A Non-Contact Device for Fast Screening of Wound Infections

Abstract

Screening for wound infection relies on the expertise of the provider. Clinical diagnosis of infections based on wound swab/biopsy results often take a few days and may not assess the full wound. There is a need for a non-invasive tool that can quickly and accurately diagnose wound infection. Leukocyte esterase strips are used to identify various infectious diseases. However, it is not clear whether infected wounds also have elevated leukocyte esterase activities as compared with non-infected wounds. To achieve the objective, a device was developed to detect elevated leukocyte esterase activities in wounds by measuring wound exudates adsorbed onto wound dressings in 3 minutes. The efficacy of the device in assessing leukocyte esterase activities across various chronic wounds was tested. Such measurements were unaffected by the type of underlying wound dressing. By correlating the device outputs with clinical adjudication of infection, we found that this device had high positive predictive values for diagnosing wound infection in a wide variety of chronic wounds. In addition, a positive device output increases the probability of detecting infected wounds while the negative device output reduces the probability of detecting infected wounds. This rapid non-contact and disposable diagnostic tool may serve as a rapid and accurate indication of infection in the chronic wound.

1. BACKGROUND

Clinical diagnosis of wound infections that are often associated with chronic wounds is based on the provider’s assessment of clinical evidence of infection and laboratory wound culture results. Wound culture results require at least 2 to 3 days by when the wound environment and infection could worsen.^1,2 Procedures to collect samples for culture involve contacting the wound which can be both inconvenient for the patient and subject to sampling errors. Since no device can rapidly confirm a doctor’s suspicion of infection at the patient’s bedside, it is imperative to develop a fast screening tool that can identify infected and non-infected wounds to assist the doctor in effectively treating the wound, ideally without contacting the patient.

Leukocytes, although generally associated with acute inflammation, are found in large numbers in chronically infected tissues due to the fact that they do not undergo apoptosis as they normally should, thereby adding to the chronicity of the wound. At the site of infection, some of the recruited monocytes and neutrophils may release various enzymes, cytokines, and growth factors, including non-specific esterase, leukocyte proteinase, monocyte chemoattractant proteins, myeloperoxidase etc.^3–9 Since the increase in leukocyte numbers have been associated with the elevated leukocyte esterase activity,^10,11 leukocyte esterase activities have been used to assess the incidence of some infections. In fact, several commercially available strips have been used for diagnosing urinary tract infection and peritonitis by detecting elevated leukocyte esterase within liquid milieu such as urine and peritoneal fluid, respectively^12,13 While these strips can detect leukocyte esterase that is dispersed in liquid milieu, they may not be suitable for assessing chronic wound infection for the following reasons. First, chronic wounds have a heterogeneous environment with wound infection and bacterial colonization residing only in certain areas of the wound.^14–16 Second, leukocytes recruited to the site of infection and their products, such as leukocyte esterase, may only be present at the sites with bacterial colonization.¹³ Third, application of these strips on chronic wounds may irritate the wound site, introduce contaminants to the wounds, and cause cross contamination between different areas of wounds. To overcome these drawbacks, we recently developed a technology to detect various biomolecules across the whole wounds in the wound exudates soaked on freshly discarded wound dressings, thereby avoiding any contact with the patients.¹⁶ The current study reports the use of a non-contact device – DETEC^® Esterase, which uses N-Tosyl-L-alanine 3-indoxyl ester as the chromogenic substrate that has also been reported earlier,¹⁷ as a non-contact device to assesses elevated leukocyte esterase activities on freshly discarded wound dressings for screening wound infections.

2. QUESTIONS ADDRESSED

Wound infections are clinically determined based on the provider’s judgment and laboratory tests that can take a few days to result by when the infection could become severe. Clinically used technology for urinary tract infections and peritonitis cannot be used to rapidly assess infection across large wound sizes. Hence, a portable, non-contact, quick diagnostic device has been developed that can assess various extents of leukocyte esterase activities in wound exudate adsorbed onto discarded wound dressing and thus indicate the presence or absence of wound infection.

