sensitive to infection included increased
serous exudate (sensitivity 1.0), wound breakdown/pocketing of wound base
(sensitivity 1.0),odor (sensitivity 1.0), temperature (sensitivity 0.94), discolored
granulation tissue (sensitivity 0.93), and purulent exudate (sensitivity 0.87). Those
most specific to infection included surrounding skin erythema (specificity 0.92),
delayed healing (specificity 0.86), and temperature (specificity 0.86) (Figure 8).
(*These are the number of signs and symptoms of infection present as per Gardner et
al's16 12 signs and symptoms of clinical infection.)
Figure 8. SUMMARY OF CLINICAL SIGNS AND SYMPTOMS
Fifty percent of the noninfected group and 77% of the infected group demonstrated
positive swab results of heavy growth and/or more than 1 bacterial species cultured
from the semiquantitative wound swabs (sensitivity, 0.77; specificity, 0.50; positive
predictive value, 65%; negative predictive value, 64%). This is slightly lower than the
79% sensitivity identified in literature,41-43 possibly due to clinical versus research
laboratory settings. Results were used as needed to guide antibiotic treatment (Figure
9).
Figure 9. SEMIQUANTITATIVE WOUND CULTURE SWAB RESULTS VS
INFECTION
Within the noninfected group, the bacteria most frequently cultured included
Staphylococcus aureus, either alone or in combination, followed by mixed or other
Gram-negative bacteria. The infected group most frequently had a culture positive for
mixed and/or other Gram-negative bacteria (Figure 10).
Figure 10. BACTERIOLOGY SUMMARY
The nonwounded group 2a (n = 18) demonstrated mean periwound skin temperatures
ranging from 86.8[degrees] F to 96.1[degrees] F (mean, 90.6[degrees] F [SD,
2.30[degrees] F]). The equivalent contralateral control site temperatures ranged from
85.3[degrees] F to 95.6[degrees] F (mean, 90.2[degrees] F [SD, 2.59[degrees] F]).
These were not found to be significantly different (P= .117). Skin temperature
difference between periwound and control sites ranged from -0.1[degrees] F to