3. EXPERIMENTAL DESIGN

Instruction for use of the device

The instruction for use of device can be found in the supplementary materials. Briefly, wound dressings were tested by placing them in the device (Figure 1A) after which color seen through the transparent top layer was compared with the color chart to visualize the elevated leukocyte esterase activities in wound exudates (Figure 1B). The dry dressing or dressings which could not produce visible damp patch on the devices were excluded from this study.

Figure 1.

(A) Schematic illustration of the use of DETEC^® Esterase to detect esterase activities on discarded wound dressings; and (B) the output reference chart

Evaluation of the device efficacy

The efficacy of the device was determined by testing freshly disposed wound dressings collected from anonymous patients during their visit at the Texas Health Arlington Memorial Wound Care and Limb Salvage Clinic Arlington, TX, USA. The wounds were imaged immediately upon removal of the dressings to identify infected regions of the wound and serve as a comparison for the elevated leukocyte esterase activity. The outline of wound exudate and the wound itself with infected regions were marked both on the digital image of the patient’s wound and on the leukocyte esterase distribution of the output. The presence or absence of elevated leukocyte esterase in the wound milieu could thus be identified and matched with the corresponding wound areas.

Effect of wound dressing on device performance

The effect of wound dressing on device performance was assessed using extracted leukocyte esterase solutions from discarded wound dressings. Briefly, freshly recovered wound dressings were tested with the device. Based on the device output (presence and absence of purple color), the dressings were then recorded to have high and low levels of esterase activities, respectively. To extract wound exudates with high and low leukocyte esterase activities from the dressings, 1cm × 1cm of the corresponding dressings were soaked in 2mL of simulated wound fluid (SWF) for 10 minutes.^16,18 The soaking solution was centrifuged at 2,460 rcf for 10 minutes to remove cell debris and dressing fibers. The wound exudate-containing supernatants with high and low levels of leukocyte esterase activities were then used for the following tests. To assess the influence of dressing on device measurement, 5mm × 5mm of various wound dressings (Curity Gauze sponges, PROMOGRAN PRISMA^™, AQUACEL^® Extra^™, and Hydrofera BLUE^®) were incubated in 100 μL of wound exudate at room temperature for 20 minutes. The dressing-incubated wound exudates were then tested using DETEC^® Esterase device and their results were recorded.

Clinical study design and participants

A double-blind, observational, cohort study was conducted on 888 wound dressings from patients presenting with chronic wounds at the clinic between May 2019 and January 2021 in accordance with the Declaration of Helsinki as revised in 2013. All patient identifying information was anonymized and the device output was not provided to the provider who adjudicated infections and treated the patients based on the clinical observations and laboratory results. The device output (presumptive infected or not infected) was compared with this clinical adjudication. Patients with dry dressings or without same-day laboratory results were excluded from the final analysis.

Statistical Considerations

To test the effect of wound dressings on the device output, a 2×2 contingency table was constructed to compare the device output of wound exudates with and without dressing exposure. A Fisher’s exact test for association was conducted and results were determined to be significant at p<0.01.

Based on the clinical data, the device performance characteristics were determined as follows:

$$\begin{gathered} \text{ Sensitivity }=\text{ True Positive }/\text{ (True Positive }+\text{ False Negative) } \\ \text{ Positive predictive value }(\text{PPV})=\text{ True Positive }/\text{ (True Positive }+\text{ False Positive)} \\ \text{ Specificity }=\text{ True Negative }/\text{ (True Negative }+\text{ False Positive) } \\ \text{ Negative predictive value }(\text{NPV})=\text{ True Negative }/\text{ (True Negative }+\text{ False Negative) } \\ \text{ Accuracy }=\text{ True Negative }+\text{ True Positive }/\text{ Total Population } \\ \text{ Pre-test infection probability }=(\text{ True Positive }+\text{ False Negative })/\text{ Total Population } \\ \text{ Post-test infection probability with positive test result }=\text{ True Positive }/\text{ (True Positive }+\text{False Positive) } \\ \text{ Post-test infection probability with negative test result }=\text{ False Negative }/\text{ (True Negative }+\text{ False Negative) } \\ \text{ Increase in risk of infection among positive test result }=\text{ Post-test infection probability with positive test result - Pre-test infection probability} \end{gathered}$$

4. RESULTS

Evaluation of the device efficacy

As shown in the representative images of application of DETEC^® Esterase device on wounds with and without infection (Figure 2), a clear difference in purple color intensity was seen under these two circumstances. Specifically, device output for infected wounds showed several areas with dark purple color indicating elevated esterase activities in corresponding wound areas (Figure 2A). In contrast, the device output for wound dressings from non-infected wounds did not indicate elevated leukocyte esterase. These results support the use of DETEC^® Esterase in surveying leukocyte esterase activities over the entire wound. One can also infer that sporadic distribution of elevated leukocyte esterase activity may not be accurately detected using a single-spot testing such as the Urinalysis strips.

Figure 2.

Characterization of DETEC^® Esterase device. (A) Panel of images showing application of DETEC^® Esterase device to assess leukocyte esterase activities through infected and non-infected regions on the wounds, mapped on to the wound exudate on various dressings and device outputs; (B) Influence of different types of wound dressings on the accuracy of DETEC^® Esterase device output

Effect of wound dressings on device performance

The leukocyte esterase activities in wound exudates with and without dressing exposure were determined using DETEC^® Esterase. Indeed, various wound dressings had no significant effect on device output as the p values calculated through Fisher’s exact test were 1.00 (Figure 2B). The results support that this device can be used to evaluate the esterase activity within the human wound milieu regardless of the type of wound dressing used.

Clinical evaluation of the device for diagnosing wound infection

The device was clinically evaluated through a prospective observational clinical study for which the screening and enrollment processes are illustrated in the flowchart (Figure 3). Patient demographic and wound details are summarized in Table S1. Clinical diagnosis of infection was made by providers based on wound etiology, assessment, and laboratory results. The device outcome was correlated with this clinical diagnosis of infection.

Figure 3.

Screening and enrollment of subjects in the study

We first investigated the capability of DETEC^® Esterase device to identify wound infection in those wounds with clinical signs of infection. Fifty-two wounds were included in this test. Comparison of DETEC^® Esterase device output with provider’s diagnosis showed that the device had good sensitivity (84.78%), specificity (83.33%), and accuracy (84.62%) (Table 1A). The high PPV (97.50%) and low NPV (41.67%) suggest that a positive test result has a good chance to identify wound infection but wounds with negative test results may not be truly free of wound infection. Moreover, the pre-test infection probability (based on the clinical signs of infection) was 88.46%, and with the DETEC^® Esterase device output, the post-test infection probability among patients with positive test increased by 9.04% to 97.50% (Table 1A). However, a negative test still left a 58.33% probability of wound infection.

Table 1.

Clinical evaluation of DETEC^® Esterase on dressing (A) from patients with suspected infection by comparing device output and clinical adjudication, (B) from patients without suspicion of infection, and (C) from all patients regardless of suspicion of infection.

A
Clinical adjudication	Device output		Total # of patients
	Yes	No
Yes	39	7	46
No	1	5	6
Total # of patients	40	12	52
Sensitivity: 84.78% (95% CI= 71.13% to 93.66%) Specificity: 83.33% (95% CI= 35.88% to 99.58%) Accuracy: 84.62% (71.92% to 93.12%) Positive Predictive Value: 97.50% (95% CI= 86.65% to 99.58%) Negative Predictive Value: 41.67% (95% CI = 24.85% to 60.68%) Pre-test infection probability/prevalence: 88.46% Post-test infection probability with positive test result: 97.50% Post-test infection probability with negative test result: 58.33% Increase in risk of infection among positive test result: 9.04%

B
Clinical adjudication	Device output		Total # of patients
	Yes	No
Yes	53	9	62
No	12	15	27
Total # of patients	65	24	89
Sensitivity: 85.48% (95% CI= 74.22% to 93.14%) Specificity: 55.56% (95% CI= 35.33% to 74.52%) Accuracy: 76.40% (95% CI= 66.22% to 84.76%) Positive Predictive Value: 81.54% (95% CI= 74.10% to 87.21% Negative Predictive Value: 62.50% (95% CI = 45.49% to 76.90%) Pre-test infection probability/prevalence: 69.66% Post-test infection probability with positive test result: 85.48% Post-test infection probability with negative test result: 37.50% Increase in risk of infection among positive test result: 15.82%

C
Clinical adjudication	Device output		Total # of patients
	Yes	No
Yes	92	16	108
No	13	20	33
Total # of patients	105	36	141
Sensitivity: 85.19% (95% CI= 77.06% to 91.29%) Specificity: 60.61% (95% CI= 42.14% to 77.09%) Accuracy: 79.43% (95% CI= 71.82% to 85.77%) Positive Predictive Value: 87.62% (95% CI= 82.15% to 91.59% Negative Predictive Value: 55.56% (95% CI =42.40% to 67.97%) Pre-test infection probability: 76.60% Post-test infection probability with positive test result: 87.62% Post-test infection probability with negative test result: 44.44% Increase in risk of infection among positive test result: 11.02%

D
Device outputs vs. Clinical observations	DETEC Esterase	Clinical signs of infection
Sensitivity	85.19%	42.59%
Specificity	60.61%	81.82%
Accuracy	79.43%	49.65%

We then analyzed the probability of DETEC^® Esterase device to identify infected wounds in those wounds without clinical signs of infection. Eighty-nine wounds were included in this test. By comparing DETEC^® Esterase device output with provider’s diagnosis, we find that the device had good sensitivity (85.48%) and accuracy (76.40%) while the device specificity is a bit low (55.56%) (Table 1B). We also find that the device results have high PPV (81.54%) and low NPV (62.50%) as discussed above. Moreover, the pre-test infection probability is 69.66% which represent the probability of the infected wounds among those wounds without the clinical signs of infection. As expected, with the DETEC^® Esterase device output, the post-test infection probability among patients with positive test is increased by 15.82% to 85.48% (Table 1B). However, a negative test still left a 37.50% probability of wound infection.

Finally, we explored the possibility of using the device as a tool to screen all wounds by combining all wounds (with and without signs of infection). One hundred and forty-one wounds were included in the test. Our results show that DETEC^® Esterase had good sensitivity (85.19%) and accuracy (79.43%). It had lower specificity (60.61%) for all wounds as compared to those wounds with signs of infection (Table 1C). Equally important, the device had high PPV (87.62%) with an improvement in NPV (55.56%). Moreover, the pre-test infection probability is 76.60%, and with the DETEC^® Esterase device output, the post-test infection probability among patients with positive test result increased by 11.02% to 87.62%. It should be noted that a negative test still left a 44.44% probability of wound infection.

5. CONCLUSIONS & PERSPECTIVES

Here we report the creation of a non-contact device for the diagnosis of wound infection. The device is designed to assess leukocyte esterase activity that is often associated with leukocyte recruitment during the course of infection. In fact, leukocyte esterase has been used to detect infection in various clinical and research settings.^{12,13,19–23} While commercial esterase strips such as Mission Urinalysis Strips and Serim^® GUARDIAN^™ PeriScreen are available for infectious diseases, such as urinary tract infection, and peritonitis respectively,^19,24 they are limited by the measurable area and inability to present a planar view of infection on the wound landscape. Infected chronic wound environments are non-homogenous with localized bacterial colonization.^14–16 To fill the gap, DETEC^® Esterase device was designed specifically to distinguish between infected and non-infected wounds without directly contacting the wounds. As the wound dressings had direct contact with wound exudates in the wound bed prior to testing with our device, the esterase activity on the wound dressings is likely similar to that in the wound bed. By testing wound exudates absorbed on wound dressing, our test can potentially bypass the need for wound swab and biopsy which can be invasive and are not always reliable.^1,2 Furthermore, it produces results in less than 3 minutes without specialized equipment or facilities making it a powerful tool for quick diagnosis of wound infection at the bed side.

Since the devices contain identical amounts of the chromophore, the production of “purple” color has direct relationship with the presence of “enzymes”- leukocyte esterase – in wound exudates. There is a linear relationship between the enzyme amounts/concentrations and the “purple” color intensity. In fact, similar working principle has also been used by Mission Urinalysis Strips and Serim® GUARDIAN^™ PeriScreen which are available for urinary tract infection, and peritonitis respectively. Even though the device is designed to assess the extent of leukocyte esterase activities in wound exudates, the device chromogenic substrate, N-Tosyl-L-alanine 3-indoxyl ester, can also react with many other esterase enzymes, such as lymphocyte esterase, macrophage non-specific esterase, eosinophil esterase, basophil esterase, megakaryocyte esterase, platelet esterase, and blast esterase.²⁵ Furthermore, three types of non-specific esterase and a serine protease –elastase, that is expressed by neutrophils, participates in the wound healing process.^3,26,27 Due to our current knowledge gap on the specific enzymes present in the wound exudates that can react with the chromogenic substrate, we are unable to determine the specific enzyme threshold at the present time.

Our results support the overall objective of designing DETEC^® Esterase device as a diagnostic tool to compliment standard clinical diagnosis of wound infection. The sensitivity and PPV of the device highlight its clinical relevance in the detection of infection and its service as an aid to confirm the clinical determination of wound infection. We also found that regardless of the wound having signs of infection, DETEC^® Esterase had high PPV for identifying such infected wounds. These observations are also corroborated by an earlier study which used leukocyte esterase test strips for detection of infection, albeit in the synovial fluid.¹² Our observation of higher sensitivity compared to device specificity is typical of leukocyte esterase tests.¹⁹ Furthermore, our positive output can improve the diagnosis of infected wounds for 9–16%. Since there is no device that can diagnose wound infection at point-of-care, we compared the accuracy of our device output with the clinical observations which are based on the clinical signs of infection. As shown in Table 1D, our device output has better sensitivity than diagnosis based on the clinical signs of infection. However, diagnosis based on the clinical signs of infection has better specificity than our device output. Most importantly, the new results show that the accuracy of wound infection diagnosis based on the output of our device is substantially better (79.43% vs. 49.65%) than that based on the clinical signs of infection. The results lend support to the use of the device in improving the diagnosis of infected wounds in clinical settings. While the wound status cannot be concluded based on this device alone, DETEC^® Esterase devices can be used as an aid for providing rapid infection diagnosis to assist wound care providers with their clinical assessment of infection in wounds. The combined enhanced infection diagnosis and fast device output may significantly improve wound care by shortening the wait time required for laboratory report/clinical adjudication prior to the treatment. The identification of the areas with infected wounds may be validated in a future study for helping the wound care specialists to pay special attentions to the areas with elevated esterase when the treatments will be applied.

Even though the device showed convincing performance in wound infection diagnosis, there are still scenarios where the device is not applicable. For example, since wound infection is diagnosed through the detection of esterase within the superficial wound fluid adsorbed on the wound dressings, future studies will evaluate the usability of the device with deeper tissue infection, such as fasciitis, cellulitis, sepsis, and osteomyelitis. Moreover, dry dressings were not suitable for the device since it could possibly lead to a false negative result. In fact, about 24% of the patients were excluded from the study due to dry wound dressings. It must be noted that the device only presumptively indicates if the wound is infected or not and therefore, informs the provider who can decide on further treatment. The reliability of mapping the output to the wound surface has not been validated in this study. Importantly, the use of the device does not impinge on any routine procedures in the wound clinic and can be used with various wound dressings. Overall, the further development of this device can provide an invaluable tool for active management of wound infections.